PSYCHOTHERAPY

RELATIONSHIPS
THAT WORK
V olum e 2: E v id e n c e -B a s e d
T h e r a p is t R esp o n siv en ess

THIRD EDITION

Edited By
JOHN C. NORCROSS
BRUCE E. WAMPOLD
Psychotherapy Relationships
That Work
Psychotherapy Relationships
That Work
Volume 2: Evidence-Based

Therapist Responsiveness
TH IRD E D IT io N

Edited by John C. Norcross

and
Bruce E. Wampold

OXFORD
U N IV E R SITY PRESS
 0^ 0^
U N IV E R S IT Y PRESS

Oxford University Press is a department of the University of Oxford. It furthers

the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.

Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States o f America.

© John C. Norcross 2019

Second Edition published in 2011

Third Edition published in 2019

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.

You must not circulate this work in any other form

and you must impose this same condition on any acquirer.

CIP data is on file at the Library of Congress

ISBN 9 7 8 -0-19-084396-0

1 3 5 7 9 8 6 4 2
Printed by Sheridan Books, Inc., United States of America
 Dedicated to
Larry E. Beutler and Jerome D. Frank
pioneers in fitting psychotherapy to each patient

“My position is not that technique is irrelevant to outcome. Rather, I maintain

that . . . the success of all techniques depends on the patient’s sense of alliance with
an actual or symbolic healer. This position implies that ideally therapists should
select for each patient the therapy that accords, or can be brought to accord, with the
patient’s personal characteristics and view of the problem.”
—Jerome D. Frank, Persuasion and Healing (1991, p. xv)
 CONTENTS

Preface ix

About the Editors xv

Contributors xvii

1. Evidence-Based Psychotherapy Responsiveness: The Third Task Force 1

John C. Norcross and Bruce E. Wampold

2. Attachment Style 15
Kenneth N. Levy, Benjamin N. Johnson, Caroline V. Gooch, and Yogev Kivity

3. Coping Style 56
Larry E. Beutler, Christopher J. Edwards, Satoko Kimpara, and Kimberley Miller

4. Cultural Adaptations and Multicultural Competence 86

Alberto Soto, Timothy B. Smith, Derek Griner, Melanie Domenech Rodríguez, and
Guillermo Bernal

5. Gender Identity 133

Stephanie L. Budge and Bonnie Moradi

6. Preferences 157
Joshua K. Swift, Jennifer L. Callahan, Mick Cooper, and Susannah R. Parkin

7. Reactance Level 188

Christopher J. Edwards, Larry E. Beutler, and Kathleen Someah

8. Religion and Spirituality 212

Joshua N. Hook, Laura E. Captari, William Hoyt, Don E. Davis, Stacey E. McElroy, and
Everett L. Worthington Jr.

Vii
 viii CONTENTS

9. Sexual Orientation 264

Bonnie Moradi and Stephanie L. Budge

10. Stages of Change 296

Paul Krebs, John C. Norcross, Joseph M. Nicholson, and James O. Prochaska

11. Personalizing Psychotherapy: Results, Conclusions, and Practices 329

John C. Norcross and Bruce E. Wampold

Index 343
 PREFACE

A warm welcome to the third edition of Psychotherapy Relationships That Work. This
book seeks, like its predecessors, to identify effective elements of the psychotherapy re­
lationship and to determine effective methods of adapting or tailoring that relationship
to the individual patient. That is, we summarize the research evidence on what works
in general as well as what works in particular.
This dual focus has been characterized as “two books in one,” one book on relation­
ship behaviors and one book on adapting therapy to patients, under the same cover. In
this third edition, we separate those “two books” into two volumes as the number of
chapters and the amount of research have grown considerably over the past decade. This
volume 2 features evidence-based therapist responsiveness to patient transdiagnostic
characteristics; volume 1 addresses evidence-based therapist contributions to the re­
lationship. As we move from volume 1 to volume 2, we transition from relationship
elements to treatment adaptations or interpersonal responsiveness. We also move from
primarily correlational research designs to randomized clinical trials.
Our hope in this book, as with the earlier editions, is to advance a rapprochement
between the warring factions in the culture wars of psychotherapy and to demonstrate
that the best available research clearly shows the efficacy of tailoring psychotherapy
to the individual client and context. Adapting brand-name treatments to particular
disorders does not begin to address the complexity of psychotherapy nor does it lev­
erage the central sources of healing: the therapeutic relationship and the patient him-
or herself.

CHANGES IN THE NEW EDITION

The aims of this third edition of Psychotherapy Relationships That Work remain the
same as its predecessors, but its sponsorship, format, and editorship differ some­
what. This edition was overseen by an interdivisional Task Force on Evidence-Based
Relationships and Responsiveness co-sponsored by the Society for the Advancement
of Psychotherapy (Division 29 of the American Psychological Association [APA]) and
the Society of Counseling Psychology (Division 17 of the APA). This edition also boasts
a practice-friendlier smaller trim size (the physical size of the book). As noted, we have
expanded the book into two volumes, each now co-edited by a prominent psycho­
therapy research (Bruce Wampold on this volume and Michael Lambert on volume 1).
We have expanded the breadth of coverage. New reviews were commissioned on tai­
loring psychotherapy to the client’s gender identity, sexual orientation, and functional

ix
 x PREFACE

impairment, although the latter did not survive the rigors of the meta-analytic pro­
cess and the pressing timeline of the project. O f course, updated meta-analyses were
conducted on all returning chapters.
The content o f each chapter has also expanded. Five new sections appear in
each chapter: landm ark studies, results o f previous meta-analyses, evidence for
causality, diversity considerations, and training implications. These additions help
readers appreciate the context of the research evidence and unpack its results, espe­
cially for treating diverse clients and training psychotherapy students. The causal
evidence for these treatm ent adaptations or therapist responsiveness has increased
steadily over the years; the authors of the m ajority of these meta-analyses can
now legitimately claim that these adaptations cause improvements in treatm ent
outcomes.
The net result is a compilation of nine, original cutting-edge meta-analyses on what
works in personalizing psychotherapy to the individual client and his or her singular
situation. This new edition, appearing nine years after the last incarnation, also proves
more practical and reader- friendly: more clinical examples, sections on landmark
studies, and ending with bulleted recommendations for clinical practice at the end of
each chapter.

PROBABLE AUDIENCES

In planning the first edition of the book more than 20 years ago, we struggled to
identity the intended audiences. Each of psychotherapy’s stakeholders—patients,
practitioners, researchers, educators, students, organizations, insurance companies,
and policymakers— expressed different preferences for the content and length of the
volume.
We prepared Psychotherapy Relationships That Work for multiple audiences but
in a definite order of priority. First came clinical practitioners and trainees of di­
verse theoretical orientations and professional disciplines. They need to address
urgent pragmatic questions: W hat do we know from the research evidence about
cultivating and maintaining the therapeutic relationship? W hat are the research-
supported means of adapting treatment to the individual patient beyond his or her
diagnosis?
Our second priority was accorded to the mental health disciplines themselves,
specifically those committees, task forces, and organizations promulgating lists of
evidence-based practices or treatment guidelines. We hope our work will inform and
balance any efforts to focus exclusively on techniques or treatments to the neglect of
the humans involved in the enterprise.
Our third priority were payors (health insurance companies, managed care orga­
nization, and government entities) and accreditation organizations, many of which
have unintentionally devalued the person of the therapist and the personalization of
psychotherapy. Finally, this book is intended for psychotherapy researchers seeking a
central resource on these urgent matters.
 xi PREFACE

ORGANIZATION OF THE BOOK

Our opening chapter introduces the book by outlining the purpose and history of the
interdivisional task force and its relation to previous efforts to identify evidence-based
practices in psychotherapy. That chapter also presents the key limitations of our work.
The heart of the book is composed of original meta-analyses and systematic reviews
on adapting psychotherapy to patient’s transdiagnostic characteristics, or what we call
“creating a new therapy for each patient.” It features nine chapters on fitting treatment
methods and relational behaviors to a client’s attachment style, coping style, racial/
ethnic culture, gender identity, therapy preferences, reactance level, religious/spiritual
commitment, sexual orientation, and stage of change. For good measure, the chapter
on cultural identity also explores the relation of the therapist’s cultural competence to
treatment outcome.
The concluding chapter presents the Task Force conclusions on evidence-based
relationships and responsiveness and our reflections on what works and what doesn’t.
We also feature 28 Task Force recommendations, divided into general, practice,
training, research, and policy recommendations.

CHAPTER GUIDELINES
Except for the bookends (Chapters 1 and 11), all chapters use the same section headings
and follow a consistent structure, as follows:

♦ Introduction (untitled). Introduces the patient transdiagnostic characteristics and the

chapter in a few, reader-friendly paragraphs.
♦ Definitions. Defines in theoretically neutral language the patient characteristic
and identifies any highly similar or equivalent constructs from diverse theoretical
traditions.
♦ Measures. Reviews the popular measures used in the research and included in the
ensuing meta-analysis.
♦ Clinical Examples. Provides several concrete examples of the patient characteristic
being reviewed. Portions of psychotherapy transcripts are encouraged here while
protecting the privacy of patients.
♦ Landmark Studies. Walks the reader through two to four landmark studies on the
topic (including one qualitative study), describing their design, participants, and
results.
♦ Results of Previous Meta-Analyses. Offers a quick synopsis of the findings of
previous meta-analyses and systematic reviews on the topic.
♦ Meta-Analytic Review. Conducts an original meta-analysis of all available studies
employing a random effects model. Systematically compiles all available empirical
studies on the psychotherapy adaptation of the client characteristic to distal, end-
of-treatment outcome in the English language (and other languages, if possible).
Includes only actual psychotherapy studies and no analogue studies. Uses the Meta­
Analysis Reporting Standards as a general guide for the information to include.
 xii PREFACE

Performs and reports a test of heterogeneity. Reports the effect size as d or g (or other
standardized mean difference). Includes a fail-safe statistic to address the file-drawer
problem and provides a table or funnel plot for each study in the meta-analysis.
♦ Moderators and Mediators. Presents the results of the potential mediators and
moderators in the meta-analysis. Examples include year of publication, rater
perspective (assessed by therapist, patient, or external raters), therapist variables,
patient factors (including cultural diversity), different measures of the relationship
element and treatment outcome, time of assessment (when in the course of therapy),
and type of psychotherapy/theoretical orientation.
♦ Evidence for Causality. Summarizes the evidence demonstrating that the adaptation
or match causally contributes to treatment outcome. How strong is the extant
research for a causal link?
♦ Limitations of the Research. Points to the major limitations of the research
conducted to date. A concise paragraph or two here on future research directions is
also sometimes included.
♦ Diversity Considerations. Addresses how dimensions of diversity (e.g., gender, race/
ethnicity, sexual orientation, socioeconomic status) fare in the research studies and
the meta-analytic results.
♦ Training Implications. Explicates briefly the take-home points of the meta-analysis
for clinical educators and supervisors. Mentions any training resources or programs
that have a proven record of teaching the treatment adaptation or responsiveness.
♦ Therapeutic Practices. Places the emphasis here squarely on what works. Practice
implications from the foregoing research, primarily in terms of the therapist’s
contribution and secondarily in terms of the patient’s perspective, are set as bulleted
points.

ACKNOWLEDGMENTS

Psychotherapy Relationships That Work would not have proved possible without two
decades of organizational and individual support. On the organizational front, the
board of directors of the Society for the Advancement of Psychotherapy (APA Division
29) consistently supported the interdivisional task force, and the Society of Counseling
Psychology (APA Division 17) co- sponsored this iteration. At Oxford University
Press, Sarah Harrington and Joan Bossert shepherded these books through the pub­
lishing process and recognized early on that they would complement Oxford’s land­
mark Treatments That Work. We are particularly appreciative of Oxford’s flexibility
in copyright matters that have enabled us to disseminate farther these consequential
meta-analytic results. Their commitment to spreading the knowledge in this book
through special journal issues and other practitioner publications, even at the expense
of their occasional loss of book sales, is noble and probably unprecedented in pub­
lishing circles.
On the individual front, many people modeled and manifested the ideal thera­
peutic relationship throughout the course of the project. The authors of the respective
 xiii preface

chapters, of course, were indispensable in generating the research reviews and were
generous in sharing their expertise. The Steering Committee of the previous task forces
assisted in canvassing the literature, defining the parameters of the project, selecting
the contributors, and writing the initial conclusions. We are grateful to them all: Steven
J. Ackerman, Lorna Smith Benjamin, Guillermo Bernal, Larry E. Beutler, Franz Caspar,
Louis Castonguay, Charles J. Gelso, Marvin R. Goldfried, Clara Hill, Mark J. Hilsenroth,
Michael J. Lambert, David E. Orlinsky, Jackson P. Rainer, and Bruce E. Wampold.
For this task force and this edition of the book, we thank the following for serving
on the Steering Committee:

Franz Caspar, PhD, University of Bern

Melanie M. Domenech Rodriguez, PhD, Utah State University
Clara E. Hill, PhD, University of Maryland
Michael J. Lambert, PhD, Brigham Young University
Suzanne H. Lease, PhD, University of Memphis (representing APA Division 17)
James W. Lichtenberg, PhD, University of Kansas (representing APA Division 17)
Rayna D. Markin, PhD, Villanova University (representing APA Division 29)
John C. Norcross, PhD, University of Scranton (chair)
Jesse Owen, PhD, University of Denver
Bruce E. Wampold, PhD, University of Wisconsin and Modum Bad Psychiatric
Center

We gratefully dedicate this book to Larry Beutler and Jerome Frank, pioneers in
researching and advancing responsiveness in mental health. In both word and deed,
they have inspired others and us to personalize or fit the treatment to each client.
Last but never least, our immediate families tolerated our absences, preoccupations,
and irritabilities associated with editing this book. They did so with a combination of
empathy and patience that would do any seasoned psychotherapist proud.

John C. Norcross and Bruce E. Wampold

 A B O U T TH E EDITORS

John C. Norcross, PhD, ABPP, is Distinguished Professor of Psychology at the University

of Scranton, Adjunct Professor of Psychiatry at SUNY Upstate Medical University,
and a board-certified clinical psychologist. His recent books include Clinician’s
Guide to Evidence-Based Practice in Behavioral Health and Addictions, Systems o f
Psychotherapy: A Transtheoretical Analysis, Self-Help That Works, Psychologists’ Desk
Reference, History o f Psychotherapy, Changeology, and a dozen editions of the Insider’s
Guide to Graduate Programs in Clinical & Counseling Psychology. He has served as
president of the American Psychological Association (APA) Division of Clinical
Psychology, the APA Division of Psychotherapy, and the Society for the Exploration
of Psychotherapy Integration. Dr. Norcross has received multiple professional awards,
such as APA’s Distinguished Career Contributions to Education & Training Award,
Pennsylvania Professor of the Year from the Carnegie Foundation, and election to the
National Academies of Practice.

Bruce E. Wampold, PhD, ABPP, is Director of the Research Institute at Modum Bad
Psychiatric Center in Vikersund, Norway, and Emeritus Professor of Counseling
Psychology at the University of W isconsin-Madison. He is a Fellow of the APA
(Divisions 12, 17, 29, 45), is board certified in counseling psychology by the American
Board of Professional Psychology, and is 2019 president of the Society for Psychotherapy
Research. He is the recipient of the Distinguished Professional Contributions to
Applied Research Award from the APA and the Distinguished Research Career Award
from the Society for Psychotherapy Research. Currently his work, summarized in The
Great Psychotherapy Debate: The Evidence fo r What Makes Psychotherapy Work (with
Z. Imel), involves understanding psychotherapy from empirical, historical, social, and
anthropological perspectives.

xv
 CONTRIBUTORS

Guillermo Bernal, PhD Caroline V. Gooch, BS

Clinical Psychology Program, Carlos Department of Psychology, Ohio
Albizu University University

Larry E. Beutler, PhD Derek Griner, PhD

Department of Clinical Psychology, Palo Counseling and Psychological Services,
Alto University Brigham Young University
School of Education, University of
Joshua N. Hook, PhD
California, Santa Barbara
Department of Psychology, University of
Stephanie L. Budge, PhD North Texas
Department of Counseling Psychology,
William Hoyt, PhD
University of Wisconsin-Madison
Department of Counseling Psychology,
Department of Orthopedics and
University of Wisconsin-Madison
Rehabilitation, University of
Wisconsin Hospitals and Clinics Benjamin N. Johnson, M S
Department of Psychology, Pennsylvania
Jennifer L. Callahan, PhD
State University
Department of Psychology, University of
North Texas Satoko Kimpara, PhD
Department of Clinical Psychology, Palo
Laura E. Captari, M A, MS
Alto University
Department of Psychology, University of
North Texas Yogev Kivity, PhD
Department of Psychology, Pennsylvania
Mick Cooper, PhD
State University
Department of Psychology, University of
Roehampton Paul Krebs, PhD
Department of Population Health,
Don E. Davis, PhD
New York University School of
Department of Counseling and
Medicine
Psychological Services, Georgia State
Department of Psychology, VA
University
New York Harbor Healthcare System
Christopher J. Edwards, M A, MS
Department of Clinical Psychology, Palo
Alto University

x v ii
 xvill CONTRIBUTORS

Kenneth N. Levy, PhD Melanie Domenech Rodríguez, PhD

Department of Psychology, Pennsylvania Department of Psychology, Utah State
State University University

Stacey E. McElroy, PhD Timothy B. Smith, PhD

Department of Counseling and Department of Counseling and Special
Psychological Services, Georgia State Education, Brigham Young University
University
Kathleen Someah, MS
Kimberley Miller, BA Department of Clinical Psychology, Palo
Department of Clinical Psychology, Palo Alto University
Alto University
Alberto Soto, PhD
Bonnie Moradi, PhD Department of Psychology &
Department of Psychology, University of Counseling, University of Central
Florida Arkansas
Center for Gender, Sexualities, and
Joshua K. Swift, PhD
Women’s Studies, University of Florida
Department of Psychology, Idaho State
Joseph M. Nicholson, MPH University
Medical Library, New York University
Bruce E. Wampold, PhD
School of Medicine
Modum Bad Psychiatric Center, Norway
John C. Norcross, PhD Department of Counseling Psychology,
Department of Psychology, University of University of Madison-Wisconsin
Scranton
Everett L. Worthington Jr., PhD
Department of Psychiatry, SUNY
Department of Psychology, Virginia
Upstate Medical University
Commonwealth University
Susannah R. Parkin, BS
Department of Psychology, Idaho State
University

James O. Prochaska, PhD

Department of Psychology and
Cancer Prevention Research Center,
University of Rhode Island
Psychotherapy Relationships
That Work
 1

e v id e n c e -b a sed p s y c h o t h e r a p y

r e s p o n s iv e n e s s : t h e t h ir d task fo r c e

John C. Norcross and Bruce E. Wampold

The need to adapt or fit psychotherapy to the individual patients has been univer­
sally recognized from the beginning of modern psychotherapy. As early as 1919, Freud
introduced psychoanalytic psychotherapy as an alternative to classical analysis based
on the recognition that the more rarified approach lacked universal applicability and
that many patients did not possess the requisite psychological-mindedness (Wolitzky,
2011). The mandate for individualizing psychotherapy was embodied in Gordon Paul’s
(1967) iconic question: What treatment, by whom, is most effective for this individual
with that specific problem, and under which set of circumstances? Every psychother­
apist recognizes that what works for one person may not work for another; we seek
“different strokes for different folks” (Blatt & Felsen, 1993).
To many, the means of such matching was to tailor the psychotherapy to the patient’s
disorder or presenting problem—that is, to find the best treatment method for a par­
ticular disorder. The research suggests that Treatment A for Disorder Z may prove dif­
ferentially useful for a handful of disorders, such as some form of exposure for trauma
and parent management training for childhood externalizing disorders. Although
some psychotherapies may make better marriages with some mental health disorders
(Barlow, 2014; Nathan & Gorman, 2015), the repeated Dodo Bird conclusion indicates
that bona fide psychotherapies produce similar outcomes, once the researchers alle­
giance effect is identified and controlled (Wampold & Imel, 2015).
Still, the overwhelming majority of randomized clinical trials in psychotherapy
compare the efficacy of specific treatments for specific disorders. Those research
studies problematically collapse numerous clients under a single diagnosis. It is a false
and, to be blunt, misleading presupposition in randomized controlled/clinical trials
(RCTs) that the patient sample is homogenous (Beutler & Clarkin, 1990). Perhaps the
patients are diagnostically homogeneous, but nondiagnostic variability is the rule. It
is precisely the unique individual and the singular context that many psychotherapists
attempt to “treat” (Norcross & Beutler, 2014).

i
 2 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

As every clinician knows, matching psychotherapy solely to a disorder is incom­

plete and not always effective. Particularly absent from much of the controlled research
and clinical training has been adapting psychotherapy to the person of the patient,
beyond his or her disorder. As Sir William Osler (1906), father of modern medicine,
wrote: “It is much more important to know what sort of a patient has a disease than
what sort of disease a patient has.” The accumulating research demonstrates that it
is indeed frequently effective to tailor or match psychotherapy to the entire person
(Norcross, 2011).
Hundreds of potential client characteristics have been proposed as markers for
using one type of treatment or style rather than another (Clarkin & Levy, 2004); how­
ever, it has been only in the past 20 years that the perennial quest for adapting psycho­
therapy to transdiagnostic patient characteristics on sound research has been fulfilled.
As manifested in the meta-analyses and systematic reviews in this book, multiple
methods of relational responsiveness or treatment adaptations have proven effective.
These rightfully carry the designation of evidence-based practices.
In this chapter, we introduce volume 2 of Psychotherapy Relationships That
Work devoted to evidence-based means of adapting psychotherapy to the patient’s
transdiagnostic characteristics. (The first volume features evidence-based elements
of the psychotherapy relationship.) We begin by reviewing the innumerable terms
accorded to this process of matching therapy and client. We summarize the purposes
and processes of the third interdivisional task force cosponsored by APA Division of
Psychotherapy and the APA Division of Counseling Psychology. We then discuss the
clinical and research process of determining what works for whom. The latter part of
the chapter features the limitations of the task force’s work and responds to frequently
asked questions.

A ROSE BY ANY NAME

The process of creating the optimal match in psychotherapy has been accorded multiple
names over the years. In alphabetical order, these terms include aptitude by treatment
interaction (a research design), attunement, customizing, differential therapeutics, fit­
ting, individualizing, matchmaking, personalizing, prescriptionism, responsiveness,
specificity factor, tailoring, therapy fit, treatment adaptation, and treatment selection.
In the professional literature, treatment adaptation and responsiveness tend to prevail;
we employ both terms interchangeably here in the interest of theoretical neutrality (ad­
aptation is favored by cognitive-behavior therapists, while responsiveness is favored by
relational, humanistic, and psychotherapies therapists). In clinical work, clients tend
to prefer the terms individualizing and personalizing as they are self-explanatory and
parallel language in personalized medicine (these conclusions hail from focus groups
on actual and potential psychotherapy clients).
By whatever name, the goal is to enhance treatment effectiveness by tailoring it to
the unique individual and his/her singular situation. In other words, psychotherapists
endeavor to create a new therapy for each client. They do so by capitalizing on both
the nomothetic and idiographic traditions: attuning psychotherapy to the particulars
 3 Evidence-Based Psychotherapy Responsiveness

of the individual according to the generalities of the research findings. And when we
speak of the individual, we naturally recognize that the person may be in individual
therapy or in a larger treatment format, such as a couple or in a group.
This position can be effortlessly misunderstood as an authority-figure therapist
prescribing a specific form of psychotherapy for a passive client. Far from it; the goal
is for an empathic therapist to collaboratively create an optimal relationship with an
active client on the basis of the client’s personality, culture, and preferences. When a
client frequently resists in session, for example, then the therapist considers whether
he or she is pushing something that the client finds incompatible (preferences) or the
client is not ready to make those changes (stage of change) or is uncomfortable with
a directive style (reactance). Clinicians strive to offer a therapy that fits or resonates
to the patient’s characteristics, proclivities, and worldviews—in addition to diagnosis.

THE THIRD INTERDIVISIONAL TASK FORCE

The dual purposes of the Interdivisional APA Task Force on Evidence-Based
Relationships and Responsiveness were to identify effective elements o f the therapy
relationship and to determine effective methods of adapting or tailoring therapy to the
individual patient on the basis of transdiagnostic characteristics. In other words, we
were interested in both what works in general and what works for particular patients.
The task force applies psychological science to the identification and promulgation
of effective psychotherapy. It does so by expanding or enlarging the typical focus of
evidence-based practice to treatment adaptations and therapy relationships. Focusing
on one area—in this volume, responsiveness—may unfortunately convey the impres­
sion that this is the only area of import. Thus, we take pains here and elsewhere in the
two volumes to repeat that that our work acknowledges the simultaneous contribution
of treatment methods to client success.
At the same time, decades o f careful research indicate that the patient, the therapy
relationship, and these transdiagnostic adaptations exercise more influence on outcome
than the particular treatment method. Put differently, culture eats strategy, as the famed
management consultant Peter Drucker has said. The relational ambience of psycho­
therapy and responsiveness to clients prove typically more powerful than the partic­
ular therapeutic method or strategy. We endorse Jerome Frank’s position, in his classic
Persuasion and Healing (Frank & Frank, 1991, p. xv), which we selected for our epigraph.

My position is not that technique is irrelevant to outcome. Rather, I maintain

that . . . the success of all techniques depends on the patient’s sense of alliance with
an actual or symbolic healer. This position implies that ideally therapists should
select for each patient the therapy that accords, or can be brought to accord, with
the patient’s personal characteristics and view of the problem.

We consulted psychotherapy experts, the research literature, and potential authors

to discern whether there were sufficient numbers of studies on a treatment adaptation
or matching strategy to conduct a systematic review and meta-analysis. Seven such
 4 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

client characteristics—attachment style, coping style, culture, preferences, reactance

level, religion/spirituality, and stages of change—returned from the second edition of
the book. We requested that a meta-analysis on functional impairment style also re­
turn from the first edition, but the authors could not accommodate the publication
schedule. Two client characteristics not represented in the previous edition—gender
identity and sexual orientation—were added, resulting in nine systematic reviews and
meta-analyses on their respective effectiveness.
Once these decisions were finalized, we commissioned original meta-analyses on
the relationship elements (volume 1) and the adaptation/responsiveness methods
(volume 2). Authors followed a comprehensive chapter structure (provided in the
Preface) and specific guidelines for their meta-analyses. The analyses quantitatively
analyzed the efficacy of adapting psychotherapy to that patient quality. Outcome was
primarily defined as distal posttreatment outcomes, which sharpened our focus on
“what works.”
The chapters and the meta-analyses therein were peer reviewed by at least two
editors and subsequently underwent at least one substantive revision. In particular,
the review established that the meta-analyses adhered to the Meta-Analysis Reporting
Standards and reported the requisite information (outlined in the Preface).
When the chapters were finalized, a 10-person expert panel (identified in the
Preface and in chapter 11) reviewed and rated the evidentiary strength of the effec­
tiveness of fitting psychotherapy to that client dimension. They did so according to
the following criteria: number of empirical studies, consistency of empirical results,
independence of supportive studies, magnitude of the effect size, evidence for causal
link, and the ecological or external validity of research. Using these objective criteria,
experts independently judged the strength of the research evidence as

Demonstrably Effective
Probably Effective
Promising but Insufficient Research to Judge
Important but Not Yet Investigated or
Not Effective

We then aggregated the individual ratings to reach a consensus conclusion on each ad­
aptation method. These conclusions are summarized in the last chapter of this volume,
as are 28 recommendations approved by all members of the steering committee.

WHAT WORKS FOR WHOM

As the field of psychotherapy has matured, the identical psychosocial treatment for all
patients is now recognized as inappropriate and, in select cases, perhaps even uneth­
ical. We will not progress, and our patients will not benefit, by imposing a Procrustean
bed onto unwitting consumers of psychological services. In his Foreword to the land­
mark Differential Therapeutics in Psychiatry (Frances et al., 1984), Robert Michels
(1984, p. xiii) summed it as follows:
 5 Evidence-Based Psychotherapy Responsiveness

The easiest way to practice psychiatry is to view all patients and problems as ba­
sically the same, and to apply one standard therapy or mix of therapies for their
treatment. Although some may still employ this model, everything we have learned
in recent decades tells us that it is wrong—wrong for our patients in that it deprives
them of the most effective treatment, and wrong for everyone else in that it wastes
scarce resources.

The clinical reality is that no single psychotherapy is effective for all patients and
situations, no matter how good it is for some. Evidence-based practice has come to
demand a flexible, if not integrative, perspective. One size fits all therapy is proving
impossible. O f course, as Michels notes, that would simplify treatment selection—give
every client the same psychotherapy!—but it flies in the face of what we know about
individual differences, patient preferences, and disparate cultures.
Imposing a parallel situation onto other healthcare professions drives the point
home (Norcross & Beutler, 2014). To take a medical metaphor, would you entrust
your health to a physician who prescribed the identical treatment (say, antibiotics or
neurosurgery) for every patient and illness encountered? Or, to take an educational
analogy, would you prize instructors who employed the same pedagogical method
(say, a lecture) for every educational opportunity? Or would you entrust your child
to a child care worker who delivers the identical response (say, a nondirective attitude
or a slap on the bottom) to every child and every misbehavior? “No” is probably your
resounding answer. Psychotherapy clients deserve no less consideration.
Concisely put, no theory is uniformly valid and no mechanism of therapeutic ac­
tion is equally applicable to all individuals. As a consequence, the goal is to select dif­
ferent methods, stances, and relationships according to the patient and the context.
The result is a more efficient and efficacious therapy— and one that fits both the client
and the clinician.
On the face of it, of course, virtually every clinician endorses matching the therapy
to the individual client. After all, who can seriously dispute the notion that psycholog­
ical treatment should be tailored to the needs of the individual patient? However, the
treatment adaptations in this volume go beyond this simple acknowledgment of the
need for flexibility in several ways (Norcross & Beutler, 2014). These adaptation or
responsiveness methods

♦ are derived directly from outcome research rather than from an idiosyncratic theory
or seat-of-the-pants experimentation.
♦ embrace the potential contributions of multiple systems of psychotherapy rather than
working from within a single theoretical system.
♦ are predicated on multiple transdiagnostic and occasionally diagnostic client
characteristics, in contrast to relying on patient diagnosis alone.
♦ offer optimal treatment methods and healing relationships, whereas many matches
historically focused narrowly on selecting methods.
♦ occur throughout the course of therapy, not only at pretreatment as a case
formulation.
 6 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

Psychotherapy matching is an old idea come to evidence-based fruition. The

decision-making may be expressed in a series of when . . . then statements. When the
client presents with this (feature), then consider doing this. In that way, the clinician
matches to the entire person, not simply diagnoses.
One of our favorite clinical tales of responsiveness comes from the late Arnold
Lazarus (1993), who related an illuminating incident with a Mrs. Healy, a middle-aged
woman who indicated she sought an assertive, humorous, and challenging relation­
ship with her therapist. When she first entered the office, she looked Lazarus up and
down and asked, “Why do you have graves outside your office?” In perfect Rogerian
style Lazarus responded, “I have graves outside my office?” “Look out the window,
dummy!” she replied. He went to the office window and looked out. Two new flower
beds had been installed alongside the front walk on the grass. It was early spring and
the shoots had yet to emerge from the soil. “Well, since you ask,” Lazarus replied, “I
have just buried one of my clinical failures in the one grave and the other is earmarked
for you Mrs. Healy if you turn out to be an uncooperative client.” The twinkle in her eye
told Lazarus that the response was an appropriate one. Had he responded in a stodgy
or serious way—“Oh, those are merely newly planted flower beds,”—he doubted
whether the necessary rapport would have developed, because she strongly valued
“people with a sense of humor.” Indeed, each session would start with some friendly
banter and jesting, followed by attention to the serious problems for which she sought
psychotherapy.
Others of his patients, Lazarus discovered, sought a good listener and only a good
listener. No interpretations, no advice, no self-disclosure. Different strokes for different
folks indeed. We all have similar clinical tales in which we met the patient where he
or she was.
The adaptation or responsiveness can be based on client-therapist similarity or com­
plementarity. In general, the research favors therapists adopting a complimentary style
of interaction—a theory of opposites for interpersonal dimensions. Highly resistant
clients usually benefit more from lower therapist directiveness, whereas low resistant
clients benefit from more guidance and direction (Chapter 6). In the Vanderbilt psy­
chotherapy research studies, for instance, Hans Strupp and associates (1986; Strupp,
1993) found that increasing anticomplementarity between therapist self-concept and
therapists’ perceptions of patient behavior was negatively associated with outcome. At
other times, research favors similarity for more direct, conscious patient requests, such
as cultural adaptations (Chapter 4), therapy preferences (Chapter 6), and religious/
spiritual accommodations (Chapter 8).
The number of permutations for every possible interaction or matching algo­
rithm among patient, therapist, and relationship would prove endless without some
empirical-driven guidance. The meta-analyses in this book delimit the universe of
possible client variables to a manageable number for clinical purposes. Research has
isolated a handful of readily assessed client features that are reliably associated with dif­
ferential responses to various styles of therapeutic relating. While consensus is no epi-
stemic warrant and while research will never provide definitive answers to all matching
questions (Mahoney & Norcross, 1993), we can enhance psychotherapy efficacy and
 7 Evidence-Based Psychotherapy Responsiveness

efficiency by means of the nine adaptation/responsiveness methods highlighted in this

volume.

EFFECT SIZES

The subsequent chapters feature original meta-analyses on the effectiveness of adapting

or responsively matching psychotherapy to a particular patient characteristic. Insisting
on quantitative meta-analyses for all the chapters (with the exceptions of the chapters
on gender identity and sexual orientation) enables direct estimates of the magnitude or
strength of effectiveness in the form of effect sizes. These are standardized differences
between two group means, say, psychotherapy and a control, divided by the (pooled)
standard deviation. The resultant effect size is in standard deviation units. Both Cohen’s
d and Hedges’ g estimate the population effect size.
All of the meta-analyses in this volume employed d or g. This increased the con­
sistency among the meta-analyses, enhancing their interpretability, and enabled di­
rect comparisons of the meta-analytic results to one another. In all of these analyses,
the larger the magnitude of d or g , the higher the probability of client success in
psychotherapy.
Table 1.1 presents several practical ways to interpret the effect sizes d and g in be­
havioral health. By convention (Cohen, 1988), a d of .30 is considered a small effect, .50
a medium effect, and .80 a large effect. For the sake of comparison, across thousands
of studies, average d for psychotherapy versus no psychotherapy is .80 to .85 and the
average for one bona fide treatment method versus another (controlled for researcher
allegiance) is 0 to .20 (Wampold & Imel, 2015).
O f course, these general rules or conventions cannot be dissociated from the context
of decisions and comparative values. There is little inherent value to an effect size of
2.0 or 0.2; it depends on what benefits can be achieved at what cost (Smith et al., 1980).
For example, the authors of Chapter 7 conducted a meta-analysis of 13 RCTs
that investigated the effectiveness o f matching therapist directiveness to the client’s
reactance level. Their meta-analysis, involving a total o f 1,208 patients, found a
weighted mean d o f .78. As shown in Table 1.1, this is a medium to large effect
size. In concrete terms, this effect size indicates that matching versus not increases
success rates by 18% to 20%. Such numbers translate into happier and healthier
clients; responsively adapting or tailoring leads more progress and fewer dropouts
in psychotherapy.

LIMITATIONS OF THE WORK

A single task force can accomplish only so much work and cover only so much content,
even in two volumes. As such, we wish to publicly acknowledge early in the book sev­
eral necessary omissions and unfortunate truncations in our work.
First, this volume probably suffers from content overlap. A client’s therapy
preferences probably reflect in part his or her cultural identities, although these are
considered in separate chapters. A client’s level of trait reactance correlates in the .30
 8 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

Table 1 .1. P ractical Interpretation o f d and g Values

d or g Cohens Type of Percentile of Success Rate of
Benchmark Effect Treated Patientsa Treated Patientsb
1.00 Beneficial 84 72%
.90 Beneficial 82 70%
.80 Large Beneficial 79 69%
.70 Beneficial 76 66%
.60 Beneficial 73 64%
.50 Medium Beneficial 69 62%
.40 Beneficial 66 60%
.30 Beneficial 62 57%
.20 Small Beneficial 58 55%
.10 No effect 54 52%
.00 No effect 50 50%
-.1 0 No effect 46 48%
-.2 0 D etrim ental 42 45%
-.3 0 D etrim ental 38 43%

Adapted from Cohen (1988), Norcross et al. (2017), and Wampold and Imel (2015).
a Each effect size can be conceptualized as reflecting a corresponding percentile value; in
this case, the percentile standing of the average treated patient after psychotherapy relative
to untreated patients.
b Each effect size can also be translated into a success rate of treated patients relative to un­
treated patients; a d of .80, for example, would translate into approximately 70% of patients
being treated successfully compared to 50% of untreated patients.

range with the precontemplation stage of change for a particular problem, but these
two client dimensions are treated in different chapters. We also suspect that a person’s
attachment style and coping style are intertwined. The field requires a gigantic factor
analysis of sorts to identify the construct overlap and to determine superordinate ad­
aptation methods.
Researcher allegiance may have also posed a problem in conducting and interpreting
the meta-analyses. We invited authors with an interest and expertise in a particular
client characteristic and adaptation method, but, in some cases, the authors might have
experienced conflicts of interest due to their emotional, academic, or financial interests.
In dozens of implicit ways and explicit decisions, authors may have favored the effec­
tiveness of their scholarly offspring. The use of objective meta-analytic guidelines, peer
review, and transparent data reporting probably attenuated effects of their allegiance,
but it remains a strong human propensity in any discipline.
Another prominent limitation of the work as a whole is the relatively small number
of research-supported methods to responsively match psychotherapy to the indi­
vidual. There are but nine represented in this volume, and two of those did not pos­
sess any controlled studies to meta-analyze. Seven probably overlapping methods is
a modest number, although they all possess robust meta-analytic research support.
 9 Evidence-Based Psychotherapy Responsiveness

Moreover, the meta-analyses were conducted on mostly RCTs and are capable of
causal conclusions.
As with the previous task forces, the overwhelming majority of research studies
analyzed were conducted in Western developed nations and published in English-
language journals. The literature searches are definitely improving in accessing studies
conducted internationally, but most chapter authors did not translate articles published
in other languages. The meta-analytic results are, therefore, English-centric.

FREQUENTLY ASKED QUESTIONS

The third Interdivisional APA Task Force on Evidence-Based Relationships and

Responsiveness has generated considerable enthusiasm, but it has also engendered
misunderstandings and reservations. Here we address frequently asked questions
about the task force’s goals and results.

♦ Are you saying that matching treatment methods to the particular disorder is
ineffective?

Not quite. We are saying that matching psychotherapy to the entire person, princi­
pally to the transdiagnostic characteristics identified in this volume, typically proves
more effective than matching psychotherapy to the disorder. Culture eats strategy, so to
speak. Practitioners are not forced to select between one way or the other; use all that
work. Engage and customize to the full range of the human condition, refusing to be
restricted by a single clinical method or patient characteristic.

♦ Isn’t is premature to launch a set o f research-based conclusions on patient matching?

Science is not a set of answers; science is a series of processes and steps by which
we arrive closer and closer to elusive answers. Sophisticated research over the past two
decades has been conducted on means of adapting psychotherapy to individual clients.
It is premature to proffer the last word, but it is time to codify and disseminate what we
do know. We look forward to regular updates on our research conclusions and practice
recommendations.

♦ We are seasoned therapists and understand that we must attune treatment to each
person. But there are few resources (beyond the book) to train and supervise students in
evidence-based responsiveness. What gives?

What gives, we suspect, is that training invariably lags behind cutting-edge research
and practice. Studies must be conducted, published, and aggregated to create a strong
evidence base, and only then does it filter down to widespread training. Researchers
are enamored with, perhaps even addicted to, discovery, but not so much with the im ­
plementation of and training of their discoveries (Norcross et al., 2017).
 10 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

As one step toward training resources, all contributors to this volume addressed
training practices for their respective client dimensions. Those practices include guid­
ance on how to assess the client characteristic reliably in session, how to determine
its relative salience for a particular case, how to match specifically in accord with the
research, and sometimes how to avoid drifting back to the clinician’s baseline style of
psychotherapy. In the next edition, we plan to ask contributors to describe even more
what can be done to develop and train students.

♦ My favorite patient characteristic— say, symptom severity—is not reviewed in this book.
Is that not an effective way to match? Are there others that work?

You bet! We have neither completed the search nor exhausted the matching
possibilities. Along with symptom severity and its associated functional impairment,
we suspect that the client’s attachment style, cognitive complexity, and psychological
mindedness probably “work” as markers of effective adaptation. Unfortunately, suf­
ficient controlled research has not yet been published (at least in English) to include
them in the book. The absence of research evidence does not mean the absence of
clinical effectiveness.

♦ Can psychotherapists really adapt their relational style to fit the proclivities and
personalities o f their patients?

Relational flexibility conjures up many concerns, but two particular import to this
question: the limits of human capacity and the possibility of capricious posturing.
Although the psychotherapist can, with training and experience, learn to relate in a
number of different ways, there are limits to our human capacity to modify relation­
ship stances. It may be difficult to change interaction styles from client to client and
session to session, assuming one is both aware and in control of one’s styles of relating
(Lazarus, 1993).
Years of training experience and some research evidence supports the assertion
that psychotherapists can authentically differ from their preferred or habitual style of
relating. Effective therapists are capable of more malleability, more flexible repertoires,
and “mood transcendence” (Hill et al., 2017; Gurman, 1973; Tracey et al., 2014). The
research on the therapist’s experience suggests that experience begets heightened at­
tention to the client (less self-preoccupation), an innovative perspective, and, in ge­
neral, more endorsement of an “integrative” orientation predicated on client need
(Auerbach & Johnson, 1977; Norcross & Goldfried, 2019). Indeed, several research
studies have demonstrated that therapists can consistently use different treatments in a
discriminative fashion. Experienced therapists are able to help clients respond sooner
and to provide a smoother course to recovery (Lambert, 2010).
Thus, our clinical experience and a modest amount of research attest that
practitioners can shift back and forth among different relationship styles for a given
case. At the same time, we caution therapists that the blending of stances and strategies
should ever deteriorate into play-acting or capricious posturing.
 11 Evidence-Based Psychotherapy Responsiveness

♦ But what about behavioral drift—the ubiquitous tendency to revert to old behavioral
patterns? Don’t we return over time to our practice baseline or pet methods?

A definite possibility. We endorse responsiveness adherence checks, ongoing case

reformulation, and deliberate practice to maintain flexibility with each patient.
Moreover, clients and clinicians reciprocally shape each other during the course of
psychotherapy, and responsiveness evolves over their time together. Each patient, each
session, requires something different or new.

♦ What should we do if we are unable or unwilling to adapt our therapy to the patient in
the manner that research indicates is likely to enhance psychotherapy outcome?

Four possibilities spring to mind. First, address the matter forthrightly with the pa­
tient as part of the evolving therapeutic contract and the creation of respective tasks,
in much the same way one would with patients requesting a form of therapy or a type
of medication that research has indicated would fit particularly well in their case but
which is not in your repertoire. Second, treatment decisions are the result of multiple
and recursive considerations on the part of the patient, the therapist, and the context.
A single evidence-based guideline should be seriously considered but only as one of
many determinants of treatment itself. Third, an alternative to the one-therapist-fits-
most-patients perspective is practice limits. Without a willingness and ability to en­
gage in a range of interpersonal stances, the therapist may limit his or her practice to
clients who fit that practice. Mental health professionals need not offer all services to
all patients. Fourth, consider a judicious referral to a colleague who can offer the re­
lationship stance (or treatment method or medication) indicated in a particular case.

♦ Your relational responsiveness seems at odds with what managed care and
administrators ask o f me in my practice. How do you reconcile these?

We do not reconcile these views, but we hope to influence managed care and be­
havioral health administrators with the compelling meta-analytic findings in these
two volumes. Among payers and policymakers, the dominant image o f modern
psychotherapy is as a mental health treatment. This “treatm ent” or “m edical”
model inclines people to define process in terms of method, therapists as providers
applying techniques, treatment in terms of number of contact hours, patients as
embodiments of mental disorders, and outcome solely as symptom reduction
(Orlinsky, 1989).
The steering committee believes this model to be restricted and inaccurate. The psy­
chotherapy enterprise is far more complex and interactive than the linear “Treatment
operates on disorders to produce effects” We prefer a broader, integrative view that
aligns with the tripartite evidence-based practice model that privileges best avail­
able research, clinician expertise, and client characteristics, cultures, and values. That
model incorporates the relational and educational features of psychotherapy, one that
recognizes both the interpersonal and instrumental components of psychotherapy,
 12 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

that appreciates the bi-directional process of therapy, and in which the therapist and
patient co-create an optimal process and outcome.

♦ Are the task forces conclusions and recommendations intended as practice standards?

No. These are research-based conclusions that can lead, inform, and guide
practitioners toward responsiveness or treatment adaptations (and, in volume 1, to
evidence-based therapy relationships). They are not legal, ethical, or professional
mandates.

♦ Well, don’t these represent the official positions o f APA Division 29 (Psychotherapy),
Division 17 (CounselingPsychology), or the APA?

No, no, and no.

♦ So, are you saying that the therapy relationship (in addition to the treatment method) is
crucial to outcome, that it can be improved by certain therapist actions, and that it can
be effectively tailored to the individual patient?

Precisely. And this two-volume book shows specifically how to do so on the basis of
the research evidence.

IN CLOSING
Decades of research now scientifically support what psychotherapists have long
known: different types of clients require different treatments and relationships.
And the research has now identified specific client characteristics and optimal
matches by which to tailor or adapt psychotherapy. In the tradition of evidence-
based practice, psychotherapists can create a new, responsive psychotherapy
for each distinctive patient and his or her singular situation— in addition to
disorder.
The future of psychotherapy portends the integration of the instrumental and the
interpersonal, of the technical and the relational in the tradition of evidence-based
practice (Norcross et al., 2011). Evidence-based responsiveness aligns with this future
and embodies a crucial part of evidence-based practice, when properly conceptualized.
We can imagine few practices in all of psychotherapy that can confidently boast that
they integrate as well “the best available research with clinical expertise in the con­
text of patient characteristics, culture, and preferences” (APA, 2006) as the relational
behaviors and treatment adaptations presented in these two volumes. We are reminded
daily that research can guide how to create, cultivate, and customize that powerful
human relationship.
 13 Evidence-Based Psychotherapy Responsiveness

REFERENCES
A m erican Psychological A ssociation Task Force on Evidence-Based Practice. (2006).
Evidence-based practice in psychology. American Psychologist, 61, 2 7 1 -2 8 5 .
Auerbach, A. H., & Johnson, M . (1977). Therapist experience. In A. S. G urm an & A. M. Razin
(Eds.), Effective psychotherapy: A handbook o f research. New York, NY: Pergamon.
Barlow, D. H. (Ed.). (2014). Clinical handbook o f psychological disorders: A step-by-step treat­
ment manual (5th ed.). New York, NY: Guilford.
Beutler, L. E., & Clarkin, J. (1990). Systematic treatment selection: Toward targeted therapeutic
interventions. New York, NY: Brunner/Mazel.
Blatt, S. J., & Felsen, I. (1993). Different kinds o f folks may need different kinds o f strokes: The
effect o f patients’ characteristics on therapeutic process and outcome. Psychotherapy
Research, 3, 2 4 5 -2 5 9 .
Clarkin, J. F., & Levy, K. N. (2004). The influence o f client variables on psychotherapy. In M.
J. Lam bert (Ed.), Handbook o f psychotherapy and behavior change (5th ed., pp. 194-2 2 6 ).
New York, NY: Wiley.
Cohen, J. (1988). Statistical power analysis fo r the behavioral sciences (2nd ed.). Hillsdale,
NJ: Erlbaum.
Frances, A., Clarkin, J., & Perry, S. (1984). Differential therapeutics in psychiatry. New York,
NY: Brunner/Mazel.
Frank, J. D., & Frank, J. B. (1991). Persuasion and healing: A comparative study o f psycho­
therapy. JH U Press.
Gurm an, A. S. (1973). Effects o f therapist and patient m ood on the therapeutic functioning
o f high-and low-facilitative therapists. Journal o f Consulting and Clinical Psychology,
4 0 (1 ), 48.
Henry, W. P., Schacht, T. E., & Strupp. H. H. (1986). Structural analysis o f social beha­
vior: Application to a study o f interpersonal process in differential psychotherapeutic out­
com e. Journal o f Consulting and Clinical Psychology, 54, 2 7 -3 1 .
Hill, C. E., Spiegel, S., Hoffm an, M . A., Kivlighan, D., & Gelso, C. (2017). Therapist expertise
in psychotherapy revisited. The Counseling Psychologist, 45, 7 -5 3 .
Lam bert, M. J. (2010). Prevention o f treatment failure: The use o f measuring, monitoring, &
feedback in clinical practice. W ashington, DC: A m erican Psychological Association Press.
Lazarus, A. A. (1993). Tailoring the therapeutic relationship, or being an authentic chameleon.
Psychotherapy, 30, 4 0 4 -4 0 7 .
Mahoney, M. J., & Norcross, J. C. (1993). Relationship styles and therapeutic choices: A co m ­
mentary. Psychotherapy, 30, 4 2 3 -4 2 6 .
M ichels, R. (1984). Foreword. In A. Frances, J. Clarkin, & S. Perry (Eds.), Differential
therapeutics in psychiatry. New York, NY: Brunner/Mazel.
Nathan, P. E., & G orm an, J. M. (Eds.). (2015). A guide to treatments that work (4th ed.).
New York, NY: Oxford University Press.
N orcross, J. C. (Ed.). (2011). Psychotherapy relationships that work (2nd ed.). New York,
NY: O xford University Press.
N orcross, J. C., & Beutler, L. E. (2014). Evidence-based relationships and responsiveness for
depression and substance abuse. In D. H. Barlow (Ed.), Clinical handbook o f psychological
disorders (5th ed.). New York, NY: Guilford.
 14 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

Norcross, J. C., Freedheim , D. K., & VandenBos, G. R. (2011). Into the future: Retrospect and
prospect in psychotherapy. In J. C. Norcross, G. R. Vanderbos, & D. K. Freedheim (Eds.),
History o f psychotherapy (2nd ed.). W ashington, DC: A m erican Psychological Association.
Norcross, J. C., & Goldfried, M. R. (Eds.). (2019). Handbook o f psychotherapy integration (3rd
ed.). New York, NY: Oxford University Press.
Norcross, J. C., Hogan, T. P., Koocher, G. P , & Maggio, L. A. (2017). Clinician’s guide to
evidence-based practices: Behavioral health and addictions (2nd ed.). New York, NY: Oxford
University Press.
Norcross, J. C., & Wampold, B. E. (Eds.). (2011). Adapting psychotherapy to the individual
patient [Special issue]. Journal o f Clinical Psychology, 67, 1 43-154.
Orlinsky, D. E. (1989). Researchers’ images o f psychotherapy: Their origins and influence on
research. Clinical Psychology Review, 9, 4 1 3 -4 4 1 .
Osler, W. (1906). Aequanimatas. New York, NY: McGraw-Hill.
Paul, G. L. (1967). Strategy o f outcom e research in psychotherapy. Journal o f Consulting, 31(2),
1 0 9 -1 1 8 .
Sm ith, M. I., Glass, G. W. V., & M iller T. L. (1980). The benefits o f psychotherapy. Baltimore,
M D : Johns Hopkins University Press.
Strupp, H. H. (1993). The Vanderbilt psychotherapy studies: Synopsis. Journal o f Consulting
and Clinical Psychology, 61, 4 3 1 -4 3 3 .
Tracey, T. J. G., Wampold, B. E., Lichtenberg, J. W., & Goodyear, R. K. (2014). Expertise in
psychotherapy: An elusive goal. American Psychologist, 69, 2 1 8 -2 2 9 .
W ampold, B. E., & Imel, Z. (2015). The great psychotherapy debate (2nd ed.). Mahwah,
NJ: Erlbaum.
Wolitzky, D. L. (2011). Psychoanalytic theories o f psychotherapy. In J. C. Norcross, G. R.
VandenBos, & D. K. Freedheim (Eds.), History o f psychotherapy (2nd ed.). W ashington,
DC: A m erican Psychological Association.
 2

ATTACHMENT STYLE

Kenneth N. Levy, Benjamin N. Johnson, Caroline V. Gooch, and

Yogev Kivity

Attachment theory, originally developed by John Bowlby to explain human bonding,

has profound implications for conducting and adapting psychotherapy. The concept
of attachment style (also sometimes referred to as attachment pattern, organization,
type, or category) derives from Bowlby and Mary Ainsworth’s attachment theory and
refers to a person’s characteristic ways of relating in intimate care giving and receiving
relationships, particularly with one’s parents, children, and romantic partners. From
an attachment perspective, these individuals are called attachment figures. Attachment
style involves one’s confidence in the availability of the attachment figure so as to use
that person as a secure base from which the individual can freely explore the world when
not in distress, as well as the use of this attachment figure as a safe haven from which
the individual seeks support, protection, and comfort in times of distress. Exploration
of the world includes not only the physical world but also the examination of the “rela­
tional world,” of relationships with other people, and the “internal world,” with regard
to one’s capacity for reflection about one’s own internal experience and the capacity to
make relatively accurate inferences about other people’s internal experiences.
Attachment theory and the research generated from it has resulted in it be­
coming one of the most influential frameworks in psychology (Haggbloom et al.,
2002). However, from the theory’s inception, psychiatrist and psychoanalyst Bowlby
conceptualized attachment theory as a clinical guide. This is not surprising given that
attachment theory emerged from Bowlby’s clinical observations in his classic study
of 44 delinquent children. He noted the pervasiveness of interpersonal loss in these
children’s lives, which was elaborated in his work with institutionalized children in
orphanages, particularly during World War II. Additionally, these ideas were further
stoked by his own clinical work with children, parents, and adults.
Bowlby integrated his clinical observations with principles from other disciplines
to explain affectional bonding between infants and their caregivers and the long-term
effects of early attachment experiences on personality development, interpersonal
functioning, and psychopathology. Bowlby postulated that the attachment system was
operative throughout the lifespan—“from the cradle to the grave”— and across a wide
variety of relationships, including the therapeutic relationship (Bowlby, 1977; Eagle,

15
 16 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

1997; Farber et al., 1995). Ainsworth, a clinical psychologist, also saw attachment
theory as relevant for clinical work, although most of her and her students’ research
focused on developmental psychology.
Bowlby (1975) suggested that the chief role of the psychotherapist is “to provide the
patient with a temporary attachment figure”. He thought that doing so would “provide
the patient with a secure base from which to explore both himself and also his relations
with all those with whom he has made or might make, an affectional bond” (Bowlby,
1977, p. 423). Secure attachment behaviors in psychotherapy include the use of the
therapist as a secure base from which the individual can freely reflect on his or her
experience, reflect on the possible contents of the minds of significant others, and try
new experiences and engage in novel behaviors. A number of clinical theorists have
elaborated upon Bowlby’s ideas about the function of attachment within the thera­
peutic relationship (e.g., Farber et al., 1995; Levy & Johnson, 2018).
In this spirit, Bowlby (1988) formulated five key tasks for psychotherapy: (a)
establishing a secure base, which involves providing patients with a strong internal felt
sense of trust, care, and support and which allows the patient to more fully and safely
explore the world and the contents of his or her mind; (b) exploring past attachment
experiences, which involves helping patients explore past and present relationships;
(c) exploring the therapeutic relationship, which involves how it may relate to
relationships or experiences outside of therapy; (d) linking past experiences to present
ones; and (e) revising internal working models, which involves helping patients to feel,
think, and act in new ways. We have also elaborated on a sixth function: to provide a
safe haven, a place the patient can “go” or “envision” in times of distress (Levy, 2013).
Adult attachment has been examined in psychotherapy research as both an out­
come variable and a moderator of treatment outcome. Early findings from this body of
research suggest that patient attachment status is relevant to the course and outcome
of psychotherapy and may also change as a result of psychotherapy. A review of this
literature (Berant & Obegi, 2009) concluded that securely attached clients tend to ben­
efit more from psychotherapy than insecurely attached clients. However, the findings
across these studies have been inconsistent, with some studies suggesting that securely
attached clients may not necessarily show more improvement in treatment compared
to insecurely attached clients (Fonagy et al., 1996).
In addition, the strength of the relation between attachment security and treatment
outcome remains unclear. Our previous meta-analysis determined a small to moderate
effect of attachment security on psychotherapy outcome (Levy et al., 2011), although
the results of this study were mixed depending on attachment style. However, given
the small number of studies included in this meta-analysis (k = 14), changing trends
in attachment styles over the past decades (Konrath et al., 2014), and recent statistical
advances, an update to this study is warranted.
This chapter examines the relation between clients’ attachment styles and their suc­
cess in psychotherapy (outcome) and whether certain attachment styles prove more
effective with certain types of psychotherapy (moderator). First, we review definitions
and measures of attachment and provide clinical examples of attachment patterns in
psychotherapy. Second, to draw an overall conclusion about the relation between at­
tachment and treatment outcome, we present a meta-analysis of the research on the
 17 Attachment Style

association between clients’ pretreatment attachment style, change in attachment, and

psychotherapy outcome, as well as an examination of potential moderators of these
effects. We conclude with limitations of the extant research, diversity considerations,
training implications, and therapeutic practices based on the meta-analytic findings.

DEFINITIONS
Attachment style is a term used to describe one’s characteristic way of viewing, relating
to, and interacting with significant others such as parents, children, and romantic part­
ners. Bowlby and Ainsworth tended to refer to these styles as “patterns,” whereas Mary
Main referred to these styles as attachment “organization.” In the social psychological
literature, authors have tended to use the concept of “style,” particularly in reference to
attachment measured from self-report measures. Bowlby described three main attach­
ment patterns: secure, anxious-ambivalent, and avoidant. Later in her study of infant-
parent dyads, Ainsworth renamed the anxious-ambivalent pattern anxious-resistant and
later identified a fourth pattern—disorganized. Over the years, a number of researchers
and theorists have referred to these basic patterns using similar but slightly different
names, for example, dismissing for avoidant and preoccupied for anxious-ambivalent.
The caregiver’s reliable and sensitive provision of loving care is believed to result
in what Bowlby called a secure bond between the infant and the caregiver. This at­
tachment security is conceptualized as deriving from repeated transactions with pri­
mary caregivers, through which the infant is believed to form internal working models
(IW M s) of attachment relationships. These IW Ms include expectations, beliefs, emo­
tional appraisals, and rules for processing or excluding information. They can be
partly conscious and partly unconscious and need not be completely consistent or
coherent. IWMs are continually elaborated; with development, they organize person­
ality and subsequently shape thoughts, feelings, and behaviors in future relationships.
Thus differences in caregiver behavior result in differences in infants’ IW Ms, which in
turn are the basis for individual differences in the degree to which relationships are
characterized by security.
Based on Bowlby’s attachment theory, Ainsworth and colleagues (1978) developed a
laboratory method called the Strange Situation to evaluate individual differences in at­
tachment security. The Strange Situation involves a series of short laboratory episodes
staged in a playroom through which the infant, the caregiver, and a stranger interact
in a comfortable setting and the behaviors of the infant are observed. Ainsworth
and colleagues paid special attention to the infant’s behavior upon reunion with the
caregiver after a brief separation. Ainsworth and colleagues identified three distinct
patterns or styles of attachment, which have since been termed secure (63% of the
dyads tested), anxious-resistant or am bivalent (16%), and avoidant (21%).
In the Strange Situation, secure infants can find the brief separation from the care­
giver and the entrance of the stranger to be upsetting, but they approach the care­
giver upon his or her return for support, calm quickly upon the caregiver’s return, are
easily soothed by the caregiver’s presence, and go back to exploration without fuss.
In contrast, anxious-resistant infants tend to become extremely distressed upon the
caregiver’s departure, and they ambivalently approach the caregiver for attention and
 18 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

comfort upon the caregiver’s return. They are clingy and dependent, often crying, but
they also seem angry and resist their caregiver’s efforts to soothe them. Avoidantly at­
tached infants frequently act unfazed or unaware of the caregiver’s departure and often
avoid the caregiver upon reunion. Sometimes these infants appear shut down and de­
pressed and at other times indifferent and overinvested in play (although the play has
a rote quality rather than a rich symbolic quality). Despite their outward appearance of
calmness and unconcern, research has shown that avoidant infants are quite distressed
in terms of physiological responding, similar to the anxious-resistant babies (Sroufe &
Waters, 1977).
Despite the obvious resemblance of these patterns to temperament types (Kagan,
1998), these attachment behaviors in the Strange Situation experiment are not simply
a result of infant temperament (see Levy, 2005, for a review). Temperament may af­
fect the manner in which attachment security is expressed, but temperament does not
affect the security of the attachment itself (Belsky & Rovine, 1987). Ainsworth’s orig­
inal work has been replicated and extended in hundreds of studies with thousands of
infants and toddlers (Fraley, 2002).
A growing body of research (e.g., Grossmann et al., 2005; Waters, Hamilton, &
Weinfield, 2000) examining attachment continuity suggests that patterns of attachment
are both relatively stable over long periods of time and subject to change, influenced
by a variety of factors including ongoing relationships with family members, new ro­
mantic relationships, traumatic life events, and possibly psychotherapy (e.g., Fraley,
2002). These findings are consistent with Bowlby’s (1982) idea that attachment theory
was not limited to infant-parent relationships.

MEASURES
A number of measures have been developed to assess attachment beginning with
Ainsworth’s Strange Situation and evolving into efforts to develop measures for assessing
attachment in adults. These measures derived from two main traditions: develop­
mental psychologists focused on interviews about childhood attachment relationships
with caregivers (Main et al., 1985), and social psychologists (e.g., Hazan & Shaver,
1987) focused on self-report measures of romantic and other adult relationships. In
addition, a number of clinical psychologists developed self-report measures, typically
focused on adult relationships, often with romantic partners (West & Sheldon, 1988).
From a developmental tradition, Main and her colleagues developed the Adult
Attachment Interview (AAI; George et al., 1985; Main et al., 1985), a one-hour
attachment-history interview, noting that features in interviews with parents of
infants reliably predicted the Strange Situation behavior of their children. The in­
terview inquiries into “descriptions of early relationships and attachment and adult
personality,” by probing for both specific corroborative and contradictory memories
of parents and one’s relationship with parents (Main et al., 1985, p. 98). Three
major patterns of adult attachment were initially identified: secure/autonomous,
dismissing, and enmeshed/preoccupied. More recently, two additional categories have
been identified: unresolved and cannot classify. The first three categories parallel the
 19 Attachment Style

attachment classifications originally identified in childhood of secure, avoidant, and

anxious-resistant (Ainsworth et al., 1978), and the unresolved classification parallels a
pattern Main later described in infants that she called disorganized/disoriented (Main
& Solomon, 1986). A number of studies have found that AAI classifications based on
individuals’ reports of interactions with their own parents can predict their children’s
Strange Situation classifications (van IJzendoorn, 1995).
A 100-item Adult Attachment Q-set was derived from the AAI scoring system and
has been applied to AAI transcripts (Kobak et al., 1993). This system identifies secure,
preoccupied, and dismissing categories based on ratings of two dimensions: security
versus anxiety and deactivation versus hyperactivation. Hyperactivating emotional
strategies are typical of preoccupied individuals, whereas deactivating strategies are
typical of dismissing individuals. Scores are compared to a criterion or “ideal” proto­
type sort to identify the three organized attachment categories. One notable disadvan­
tage of the Q-set is that there is no rating for a disorganized attachment dimension, nor
can it identify the cannot classify category.
In contrast to Main’s focus on relationships with parents, Hazan and Shaver (1987,
1990), from a social psychological perspective, extrapolated the childhood attach­
ment paradigm to study attachment in adulthood by conceptualizing romantic love
as an attachment process. They translated Ainsworth’s secure, avoidant, and anxious-
ambivalent attachment patterns into a paper-and-pencil prototype-matching measure
of adult attachment styles (preferring the term anxious-ambivalent to anxious-
resistant). Several other researchers have altered and extended the original Hazan and
Shaver measure by breaking out the sentences in the prototypes into separate items.
Factor analyses of these multi-item measures found a three-factor solution (desire for
closeness, comfort with dependency, and anxiety about abandonment; Collins & Read,
1990), as well as a two-factor solution (desire for closeness and anxiety about abandon­
ment; Simpson, 1990). A number of empirical studies using Hazan and Shaver’s (1987)
measure or derivative measures of adult attachment have found that the distribution
of adult attachment styles is similar to those found for infants. Approximately 55% of
individuals are classified as secure, 25% as avoidant, and 20% as anxious (Shaver &
Clark, 1994; Shaver & Hazan, 1993).
Bartholomew (1990; Bartholomew & Horowitz, 1991) revised Hazan and
Shaver’s three-category classification scheme and proposed a four-category model
that differentiated between two types of avoidant styles—fearful and dismissing.
Bartholomew’s key insight was an incongruity between Main’s (Main & Goldwyn,
1998) and Hazan and Shaver’s conceptions of avoidance. Main’s prototype of the
adult avoidant style (assessed in the context of parenting) is more defensive, denial-
oriented, and overly unemotional than Hazan and Shaver’s avoidant romantic at­
tachment prototype, which seems more vulnerable, conscious of emotional pain, and
“fearful.” Thus Main’s avoidant style is predominantly dismissing, whereas Hazan and
Shaver’s avoidant style is predominantly fearful. Bartholomew’s four categories could
be arrayed in a two-dimensional space, with one dimension being model o f self (pos­
itive vs. negative) and the other being model o f others (positive vs. negative). For se­
cure individuals, models of self and others are both generally positive. For preoccupied
 20 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

or anxious-ambivalent individuals, the model of others is positive (i.e., relationships

are attractive) but the model of self is not. For dismissing individuals, the reverse is
true: the somewhat defensively maintained model of self is positive, whereas the model
of others is not (i.e., intimacy in relationships is regarded with caution or avoided).
Fearful individuals have relatively negative models of both self and others.
Bartholomew also developed an interview measure of attachment along with her
self-report measure. The interview measure, initially referred to as the Bartholomew
Attachment Interview and later the Family Attachment Interview (Bartholomew &
Horowitz, 1991), covers both relationships with parents (in line with the AAI) and
relationships with close friends and romantic partners (in line with Shaver and
Hazan’s work).
In an effort to develop a more definitive measure of adult attachment and respond
to the proliferation of attachment measures, Brennan and colleagues (1998) created
the Experiences in Close Relationships (ECR) scale. It was derived from a factor anal­
ysis of 60 attachment constructs representing 482 items extracted from a thorough
search of measures used in and developed for previous attachment research. The ECR
factor structure was consistent with the Relationship Questionnaire (RQ) but showed
stronger relations with other relevant constructs than did the RQ. Two short forms
of the ECR have also been published (Fraley et al., 2000; Wei et al., 2007), with both
highly related to the original ECR.
More recently, the Patient Attachment Coding System (Talia & M iller-Bottom e,
2012; Talia, M iller-Bottom e, & Daniel, 2017) was developed as an observer­
rated assessment to measure several aspects o f in-session attachment behavior
(e.g., patient proximity seeking toward the therapist). The authors found theo­
retically consistent patterns of patient behavior in relation to the therapist (e.g.,
preoccupied patients although seeking closeness displayed more resistance to help,
and dismissing patients avoided em otional connection m ore). Such an approach,
if utilized to examine m om ent-by-m om ent attachment behaviors, could examine
contemporaneous as well as cross-l agged dynamic relations between attachment
and outcome.

Measures Used in Our Meta-Analysis

Research groups have approached the assessment of adult attachment patterns through
a variety of measures. Interestingly, however, these measures, regardless of type or at­
tachment category, tend to have the same underlying dimensions: attachment anxiety
and attachment avoidance. The measures described here are those used in the studies
included in the current meta-analysis and are presented in chronological order of their
development.
The AAI (George et al., 1985) is a semi-structured interview of 18 questions that
probe an individual’s mental representations of early attachment relationships and
their effect on one’s adulthood. Main et al. (1985) identified three major patterns of
adult attachment— secure/autonomous, dismissing, and enmeshed/preoccupied—
and two additional codes of unresolved/disorganized and cannot classify.
 21 Attachment Style

The Adult Attachment Prototype Rating (Pilkonis, 1988) is a set of 88 items that can
be applied to interview data or used to rate an individual’s attachment style. The rating
system focuses on two dimensions with multiple facets. The excessive dependency
dimension, corresponding to attachment anxiety, subsumes excessive dependency,
borderline features, and compulsive caregiving prototypes. The excessive autonomy
dimension, corresponding to attachment avoidance, subsumes defensive separation,
antisocial features, and obsessive-compulsive features. A secure prototype was later
added to the system (Strauss et al., 1999).
The Adult Attachment Scale (Collins & Read, 1990) is a self-report instrument de­
veloped by breaking Hazan and Shaver’s (1987) prototype statements into 21 items,
later shortened to 18 (Collins, 1996). Individuals rate these statements on a 5-point,
Likert-type scale. The subscales include comfort with closeness and intimacy, comfort
depending on others, and anxiety about abandonment, which can be combined to pro­
duce scores for secure, anxious, and avoidant styles. There is strong evidence for the
scales reliability and validity (Ravitz et al., 2010).
The Perceptions o f Adult Attachment Questionnaire (PAAQ; Lichtenstein & Cassidy,
1991) is a 60-item self-report measure designed to parallel the AAI. As such, the in­
strument was designed to assess both current mental states with regard to caregivers
and perceptions of childhood attachment relationships. The PAAQ has shown good
psychometric properties (Cassidy et al., 2009).
The Relationship Questionnaire (RQ; Bartholomew & Horowitz, 1991) is a self­
report questionnaire based on Bartholomew’s (1990) four-category model of at­
tachment. The RQ consists of four paragraphs describing each of the attachment
prototypes— secure, fearful, preoccupied, and dismissing. Participants rate how well
each corresponds to their romantic relationship pattern. Participants then select the
one paragraph that best describes them. This measure has also been referred to as the
Bartholomew Attachment Rating Scale (Travis et al., 2001).
The Relationship Style Questionnaire (Bartholomew & Horowitz, 1991) contains 30
short statements on a 5-point Likert scale indicating the extent to which each statement
describes one’s characteristic style in close relationships. Five statements contribute to
the secure and dismissing attachment patterns, and four statements contribute to the
fearful and preoccupied attachment patterns. Scores for each attachment pattern are
calculated by taking the mean of the four or five items representing each attachment
prototype. Two underlying dimensions can be derived by using the scores from the
four prototype items to create linear combinations representing the self- and other-
model attachment dimensions.
The Attachment Style Questionnaire (Feeney et al., 1994) is a 40-item self-report
questionnaire rated on a 6-point, Likert-type scale. It includes subscales to measure
Self-Confidence, Discomfort with Closeness, Need for Approval, Preoccupation, and
Relationships as Secondary. The instrument has adequate reliability and has been
found to converge with other attachment measures and to have predictive validity
(Ravitz et al., 2010).
The Reciprocal Attachment Questionnaire (RAQ; West & Sheldon-Keller, 1994) is
a 43-item 5-point Likert-type self-report questionnaire designed to assess nine
 22 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

dimensions of adult attachment patterns with significant others. Four subscales—

Compulsive Self-Reliance, Compulsive Care-Giving, Compulsive Care-Seeking, and
Angry Withdrawal— assess dysfunctional patterns of adult attachment. There are
also five attachment dimension subscales: Separation Protest, Feared Loss, Proximity
Seeking, and Use and Perceived Availability of the attachment figure. The validity and
reliability of the RAQ have been established in both clinical and nonclinical adult
populations (West & Sheldon-Keller, 1994).
The ECR (Brennan et al., 1998) is a 36-item, self-report questionnaire that meas­
ures anxious and avoidant attachment domains. Participants rate the extent to which
each item is descriptive of their feelings in close relationships on a 7-point scale. The
reliability and validity of the scales have been demonstrated (Brennan et al., 1998).
Since the development of the original ECR, there have been a number of revised
versions, including the ECR-Revised, the ECR-Relationship Structures, and the
EC R -Short Form.
The Psychosis Attachment Measure (Berry et al., 2008) is a 16-item self-report measure
that assesses the attachment dimensions of anxiety and avoidance. Participants are
asked to rate how characteristic each item is of them on a 4-point scale. The measure
has shown good reliability and validity in both clinical and nonclinical samples (Berry
et al., 2008).

CLINICAL EXAMPLES

Next we provide clinical examples for adults with secure, preoccupied, and dismissing
attachment styles. Although we discuss clinical examples through the lens of attach­
ment categories, research suggests that attachment is better conceptualized dimen­
sionally as a function of level of attachment anxiety and level of attachment avoidance.
The attachment categories described here can be captured by arraying an individual
in a quadrant based on one’s level on the dimensions. Thus those low in attachment
anxiety and low in attachment avoidance would fall into a quadrant representing se­
cure attachment. In general, patients with secure attachment styles have been found to
be more collaborative, more receptive, and better able to utilize treatment (Mikulincer
& Shaver, 2007). In contrast, those who score high on attachment avoidance but low on
attachment anxiety would be placed in a quadrant characterized by dismissing attach­
ment. Those with dismissing styles have been found to be less engaged in treatment.
Those high in attachment anxiety but low in attachment avoidance are considered
preoccupied with attachment (also referred to as anxious-ambivalent attachment).
Those with preoccupied states of mind with regard to attachment have been found to
present as more needy in therapy but not necessarily compliant with treatment (e.g.,
Dozier, 1990; Riggs et al., 2002). Those high in attachment anxiety and high in at­
tachment avoidance are considered fearfully avoidant in attachment. These individuals
often desire attachment like preoccupied individuals but are fearful of getting too close
and act avoidant. In psychotherapy, these individuals may want to confide in the ther­
apist but often have a difficult time doing so.
 23 Attachment Style

Secure Attachment
Sandy, who was securely attached, entered treatment due to feelings of depression fol­
lowing the birth of her daughter. She had considered her marriage happy but with
the birth of her daughter there had been increasing stress and fights with her hus­
band. Additionally, she was anxious about parenting and, while thrilled to be a mother,
she also felt a sense of disappointment and concern about her marriage. At times she
cried and, although clearly distressed, she related easily with the therapist and showed
trust in the therapist. At times she would become quiet in session but when the thera­
pist asked her about it, she shared her concerns, even when it was about the therapist
or what the therapist might think of her. She was able to tolerate discussing difficult
topics, was relatively nondefensive, had access to her thoughts and feelings, and, most
importantly, tended to be able to integrate and utilize the therapist’s comments. Sandy
spoke openly about her ambivalence toward herself, the baby, her husband, and the
therapist when such feelings arose. As the therapy proceeded, she tolerated her am­
bivalence better and felt more positive. Consistent with those with secure attachment,
Sandy showed gratitude toward the therapist for providing treatment and was tolerant
of the therapist’s moments of empathic failure.

Preoccupied Attachment
Penny, who was preoccupied in her attachment, entered psychotherapy after the
breakup with a boyfriend and subsequent complaints of anxiety and depression.
Because preoccupied individuals can be so interpersonally engaged, they often initially
appear easy to treat. Penny was frequently distressed and eager to discuss her worries
and relationship difficulties. She often lamented her own role in these problems.
However, other times Penny presented in sessions with a friendly cavalier demeanor
toward the issues that brought her to therapy, frequently chatting about trivial matters.
Other times she came into session very upset and angry about interpersonal slights
and grievances for which she actively sought the therapist’s agreement. Like many
preoccupied patients, Penny presented with chaotic and contradictory representations
of herself and others. Although Penny readily shared her internal experience with the
therapist, it often felt confusing and entangled. Most of the time, Penny presented as
compliant, dependent, and even needy of the therapist’s approval, but other times she
was testy with the therapist and would lash out at her.
Both clinical and empirical evidence suggests that these individuals are difficult to
treat (Dozier, 1990). Those classified as preoccupied, as compared with those classified
as dismissing, tend to show less improvement (Fonagy et al., 1996). It is hypothesized
that the preoccupied patients are more difficult to treat because their representational
systems are intricately linked with emotions that are entrenched in a preoccupation
with difficult events in their lives (Slade, 1999, 2004), and thus behavior change tends
to occur over a long period of time from the therapist’s long-term emotional availa­
bility and tolerance for chaos.
 24 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

Dismissing Attachment
Dismissing patients are often resistant to treatment, have difficulty asking for help,
and retreat from help when it is offered (Dozier, 1990). Indeed, dismissing patients
often evoke countertransference feelings of being excluded from the patients’ lives
(Diamond et al., 1999, 2003). A patient classified as dismissing came into session one
morning and announced, to her therapist’s surprise, that she was getting married that
afternoon. Although he had known of her engagement, it had been many months since
she had brought up any aspect of her upcoming marriage. Additionally, dismissing
individuals often become more distressed and confused when confronted with emo­
tion in therapy (Dozier et al., 2001). Another dismissing patient, when reflecting on
her experience in therapy, stated:

He (the therapist) would start digging into things and find out why I was angry,
and then I would realize something really made me mad, but I didn’t want to be
mad. With my parents, for example, I didn’t want to be angry at them.

Psychotherapists working with dismissing patients may be pulled into enactments,

where they find themselves in a situation analogous to a “chase and dodge” sequence
with mothers and infants (Beebe & Lachmann, 1988). That leaves the patient feeling
intruded upon only to withdraw further. Conversely, those with dismissing attachment
may effectively curtail the therapist’s capacity to engage with, visualize, or evoke the
individual’s representational world or identify with the patient.

LANDMARK STUDIES
We focus on three landmark studies examining attachment in adults. In the first study,
Fonagy and colleagues (1996) conducted the first large-scale treatment study examining
change in attachment. They compared pre- and posttreatment AAI as a function of
diagnosis in 82 nonpsychotic inpatients treated at Cassel Hospital with psychoana­
lytic therapy and 85 case-matched controls. Much to their surprise, they found that
individuals classified as dismissing on the AAI were more likely to display clinically
significant improvements (93%) on the Global Assessment of Functioning scale (av­
erage increase of 16 points). In comparison to the dismissing patients, only 43% of the
preoccupied (average increase 6 points) and 33% of secure patients (average increase
of 5 points), respectively, showed significant clinical improvement. Part of the reason
for this finding was that the dismissing patients began with lower Global Assessment
of Functioning (GAF) scores (20 compared with 25 and 36 for preoccupied and se­
cure patients, respectively. Thus secure patients had the lowest chance of showing large
changes, even though by discharge secure participants continued to have higher GAF
scores than dismissing and preoccupied patients. Fonagy and colleagues also found
preoccupied patients were more likely to drop out of treatment.
In an earlier book chapter, Fonagy and colleagues (1995) reported partial findings
from a subset of 35 of the 82 inpatients in the Cassel Hospital inpatient study. All 35
 25 Attachment Style

inpatients were classified insecure during their initial interview. However, 14 (40%)
o f the 35 inpatients were assigned a secure classification upon discharge. This in­
crease in the proportion of secure classification was highly significant (p < .001). On
the individual scale ratings, bland or idealized pictures of parents and a pattern of
pervasive memory blockages were more characteristic o f the AAIs at intake than at
discharge, and they appear to have been changed by treatment. These findings are
important because they show that attachment patterns can change as a function of
treatment. However, neither the specific psychopathology nor the treatment was well
specified. Additionally, no more detailed description of the changes in AAI status
observed in this study has been published, making reports of these findings difficult
to interpret.
In a second landmark set of studies at the Personality Disorders Institute at Cornell
University, we conducted a randomized controlled trial (Levy et al., 2006) with 90
patients with borderline personality disorder. They were randomized to a psycho­
dynamic treatment called transference-focused psychotherapy (TFP), an integrative
cognitive behavioral therapy called dialectical behavior therapy, or a modified psycho­
dynamic supportive psychotherapy. Attachment organization was assessed using the
AAI, as were narrative coherence and reflective function, facets related to attachment
security. At baseline, only 2 of the 90 patients were rated as securely attached based on
the AAI. After 12 months of treatment, there was a significant increase in the number
of patients classified as secure in the TFP group but not in the other two treatment
groups. Significant changes in narrative coherence and reflective function were found
as a function of treatment, with TFP showing increases in both during the course of
treatment. These findings were confirmed in another sample (Buchheim et al., 2017;
Fischer-Kern et al., 2015). Thus there are now several studies demonstrating that a
year of interpersonal-focused psychodynamic psychotherapy produces improvements
in attachment style and reflective function.
In the third set of landmark studies, Tasca and colleagues (Illing et al., 2010; Tasca
et al., 2006, 2013) evaluated the efficacy of group psychodynamic interpersonal
therapy (GPIP) versus group cognitive-behavioral therapy (CBT) for primarily women
with binge eating disorder. GPIP is based on the idea that “cyclical relational patterns”
(CRPs) underlie maladaptive behaviors such as engagement in binge eating. As GPIP
was developed with an explicit focus on attachment, CRPs are modified in the course
of group interactions via the therapist’s assessment and elucidation of these patterns to
group members using interactions among the members themselves.
We report three findings of note. First, Need for Approval (attachment anxiety)
predicted worse outcome for women in group CBT but better outcome in GPIP
(Tasca et al., 2006). Relatedly, women high on the Relationships as Secondary (attach­
ment avoidance) were associated with risk for dropout only in group CBT. Second,
although attachment anxiety did not influence improvement across multiple GPIP
groups (divided into groups of low and high attachment anxious individuals), high
anxiety groups showed a stronger relationship between group alliance and outcome.
Finally, after one year of GPIP, improvements in attachment security co-occurred with
improved interpersonal functioning and depressive symptoms up through 12 months
 26 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

of follow-up. Together, this set of studies suggests attachment insecurity may moderate
treatment effects and attachment improvements may correspond to symptom change.

RESULTS OF PREVIOUS META-ANALYSES

In our original meta-analysis (Levy et al., 2011) we identified 14 studies containing

19 separate samples and 1,467 patients. Because these studies used 11 different at­
tachment measures, we focused on analyses on attachment anxiety and attachment
avoidance as the two dimensions underlying attachment style in these measures. Our
findings showed that higher anxiety about attachment predicted worse psychotherapy
outcomes (r = - .22), whereas higher attachment security predicted more favorable
psychotherapy outcomes (r = .18). Attachment avoidance had an insignificant effect
on psychotherapy outcome. The effects sizes for attachment anxiety and attachment
security were in the low to moderate range (Cohen, 1988).
However, one limitation of our findings was that an insufficient number of studies
available for inclusion at the time provided baseline levels of outcome variables
needed to determine change in outcome over the course of treatment as well as the
posttreatment levels of attachment needed to determine change in attachment over the
course of treatment. This limitation precluded an examination of effects of attachment
on change in outcome and the effects of change in attachment on change in outcome.
Fortunately, such an examination is now possible due to the move in recent year to­
ward multiple assessments of process measures in psychotherapy research.

META-ANALYTIC REVIEW
To characterize the relation between adult attachment and psychotherapy outcome,
we conducted several meta-analyses. We focused on attachment anxiety and avoid­
ance, because these are the most established and prevalent dimensions of attachment
in the literature. In addition, we also examined attachment security, which can be
conceptualized as a blend of low avoidance and low anxiety dimensions.
Specifically, we examined whether pretreatment attachment predicts treatment
outcome. We hypothesized that lower attachment anxiety, lower attachment avoid­
ance, and greater attachment security would each be related to better outcome and
lower likelihood of dropout. In addition, we examined whether changes in attachment
from pre- to posttreatment predict changes in outcome. Here, we hypothesized that
decreases in attachment anxiety, decreases in attachment avoidance, and increases in
attachment security would be each related to better outcome. To examine whether
treatment adaptation may moderate these relations, we examined treatment type as a
possible moderator. We focused on whether the treatment had an interpersonal compo­
nent (vs. not) as a proxy for a possible adaptation, under the assumption that an inter­
personal component better addresses the attachment and therapeutic needs of patients
with insecure attachment. Thus, although attachment insecurity is expected to confer
disadvantage in terms of treatment outcome in general, we expected that this disad­
vantage would be smaller in interpersonal treatments compared to noninterpersonal
 27 Attachment Style

treatments. This, we predicted, would be evidenced by smaller effects of attachment on

outcome in interpersonal treatments.

Inclusion Criteria and Search Strategy-

Eligible studies were reports of psychotherapy outcome in samples of adult
psychotherapy-seeking individuals (see Figure 2.1 for a flow chart). These studies were
found first through articles reviewing the literature (e.g., Berant & Obegi, 2009) and
second through a series of PsycINFO searches. These searches, conducted in May 2015
and July 2016, used the intersections of the terms attachment, interpersonal style, re­
lation* style, or the name of an adult attachment measure (see Ravitz et al., 2010, for
a list of adult attachment measures) with either therap* outcome, psychotherap* out­
come, or outcome. The search also specified either treatment outcome/clinical trial or
empirical study methodology.
The combination of these literature searches returned 2,887 results. For a study
to be eligible for the current meta-analysis, it had to meet the following criteria: (a)
presented in English; (b) published report of individual, group, couples, and/or
family psychotherapy (e.g., not psychopharmacology) outcome in samples of treat­
ment seeking adults; (c) measures both patients’ pretreatment attachment security,
anxiety, and/or avoidance as well as outcome post-treatment; (d) must not be a case
study; (e) must not be a dissertation. To avoid confounding attachment with ther­
apeutic alliance, reports were not included if the measure of attachment concerned

f ig u r e 2 . 1 Flow chart for screening and inclusion of studies in the meta-analyses.

 28 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

client attachment to therapist. Once removing studies that did not meet these criteria,
49 primary studies remained.
We aimed to obtain relational statistics for pretreatment attachment to posttreatment
outcome and pre-post change in outcome, as well as pre-post change in attachment to
pre-post change in outcome. Due to the breadth of data we were looking to obtain, all
but one (study by first author) of the corresponding authors of 49 primary studies were
contacted. Twenty of these authors responded with suitable statistics for 25 studies. An
additional 9 authors responded and reported that additional statistics were not avail­
able for 10 studies, though 6 of these studies had originally reported suitable statistics
for at least one of the analyses. O f the 12 authors of 13 studies who either did not re­
spond, or we lost further contact with, only 4 studies had originally reported suitable
statistics for at least one analysis. Our final pool of studies in the meta-analysis thus
consisted of 36 studies, which contained 48 separate therapy samples with a combined
N of 3,158 patients. Table 2.1 lists the studies included in the meta-analysis along with
relevant characteristics of their designs and samples.

Effect Size Estimates

Pearson’s correlation coefficient was chosen as the principle measure of effect size be­
cause most studies utilized dimensional measures of attachment and outcome. When
dichotomous measures were reported (e.g., dropout status or attachment classifica­
tion), we used point biserial (for correlations between dimensional and dichotomous
measures) or Cramer’s V (for correlations between two dichotomous measures) cor­
relation coefficients, which are standardized and have the same metric as Pearson’s
correlation. Prior to inclusion in the meta-analysis, all effect sizes were rescaled such
that positive effects express greater attachment security (or lower anxiety/avoidance in
attachment) predicting better treatment outcome/less dropout. For these effect sizes,
we also present Cohen’s d in the text to ease interpretation and to enhance comparison
among the results in the chapters in this volume.

Study Coding
Coding of the 36 studies was conducted by trained graduate and undergraduate
students. Several patient characteristics were coded, including the proportion of the
sample that was female, mean age of the sample, proportion of the sample that was
White or Caucasian, mean level of education, proportion of the sample that dropped
out, and whether the primary diagnosis of the sample was a personality disorder. The
treatment characteristics coded included whether the treatment was individual and/
or group therapy, if the treatment included inpatient care, and length of treatment
in weeks. Because the 48 samples included in the current study were offered 31 dif­
ferent types of psychotherapy, the specific treatment was not formally coded. Instead,
we classified treatment arms as “interpersonal” in nature (any component directly
addressing interpersonal difficulties and concerns; e.g., psychodynamic and inter­
personal therapies) and those “non-interpersonal” in nature (e.g., CBT for bulimia
Table 2.1. Summary of Studies Included in Meta-Analysis of Patient Attachment and Treatment Outcome
Psychotherapy Patients Attachment Outcome
Study Type IP Duration N %F Age Diagnosis Measure Styles Timepoints Domains Timepoints
Belanger et al., 2011 PS IP 14 44 66 40 PDA RQ Sec Pre Sym/Fun Post
Benson et al., 2013 IC B T (Male) IP 22 44 0 43 m arital AAS Anx/Avo Pre Sym/Fun Pre/Post
IC B T (Female) IP 22 45 100 42 m arital AAS Anx/Avo Pre Sym/Fun Pre/Post
T C B T (Male) IP 24 41 0 43 m arital AAS Anx/Avo Pre Sym/Fun Pre/Post
T C B T (Female) IP 24 42 100 42 m arital AAS Anx/Avo Pre Sym/Fun Pre/Post
B ernecker et al., 2016 CBT NIP 16 26 73 34 MDD ECR Anx/Avo Pre/Post Sym/Dro Pre/Post
IP T IP 16 27 78 43 MDD ECR Anx/Avo Pre/Post Sym/Dro Pre/Post
B erry et al., 2015 M I+ C B T NIP 52 72 87 38 Psych.+SU PAM Anx/Avo Pre Sym/Fun Pre/Post
Byrd et al., 2010 M ixed N/A 12 66 59 23 M ixed A A S-R Anx/Avo Pre Sym Pre/Post
Daniel et al., 2016 CBT NIP 20 32 99 26 BN AAI Sec/Anx/Avo Pre Sym/Dro/Per Pre/Post
PD T IP 20 24 99 26 BN AAI Sec/Anx/Avo Pre Sym/Dro/Per Pre/Post
Diam ond et al., 2016 A BFT IP 11 16 61 25 UA ECR-RS Anx/Avo Pre/Post Sym Pre/Post
E FT IP 11 16 61 26 UA ECR-RS Anx/Avo Pre/Post Sym Pre/Post
Forbes et al., 2010 CBT NIP 12 103 0 53 PTSD RSQ Sec/Anx/Avo Pre Sym Pre/Post
Gois et al., 2014 IP T IP 26 11 100 57 MDD A A S-R Sec/Anx/Avo Pre/Post Sym/Fun Pre/Post
Hoyer et al., 2016 CT NIP 39 244 54 35 SAD EC R -R Anx/Avo Pre Sym/Dro Post
Illing et al., 2010 Int. IP 12 147 100 27 ED ASQ Sec/Anx/Avo Pre Sym/Dro Pre/Post
Joyce et al., 2009 PD T IP 18 133 64 37 M ixed RAQ Sec/Anx/Avo Pre Sym/Dro/Fun/Per Pre/Post
Joyce et al., 2013 PD T IP 18 48 67 41 M ixed ECR Anx/Avo Pre Sym/Dro/Fun/Per Pre/Post
Kowal et al., 2015 CBT NIP 4 235 61 48 Pain ECR Anx/Avo Pre Sym/Fun Post
Lawson & Brossart, 2009 Int. C B T -P D T IP 17 49 0 32 IPV AAS Anx/Avo Pre Sym Post
Levy et al., 2006 D BT NIP 52 29 96 32 BPD A A I, EC R Sec/Anx/Avo Pre/Post Sym/Dro/Fun Pre/Post
(continued )
Table 2.1. Continued
Psychotherapy Patients Attachment Outcome
Study Type IP Duration N %F Age Diagnosis Measure Styles Timepoints Domains Timepoints
SPT IP 52 30 93 33 BPD AAI, EC R Sec/Anx/Avo Pre/Post Sym/Dro/Fun Pre/Post
T FP IP 52 30 96 28 BPD AAI, EC R Sec/Anx/Avo Pre/Post Sym/Dro/Fun Pre/Post
Lindgren et al., 2008 PD T IP 12 32 73 44 M ixed RSQ Sec/Anx/Avo Pre Sym/Dro Pre/Post
M arm arosh et al., 2009 M ixed N/A 15 31 71 25 unspec. EC R Anx/Avo Pre Sym Post
M cBride et al., 2006 CBT NIP 17 28 74 40 MDD RSQ Anx/Avo Pre Sym Post
IP T IP 17 27 72 41 MDD RSQ Anx/Avo Pre Sym Post
M eyer et al., 2001 M ixed N/A 52 104 57 35 PDNOS AA PR Anx/Avo Pre Sym/Fun Post
M uller & Rosenkranz, Int. IP 8 61 64 43 PTSD RSQ Sec/Anx/Avo Pre/Post Sym/Dro Pre/Post
2009
Newman et al., 2015 C B T + IEP IP 14 42 72 37 GA D PAAQ Anx/Avo Pre/Post Sym/Dro Pre/Post
C B T + SL NIP 14 39 80 37 GA D PAAQ Anx/Avo Pre/Post Sym/Dro Pre/Post
Parker et al., 2012 M F T (Female) IP 4 594 100 m arital EC R Anx/Avo Pre Sym Post
M F T (Male) IP 4 594 0 m arital EC R Anx/Avo Pre Sym Post
R einer et al., 2016 PD T IP 8 41 100 30 MDD AAI Sec Pre/Post Sym Pre/Post
Reis and Grenyer, 2004 PD T IP 16 58 59 46 MDD RQ Sec/Anx/Avo Pre Sym Pre/Post
Sauer et al., 2010 M ixed N/A 7 50 68 28 M ixed EC R Anx/Avo Pre Sym Pre/Post
Sm ith et al., 2012 IP T IP 36 30 100 37 M D D + C SA EC R Anx/Avo Pre Sym/Dro Pre/Post
TAU N/A 36 20 100 37 M D D + C SA EC R Anx/Avo Pre Sym/Dro Pre/Post
Stalker et al., 2005 Int. IP 6 112 100 41 PTSD RAQ Anx/Avo Pre/Post Sym/Dro Pre/Post
Strauss et al., 2006 PD T IP 10 504 70 34 M ixed AA PR Sec Pre Sym/Per Post
Strauss et al., 2011 Int. P D T -P C T IP 7 19 100 31 AVPD AA PR Sec/Anx/Avo Pre/Post Sym/Fun/Per Pre/Post
Int. P D T -P C T IP 7 21 100 31 BPD AA PR Sec/Anx/Avo Pre/Post Sym/Fun/Per Pre/Post
Tasca et al., 2006 CBT NIP 16 46 91 43 BED ASQ Sec/Anx/Avo Pre/Post Sym/Dro/Per Pre/Post
 PD T IP 16 47 91 43 BED ASQ Sec/Anx/Avo Pre/Post Sym/Dro/Per Pre/Post
Tasca et al., 2013 PD T IP 16 83 100 42 BED ASQ Sec/Anx/Avo Pre/Post Sym/Dro/Per Pre/Post
Taylor et al., 2015 CBT NIP 13 56 62 41 M ixed EC R Anx/Avo Pre/Post Sym/Dro Pre/Post
Travis et al., 2001 PD T IP 21 32 77 41 unspec. BARS Sec/Anx/Avo Pre/Post Sym/Dro Pre/Post
W atson et al., 2014 CBT NIP 16 26 67 38 MDD ASQ Sec/Anx/Avo Pre/Post Sym/Dro/Per Pre/Post
E FT IP 16 29 67 38 MDD ASQ Sec/Anx/Avo Pre/Post Sym/Dro/Per Pre/Post
Zalaznik et al., 2017 CBT NIP 12 31 61 33 PDA EC R Anx/Avo Pre/Post Sym/Dro Pre/Post

Notes. Psychotherapy type: ABFT = attachment-based family therapy, CBT = cognitive-behavioral therapy, CT = cognitive therapy, DBT = dialectical behavior therapy, EFT = emotion-
focused therapy, ICBT = integrative couples’ behavior therapy, IEP = interpersonal and emotional processing, Int. = integrative, IPT = interpersonal therapy, MFT = marriage and family
therapy, MI = motivational interviewing, PDT = psychodynamic therapy, PS = problem-solving, SL = supportive listening, SPT = supportive-psychodynamic therapy, TAU = treatment-as-
usual, TCBT = traditional couples’ behavior therapy, TFP = transference-focused psychotherapy.
Interpersonal psychotherapy type: IP = interpersonal, NIP = non-interpersonal.
Duration: Psychotherapy duration in weeks.
Patient variables: %F = percentage female; Age = mean age in years.
Diagnosis: AVPD = avoidant personality disorder, BED = binge eating disorder, BN = bulimia nervosa, BPD = borderline personality disorder, CSA = child sexual abuse, ED = eating dis­
order, GAD = generalized anxiety disorder, IPV = intimate partner violence, MDD = major depressive disorder, PDA = panic disorder and agoraphobia, PDNOS = personality disorder
not-otherwise specified, Psych. = psychosis, PTSD = posttraumatic stress disorder, SAD = social anxiety disorder, SU = substance use, UA = unresolved anger, unspec. = unspecified.
Attachment measure: AAI = Adult Attachment Interview, AAPR = Adult Attachment Prototype Rating, AAS/-R = Adult Attachment Scale/-Revised, ASQ = Attachment Style Questionnaire,
BARS = Bartholomew Attachment Rating Scale, ECR/-R/-RS = Experiences in Close Relationships scale/-Revised/-Relationship Structures, PAAQ = Perceptions of Adult Attachment
Questionnaire, PAM = Psychosis Attachment Measure, RAQ = Reciprocal Attachment Questionnaire, RQ = Relationship Questionnaire, RSQ = Relationship Scales Questionnaire.
Attachment style: Anx = anxiety, Avo = avoidance, Sec = security.
Attachment/outcome measurement timepoints: Pre = pretreatment, Post = posttreatment.
Outcome domain: Dro = dropout, Func = functioning, Per = personality, Sym = symptoms.
 32 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

nervosa; Daniel et al., 2016). Five treatment arms were not included in these analyses
because they included mixed treatment types that differed by patient and could not be
reliably classified as either “interpersonal” or “non-interpersonal.” The classification of
treatment type was done independently by the second and fourth authors (postdoctoral
fellow and advanced graduate student in clinical psychology) based on the description
of the treatments that was included in the papers. The authors agreed on 42 of the 43
coded treatment arms (kappa = .95); the one disagreement was resolved by discussion.

Meta-Analytic Procedures
Data were analyzed in R Version 3.4.2 package “metafor” (Viechtbauer & R Core
Team, 2017). Prior to analyses, correlations were transformed to a Fisher’s z
scale to ensure normality o f the underlying distribution (Borenstein et al., 2009).
Estimated z scores were transformed back to r values following analyses for ease of
interpretation.
Some of the studies included multiple treatment arms (e.g., CBT and psychody­
namic therapies). In addition, most studies reported multiple outcome measures as
well as included multiple attachment scales (e.g., avoidance and anxiety). Therefore,
we used the random-effects multivariate multilevel option for meta-analyses with mul­
tiple effect sizes nested within treatment groups, which in turn were nested within
studies. Multivariate meta-analysis takes into account the correlations among meas­
ures to adjust the weights of individual effect sizes. However, given that the majority of
the studies did not report the correlations among these measures, missing correlations
were imputed at r = .30 as a conservative estimate (Borenstein et al., 2009).
Separate meta-analyses were conducted to test whether (a) pretreatment attach­
ment styles predicted treatment outcome at post-treatment (“Pre-to-Post” effect sizes);
(b) pretreatment attachment styles predicted changes in outcome during treatment
(“Pre-to-Change”); (c) changes in attachment during treatment predicted changes
in outcome during treatment (“Change-to-Change”); and (d) pretreatment level of
attachment predicted dropout (“Pre-to-Dropout”). Thus treatment outcome was
operationalized in two different ways: First, posttreatment scores of outcome measures
were used as dependent variables. Second, because posttreatment scores do not take
into account pretreatment severity, we also looked at the changes in outcome from
pre- to posttreatment that were available for a subset of the studies in the meta-analysis
(k = 25; 69% of studies in the meta-analysis).
Publication biases were examined using funnel plots (Light & Pillemer, 1984) and
fail-safe N calculations (i.e., the number of studies with an effect size of r = 0 that
would need to be added to the pool of studies to reduce the effect size to a trivial
level of r = .05; Orwin, 1983). Heterogeneity of effects was examined by estimating
the percentage of variance attributed to real differences in effect sizes (a multi-tevel
I2; Viechtbauer & R Core Team, 2017). Because attachment anxiety and attachment
avoidance somewhat overlap but largely orthogonal to each other while attachment
security is nonorthogonal to either anxiety or avoidance, we fitted separate models
for effect sizes involving attachment anxiety/avoidance and for effect sizes involving
 33 Attachment Style

attachment security. An omnibus model in each category was followed up with sub­
group analyses examining the effects of attachment dimension (anxiety vs. avoidance),
outcome domain (symptoms, functioning, and personality), and treatment type (inter­
personal vs. non-interpersonal) on the correlation between attachment and outcome/
dropout. Finally, we also examined whether treatment type or other study characteris­
tics moderated these effect sizes.

Overall Effect Size on Psychotherapy Outcome

In total, 36 studies were included, representing a total of 3,158 patients (M = 88,
SD = 104), producing 827 effect sizes for inclusion in the meta-analyses. Pre-to-Post
effect sizes were the most commonly available (317 effects), followed by Pre-to-Change
(277), Change-to-Change (169), and finally Pre-to-Dropout (64). No effect sizes were
found to be outliers (falling 3 or more standard deviations away from the mean effect
size; Kline, 2016) among any effect size type.

Pretreatment Attachment as a Predictor of Treatment Outcome

Based on 32 studies, we found that pretreatment attachment was a small to moderate
predictor of posttreatment outcome, regardless of attachment style (r = .17, d = .35,
p < .001, 95% confidence interval [CI] = .1 3 - .22). Similar to the results of our prior
meta-analysis (Levy et al., 2011), greater attachment security/less insecurity predicted
better posttreatment outcome. A funnel plot (Figure 2.2, top left panel) did not show
evidence for a publication bias, and fail-safe N calculations indicated that 550 studies
would be needed to trivialize this effect. Heterogeneity estimates showed that 53% of
the variance could be attributed to real differences among effect sizes.
However, controlling for pretreatment levels on outcome measures by examining
pre-to-post changes in outcome showed that, based on a total of 25 studies, 277 effect
sizes, and 1,671 participants, the effect of baseline attachment on change in outcome
variables was nonsignificant across all attachment styles (r = .03, d = .06, p = .13, 95%
CI = -.0 1 -.0 7 ). A funnel plot (Figure 2.2, top right) did not show evidence for a pub­
lication bias, and heterogeneity estimates suggested that 34% of the variance could be
attributed to real differences among effect sizes.

Pre-to-Post Change in Attachment as a Predictor of Pre-to-Post

Change in Outcome
Examining whether change in attachment security/insecurity predicts change in out­
come we found that, based on a total of 15 studies, greater improvement in attachment
security predicted greater improvement in outcome (r = .16, d = .32, p < .001, 95%
CI = .07-.25). A funnel plot (Figure 2.2, bottom left) did not show evidence for a pub­
lication bias, and fail-safe N calculations indicated that 446 studies would be needed to
trivialize this effect. Heterogeneity estimates showed that 52% of the variance could be
attributed to real differences among effect sizes.
 34 P S Y C H O T H E R A P Y R E L A T I O N S H I P S THAT W ORK

f ig u r e 2 . 2 Funnel plots of effect sizes for the prediction of treatment outcome and dropout from
attachment style, shown separately for pretreatment level of attachment as a predictor of treatment
outcome at posttreatment (top left panel); pretreatment level of attachment as a predictor of change
in outcome during treatment (top right); change in attachment during treatment as a predictor
of change in outcome during treatment (bottom left); and pretreatment level of attachment as a
predictor of dropout (bottom right).

Pretreatment Attachment as a Predictor of Treatment Dropout

Based on a total of 18 studies, we found that baseline attachment did not predict
rates of dropout (r = .04, d = .08, p = .15, 95% CI = - .0 1 - .09). A funnel plot (Figure
2.2, bottom right) did not show evidence for a publication bias, and heterogeneity
estimates suggested that 33% of the variance could be attributed to real differences
among effect sizes.

Effect Size by Attachment Style and Outcome Domain

The next step was to examine whether the effect of attachment differs by attachment
style (security, anxiety, and avoidance) and outcome domain (symptoms, personality,
 35 Attachment Style

Table 2 .2 . M ean E stim ated E ffect Sizes for th e P rediction o f Treatm ent
O utcom e from A ttachm ent by O utcom e D om ain and A ttachm ent Style

Pre-to-Post M (95% CI) M (95% CI) M (95% CI)

Symptoms Personality Functioning
Security .16 (.0 7 -.2 5 )** .23 (.0 9 -.3 5 )** .19 (.0 2 -.3 4 )*
Anxiety .19 (.1 4 -.2 4 )** .22 (.1 0 -.3 4 )** .20 (.1 1 -.3 0 )**
Avoidance .16 ( .1 1 -.2 1 )* * .25 (.1 3 -.3 7 )** .08 (-.0 2 - .1 8 )

Pre-to-Change
Symptoms Personality Functioning
p
co

Security - .0 7 (-.2 1 - .0 7 ) .04 (-.1 4 - .2 1 )

i*

Anxiety .04 ( - .0 1 - .0 9 ) - .0 7 (-.1 8 - .0 4 ) .08 (-.0 2 - .1 8 )

Avoidance .05 (.0 0 -.1 0 ) -.0 8 (-.1 9 - .0 3 ) -.0 1 (-.1 0 - .0 9 )

Change-to-Change
Symptoms Personality Functioning
Security .19 (.0 3 -.3 4 )* .10 ( - . 1 1-.30) .02 (-.2 5 - .2 9 )
Anxiety .18 (.0 8 -.2 7 )** .26 (.0 9 -.4 2 )** .16 (-.0 5 - .3 6 )
Avoidance .15 (.0 5 -.2 4 )** .29 (.1 2 -.4 5 )** .00 (-.2 1 - .2 1 )

Notes. CI = confidence interval. Pre-to-Post = pretreatment level of attachment as

a predictor of treatment outcome at posttreatment; Pre-to-Change = pretreatment
level of attachment as a predictor of change in outcome during treatment; Change-
to-Change = change in attachment during treatment as a predictor of change in
outcome during treatment. Numbers in parentheses represent 95% Confidence
Intervals.
* Significantly different from zero at the p < .05 level; ** Significantly different from
zero at the p < .01 level.

and functioning). The estimated effect sizes, broken down by attachment style and
outcome domain, are reported in Table 2.2. Forest plots of the distributions of the
main effect sizes are presented in Figures 2.3 to 2.6. Effect sizes and 95% CI in the plots
may differ slightly from those included in the text due to differences in estimation
procedures used to generate multilevel forest plots.

Pretreatment Attachment as a Predictor of Treatment Outcome

A forest plot of the effect of pretreatment attachment, across dimensions, on
posttreatment outcome is included in Figure 2.3. Looking at attachment security,
we found that the effect sizes of pretreatment attachment security on outcome at
posttreatment were all significant, in the small-moderate range (r = .16-.23, d = .32­
.47) and did not differ by outcome domain (X = 1.31, p = .52). In addition, looking
at attachment anxiety and attachment avoidance, we found that the effect size of
pretreatment attachment on outcome at posttreatment did not differ by attachment
style (X = 1.45, p = .23), outcome domain (X = 2.85, p = .24), or their interaction
 N r [95% Cl]

B e r n e c k e r e t a l. 2 0 1 5 53 1------------- T“ 1------------------1 0 .0 6 [ - 0 . 1 9 , 0 .2 9 ]
B e n s o n e t a l. 2 0 1 3 172 I I - — ■---------1 0 .1 3 [ - 0 . 0 1 , 0 .2 6 ]
G o i s e t a l. 2 0 1 4 11 h-------------------- 1— -----------■------------------------------------- 1 0 .2 1 [ - 0 . 2 9 , 0 .6 1 ]
I l l i n g e t a l. 2 0 1 0 147 1— ■— 1 0 .2 3 [ 0 .1 1 , 0 .3 4 ]
J o y c e e t a l. 2 0 0 9 133 1---------- " r - ------1 -0 .0 2 [ - 0 . 1 7 , 0 .1 3 ]
J o y c e e t a l. 2 0 1 3 48 1------- h - -------- ■-------------------- 1 0 .1 8 [ - 0 . 1 0 , 0 .4 3 ]
L a w s o n e t a l. 2 0 0 9 49 y ----------- ■----------------1 0 .2 1 [ - 0 . 0 0 , 0 .4 1 ]
L e v y e t a l. 2 0 0 6 88 1--------- r - ■----------------- 1 0 .1 1 [ - 0 . 1 3 , 0 .3 3 ]
L i n d g r e n e t a l. 2 0 0 8 32 1--------- h - — ■------------------- 1 0 .1 4 [ - 0 . 1 3 , 0 .3 9 ]
D a n i e l e t a l. 2 0 1 6 36 1— j— — »---------------1 0 .1 2 [ - 0 . 0 9 , 0 .3 1 ]
B e l a n g e r e t a l. 2 0 1 1 44 1---------------r * ---------------- 1 0 .0 3 [ - 0 . 2 0 , 0 .2 6 ]
M a r m a r o s h e t a l. 2 0 0 9 31 1---------------<—— »----------------------- 1 0 .1 2 [ - 0 . 2 0 , 0 .4 1 ]
M c B r i d e e t a l. 2 0 0 6 55 1----------------- j—--------- ■------------------------------- 1 0 .1 9 [ - 0 . 2 3 , 0 .5 5 ]
M e y e r e t a l. 2 0 0 1 104 1— |— 1■----------1 0 .0 6 [ - 0 . 0 9 , 0 .2 0 ]
M u lle r & R o s e n k r a n z 2 0 0 9 78 1-------- H - ■-------------- 1 0 .0 8 [ - 0 . 1 2 , 0 .2 7 ]
P a r k e r e t a l. 2 0 1 2 594 1-------------■-------------1 0 .3 2 [ 0 .1 4 , 0 .4 7 ]
S a u e r e t a l. 2 0 1 0 50 1----------------- ■------------------1 0 .3 8 [ 0 .1 5 , 0 .5 7 ]
S m i t h e t a l. 2 0 1 2 50 1----------------------- ■---------------------- 0 .4 1 [ 0 .1 1 , 0 .6 4 ]
S t a lk e r e t a l. 2 0 0 5 125 --------- * --------- 1 0 .1 9 [ 0 .0 5 , 0 .3 2 ]
S t r a u s s e t a l. 2 0 0 6 504 1— ■ — 1 0 .1 7 [ 0 .0 9 , 0 .2 4 ]
T a s c a e t a l. 2 0 0 6 66 1------------- ■--------------1 0 .3 0 [ 0 .1 1 , 0 .4 7 ]
T a s c a e t a l. 2 0 1 3 83 ------■----------- 1 0 .1 4 [ - 0 . 0 2 , 0 .3 0 ]
T r a v i s e t a l. 2 0 0 1 59 1----------r - -------■--------------------- 1 0 .1 5 [ - 0 . 1 4 , 0 .4 2 ]
W a t s o n e t a l. 2 0 1 4 61 1-------------- ■-------------- 1 0 .3 3 [ 0 .1 3 , 0 .5 0 ]
R e in e r 2 0 1 6 41 1— y — ■----------------------- 1 0 .1 6 [ - 0 . 1 6 , 0 .4 5 ]
D i a m o n d e t a l. 2 0 1 6 32 1---------------- r - ■----------------------1 0 .0 8 [ - 0 . 2 2 , 0 .3 6 ]
H o y e r e t a l. 2 0 1 6 244 H - ■— 1 0 .1 1 [ - 0 . 0 3 , 0 .2 3 ]
N e w m a n e t a l. 2 0 1 5 81 1------------T * ------------- 1 0 .0 3 [ - 0 . 1 6 , 0 .2 3 ]
T a y l o r e t a l. 2 0 1 5 21 I— r - ---------------------■----------------------------- 0 .3 3 [ - 0 . 0 6 , 0 .6 3 ]
B e r r y e t a l. 2 0 1 5 135 1-------------■-------------1 0 .4 0 [ 0 .2 3 , 0 .5 4 ]
K o w a l e t a l. 2 0 1 5 235 1— ■ — 1 0 .1 7 [ 0 .0 7 , 0 .2 6 ]
Z a l a z n i k e t a l. 2 0 1 7 31 1--------- i—------ ■-------------------- 1 0 .1 5 [ - 0 . 1 3 , 0 .4 1 ]

R E M odel ; ♦ 0 .1 7 [ 0 .1 4 , 0 .2 0 ]

1 1 1 1 1 1
-0 .3 8 -0 .2 0 0 .2 0 .3 8 0 .5 4 0 .6 6

O b se rv e d O u tco m e

fig u r e 2.3 Forest plot o f the distribution of effect sizes of pretreatment attachment predicting posttreatment outcome.
Positive effect sizes mean greater attachment security, lower attachment avoidance and lower attachment anxiety predict better outcome.
 37 Attachment Style

(X = 2.66, p = .26). These effects were all significant and in the small-moderate range
(r = .16-.25, d = .32-.52) except for the effect of pretreatment attachment avoidance on
posttreatment functioning (r = .08, d = .16), which was not significant.
We then controlled for the effect of pretreatment levels on outcome measures by
examining pre-to-post changes in outcome. A forest plot of these effect sizes across
attachment dimensions is included in Figure 2.4. Looking at attachment security, we
found that the effect sizes of pretreatment attachment security on change in outcome
were all small in size, nonsignificant (r = -.0 7 -.0 4 , d = -.14-.08) and did not differ by
outcome domain ( X = 2.30, p = .32). In addition, looking at attachment anxiety and
attachment avoidance, we found that the effect sizes of pretreatment attachment on
change in outcome were all small in size and nonsignificant (r = -.0 7 -.0 8 , d = -.16-.14).
X = .93, p = .33), or the interaction
These effects did not differ by attachment style (
between attachment style and outcome domain (X = 2.37, p = .31), but there was a
main effect of outcome domain (X = 9.49, p < .01). Follow-up analyses showed that the
effects of pretreatment attachment on change in personality were smaller than those
for change in symptoms ( X = 9.48, p < .01) and for change in functioning (X = 5.02,
p = .03). There were no differences between the effect sizes for change in symptoms
and functioning ( X = .03, p = .86). However, given that none of the individual effect
sizes were significantly different from zero, the main effect of outcome domain does
not seem to be of practical significance.

Pre-to-Post Change in Attachment as a Predictor

of Pre-to-Post Change in Outcome
A forest plot of the effect change in attachment, across dimensions, on change in out­
come is included in Figure 2.5. First looking at attachment security, we found that
improvements in attachment security during treatment were positively correlated with
improvements in symptoms with a small-medium significant effect (r = .19, d = .39)
but not with improvements in personality or functioning. The main effect of outcome
domain was not significant (X = 2.25, p = .33). Looking at attachment anxiety and
attachment avoidance, we found that improvements in attachment during treatment
were positively correlated with improvements in symptoms and with improvements in
personality with a small-medium significant effect (r = .15-.29, d = .30-.61) but not
with improvements in functioning. The effects of attachment style ( X = 0.90, p = .34),
outcome domain ( X = 5.01, p = .08), or their interaction (X = 1.53, p = .46) were not
significant.

Pretreatment Attachment as a Predictor of Treatment Dropout

Treatment dropout was not significantly predicted by pretreatment levels of attach­
ment security (r = .04, d = .08, p = .17, 95% CI = -.0 2 -.0 9 ), anxiety (r = .04, d = .08,
p = .25, 95% CI = -.0 3 -.1 1 ) or avoidance (r = .04, d = .08, p = .29, 95% CI = -.0 3 -.1 1 ).
In addition, the effects of attachment anxiety and avoidance were not significantly dif­
ferent from one another ( X = 0.00, p = .94).
 N r [9 5 % Cl]

Bernecker et al. 2015 53 1--------- — ---------------1 -0.00 [-0.23, 0.23]

Benson et al. 2013 172 1— ■ — i -0.01 [-0.14, 0.12]
Byrd et al. 2010 66 1------------- ■------------- 1 0.24 [0.04, 0.43]
Forbes et al. 2010 103 1---- ■--------- 1 0.09 [-0.07, 0.24]
Gois et al. 2014 11 1------------------- ------ ■--------------------------------- 1 0.11 [-0.37, 0.55]
Illing et al. 2010 147 -H ■ 0.02 [-0.10, 0.14]
Joyce et al. 2009 133 1— -----* -------- 1 0.10 [-0.06, 0.25]
Joyce et al. 2013 48 1------------------■------ ----- 1 -0.20 [-0.45, 0.08]
Levy et al. 2006 88 1------- ■--------------- 1 0.04 [-0.20, 0.28]
Lindgren et al. 2008 32 1-------------- --------------- 1 -0.04 [-0.30, 0.23]
Daniel et al. 2016 36 ----- ■------------ 1 0.10 [-0.11, 0.29]
Muller & Rosenkranz 2009 78 1----- ------ * ----------- 1 0.12 [-0.08, 0.30]
Reis & Grenyer 2004 58 1---- -------- ■------------- 1 0.14 [-0.07, 0.34]
Sauer et al. 2010 50 1--------------- ■------ H -0.20 [-0.42, 0.05]
Smith et al. 2012 50 1------------ ------ ■------------------------- 1 0.12 [-0.27, 0.47]
Stalker et al. 2005 125 ■— i 0.04 [-0.10, 0.18]
Tasca et al. 2006 66 1--- ■-------------1 0.06 [-0.15, 0.26]
Tasca et al. 2013 83 1--------- ■ — -0.17 [-0.32, -0.00]
Travis et al. 2001 59 1------------ ■-------------------- 1 0.04 [-0.26, 0.34]
Watson et al. 2014 61 1------------ ■-------------1 -0.01 [-0.21, 0.20]
Diamond et al. 2016 32 1------------------- ---------------- 1 -0.07 [-0.37, 0.24]
Newman et al. 2015 81 1— ■------------1 0.05 [-0.14, 0.23]
Taylor et al. 2015 21 1------------------------- --------------------- 1 -0.08 [-0.44, 0.31]
Berry et al. 2015 135 1----------- * ----------- 1 0.23 [0.04, 0.40]
Zalaznik et al. 2017 31 1--------------- ----------------1 -0.04 [-0.31, 0.24]

RE Model 0.03 [-0.01, 0.08]

-0.54 -0.38 -0.2 0 0.2 0.38 0.54 0.66

Observed Outcome

fig u r e 2.4 Forest plot of the distribution o f effect sizes o f pre-treatment attachment predicting pre-to posttreatment change in outcome. Positive effect sizes mean greater
attachment security, lower attachment avoidance and lower attachment anxiety predict better outcome.
 N r [9 5 % Cl]

Bernecker et al. 2015 53 1---------- ■----------- 1 0.02 [-0.21, 0.24]

Gois et al. 2014 11 1--------- ------ 1 -0.02 [-0.48, 0.45]

Levy et al. 2006 88 1----- H 0.13 [-0.12, 0.36]

Muller & Rosenkranz 2009 78 h ------- ■--------- 0.16 [-0.04, 0.34]

Stalker et al. 2005 125 h— ■ — i 0.10 [-0.04, 0.24]

Strauss et al. 2011 40 1--------- ■------------- 0.07 [-0.20, 0.33]

Tasca et al. 2006 66 1--------- ■— -----1 0.25 [0.05, 0.43]

Tasca et al. 2013 83 i— ■— H 0.22 [0.06, 0.37]

Travis et al. 2001 59 1------ -------1 0.18 [-0.13, 0.46]

Watson et al. 2014 61 1— ----- ■---------- 1 0.45 [0.27, 0.61]

Reiner 2016 41 1--------------- ■--------------- 1 -0.01 [-0.32, 0.30]

Diamond et al. 2016 32 1----- -------1 0.19 [-0.11, 0.46]

Newman et al. 2015 81 1--------- -----------1 -0.15 [-0.48, 0.22]

Taylor et al. 2015 21 1------------- ------- 1 0.46 [0.10, 0.71]

Zalaznik et al. 2017 31 b ---------1 0.25 [-0.03, 0.49]

RE Model 0.17 [0.11, 0.23]

♦
1 1 1 1 1
-0.76 -0.46 0 0.46 0.76
Observed Outcome

fig u r e 2.5 Forest plot o f the distribution o f effect sizes of pre- to posttreatment change in attachment predicting pre- to posttreatment change in outcome. Positive effect sizes
mean larger increase in attachment security, larger decrease in attachment avoidance, and larger decrease in attachment anxiety predict larger improvement in outcome.
 40 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Thus, given that outcome domain was not a robust moderator in the previous
analyses, it was removed from subsequent models.

Effect Size by Attachment Style and Treatment Type

We also examined potential differences in the effects of attachment style by treatment
type (interpersonal vs. non-interpersonal). The estimated effect sizes, broken down by
attachment style and treatment type, are reported in Table 2.3.

Pretreatment Attachment as a Predictor of Treatment Outcome

First looking at attachment security, we found that the effect of treatment type
approached significance (x2 = 3.34, p = .07). Separate forest plots for these effects
are included in Figure 2.6. Consistent with our hypothesis, the effect of pretreatment

Table 2 .3 . Mean Estimated Effect Sizes for the Prediction of Treatment Outcome and
Dropout from Attachment by Treatment Type and Attachment Style
Pre-to-Change M (95% CI) M (95% CI)
Interpersonal Psychotherapy Non-interpersonal Psychotherapy
Security .15 (.06-.23)** .33 (.14-.50)**
Anxiety .18 (.11-.24)** .22 (.12-.31)**
Avoidance .17 (.11-.23)** .15 (.06-.24)**
Pre-to-Change
Interpersonal Psychotherapy Non-interpersonal Psychotherapy
Security -.01 (-.09-.07) .14 (-.02-.27)
Anxiety .03 (-.03-.08) .09 (-.01-.19)
Avoidance .02 (-.04-.07) .08 (-.03-.17)
Change-to-Change
Interpersonal Psychotherapy Non-interpersonal Psychotherapy
Security .15 (-.02-.31) .14 (-.17-.43)
Anxiety .19 (.09-.30)** .18 (.01-.33)*
Avoidance .15 (.03-.25)* .17 (.01-.33)*
Dropout
Interpersonal Psychotherapy Non-interpersonal Psychotherapy
Security .06 (-.03-.15) -.12 (-.32-.09)
Anxiety .05 (-.04-. 14) .02 (-.1 1-.15)
Avoidance .02 (-.07-.11) .05 (-.08-.17)

Notes. CI = confidence interval. Pre-to-Post = pretreatment level of attachment as a predictor of treat­

ment outcome at posttreatment; Pre-to-Change = pretreatment level of attachment as a predictor of
change in outcome during treatment; Change-to-Change = change in attachment during treatment as
a predictor of change in outcome during treatment.
* Significantly different from zero at thep < .05 level; ** Significantly different from zero at thep < .01 level.
 N [95% Cl]

Interpersonal Psychotherapies
Belanger et al. 2011 44 ^----------- 1 0.03 [-0.20, 0.26]
Daniel et al. 2016 (PDT) 36 H 0.06 [-0.26, 0.37]
Gois et al. 2014 11 1-------------------------- — ------ 1 -0.05 [-0.53, 0.45]
Illing et al. 2010 147 1----- * ----- 1 0.33 [0.20, 0.44]
Joyce et al. 2009 133 -------1 -0.02 [-0.18, 0.14]
Levy et al. 2006 88 ■--------------1 0.01 [-0.27, 0.30]
Lindgren et al. 2008 32 1------------- ■*-------------- 1 0.03 [-0.26, 0.32]
Muller & Rosenkranz 2009 78 *---------- 1 0.02 [-0.20, 0.23]
Reiner 2016 41 ------1 0.16 [-0.16, 0.45]
Strauss et al. 2006 504 1 ■ 1 0.17 [0.09, 0.24]
Tasca et al. 2006 (PDT) 66 -------1 0.21 [-0.07, 0.46]
Tasca et al. 2013 83 1-------- ■----- —1 0.23 [0.05, 0.39]
Travis et al. 2001 59 1----------------*---------------- 1 0.37 [0.06, 0.62]
Watson et al. 2014 (EFT) 61 1------------ -■-------------- 1 0.38 [0.11, 0.61]

Total Effect for Interpersonal Psychotherapies ♦ 0.15 [0.07, 0.23]

Non-Interpersonal Psychotherapies
Daniel et al. 2016 (CBT) 36 -------1 0.22 [-0.06, 0.46]
Tasca et al. 2006 (CBT) 66 1---------- — ■-------------- 0.42 [0.14, 0.63]
Watson et al. 2014 (CBT) 61 1----------- 0.42 [0.13, 0.64]

Total Effect for Non-Interpersonal Psychotherapies 0.34 [0.19, 0.49]

RE Model 0.18 [0.10, 0.25]

—
1 1 1 1 1
0.76 -0.46 0 0.46 0.76
Observed Outcome

fig u r e 2.6 Forest plots o f the distribution of effect sizes of pretreatment attachment security predicting post-treatment outcome, separately for interpersonal (top panel) and
non-interpersonal (bottom panel) psychotherapies. Positive effect sizes mean greater security predicts better outcome.
 42 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

attachment in non-interpersonal psychotherapies was in the medium-large range

(r = .33, d = .70) and tended to be larger than the effect in interpersonal psychotherapies,
which was in the small-medium range (r = .15, d = .30). In addition, looking at attach­
ment anxiety and attachment avoidance, we found that the effect size of pretreatment
attachment on outcome at posttreatment did not differ by treatment type (X = 0.06,
p = .81), attachment style (X = 2.16, p = .14), or their interaction (X = 1.21, p = .27).
When controlling for pretreatment levels of outcome measures, these effects did not
differ by treatment type (X = 1.35, p = .25), attachment style (X = 0.27, p = .61), or
their interaction (X = 0.02, p = .88).

Pre-to-Post Change in Attachment as a Predictor

of Pre-to-Post Change in Outcome
First looking at attachment security, we found that the effect sizes of improvements
in attachment security on improvements in outcome were nonsignificant for both
treatment types and did not differ by treatment type (X = 0.00, p = .95). Looking
at attachment anxiety and attachment avoidance, we found that the effect sizes of
improvements in attachment on improvements in outcome were all small-moderate
in size and significant (r = .15-.19, d = .30-.39). These effects did not differ by treat­
ment type (X = 0.00, p = .95), attachment style (X = 0.39, p = .53), or their interac­
tion (X = 0.29, p = .59).

Pretreatment Attachment as a Predictor of Treatment Dropout

Looking at attachment security, we found that the effect sizes of pretreatment attach­
ment on dropout were all nonsignificant (r = -.1 2 -.0 6 , d = -.24-.12) and did not differ
by outcome domain (X = 0.03, p = .95). In addition, looking at attachment anxiety
and attachment avoidance, we found that the effect sizes of pretreatment attachment
on dropout were all small in size and nonsignificant (r = .02-.05, d = .04-.10). These
effects did not differ by treatment type (X = 0.00, p = 1.00), attachment style (X = 0.00,
p = 1.00), or their interaction (X = 0.32, p = .57).

Summary of Meta-Analytic Results

The effect size for the association between attachment security, regardless of at­
tachm ent style, and psychotherapy outcome was in the small to moderate range
(r = .17, d = .35). Thus, in these 36 studies, clients’ attachment style reliably
predicted psychotherapy outcome. Moreover, in-ireatm ent improvement in at­
tachm ent security was related to improvement in treatm ent outcomes (r = .16,
d = .32) such that larger improvements in attachment security predicted larger
improvement in treatment outcome. Interpersonal psychotherapies may be less
hindered by attachment insecurity (r = .15, d = .30) than non-interpersonal
therapies (r = .33, d = .70).
 43 Attachment Style

MODERATORS

We examined the moderating effect of several treatment and patient characteristics

at the study level on the effects of attachment anxiety, attachment avoidance, and at­
tachm ent security on outcome. We used the I 2 statistic to determine the feasibility of
moderator analysis. In the present study, only one estimate showed I 2 values less than
25% (the effect of attachment security on dropout [I2 = .23]). Due to this finding,
and since this effect was not significant, we did not pursue moderator analysis o f this
effect. Furthermore, we only tested moderations where at least 20 effect sizes were
available.
In sum, only one significant main effect of a moderator on attachment effects
(averaged across both attachment anxiety and avoidance) appeared. In addition, five
study-level moderation effects on the differential effect of attachment anxiety versus
avoidance were significant. No moderation effects of the security-outcome relations
were revealed. Here we report only the results of significant moderator effects.
Both year of publication and treatment format predicted the difference between
pretreatment anxiety and avoidance in predicting posttreatment outcome. Specifically,
the older the study was, the stronger anxiety was as a predictor of outcome compared
to avoidance (b = -0.01, z = -2.74 , p < .01). Follow-up analyses suggested that the effect
of avoidance on outcome became stronger over time while the effect of anxiety tended
not to change. By 2011, there was no difference between the effect size of avoidance
and that of anxiety. Treatment format also appeared to produce differential effects of
anxiety versus avoidance on outcome (X = 14.20, p < .01). Follow-up analyses showed
that in treatments involving both individual and group therapy formats, attachment
anxiety predicted worse outcome (r = .34, p < .01) while attachment avoidance did not
(r = .07, p = .94), and the difference between these effects was significant (X = 12.87,
p < .01). No differences emerged between the effects of attachment anxiety and avoid­
ance for any of the other treatment formats (X = 0.11-1.40, p = .24 -.74).
Mean patient age was a significant moderator of the difference between pretreat­
ment anxiety and avoidance in predicting pre-post change in outcome (b = 0.01,
z = 2.07, p = .04). Attachment avoidance was a stronger predictor of change than anx­
iety in the youngest samples, but the difference in attachment styles was nullified in
samples roughly between 31 and 49, with older samples displaying stronger effects of
attachment anxiety on change than avoidance.
Both treatment setting (b = 0.16, z = 2.01, p = .04) and patient educational level
(b = 0.21, z = 1.96, p < .05) predicted the difference between the effect of change in anx­
iety on change in outcome and the effect of change in avoidance on change in outcome.
Follow-up analyses of the treatment setting effect showed that in inpatient settings,
change in attachment anxiety predicted change in outcome (r = .19, p < .01) while
change in attachment avoidance did not (r = .06, p = .52), and the difference between
these effects was significant (X = 5.20, p = .03). No differences emerged between the
effects of attachment anxiety and avoidance in outpatient settings (X = 0.05, p = .82).
Follow-up analyses of the educational level effect showed that, among individuals who
completed college, change in attachment anxiety predicted change in outcome (r = .19,
 44 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

p < .01) while change in attachment avoidance did not (r = .09, p = .10), and the dif­
ference between these effects bordered statistical significance (x2 = 3.83, p = .05). No
differences emerged between the effects of attachment anxiety and avoidance among
those without college degrees (X = 1.31, p = .25).
Finally, educational status was a moderator o f the effect of pretreatment attach­
ment insecurity on pre-post change in outcome (z = 2.78, p < .01), regardless of
attachment dimension (anxiety or avoidance). Follow-up evaluation suggested that
those without a college degree showed a significant positive effect of attachment se­
curity (i.e., low anxiety and avoidance) on pre-post change (r = .15, z = 2.83, p <
.01), while college graduates experienced no significant effect of attachment insecu­
rity on outcome (r = -.03, z = -0.75, p = .46). This finding is especially interesting,
given the nonsignificant effect above of pretreatment attachment insecurity on treat­
ment change, indicating the importance of educational attainment in detecting such
effects.

EVIDENCE FOR CAUSALITY

The current meta-analysis examined the correlation between attachment at pre­
treatment and pre-to-post change in attachment with outcome at posttreatment and
pre-to-post change in outcome. Given the correlational nature of these effects, our
findings do not provide evidence for a causal effect o f attachment on outcome.
At the same time, several lines of evidence inform the question of causality. First, the
effect size of pretreatment attachment security on posttreatment outcome tended to be
smaller among psychotherapies that included an interpersonal component compared
to psychotherapies that did not. This finding suggests that attachment insecurity is
less associated with worse outcome in psychotherapies that include a component that
might address attachment insecurity directly.
Second, recent trends toward utilizing intensive data collection designs as well
as advances in data-analytic methods allow for more confidence in inferences re­
garding causality. For example, Zalaznik and colleagues (2017) examined session
by session changes in attachment, biased cognitions, avoidance behaviors, and ex­
pressive suppression during CBT for panic disorder. They found that, for the most
part, reductions in biased cognitions and avoidance behaviors predicted subse­
quent reductions in attachment anxiety but not the other way around. In addition,
reductions in avoidance behaviors and expressive suppression predicted subsequent
reductions in attachment avoidance but not the other way around. As well, although
not included in our current meta-analysis because the patient sample was composed
of adolescents, Thum (2016) found that increases in attachment security during a
residential treatment program for adolescents co-occurred with improvements in
outcome but lagged relations were not examined. Advances in research design and
data analysis might in the future provide more direct evidence on this important
question of causality.
 45 Attachment Style

LIMITATIONS OF THE RESEARCH

There are still relatively few empirical studies that have examined how client attach­
ment influences psychotherapy outcome. The relative paucity of studies reduces the
power to detect moderation. In addition, no prospective investigations that we know
of have matched patients to treatments or therapists based on attachment patterns.
Additional findings are needed before conclusions can be rendered.
Furthermore, many studies used attachment measures that do not correlate well
with other measures of attachment and that do not appear to converge with underlying
dimensions of adult attachment (anxiety and avoidance). To produce findings that are
comparable to one another and that can be combined to yield clinical conclusions, it
is important for investigators to use measures of attachment that are well validated.
Another limitation of our meta-analyses is that we excluded dissertations and
publications reported in languages other than English. As a consequence, we may have
missed important studies published in other languages or those that did not find their
way into publication.
When taking into account pretreatment levels of outcome measures, the correlation
between pretreatment attachment and posttreatment outcome became statistically
nonsignificant. This finding suggests that pretreatment attachment and pretreatment
outcome measures share some of the variance that is predictive of treatment outcome.
However, it is unclear whether this shared variance is a reflection of pretreatment
attachment leading to higher levels of pretreatment severity, pretreatment severity
leading to higher levels of pretreatment attachment, both being caused by a third var­
iable, or some combination of these options. Thus it is important for future studies to
disentangle these possibilities by looking at the longitudinal associations among these
variables before treatment.

DIVERSITY CONSIDERATIONS
Unfortunately, the results of the current meta-analysis are difficult to interpret with
regard to diversity. We examined patient gender and race as potential study- level
moderators since separate analyses by these variables were rarely available. The only
significant finding was an interaction between age and attachment dimension (anxiety
vs. avoidance) in predicting the effect to pretreatment attachment on pre-post change
in outcome. In the youngest samples, attachment avoidance was a stronger predictor
of change than anxiety, but the difference became nonsignificant in samples roughly
between 31 and 49, with older samples displaying stronger effects of attachment anx­
iety on change than avoidance.
The percentage of women in the analyzed studies averaged 71%, which reflects
national numbers in patients receiving psychotherapy in the United States (Gibbons
et al., 2011). The vast majority of the analyzed studies were conducted in the United
States and all were published in English, both possibly restricting generalizability of
the results.
 46 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

In terms of therapist diversity, 68% of therapists were women, 28% were students,
and 83% were in supervision, either because they were students or because they were
treating patients as part of controlled studies. These percentages are derived from only
those studies that reported such data and may be subject to sampling bias.
Although developmental research on attachment has been carried out in diverse
and cross-cultural samples, finding considerable evidence for secure base and safe
haven behavior, less work is available at the adult level. Although this work has been
carried out in the United States, Canada, Germany, Denmark, and Australia, among
other nations, it has been carried out in predominantly Caucasian samples, and few
psychotherapy studies regularly report their outcome analyses as a function of age,
gender, ethnicity, race, sexual orientation, or other intersecting dimensions of cultural
identity.

TRAINING IMPLICATIONS
One main training implication from the meta-analysis is that patient attachment style
will influence treatment selection and outcome; accordingly, psychotherapists can be
trained to assess their client’s level of attachment security and particular style. Doing
so will probably guide improved case formulation, treatment selection, and more ef­
fective use of session time.
This chapter has focused on the patient’s attachment style, but clinical trainers are
encouraged to identify the neophyte therapist’s attachment style as well. That will
prove instructive for providing feedback, recommending self-care, managing coun­
tertransference, and the like. Educators can tailor education to the individual trainee’s
attachment patterns just as that student, in parallel fashion, attends to the patient’s
attachment.
There is evidence that therapist attachment patterns interact with that of the patient
to predict outcome (e.g., Dinger et al., 2009; Dozier, 1990), although the evidence is
mixed about how. Some work suggests that complementary styles result in better out­
come, with dismissing clients doing better with preoccupied therapists and vice versa.
Other evidence suggests no such clear relation, although therapist insecure attachment
is a risk factor that may negatively affect the therapeutic alliance and outcome, par­
ticularly in more symptomatic patients. Thus therapists should be aware of their own
personal attachment style and consider how it might be expressed during treatment
and its effect on the patient.
Training can involve didactic education about attachment theory and attach­
ment patterns and how these patterns interdigitate with relationship dynamics in
general and the unique attachment patterns o f the therapist. Following this, clinical
educators and supervisors should assign readings about attachment theory in ge­
neral as well as specific studies included in this review to help trainees gain greater
awareness and appreciation of the role and dynamics related to attachment theory.
Supervisors and trainees can consider using attachment measures to assess patient
attachment style as part of the assessment process. Self-report measures such as the
ECR are easy to use and can be used periodically or even on a session-by-session
 47 Attachment Style

basis in order to obtain a more stable sense o f the patient’s attachment style and to
track fluctuations and changes in attachment style and dimensions. If it is difficult
to utilize self-report measures, there are a number o f prototype measures that ther­
apist can use to rate patients. These include the Adult Attachment Prototype Rating
measure described earlier in this chapter. Interview measures are particularly useful
but require special training, are more time consuming and labor intensive, and are
costly to code. However, through reading, familiarity, and training, a clinician can
learn to listen to the patient narrative discourse through an attachment-sensitive ear
and code the patient’s narrative production for attachment style based on the content
and the structure of such narrative. Supervisors can develop videotapes of psycho­
therapy with patients exhibiting prototypical attachment styles or develop role plays
to use in deliberate practice in class or in supervision sessions for responding to
attachment style.

THERAPEUTIC PRACTICES
We derive several practice implications from our meta-analysis on attachment style
that can guide psychotherapists.

♦ Assess the patients attachment style. Attachment style or organization can influence
the psychotherapy process, the responses of both patients and therapists, the quality
of the therapeutic alliance, and the ultimate outcome of treatment. Thus therapists
should be attuned to indicators of a patient’s attachment style. Formal interviewing
or use of reliable self-report measures can be useful as part of the assessment process.
♦ Understand that a patients attachment organization will provide important clues
as to how the patient is likely to respond in treatment and to the therapist. Expect
longer and more difficult treatment with anxiously attached patients but quicker and
more positive outcome with securely attached patients.
♦ Beware that patients preoccupied in their attachment may be deceptively difficult
to treat, despite initially appearing engaged and cooperative. These patients may
be quick to anger or feel rejected by the therapist and are at a high risk of dropout.
Consider a stance designed to help the preoccupied patient contain his or her
emotional experience (Daly & Mallinckrodt, 2009). This may include explicit
articulations of the treatment frame, the provision of more structure to compensate
for the patient’s tendency to feel muddled, and avoidance of collusion with the
patient who may pull the therapist to engage in more emotional/experiential
techniques that only contribute to the patient feeling overwhelmed.
♦ Understand that clients with dismissing styles may require the therapist to walk a thin
line between being active but not too active, engaged but with enough distance to
provide space. Do not enact a chase and dodge dynamic with the dismissing patient.
♦ Avoid, at the same time, going too far in acting in contrast (complementarity) to
patients’ attachment styles. Practice and research suggest that therapists should titrate
their interpersonal styles so as not to overwhelm dismissing patients or to appear
disengaged, aloof, or uninterested to preoccupied patients.
 48 P S Y C H O T H E R A P Y R EL AT IO N S H IP S THAT WORK

♦ Consider providing patients low in attachment security an interpersonally focused

therapy to maximize outcome. There is some preliminary evidence that individuals
low in attachment security (these may potentially be considered “fearful avoidant”
individuals) may have significant difficulties in treatments that do not include an
interpersonal component, while these patients’ attachment difficulties may prove less
of a barrier to treatment in interpersonal therapies.
♦ Know that attachment style can be modified during psychotherapy, even in brief
treatments and for patients with severe attachment difficulties, such as those
suffering from borderline personality disorder. Therefore, change in attachment can
be conceptualized as a proximal outcome, not just a predictive patient characteristic,
and could be considered a goal of treatment.
♦ Consider intervening with patients to change attachment style. Early findings suggest
that the focus on the relationship between the therapist and patient and/or the use
of interpretations may be efficacious change mechanisms, at least for personality
disordered patients (Johansson et al., 2010; Levy et al., 2006). However, other
research also demonstrates that a range of treatments may be useful for achieving
changes in attachment representations in less disturbed patients with neurotic-level
difficulties (Levy et al., 2015).

REFERENCES
References m arked with an asterisk indicate studies included in the m eta-analysis.
Ainsworth, M. D. S. (Ed.). (1978). Patterns o f attachment: A psychological study o f the strange
situation. Hillsdale, NJ: Erlbaum.
Bartholomew, K. (1990). Avoidance o f intimacy: An attachm ent perspective. Journal
o f Social and Personal Relationships, 7(2), 1 4 7 -1 7 8 . https://www.doi.org/10.1177/
0265407590072001
Bartholomew, K., & Horowitz, L. M . (1991). Attachm ent styles am ong young adults: A test o f a
four-category model. Journal o f Personality and Social Psychology, 61(2), 2 2 6 -2 4 4 . https://
w w w .doi.org/10.1037/0022-3514.61.2.226
Beebe, B., & Lachm ann, F. M . (1988). The contribution o f m other-infant m utual influence to
the origins o f self- and object representations. Psychoanalytic Psychology, 5(4), 3 0 5 -3 3 7 .
https://www.doi.org/10.1037/0736-9735.5.4.305
*Belanger, C., Roger Marcaurelle, M archand, A., El-Baalbaki, G., Guay, S., & Pecknold, J.
(2011). Effect o f a m arital com m unication training on treatm ent outcom e in panic dis­
order with agoraphobia. In A. Columbus (Ed.), Advances in psychology research (Vol. 77,
pp. 1 4 1 -1 6 3 ). Hauppauge, NY: Nova Science.
Belsky, J., & Rovine, M. (1987). Tem peram ent and attachm ent security in the strange situa­
tion: An em pirical rapprochem ent. Child Development, 58(3), 787. https://www.doi.org/
10.2307/1130215
*B enson, L. A., Sevier, M ., & Christensen, A. (2013). The im pact o f behavioral couple therapy
on attachm ent in distressed couples. Journal o f Marital and Family Therapy, 39(4), 40 7 ­
420. https://www. doi.org/10.1111 /jmft. 12020
Berant, E., & Obegi, J. (2009). A ttachm ent-inform ed psychotherapy research with adults. In
J. H. O begi & E. Berant (Eds.), Attachment theory and research in clinical work with adults
(pp. 4 6 1 -4 8 9 ). New York, NY: Guilford.
 49 Attachment Style

*Bernecker, S. L., Constantino, M . J., A tkinson, L. R., Bagby, R. M ., Ravitz, P., & M cBride, C.
(2016). Attachm ent style as a m oderating influence on the efficacy o f cognitive-behavioral
and interpersonal psychotherapy for depression: A failure to replicate. Psychotherapy,
5 3(1), 2 2 -3 3 .
Berry, K., Barrowclough, C., & Wearden, A. (2008). Attachm ent theory: A fram ework for u n ­
derstanding symptoms and interpersonal relationships in psychosis. Behaviour Research
and Therapy, 4 6(12), 1 2 75-1282.
*Berry, K., Gregg, L., Hartwell, R., Haddock, G., Fitzsim m ons, M ., & Barrowclough, C. (2015).
T herapist-client relationships in a psychological therapy trial for psychosis and substance
misuse. Drug & Alcohol Dependence, 152, 1 7 0-176.
Borenstein, M . (2009). Effect sizes for continuous data. In H. Cooper, L. Hedges, & J. Valentine
(Eds.), The handbook o f research synthesis and meta-analysis (2nd ed., pp. 2 2 1 -2 3 5 ).
New York, NY: SAGE.
Bowlby, J. (1975). Attachm ent theory, separation, anxiety and m ourning. In American
Handbook o f Psychiatry (2nd ed.). vol 6 , pp. 2 9 0 -3 0 8 . New York: Basic Books.
Bowlby, J. (1977). The M aking and Breaking o f Affectional Bonds: II. Som e Principles o f
Psychotherapy: The Fiftieth M audsley Lecture (expanded version). The British Journal o f
Psychiatry, 130(5), 4 2 1 -4 3 1 .
Bowlby, J. (1982). Attachm ent and loss: Retrospect and prospect. American Journal o f
Orthopsychiatry, 52(4 ), 6 6 4 -6 7 8 .
Bowlby, J. (1988). Developmental psychiatry com es o f age. The American Journal o f Psychiatry,
145(1), 1 -1 0 . Retrieved from http://dx.doi.org/10.1176/ajp. 145.1.1
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report m easurem ent o f adult attach­
ment: A n integrative overview. In J. A. Sim pson & W. S. Rholes (Eds.), Attachment theory
and close relationships (pp. 4 6 -7 6 ). New York, NY: Guilford.
Buchheim , A., Horz-Sagstetter, S., D oering, S., Rentrop, M ., Schuster, P., Buchheim , P,
. . . Fischer-Kern, M . (2017). Change o f unresolved attachm ent in borderline person­
ality disorder: R C T study o f transference-focused psychotherapy. Psychotherapy and
Psychosomatics, 86(5), 3 1 4 -3 1 6 .
*Byrd, K. R., Patterson, C. L., & Turchik, J. A. (2010). W orking alliance as a m ediator o f client
attachm ent dimensions and psychotherapy outcome. Psychotherapy: Theory, Research,
Practice, Training, 47(4), 6 3 1 -6 3 6 . https://www.doi.org/10.1037/a0022080
Cassidy, J., Lichtenstein-Phelps, J., Sibrava, N. J., Thomas, C. L., & Borkovec, T. D. (2009).
Generalized anxiety disorder: C onnections with self-reported attachm ent. Behavior
Therapy, 40, 2 3 -3 8 . https://www.doi.org/10.1016/j.beth.2007.12.004
Cohen, J. (1988). Statistical Power Analysis fo r the Behavioral Sciences (2nd ed.). Hillsdale,
NJ: Lawrence Erlbaum Associates, Publishers.
Collins, N. L. (1996). W orking models o f attachm ent: Im plications for explanation, em otion,
and behavior. Journal o f Personality and Social Psychology, 71(4), 8 1 0 -8 3 2 .
Collins, N. L., & Read, S. J. (1990). Adult attachm ent, working models, and relationship quality
in dating couples. Journal o f Personality and Social Psychology, 58(4 ), 6 4 4 -6 6 3 . https://
w w w .doi.org/10.1037/0022-3514.58.4.644
Daly, K. D., & M allinckrodt, B. (2009). Experienced therapists’ approach to psychotherapy for
adults with attachm ent avoidance or attachm ent anxiety. Journal o f Counseling Psychology,
5 6(4), 5 4 9 -5 6 3 . https://www.doi.org/10.1037/a0016695
*D aniel, S. I. F., Poulsen, S., & Lunn, S. (2016). Client attachm ent in a randomized clinical trial
o f psychoanalytic and cognitive-behavioral psychotherapy for bulim ia nervosa: O utcom e
 50 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

m oderation and change. Psychotherapy, 53(2), 1 7 4 -1 8 4 . https://www.doi.org/10.1037/

pst0000046
Diam ond, D., Clarkin, J. F., Stovall-M cClough, K., Levy, K. N., Foelsch, P. A., Levine, H., &
Yeomans, F. E. (2003). Patient-therapist attachm ent: Im pact on the therapeutic process and
outcome. In M . C ortina & M . M arrone (Eds.), Attachment theory and the psychoanalytic
process (pp. 1 2 7 -1 7 8 ). Philadelphia, PA: Whurr.
Diam ond, D., Clarkin, J., Levine, H., Levy, K., Foelsch, P, & Yeomans, F. (1999). Borderline
conditions and attachm ent: A prelim inary report. Psychoanalytic Inquiry, 19(5), 8 3 1 -8 8 4 .
https://www.doi.org/10.1080/07351699909534278
*D iam ond, G. M ., Shahar, B., Sabo, D., & Tsvieli, N. (2016). Attachm ent-based family therapy
and em otion-focused therapy for unresolved anger: The role o f productive em otional pro­
cessing. Psychotherapy, 53(1), 3 4 -4 4 .
Diener, M. J., Hilsenroth, M. J., & Weinberger, J. (2009). A prim er on m eta-analysis o f cor­
relation coefficients: The relationship between patient-reported therapeutic alliance and
adult attachm ent style as an illustration. Psychotherapy Research, 1 9 (4 -5 ), 5 1 9 -5 2 6 . https://
w w w .doi.org/10.1080/10503300802491410
Dinger, U., Strack, M ., Sachsse, T., & Schauenburg, H. (2009). Therapists’ attachm ent, patients’
interpersonal problem s and alliance development over tim e in inpatient psychotherapy.
Psychotherapy: Theory, Research, Practice, Training, 4 6(3), 2 7 7 -2 9 0 . Retrieved from http://
dx.doi.org /10.1037/a0016913
Dozier, M. (1990). Attachm ent organization and treatm ent use for adults with serious psycho­
pathological disorders. Development and Psychopathology, 2 (01), 4 7 -6 0 . https://www.doi.
org/10.1017/S0954579400000584
Dozier, M ., Lomax, L., Tyrrell, C. L., & Lee, S. W. (2001). The challenge o f treatm ent for clients
with dismissing states o f mind. Attachment & Human Development, 3 (1), 6 2 -7 6 . https://
w w w .doi.org/10.1080/14616730010000858
Falkenstrom , F., Finkel, S., Sandell, R., Rubel, J. A., & Holmqvist, R. (2017). Dynam ic models
o f individual change in psychotherapy process research. Journal o f Consulting and Clinical
Psychology, 8 5(6), 5 3 7 -5 4 9 . https://www.doi.org/10.1037/ccp0000203
Farber, B. A., Lippert, R. A., & Nevas, D. B. (1995). The therapist as attachm ent figure.
Psychotherapy: Theory, Research, Practice, Training, 32(2), 2 0 4 -2 1 2 . https://www.doi.org/
10.1037/ 0033-3204.32.2.204
Feeney, J. A., Noller, P, & Hanrahan, M. (1994). Assessing adult attachm ent. In M. B. Sperling
& W. H. Berm an (Eds.), Attachment in adults: Clinical and developmental perspectives (pp.
1 2 8 -1 5 2 ). New York, NY: Guilford.
Fischer-Kern, M ., D oering, S., Taubner, S., Horz, S., Zim m erm ann, J., Rentrop, M ., . . .
Buchheim , A. (2015). Transference-focused psychotherapy for borderline personality dis­
order: Change in reflective function. The British Journal o f Psychiatry, 20 7 (2 ), 1 73-174.
https://www.doi.org/10.1192/bjp.bp.113.143842
Foangy, P , Steele, M ., Steele, H., Leigh, T., Kennedy, R., M attoon, G ., & Target, M. (1995).
Attachm ent, the reflective self, and borderline states: The predictive specificity o f the Adult
Attachm ent Interview and pathological em otional development. In S. Goldberg, R. Muir,
& J. K err (Eds.), Attachment Theory: Social, Developmental, and Clinical Perspectives (pp.
2 3 3 -2 7 8 ). New York: Analytic Press.
Fonagy, P, Leigh, T., Steele, M ., Steele, H., Kennedy, R., M attoon, G ., . . . Gerber, A. (1996). The
relation o f attachm ent status, psychiatric classification, and response to psychotherapy.
 51 Attachment Style

Journal o f Consulting and Clinical Psychology, 64(1), 2 2 -3 1 . https://www.doi.org/10.1037/

0 0 2 2 -0 0 6 X .6 4 .1.22
*Forbes, D., Parslow, R., Fletcher, S., M cHugh, T., & Creamer, M . (2010). Attachm ent style
in the prediction o f recovery following group treatm ent o f com bat veterans with post­
traum atic stress disorder. The Journal o f Nervous and Mental Disease, 198(12), 8 8 1 -8 8 4 .
https://www.doi.org/10.1097/NMD.0b013e3181fe73fa
Fraley, R. C. (2002). Attachm ent stability from infancy to adulthood: M eta-analysis and dy­
nam ic m odeling o f developmental m echanism s. Personality and Social Psychology Review,
6 (2 ), 1 2 3 -1 5 1 . https://www.doi.org/10.1207/S15327957PSPR0602_03
Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis o f self­
report m easures o f adult attachm ent. Journal o f Personality and Social Psychology, 78(2),
3 5 0 -3 6 5 . https://www.doi.org/10.1037/0022-3514.78.2.350
George, C., Kaplan, N., & M ain, M. (1985). Attachment interview fo r adults. Unpublished
m anuscript, University o f California, Berkeley.
G ibbons, M . B. C., Rothbard, A., Farris, K. D., Stirm an, S. W., Thom pson, S. M ., Scott, K., . . .
Crits-Christoph, P. (2011). Changes in psychotherapy utilization am ong consum ers o f serv­
ices for m ajor depressive disorder in the com m unity m ental health system. Administration
and Policy in Mental Health and Mental Health Services Research, 38(6), 4 9 5 -5 0 3 . https://
w w w .doi.org/10.1007/s10488-011-0336-1
*G ois, C., Dias, V. V., Carm o, I., Duarte, R., Ferro, A., Santos, A. L., . . . Barbosa, A. (2014).
Treatm ent response in type 2 diabetes patients with m ajor depression: Treatm ent response
in type 2 diabetes patients. Clinical Psychology & Psychotherapy, 2 1 (1 ), 3 9 -4 8 . https://www.
doi. org/1 0 . 1 0 0 2 /cpp. 1817
G rossm ann, K. E., Grossm ann, K., & Waters, E. (2005). Attachment from infancy to adult­
hood: The major longitudinal studies. New York, NY: Guilford.
Haggbloom, S. J., W arnick, R., W arnick, J. E., Link to external site, this link will open in a
new window, Jones, V. K., Yarbrough, G. L., . . . M onte, E. (2002). The 100 m ost em i­
nent psychologists o f the 20th century. Review o f General Psychology, 6(2), 1 3 9 -1 5 2 . http://
dx.doi. org/10.103 7/ 1089-2680.6.2.139
Hazan, C., & Shaver, P. (1987). Rom antic love conceptualized as an attachm ent process.
Journal o f Personality and Social Psychology, 52(3), 5 1 1 -5 2 4 . https://www.doi.org/10.1037/
0 0 2 2 -3 5 1 4 .5 2 .3 .511
Hazan, C., & Shaver, P. R. (1990). Love and work: An attachm ent-theoretical perspective.
Journal o f Personality and Social Psychology, 59(2), 2 7 0 -2 8 0 . https://www.doi.org/10.1037/
0 0 2 2 -3 5 1 4 .5 9 .2 .2 7 0
*Hoyer, J., W iltink, J., Hiller, W., Miller, R., Salzer, S., Sarnowsky, S., . . . Leibing, E. (2016).
Baseline patient characteristics predicting outcom e and attrition in cognitive therapy for
social phobia: Results from a large m ulticentre trial: predictors o f outcom e in social phobia.
Clinical Psychology & Psychotherapy, 2 3(1), 3 5 -4 6 . https://www.doi.org/10.1002/cpp.1936
*Illing, V , Tasca, G. A., Balfour, L., & Bissada, H. (2010). Attachm ent insecurity predicts
eating disorder symptoms and treatm ent outcom es in a clinical sample o f women. The
Journal o f Nervous and Mental Disease, 198(9), 6 5 3 -6 5 9 . https://www.doi.org/10.1097/
N M D .0b013e3181ef34b2
Johansson, P., Hoglend, P , U lberg, R., Amlo, S., M arble, A., Bogwald, K. P , . . . Heyerdahl,
O. ( 2 0 1 0 ). The m ediating role o f insight for long-term improvements in psychodynamic
therapy. Journal o f Consulting and Clinical Psychology, 78(3), 438.
 52 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HA T W O R K

*Joyce, A. S., Ennis, L. P., O’Kelly, J. G., Ogrodniczuk, J. S., & Piper, W. E. (2009). Depressive
m anifestations and differential patterns o f treatm ent outcom e in an intensive partial hos­
pitalization treatm ent program. Psychological Services, 6(2), 1 5 4 -1 7 2 . https://www.doi.org/
10.1037/a0014126
*Joyce, A. S., Fujiwara, E., Cristall, M ., Ruddy, C., & Ogrodniczuk, J. S. (2013). Clinical
correlates o f alexithymia am ong patients with personality disorder. Psychotherapy Research,
2 3 (6 ), 6 9 0 -7 0 4 . https://www.doi.org/10.1080/10503307.2013.803628
Kagan, J. (1998). Three seductive ideas. Cam bridge, M A: Harvard University Press.
Kline, R. B. (2016). Principles and practice o f structural equation modeling.
New York: Guilford Press.
Kobak, R. R., Cole, H. E., Ferenz-Gillies, R., Fleming, W. S., & Gamble, W. (1993). Attachm ent
and em otion regulation during m other-teen problem solving: A control theory analysis.
Child Development, 64(1), 2 3 1 -2 4 5 . https://www.doi.org/10.2307/1131448
Konrath, S. H., Chopik, W. J., Hsing, C., K., & O ’Brien, E. (2014). Changes in adult attachm ent
stules in A m erican college students over tim e: A m eta-analysis. Personality and Social
Psychology Review, 18, 3 2 6 -3 4 8 . https://www.doi.org/10.1177/1088868314530516
*Kowal, J., M cW illiam s, L. A., Peloquin, K., W ilson, K. G., Henderson, P. R., & Fergusson, D.
A. (2015). Attachm ent insecurity predicts responses to an interdisciplinary chronic pain
rehabilitation program . Journal o f Behavioral Medicine, 38(3), 5 1 8 -5 2 6 . https://www.doi.
org/10.1007/s10 8 6 5 -0 1 5 -9 6 2 3 -8
*Lawson, D. M ., & Brossart, D. F. (2009). Attachm ent, interpersonal problem s, and treatm ent
outcom e in group therapy for intim ate partner violence. Psychology o f Men & Masculinity,
10(4), 2 8 8 -3 0 1 . https://www.doi.org/10.1037/a0017043
Levy, K. N. (2005). The implications o f attachm ent theory and research for understanding
borderline personality disorder. Development and Psychopathology, 17(4), 9 5 9 -9 8 6 .
Levy, K. N. (2013). Introduction: Attachm ent theory and psychotherapy. Journal o f Clinical
Psychology, 69(11), 113 3 -1 1 3 5 . https://doi.org/10.1002/jclp.22040
Levy, K. N., Ellison, W. D., Scott, L. N., & Bernecker, S. L. (2011). Attachm ent style. In J. C.
Norcross (Ed.), Psychotherapy relationships that work: Evidence-based responsiveness (pp.
3 7 7 -4 0 1 ). New York, NY: Oxford University Press.
Levy, K. N., & Johnson, B. N. (2018). Attachm ent and psychotherapy: Im plications from em ­
pirical research. Canadian Psychology/Psychologie Canadienne, No Pagination Specified-
No Pagination Specified. https://doi.org/10.1037/cap0000162
Levy, K. N., Johnson, B. N., Clouthier, T. L., Scala, J. W., & Temes, C. M . (2015). An attach­
m ent theoretical fram ework for personality disorders. Canadian Psychology/Psychologie
Canadienne, 56(2), 1 9 7 -2 0 7 . https://www.doi.org/10.1037/cap0000025
*Levy, K. N., M eehan, K. B., Kelly, K. M ., Reynoso, J. S., Weber, M ., Clarkin, J. F., & Kernberg,
O. F. (2006). Change in attachm ent patterns and reflective function in a randomized
control trial o f transference-focused psychotherapy for borderline personality disorder.
Journal o f Consulting and Clinical Psychology, 74(6), 10 2 7 -1 0 4 0 . https://www.doi.org/
10.1037/ 0022-006X .74.6.1027
Lichtenstein, J., & Cassidy, J. (1991, June). The Inventory o f Adult Attachment: Validation o f a
new measure. Paper presented at the biennial m eeting o f the Society for Research in Child
Development, Seattle, WA.
Light, R. J., & Pillemer, D. B. (1984). Summing up: The science o f reviewing research. Cambridge,
M A: Harvard University Press.
 53 Attachment Style

*Lindgren, A., Barber, J. P., & Sandahl, C. (2008). A lliance to the g ro u p -as-a-w h o le as a
predictor o f outcom e in psychodynam ic group therapy. International Journal o f Group
Psychotherapy, 58(2), 163-184.
M ain, M ., & Goldwyn, R. (1998). Adult attachment classification system. Unpublished m anu­
script, University o f California, Berkeley.
M ain, M ., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A
m ove to the level o f representation. Monographs o f the Society fo r Research in Child
Development, 5 0 (1 -2 ), 6 6 . https://www.doi.org/10.2307/3333827

M ain, M ., & Solom on, J. (1986). Discovery o f an insecure-disorganized/disoriented attach­

m ent pattern. In T. B erry Brazelton & M. W. Yogman (Eds.), Affective development in in­
fancy (pp. 9 5 -1 2 4 ). W estport, CT: Ablex.
*M arm arosh, C. L., Gelso, C. J., M arkin, R. D., M ajors, R., Mallery, C., & Choi, J. (2009).
The real relationship in psychotherapy: Relationships to adult attachm ents, working alli­
ance, transference, and therapy outcom e. Journal o f Counseling Psychology, 56(3), 3 3 7 -3 5 0 .
https: //ww w. doi.org/10.1037/a0015169
*M cBride, C., A tkinson, L., Quilty, L. C., & Bagby, R. M . (2006). Attachm ent as m oderator
o f treatm ent outcom e in m ajor depression: A randomized control trial o f interpersonal
psychotherapy versus cognitive behavior therapy. Journal o f Consulting and Clinical
Psychology, 7 4(6), 104 1 -1 0 5 4 . https://www.doi.org/10.1037/0022-006X.74.6.1041
*M eyer, B., Pilkonis, P. A., Proietti, J. M ., Heape, C. L., & Egan, M. (2001). Attachm ent styles
and personality disorders as predictors o f symptom course. Journal o f Personality Disorders,
15(5), 3 7 1 -3 8 9 . https://www.doi.org/10.1521/pedi.15.5.371.19200
M ikulincer, M ., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamics, and
change. New York, NY: Guilford.
*M uller, R. T., & Rosenkranz, S. E. (2009). Attachm ent and treatm ent response am ong adults
in inpatient treatm ent for posttraum atic stress disorder. Psychotherapy: Theory, Research,
Practice, Training, 46(1), 8 2 -9 6 . https://www.doi.org/10.1037/a0015137
*N ew m an, M. G., Castonguay, L. G ., Jacobson, N. C., & Moore, G. A. (2015). Adult attachm ent
as a m oderator o f treatm ent outcom e for generalized anxiety disorder: Com parison b e­
tween cognitive-behavioral therapy (C B T ) plus supportive listening and C B T plus inter­
personal and em otional processing therapy. Journal o f Consulting and Clinical Psychology,
8 3 (5 ), 9 1 5 -9 2 5 . https://www.doi.org/10.1037/a0039359
O rw in, R. G. (1983). A fail-safe N for effect size in m eta-analysis. Journal o f Educational
Statistics, 8 (2 ), 1 5 7-159.
*Parker, M . L., Johnson, L. N., & Ketring, S. A. (2012). Adult attachm ent and symptom dis­
tress: a dyadic analysis o f couples in therapy: Dyadic analysis o f couple therapy. Journal o f
Family Therapy, 34(3), 3 2 1 -3 4 4 . https://www.doi.org/10.1111/ j.1467-6427.2012.00598.x
Pilkonis, P. A. (1988). Personality prototypes am ong depressives: Them es o f dependency and
autonomy. Journal o f Personality Disorders, 2 (2 ), 1 4 4-152.
Ravitz, P., Maunder, R., Hunter, J., Sthankiya, B., & Lancee, W. (2010). Adult attachm ent m eas­
ures: A 25-year review. Journal o f Psychosomatic Research, 69(4), 4 1 9 -4 3 2 . https://www.
doi.org/10.1016/j.jpsychores.2009.08.006
*Reiner, I., Bakerm ans-Kranenburg, M. J., Van IJzendoorn, M . H., Frem m er-Bom bik, E., &
Beutel, M . (2016). Adult attachm ent representation m oderates psychotherapy treatm ent
efficacy in clinically depressed inpatients. Journal o f Affective Disorders, 195, 1 6 3-171.
https://www.doi.org/10.1016/j.jad.2016.02.024
 54 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

*Reis, S., & Grenyer, B. F. S. (2004). Fearful attachm ent, working alliance and treatm ent re­
sponse for individuals with m ajor depression. Clinical Psychology & Psychotherapy, 11(6),
4 1 4 -4 2 4 . https://www.doi.org/10.1002/cpp.428
Riggs, S. A., Jacobovitz, D., & Hazen, N. (2002). Adult attachm ent and history o f psycho­
therapy in a norm ative sample. Psychotherapy: Theory, Research, Practice, Training, 39(4),
3 4 4 -3 5 3 . https://www.doi.org/10.1037/0033-3204.39.4.344
*Sauer, E. M ., Anderson, M. Z., Gormley, B., Richm ond, C. J., & Preacco, L. (2010). Client at­
tachm ent orientations, working alliances, and responses to therapy: A psychology training
clinic study. Psychotherapy Research, 2 0(6), 7 0 2 -7 1 1 . https://www.doi.org/10.1080/
10503307.2010.518635
Shaver, P. R., & Clark, C. L. (1994). The psychodynam ics o f adult rom antic attachm ent. In J.
M. M asling & R. F. Bornstein (Eds.), Empirical perspectives on object relations theory (pp.
1 0 5 -1 5 6 ). W ashington, DC: A m erican Psychological Association. https://www.doi.org/
10.1037/ 11100-004
Shaver, P R., & Hazan, C. (1993). Adult rom antic attachm ent: Theory and evidence. In D.
Perlman & W Jones (Eds.), Advances in personal relationships (Vol. 4, pp. 2 9 -7 0 ). London,
England: Jessica Kingsley.
Sim pson, J. A. (1990). Influence o f attachm ent styles on rom antic relationships. Journal
o f Personality and Social Psychology, 59(5), 9 7 1 -9 8 0 . https://www.doi.org/10.1037/
0 0 2 2 -3 5 1 4 .5 9 .5 .9 71
Slade, A. (1999). Attachm ent theory and research: Im plications for the theory and practice of
individual psychotherapy with adults. In J. Cassidy & P. R. Shaver (Eds.), Handbook o f at­
tachment: Theory, research, and clinical applications (pp. 5 7 5 -5 9 4 ). New York, NY: Guilford.
Slade, A. (2004). Two therapies: Attachm ent organization and the clinical process. In L.
A tkinson & S. Goldberg (Eds.), Attachment issues in psychopathology and intervention (pp.
1 8 1 -2 0 6 ). Mahwah, NJ: Erlbaum.
*Sm ith, P N., Gamble, S. A., C ort, N. A., Ward, E. A., He, H., & Talbot, N. L. (2012). Attachm ent
and alliance in the treatm ent o f depressed, sexually abused women: Attachm ent orienta­
tion and working alliance. Depression and Anxiety, 2 9 (2 ), 1 2 3 -1 3 0 . https://www.doi.org/
10.1002/da.20913
Sroufe, L. A., & Waters, E. (1977). Heart rate as a convergent measure in clinical and devel­
opmental research. Merrill-Palmer Quarterly o f Behavior and Development, 2 3 (1 ), 3 -2 7 .
*Stalker, C. A., Gebotys, R., & Harper, K. (2005). Insecure attachm ent as a predictor o f outcom e
following inpatient traum a treatm ent for women survivors o f childhood abuse. Bulletin o f
the Menninger Clinic, 69(2), 1 3 7 -1 5 6 . https://www.doi.org/10.1521/bumc.69.2.137.66508
*Strauss, B., K irchm ann, H., Eckert, J., Lobo-D rost, A., M arquet, A., Papenhausen, R., . . .
Höger, D. (2006). Attachm ent characteristics and treatm ent outcom e following inpa­
tient psychotherapy: Results o f a m ultisite study. Psychotherapy Research, 16(5), 5 7 9 -5 9 4 .
https://www.doi.org/10.1080/10503300600608322
Strauss, B. M ., Lobo-D rost, A. J., & Pilkonis, P A. (1999). Einschätzung von Bindungsstilen bei
Erw achsenen-erste Erfahrungen m it der deutschen Version einer Prototypenbeurteilung
[Evaluation o f attachm ent styles in adults: First results with the G erm an version o f a proto­
type evaluation]. Zeitschrift Fur Klinische Psychologie Psychopathologie Und Psychotherapie,
47, 3 4 7 -3 6 4 .
*Strauss, B. M ., Mestel, R., & Kirchm ann, H. A. (2011). Changes o f attachment status among
women with personality disorders undergoing inpatient treatment. Counselling and
Psychotherapy Research, 11(4), 2 7 5 -2 8 3 . https://www.doi.org/10.1080/14733145.2010.548563
 55 Attachment Style

Talia, A., & M iller-Bottom e, M. (2012). The Patient Attachm ent Coding System: Scoring
manual. Unpublished manuscript. Department o f Psychology, University o f Copenhagen.
Talia, A., M iller-Bottom e, M ., & Daniel, S. I. (2017). Assessing attachm ent in psycho­
therapy: validation o f the patient attachm ent coding system (PACS). Clinical Psychology &
Psychotherapy, 2 4 (1 ), 149-161.
*Tasca, G. A., Ritchie, K., Conrad, G., Balfour, L., Gayton, J., Lybanon, V , & Bissada, H. (2006).
Attachm ent scales predict outcom e in a randomized controlled trial o f two group therapies
for binge eating disorder: An aptitude by treatm ent interaction. Psychotherapy Research ,
16(1), 1 0 6 -1 2 1 . https://www.doi.org/10.1080/10503300500090928
*Tasca, G. A., Ritchie, K., Dem idenko, N., Balfour, L., Krysanski, V , Weekes, K., . . . Bissada, H.
(2013). M atching women with binge eating disorder to group treatm ent based on attach­
m ent anxiety: Outcom es and m oderating effects. Psychotherapy Research, 2 3 (3 ), 3 0 1 -3 1 4 .
https://www.doi.org/10.1080/10503307.2012.717309
*Taylor, P. J., Rietzschel, J., Danquah, A., & Berry, K. (2015). The role o f attachm ent style,
attachm ent to therapist, and working alliance in response to psychological therapy.
Psychology and Psychotherapy: Theory, Research and Practice, 88(3), 2 4 0 -2 5 3 . https://www.
doi. org/10.1111 /papt. 12045
Thum, L. S. (2016). Assessing change in attachment security in adolescents at residential ther­
apeutic programs (Unpublished doctoral dissertation). W estern M ichigan University,
Kalamazoo.
*Travis, L. A., Bliwise, N. G., Binder, J. L., & Horne-M oyer, H. L. (2001). Changes in
clients’ attachm ent styles over the course o f tim e-lim ited dynamic psychotherapy.
Psychotherapy: Theory, Research, Practice, Training, 38(2), 1 4 9 -1 5 9 . https://www.doi.org/
10.1037/ 0033-3204.38.2.149
van IJzendoorn, M. H. (1995). Adult attachm ent representations, parental responsiveness,
and infant attachm ent: A meta-analysis on the predictive validity o f the Adult Attachm ent
Interview. Psychological Bulletin, 117(3), 3 8 7 -4 0 3 . https://www.doi.org/10.1037/
0 0 3 3 -2 9 0 9 .1 1 7 .3 .3 8 7
Viechtbauer, W., & R Core Team. (2017). metafor: Meta-analysis package fo r R [R package ver­
sion 2.0.0]. Retrieved from http://CRAN.R-project.org/package=metafor
W aters, E., H am ilton, C. E., & Weinfield, N. S. (2000). The stability o f attachm ent security from
infancy to adolescence and early adulthood: G eneral introduction. Child Development,
7 1(3), 6 7 8 -6 8 3 . https://doi.org/10.1111/1467-8624.00175
*W atson, J. C., Steckley, P. L., & M cM ullen, E. J. (2014). The role o f em pathy in prom oting
change. Psychotherapy Research, 2 4(3), 2 8 6 -2 9 8 . https://www.doi.org/10.1080/
10503307.2013.802823
Wei, M ., Russell, D. W., M allinckrodt, B., & Vogel, D. L. (2007). The Experiences in Close
Relationship Scale (E C R )-S h o rt Form: Reliability, validity, and factor structure. Journal o f
Personality Assessment, 88(2), 1 8 7 -2 0 4 . https://www.doi.org/10.1080/00223890701268041
W est, M ., & Sheldon, A. E. R. (1988). Classification o f pathological attachm ent patterns
in adults. Journal o f Personality Disorders, 2 (2 ), 1 5 3 -1 5 9 . https://doi.org/10.1521/
ped i.1988.2.2.153
W est, M . L., & Sheldon-Keller, A. E. (1994). Patterns o f relating: An adult attachment perspec­
tive. New York, NY: Guilford.
*Zalaznik, D., Weiss, M ., & Huppert, J. D. (2017). Im provement in adult anxious and avoidant
attachm ent during cognitive behavioral therapy for panic disorder. Psychotherapy Research.
[Advance online publication] https://www.doi.org/10.1080/10503307.2017.1365183
 3

COPING ST Y LE

Larry E. Beutler, Christopher J. Edwards, Satoko Kimpara,

and Kimberley Miller

Coping styles in psychotherapy process and outcome have been described within
many theoretical orientations. Freud (1938, 1954) offered a developmental perspective
in which dysfunctional states and personality traits were seen as the result of regres­
sion to earlier behaviors and fixation on behaviors which had, in those earlier times,
offered some escape from pain. A common pathway of conflict and defense led to an
unending list of defensive styles.
Over time, psychotherapists have tried to define and conceptualize coping styles and
their relation to the treatment of those with psychopathology. One of most influential
was David Shapiro (1965), who introduced the concept of neurotic styles. Behaviors
associated with the neurotic styles could be expected to be rigidly embraced even in
the absence of acute stress. Neurotic styles were accumulations of cross-situational,
dysfunctional coping styles, the identification of which was thought to have treat­
ment implications. For the more part, however, the links among various coping styles,
therapy procedures, and treatment outcome have not been established.
Over the years, there has been a concerted effort to reduce the multitude of coping
styles to a small set of essential constructs using various statistical methods. From
Eysenck (1947) to Costa and McRae (1985), researchers grouped the multiple styles
into two clusters, extraversion and introversion. Over time, these terms have been re­
vealed as too simple to capture the nature and impact of coping styles. Accordingly,
the complexities of coping styles have become more frequently recognized, leading to
a shift in their terminology that captures broader meanings (McRae et al., 2011).
Internalization and externalization, as descriptors of people struggling to cope, is
reminiscent of concepts offered by Freud (1938/1954), but the newer renditions occupy
a more empirically based place in understanding personality. These trait-like styles
are distributed relatively normally in the population, but when combined with mod­
erate to high levels of impairment, they produce recognizable pathological patterns.
Externalizers are recognizable clinically because they avoid and act out when stressed
or when they face change and they tend to blame their unhappiness and failure on the
environment or others. In contrast, internalizers tend to face change and threat by the
adoption of an inner-blaming “neurotic” style of coping (Costa & McRae, 1985).

56
 57 Coping Style

Internalization and externalization are sufficiently broad categories to embody the

narrow descriptions of introverts and extroverts, especially when the clinician wants
to use terms that minimize the suggestion of dysfunction. The typical descriptions
of these two coping styles include descriptors like inhibited versus expressive, low
versus high arousability, and socially reclusive versus gregarious. Individuals with
internalizing coping styles have a high level of internal arousability and are typically
avoidant when facing conflict or change. They are characterized by the traits of re­
straint, inhibition, and social withdrawal.
Internalization is often called “neuroticism” to capture its very broad-ranging
symptoms. Internalizers are more complex individuals and react strongly and incon­
sistently when stressed, unlike individuals with externalizing coping styles who are
disposed to a more rigid and repetitive manifestation of external arousal. Externalizers
tend to be outgoing and gregarious and to deal with the world by direct avoidance or
excessive confrontation (Beutler, Moos, & Lane, 2003; Costa & McCrae, 1985).
Our understanding of how coping styles evolve and manifest has been advanced
by the writings of Jerome Kagan (1998), a developmental psychologist who observed
the behavior of infants and children over long periods of time. Kagan determined that
even shortly after birth, two prominent patterns of temperament distinguished them­
selves. One group of infants was highly anxious and socially avoidant; they ultimately
tended to develop anxious, obsessive, and repetitious coping styles as ways of control­
ling themselves in environments that they found to be overwhelming. They became
intolerant of intrusions into their space and time and withdrew from intimate contact.
The second group of infants was notable for their lack of reactivity to internal states
and, paradoxically, by their strong needs for external stimulation that they addressed
in a demanding and often aggressive way. Though actively seeking stimulation, these
children, as adolescents, began taking direct action to cope, change, avoid, and escape
threatening environments in which they found themselves. As adults, they tended to
lack empathy and frequently expressed anger and rage directly with others and with
their situation.
In modern nomenclature, these two types of individuals are called internalizers and
externalizers, respectively. Internalizers are worriers. They are persistently anxious and
afraid and often become quite introverted and socially isolated. They are self-critical
and depressive. In contrast, externalizers are recognizable clinically by their impact on
their social environment. They confront others, impose force on things that obstruct
their progress, and otherwise act out.
In what follows, we focus on coping style (CS) as the distinction between
externalizing and internalizing attributes This chapter begins by providing definitions
of the relevant constructs, reviewing common measures of these constructs, furnishing
a clinical example, and describing landmark studies on patient CS as it relates to ef­
fective psychotherapy. We then review the findings of our 2011 meta-analysis (Beutler
et al., 2011). That review examined the effect of the interaction of patient coping
styles with treatment type on therapy outcome. Specifically, the meta-analysis tested
an interaction hypothesis: those patients with an externalizing CS respond best to a
symptom-oriented treatment, whereas those patients with an internalizing CS respond
 58 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

best to an insight-oriented treatment. We then report the methodology and results of

an expanded meta-analysis; indeed, our aim is to confirm the role of CS in choosing
the effective treatment that was suggested in the 2011 review. The chapter concludes
with diversity considerations, training implications, and recommended therapeutic
practices.

DEFINITIONS
Coping Style

Coping style denotes an enduring personality trait that predisposes people to deal
differentially with anticipated or experienced change. It is a characteristic way of
behaving to reduce discomfort and to adapt to a changing environment that is out­
side of one’s control (Beutler, Moos, et al., 2003). It is not a term that uniformly
connotes psychopathology because emotionally healthy individuals have coping
styles and most even have a preferred CS that predisposes them to certain habitual
responses. However, if this CS becomes extremely exaggerated, extremely variable,
or extremely rigid to the point of being applied the same way to all situations, it can
be pathological.
There are a variety of terms that include the concept of coping and from which CS
must be distinguished. For example, coping style connotes both a more general disposi­
tion and a more trait-like quality than the terms coping strategy or coping skill. Coping
strategy implies planning and thought, while coping style is quite automatic and intu­
itive. The link to planning and forethought suggests that one’s coping strategy is sub­
ject to change via education, psychotherapy, or cognitive training, and thereby coping
strategies may be less durable and habitual than those behaviors that are implied by the
designation of CS. By emphasizing the enduring and habitual repetition of behavioral
patterns, coping styles emphasize the trait-like qualities that emerge in times of change
and anticipated change.
Likewise, CS is to be differentiated from coping skill, which refers to how well the
strategies work to reduce stress or protect one against discomfort (Beutler & Moos,
2003). Unlike coping style, coping skill designates how well one copes, a characteristic
that can vary from high functioning to dysfunctional. Skill, however, ignores the na­
ture of the threat with which one is coping and the strength of one’s efforts to avoid that
threat, all of which are associated with one’s CS. Coping style refers to a durable and
reflexive tendency to respond in a particular way, a way that is exaggerated in times of
stress and that can become pathological.

Insight-Oriented and Symptom-Focused Psychotherapies

Freud (1938, 1954) developed his view of psychotherapy as a process that facilitates
patient change via insight. Such was the prevailing viewpoint for more than 30 years.
In the 1950s, non-insight views of psychotherapy emerged as viable alternatives to psy­
choanalytic psychotherapy (Orlinsky, 2017). These approaches were centered on the
 59 Coping Style

therapeutic relationship itself as a facilitator of change and the use of procedures that
aimed at teaching new skills and behaviors.
Generally, the many theories that constitute psychotherapy can be bifurcated
into those that propose insight relieves symptoms and those that directly change the
same symptoms. Insight approaches emphasize a degree of re-experiencing certain
emotions that have been repressed, suppressed, or diverted in the course of avoiding
emotional pain. This description encompasses those theories designed to achieve self­
understanding and to open one’s hidden and often guilty feelings and experiences to
one’s self and others, with the assurance and hope that these experiences will change if
one understands why they exist.
Alternatively, symptom change approaches take a view that one changes best by
intensely engaging in a systematic process of enacting new behaviors and acquiring
new perceptions, followed by social reward/reinforcement. New learning occurs if it is
rewarded; old habits are abandoned if they fail to generate reinforcement. Sometimes,
the reward that changes behavior is derived directly from engaging in a relationship
that provides a new experience of acceptance and personal value and other times not.
At those latter times, change occurs by determining what the reinforcement is for
unwanted behavior and eliminating it. A focus on direct symptom development and
change eschews the view of many insight focused therapies, that one must revisit past
pains to understand the meanings of symptoms.
This chapter emphasizes that the effectiveness of these two broad classes of
psychotherapies—insight focused and symptom focused— is greatest when they are
compatible with the patient’s CS. Psychotherapists can determine whether a particular
patient’s CS calls for a direct attack on symptoms or whether it calls for procedures that
promote insight and awareness. Unfortunately, most therapists use either an insight-
oriented approach or a symptom-focused approach, ignoring the fact that choice of
therapy type should depend on the CS of the patient.

MEASURES

Categorical Versus Continuous Scores

Accurate measurement is the key to understanding the interactive roles of externalized
and internalized coping styles and therapist insight versus symptom focus. In our dis­
cussion so far, we have presented coping styles and therapy foci as categories or dis­
crete classes. However, in clinical reality and in many measurements, therapy focus
and CS are a matter of degree. A therapist is more or less symptom focused, a pa­
tient is more or less externalized, and everyone has some of both externalizing and
internalizing tendencies.
Alas, there are times when categorical measurement is all that is possible. When
that is the case, it behooves the investigator to remember that using categories prob­
ably understates the strength of the treatment by patient interactions under study.
Measurement is more accurate and research results generally less variable when con­
tinuous scores are employed.
 60 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

Direct Versus Indirect Measures

Another consideration when selecting ways to measure coping styles and treatment
focus is the directness of the measures. The scores obtained from “direct” measures
are not filtered through a theoretical net or another person’s belief system. They are
observations; they can be seen and counted. They either can be the observations of
the patient, the therapist, or an external observer, but all observers can agree about
what they are. A direct measure asks the person involved a question and the answer
is assigned a value. The score obtained is a compilation of those responses. This is not
to say that such scores are either objective or accurate—only that they are minimally
encumbered by the observer’s theoretical framework or bias. The responses of a patient
to a structured set of questions, for example, are direct even if some of those questions
ask for “opinions.” If, however, a therapist is asked to report how a patient feels, the
measurement is indirect. It is filtered through the therapist’s assumptions as he or she
translates what the patient says to an inference of a feeling state. This is different from
asking what the patient “says” about his or her feelings, which is a direct measure.
In evaluating the joint and interactive roles of patient CS and therapist focus, there
are instances when direct measures are not available. Consider for example what
occurs when we want to know if the therapist believes in the role of insight or prefers
to move directly to an effort to change a set of symptoms. We might ask therapists
directly about their focus, but what if the therapists are not available to us and all
we know about them is their theoretical approach? Our efforts to measure therapy
might be reduced to estimating the therapists’ relative preferences from a knowledge
of what therapy they are practicing. W hat assumptions are made by the theory that
therapists hold? If we know that they prefer to use psychodynamic therapy, then we
might judge them to favor insight over direct change. But this requires an inference
that is not backed by a direct measure of the particular therapy delivered—we have to
guess what “psychodynamic therapy” means to them. If we cannot ask them directly,
we may be forced to make an educated guess, but this results in an indirect measure
of their preference—one that does not come from the identified therapists themselves.

Coping Styles
Several direct measures are available to clinicians and researchers for assessing pa­
tient coping styles. Individuals’ coping styles can be extracted from omnibus person­
ality measures, such as the Minnesota Multiphasic Personality Inventory (MMPI and
the revised M M PI-2; Butcher et al., 2011) and the NEO inventories (McCrae et al.,
2011). The MMPI (Butcher, 1990) is probably the most widely used measure of pa­
tient coping styles. Coping style is not included as one of the regular or content MMPI
scales, but validated algorithms can extract coping styles by combining relevant scales.
Several studies included in our 2011 meta-analysis used a version of the MMPI inter­
nalization ratio formula. Scores on four externalizing scales (Hy, Pd, Pa, Ma) and four
internalizing scales (Hs, D, Pt, Si) were employed to compute a ratio that indicates the
relative strength of these two coping styles.
 6l Coping Style

A second direct measure is also occasionally used to code patient CS. The
N EO-PI-R (Costa & McRae, 1989) is the original measure of the “big five” per­
sonality dimensions: Neuroticism, Extraversion, Openness, Agreeableness, and
Conscientiousness. Like the MMPI and M M PI-2, it is a direct patient self-report in­
strument. Coping style can be extracted from the N EO-PI-R by combining scales of
Extraversion, Neuroticism, and Openness in various combinations (McCrae et al.,
2011). The Neuroticism subscale is relatively complex, combining aspects of anxiety and
introversion, and is similar to what we have identified as internalization. Extraversion
is also similar to the concept of externalization but is less complex than introversion.
A direct assessment of internalizing and externalizing coping styles is afforded
by the STS/Innerlife (Beutler et al., 2009), an instrument administered and scored
via the Internet. It was developed by modifying and refining the items on the STS
Clinician Rating Form (Corbella et al., 2003; Fisher et al., 1999). The STS/Innnferlife
comprises 171 questions and assesses 22 problems domains as well as externalizing and
internalizing coping styles. It has demonstrated moderate-high internal consistency
for all the scales across cultures. The internalizing scale is comprised of items such as,
“I avoid meeting people or being around certain people because doing so makes me so
upset or angry”; “I can’t seem to say the things that go through my mind.” Contrasting
items capture externalizing patterns (e.g.,” I frequently seek out very exciting activities,
like bungee jumping, parachuting, racing, gambling, etc”.; “I have gotten into trouble
quite often because of my behavior”).
The STS/I nnerelife has certain advantages over alternative measures, primarily
because this instrument was specifically developed to assess patient characteristics
that can be used to fit treatment to particular patients. The two scales representing
externalizing and internalizing personality qualities are continuous, self-report meas­
ures. A ratio of these two CS indices can reflect the dominance of one or the other
(Harwood et al., 2011).
As already mentioned, indirect measures of CS are frequently used in psycho­
therapy research. This usually involves the assignment of a group designation trait to
all patients who share a particular diagnosis, as opposed to taking measures of all in­
dividual patients. For example, conditions such as bulimia and antisocial personality
may be indirectly coded as externalizing, while conditions like unipolar depression
and generalized anxiety may be coded as internalizing. See Table 3.1 for lists of indirect
diagnostic measures for internalizing and externalizing coping styles.

Table 3 . 1. In direct D iagnostic M easures o f Patient C oping Style

Diagnostic Indicators o f Internalization Diagnostic Indicators o f Externalization
Generalized A nxiety Disorder A ntisocial Personality Disorder
Obsessive-Com pulsive Disorder (Obsessive) Substance Abuse Disorder
Avoidant Personality Disorder H istrionic Personality Disorder
Depressive Disorder Spectrum Obsessive Compulsive Disorder (Compulsive)
Social A nxiety Disorder Hypomania/Cyclothymia
 62 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

Therapy Focus
The most frequently used direct measures of treatment focus are of a psychotherapist’s
in-therapy behavior or theoretical orientation. For example, the Therapist Process
Rating Scale (Malik et al., 2003) is a research instrument under development that
can rate in-session behaviors to identify therapeutic styles and treatment focus.
Compliance with treatment methods for increasing insight and reducing symptoms
can be directly rated from therapist video or audio samples of a sample therapy session
(e.g., Holt et al., 2015).
The clinician’s activity ratings can be obtained on therapy as it is normally applied
in a naturalistic design (e.g., Kadden et al., 1989), but such a procedure is confounded
by numerous other factors that influence what and how the therapist delivers the treat­
ment. A more reliable procedure is to subject participating therapists to advanced
training of a preferred approach (e.g., Barber & Muenz, 1996; Beutler, Moleiro, et al.,
2003; Poulsen et al., 2014).
Unfortunately, random assignment of patients to a particular treatment focus is
often not possible. Most randomized controlled trials (RCTs) incorporate categorical
designations of treatment (e.g., cognitive-behavioral therapy [CBT], psychodynamic,
experiential) and patients (depressed, anxious, psychotic). Research of this type rarely
looks at individual differences among patients or therapists within classes at all, unless
through a post hoc analysis, and this is rare as well. As an unfortunate result, patient
predictors are confined to diagnostic criteria and therapy predictors are confined to
categorical differences among brand names of therapies. Individual differences among
therapists’ applications of theories as well as individual patient differences within diag­
nostic groups are ignored, increasing error variance.
Fortunately, RCT data often can be nudged to address the optimizing role of “fit” be­
tween CS and therapy focus, but only by sacrificing sensitivity by using categorical, in­
direct measures. Thus the same group designations are applied to all individuals, with
the probable loss of specificity and sensitivity. The measurement is indirect because
the resulting classifications of patient coping styles and treatment foci are not derived
from the individuals themselves but from a group generalization. The magnitude of
psychotherapy effect using an indirect measure is usually smaller than when the same
constructs are directly measured (Beutleret al., 2011).
When indirect measures are required, researchers assign a treatment focus to the
entire group of psychotherapists sharing the same theoretical orientation or following
the same treatment manual. For example, if a therapist is identified as psychodynamic,
that person may be inferred to be vitally interested in the patient’s unconscious ex­
perience and with the role of insight in a patient’s recovery. Alternatively, a cognitive
therapist is typically inferred to be interested in stimulating symptom change. In either
case, the group label is an insensitive estimate of in-session treatment focus.

CLINICAL EXAMPLE
Mr. S. is a 42-year-old, married Vietnamese American man who immigrated to the
United States from Vietnam 17 years ago, where he had worked as a science teacher. By
 63 Coping Style

his report, since immigrating, he has not made any close friends, has lost much of his
social identity, and has been isolated from others, including his extended family. Mr.
S. lives with his wife and his two children (ages 12 and 15) in a modest and poorly kept
house near his work, a Vietnamese grocery shop.
Mr. S. was referred to the outpatient clinic by his wife for treatment of his gambling
addiction. His wife reports that Mr. S. often isolates himself in his room or escapes to a
casino, where he prefers odds-based games, such as the slot machine, over skill-based
games (e.g., black jack, poker). When his wife has asked him why he only plays the slot
machine, Mr. S. responded that the slot machine had “become his close friend who can
trust and satisfy his need for enjoyment.” Mr. S. had secretly accumulated $3,000 debt
on his credit card. His wife discovered his credit card debt, at which point she called
211 to obtain information for medical/mental health services and reached a psychol­
ogist, Dr. K., who previously had worked with gambling addictions and who ran the
training clinic at a local university.
The psychologist asked to speak with Mr. S. to arrange an appointment at the
training clinic. Dr. K. explained that he would arrange treatment with a PhD student
therapist and would supervise the treatment personally. After some intense discussion
and initial refusal, the patient agreed to keep the appointment offered.
The patient and his wife arrived on time for the appointment with the psychology
trainee (Ms. J). They waited quietly until they were called. Ms. J. saw the patient without
his wife. While beginning to establish a therapeutic relationship and gathering a social
history, Ms. J. gathered information about Mr. S’s support system and debt. The thera­
pist confirmed Mr. S’s report that he was several thousand dollars in debt, that he does
not play skill-based games, and that he avoids goal-oriented, outgoing, competitive,
and social activities. Both his behavior and the test results indicated a socially isolated
and emotionally restricted pattern. The patient indicated that he does not like his cur­
rent lifestyle, but he felt hopeless to change it. He reported some suicidal thoughts but
no intention or activity.
When asked about his home and work environment, Mr. S. acknowledged conflicts
with his wife and coworkers, to which his typical pattern was to engage in vocal
outbursts for a short time and then to withdraw and escape. He acknowledged several
instances of impulsive behaviors, most of which came when immediately confronted
with disagreement. These externalizing behaviors were quickly abandoned, however,
for fear of him being rebuked or discounted by family members and disregarded by
coworkers. He did not speak out about these difficulties but rather isolated himself to
avoid further confrontation. He also became self-blaming and felt guilty.
The therapist conjectured that Mr. S. has come to believe that his extended family
members do not like his wife; they blame her for his troubles and want him to get a
divorce. He blames himself for letting them persist in these beliefs. He is certain that if
he could try harder to quit gambling, he could repair the situation with his wife. These
strains cause much concern for Mr. S. He maintains that he does not want a divorce in
spite of the problems in his marriage. He is afraid that this extended family will never
accept his wife or permit her to visit them.
The therapist’s formulation emphasizes that Mr. S. is an internalizing individual
who uses gambling as an escape from familial threats and as a confirmation of his
 64 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

own lack of self-worth. Concomitantly, he scored high, within the clinical range, on
the STS/Innerlife internalizing scale. Externalization was also evident but barely in the
clinical range. Mr. S. was considered a self-debasing internalizer with some defensive
externalizing impulses.
Mr. S’s depression and anxiety scales were also in the clinical range on intake.
Predictably, he also scored high on scales indicating family disturbance, obsessive-
compulsive behaviors, and social avoidance. He acknowledged all of these symptoms
and indicated that, when he becomes depressed or anxious, he begins to fear what he
will do and withdraws from others to avoid their blame. Subsequently, he feels like no
one loves or wants to be around him.
The treatment plan was to focus on his recurrent depression and self-blame first,
as an intial step and motivator of his compulsive gambling. They elected to use time-
limited dynamic psychotherapy (Strupp & Binder, 1984) because of its interpersonal
focus and its emphasis on vicious cyclical patterns. This represents a match between
Mr. S’s internalized CS and an insight-oriented psychotherapy. An initial contract was
established for 25 sessions.
The plan was to link the patient’s depression and self-loathing to its historical
roots in his family and to treat his gambling as both an expression of anger and a
confirmation o f his low self-worth. The therapist initially encouraged the patient
to look at the longstanding and maladaptive patterns in his family relationships
and extended these to the self-destructive similarities in his gambling. The thera­
pist analyzed the patient’s core conflictual relationship theme, framed in terms of
his wish (his thought), expectations from others, actual response from others, and
introject.
The patient gradually came to understand that his CS was related to his father’s
tendency to discount and reject him. As a child, Mr. S. was constantly blamed for
displeasing his father and stimulating his father’s anger. Early on, he initially tried to
fight back, an approach that was quickly smashed. Then he emotionally hid as a way of
pacifying his father, a behavior that included blaming himself as an appeasement. His
anger at his father precipitated his own sense of shame, self-loathing, and doubt and
fomented an introject as one who is guilty for all family problems.
The therapist’s work centered on understanding the patient’s relationship with
his father and his wife, his efforts to protect himself and them, and the self-punitive
results of gambling. The therapy emphasized insight and its ability to generate behavior
change. The therapist encouraged self-monitoring of his core conflictual relationship
theme so he could both see how “angry” behaviors, such as gambling and withdrawal,
routinely followed a rejection or criticism and how they were accompanied by an
introjected self-loathing. The therapist believed that Mr. S. had literally directed the
anger and hate that he had at his father, internally in the form of the introject, “I am
hateful.”
A turning point for the patient occurred when he realized that he had reconstructed
his family in the therapy room. Ms J. played the role of his mother, who was both
rejecting and rescuing when conflicts emerged with his father. Dr. K. assumed the role
 65 Coping Style

of the patient’s father, remote and removed but in control of all that happened, behind
the scenes, and was he from whom the patient sought escape. Soon the patient came to
see his losing streak(s) in gambling as an expression of rebellion that arose from and
overwhelming sense that he was indebted to his father who tolerated him but neither
appreciated his suffering nor acknowledged it.
About eight months into his treatment, Mr. S. visited his widowed mother and
talked about his father. He discovered that his father had been abusive to his mother
and was a closet alcoholic. Mr. S. also recognized the punishing role that his gambling
often played. Thereafter, his gambling became less and less frequent.
Nonetheless, these changes were followed by a period of distress and depression.
But the initial disappointment he experienced in his father was followed by his gradual
realization that his father was much like him, but Mr. S. had made a better life than
had his father. He felt vindicated and, finally, even proud that he had been “a better
person” than his father. When he ended therapy after 25 sessions, Mr. S. was no longer
gambling or hiding in a remote part of the house. He asked that Dr. K. join Ms. J. and
him for the final session. Here he confronted Dr. K. with being the Oz behind the cur­
tain. When Dr. K. gave him a goodbye hug, he felt relief and pride.
In this case, Mr. S. responded well to an insight-focused psychotherapy that fit his
internalizing CS. Although mental health professionals would have been understand­
ably tempted to treat his gambling more directly with a symptom-focused therapy,
this clinical example and the research evidence favor adapting the treatment to his
transdiagnostic personality features, including CS, for maximum results.

LANDMARK STUDIES
This section describes three studies that we consider to be landmarks for their role in
drawing attention to coping styles and their interaction with psychotherapy focus. The
initial study was the first, to our knowledge, to uncover clear evidence of the interactive
effects of patient CS and therapist focus. The second study was drawn from archival
data of a historic psychotherapy study. The third investigation was a rigorous RCT that
demonstrated the differential effect of treatment focus on psychotherapy outcome as a
function of patient CS.

Beutler, Engle, et al., 1991

B eutler, Engle, and colleagues (1991) conducted a randomized trial of three manualized
treatments (cognitive therapy [CT], supportive self-directed therapy [S/SD], and expe­
riential therapy [FEP]) for 63 patients with major depression. Patient coping styles were
assessed with the MMPI internalization ratio (Butcher et al., 2011), and treatments
were selected to represent certain qualities that both fit and were incompatible with
patient coping styles. Specifically, the symptom focus of CT was expected to be most
effective among patients with externalizing coping styles, whereas both S/SD and FEP
would prove good fits with internalizing patients.
 66 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

Analyses confirmed the expectation that differences among treatments were neg­
ligible when patient CS was not considered. This finding is reminiscent of the “Dodo
bird verdict” that has become prevalent among outcome findings. Namely, that on av­
erage one well-reasoned treatment is about as effective as another across most patient
populations (Wampold & Imel, 2015). However, when patient CS was entered into
the equation, those patients provided with a treatment that fit their CS did better than
those whose treatment did not fit. Externalizing patients did much better in symptom-
focused treatment (CT) than they did in insight-focused treatments (FEP and S/
SD). On the other hand, internalizing patients responded better to insight-oriented
interventions (FEP and S/SD) than to symptom focused ones (CT). The better the fit,
the better the outcomes as indicated by a high effect size (d = .75; Beutleret al., 2011).

Barber and Muenz, 1996

Barber and Muenz (1996) used archival data from the historic Treatment for Depression
Collaborative Research Program (Elkin et al., 1989) to look for patient by treatment in­
teraction effects. The sample included patients with unipolar depression from the orig­
inal study who were subclassified as obsessive-compulsive (internalizers) or avoidant
(externalizers). Patients were randomly assigned to either CT, interpersonal therapy
(IPT), tricyclic antidepressants, or a pill placebo. Therapists were psychologists or
psychiatrists supervised and monitored to maintain treatment fidelity. The efficacy of
CT and IPT were found to be essentially equivalent.
Because the original study found few outcome differences among the treatments,
Barber and Muenz (1996) looked specifically at the interactions between patient per­
sonality (obsessive-compulsive vs. avoidant) and the two psychotherapies, CT and
IPT. The investigators reviewed the histories of patients representing avoidant and
obsessive-compulsive styles. When the clients reached adulthood, those with avoidant
(externalizing) coping patterns primarily adopted direct avoidance strategies to deal
with unpleasant activities, such as running away from work, school, or assigned social
activities. They engaged in many activities that directly avoided whatever they were
assigned to do and overall, adopted an externalizing CS pattern (Kagan, 1998). In con­
trast, clients with internalizing coping styles (obsessive-compulsive personalities), ex­
perienced preoccupation with failure, a need to be excessively orderly, and valuing
of perfectionism. They had high levels of self-doubt, approval seeking from others,
mental and interpersonal over control, and self-inhibition.
The results of this study were consistent with the predicted CS and treatment focus
interaction. obsessive-compulsive disorder/fnternalizing patients did best with IPT,
and the avoidant/externalizing patients did best with CT.

Beutler, Moleiro, et al., 2003

In a pivotal study, clients with co-occurring depression and substance abuse were
assigned to CT, narrative therapy, or prescriptive therapy. The symptom-focused CT
 67 Coping Style

was expected to be most advantageous for externalizing patients, while the insight-
oriented FEP was expected to be most effective for internalizing patients. Prescriptive
therapy was expected to do equally well with both patient types.
In a head-to-head comparison, the three psychotherapies yielded no meaningful
differences in treatment outcome, so the authors analyzed the results using direct
measures of treatment fit to CS. The insight or symptom focus was derived by direct
observations and blind ratings of therapists in-treatment. The results supported the
effectiveness of treatment fit on outcomes, across therapy types. That is, patients with
internalizing coping styles fared better with an insight-focused treatment (FEP) while
those with an externalizing style fared best with a treatment that focused on direct
symptom change (CT). The effect size of a good treatment match was d =.71, a moder­
ately strong effect (Beutler et al., 2011).

RESULTS OF PREVIOUS META-ANALYSES

To our knowledge, the only meta-analysis on CS in psychotherapy was the 2011 meta­
analysis (Beutler et al., 2011) published in the second edition of this book. The pur­
pose of that 2011 analysis was to identify research studies that provided data on the
interaction effect associated with client CS (internalizing and externalizing) and psy­
chotherapy type (insight oriented vs. symptom focused). To be included in the meta­
analysis, a study had to meet four or more of the following six criteria:

1. Inclusion of one or more reliably applied therapeutic treatments by trained

psychotherapists to represent meaningful variations among the treatments.
2. Inclusion of patients who were at least moderately impaired to represent client
variability among coping styles.
3. Clearly defined treatments and distinct patient “matching” characteristics that
permitted the derivation of fit between patient coping styles and treatment focus.
4. Employed random assignment of clients to treatment and a diverse body of
therapists to provide contrasting responses by clients to insight/awareness and
symptom-focused treatments.
5. Confirmed the reliability of treatment and CS measures.
6. Included outcome measures that permitted an analysis of outcomes as a function
of the fit between client CS and therapy focus.

The researchers searched major indexing sources, inspected the bibliographies of

included articles, and reviewed the past year of issues of prestigious psychotherapy
journals. The review identified 16 studies, 12 of which met at least 4 criteria. The most
frequently missed criteria were numbers 3 and 6. All 12 meta-analyzed studies either
produced or provided the data to produce pre- and post-effect sizes (all effect sizes
were converted to Cohen’s d) on the 42 treatments included in these studies. Mean
treatment effects were averaged to produce one estimate of mean efficacy per study.
Studies also included data to calculate an effect size based on treatment fit for each
 68 P S Y C H O T H E R A P Y R EL AT IO N S H IP S THAT WORK

treatment pair and sufficient additional data to extract the effect sizes associated with
the interaction between patient CS and therapy focus. In all cases, the effect sizes were
weighted by the study sample size.
The 2011 meta-analysis found a weighted mean effect size of d =.55 (p < .05; 95%
confidence interval [CI] = 0.52-0.58) for the interaction between patient CS and
therapy focus. The effect size is within the range identified as “medium” by Cohen
(2008) and indicates that benefits accrue from matching client CS and treatment
focus. Specifically, there was a clear pattern found in which externalizing patients did
best with symptom-focused psychotherapies and internalizing patients did best with
insight-oriented psychotherapies. The match of CS and therapist focus accounted for
20% of the overall variance in patient improvement.
The type of measurement for treatment focus proved critical. Assessment of
therapy insight and symptomatic focus overwhelmingly relied on therapy brand name
designations (indirect) rather than direct observations of therapist in-session behavior.
Although direct measures were infrequent, their mean effect sizes were higher than
those obtained using indirect indices (n = 9; ds = .73 vs. .44). Thus the overall effect
size in the meta-analysis was probably an underestimate of the actual strength of the fit
between patient CS and treatment focus.

META-ANALYTIC REVIEW

Literature Search
We began our literature review with the 12 studies included in the 2011 meta­
analysis (Beutler et al., 2011). We did not subject these studies to additional eligi­
bility analysis since the 2011 inclusion and exclusion criteria were virtually identical
to those used in this meta-analysis. To extend the number of studies, we conducted
broad-band searches, first using terms such as “coping style,” “personality,” “intro­
version” and “extraversion,” and the like to identify the studies via PsycINFO and
the most widely cited journals, such as Journal o f Consulting and Clinical Psychology,
Journal o f Counseling Psychology, and Journal o f Clinical Psychology. We also searched
for references to patient samples that carried diagnoses that were indicative of
internalizing and externalizing disorders.
We searched the years 1990 to 2018. Figure 3.1 presents a PRISMA flow diagram.
The search terms “psychotherapy outcome” and “randomized control trial” resulted
in a pool of 756 studies. Then we selectively entered key diagnostic terms that were
used in the 2011 analysis to indicate internalizing and externalizing disorders. The
diagnostic search terms for internalizing and externalizing are presented in Table 3.1.
As potential studies emerged from these searches, we examined them using the
same six inclusion criteria used in the 2011 meta-analysis (listed in the previous sec­
tion). To be considered for the final analysis, the authors of a published study needed
to include sufficient data to calculate the total sample size, mean effect size, and vari­
ance of the mean effect sizes. In addition, each study needed to have provided data in
which an effect size corresponding to the difference between good and poor treatment
fit could be calculated. We excluded Internet and online treatment studies because
 69 Coping Style

figure 3 .1 PRISMA flow diagram.

of probable confounds, such as the unknown feasibility of online application and the
uncertainty of determining the nature of the disorders and coping styles presented in
these studies.
We identified 11 new studies not included in the 2011 analysis that served as
candidates for inclusion in this meta-analysis. We examined each o f these studies
 70 P S Y C H O T H E R A P Y R EL AT IO N S H IP S THAT WORK

to ensure the presence of actual therapy and clinical populations. Two studies were
dropped from further analysis on this basis. Three more studies were dropped because
of insufficient data to calculate effect sizes.
The final sample of our meta-analysis consisted of 18 studies, 12 from the 2011
review and 6 from the literature search described here (see Figure 3.1). One study
(Barber & Muenz, 1996) excluded from the 2011 analysis was included in this one
when it was determined that the patient distinctions used by those authors could reli­
ably be mapped onto internalizing and externalizing patients. Among the 18 studies,
there was a total of 1,947 patients and 57 different treatments on which effect sizes were
computed.

The Studies
The18 studies in the meta-analysis are summarized in Table 3.2. Eight of the 18
studies used direct measures to assess client CS, and 9 studies used indirect measures.
The two remaining studies included both direct and indirect measurements of CS.
The two measures resulted in identical nominal classification, and this classification
was used for the analyses. Fourteen of the 18 studies were RCTs that tested the rela­
tive effects of two therapies that were classified by “focus” for our analysis. Fourteen
studies (74%) were conducted in the United States, three (16%) were conducted in
Germany, one (5%) was completed in China, one was carried out in Finland, and one
was conducted in both the United States and England.

Coding Studies
We coded each study with respect to four variables (outcome, coping styles, interac­
tion effects, and therapy foci—see Table 3.1). Outcome was defined by the outcome
measures used by the authors of each study. To allow comparison among studies, out­
come in each study was converted to a standard score (Cohen’s d). Only 12 of the
studies provided outcome data on the role of therapy focus (Table 3.2, column 7) and
only 9 studies provided data on patient CS (Table 3.2, column 8). All 57 treatments
contained in these studies, however, were coded as either symptom or insight focused
using three trained and independent raters. Two raters independently calculated all
of the effect sizes representing outcomes and, separately, the two patient and therapy
distinctions. A third reviewer independently reviewed the raters’ conclusions. The aim
was to achieve ratings on which at least one of the two primary raters and the third
rater could agree. If the primary raters did not agree in two tries, the third rater made
the decisive rating. To retain the study in the meta-analysis, agreement was achieved
between one of the first two raters and the third rater on the effect size as well as on the
classification of therapy focus and CS. This process was also employed to calculate the
interaction terms, as we describe later.
Table 3.2. Results of Effects of Coping Style and Treatment Focus
Study Total N Design Type of Measure (Treatment
Focus)
Beutler, Engle, et al. 63 RCT I (FEP/Ins vs. SSD/Sym)
(1991)
Litt et al. (1992) 79 RCT I (CST/Sym vs. Interact/Ins)
Beutler et al. (1993) 46 RCT I (CT/Sym FEP/Ins vs. SSD/
Ins

Longabaugh et al. (1994) 140 RCT I (CBT/Sym vs. ECBT/Ins)

Barbar & M uenz (1996) 84 RCT/MR I (CBT/Sym vs. IPT/Ins)
Calvert et al. (1998) 108 MR/Q-E D (T O Q ) (Sym vs. Interact)
Kadden et al. (1989) 96 Nat I (CBT/Sym vs. IPT/Ins)
Karno et al. (2003) 47 RCT I (CT/Sym vs. FST/Ins)
W ilson et al. (2002) 154 RCT D (C B T vs. IPT)
Beutler, M oliero, et al. 40 RCT/MR I (CT/Sym vs. NT/Ins, PT)
(2003)
M ilrod et al. (2007) 49 RCT I (PFP/Ins vs. ART/Sym)
Knekt et al. (2008) 326 RCT I (SFT/Sym vs. STD/Ins &
LTD/Ins)
Kim para (in Beutler, 12 1 Nat D (SFT/Ins vs. Sym)
2009)
Johannsen (in Beutler, 92 Q-E/MR D (TPRS/Ins vs. Sym)
2009)
Leichsenring et al. (2009) 57 RCT I (CBT/Sym vs. STD/Ins)
Stangier e ta l. (2011) 117 RCT I (CBT/Sym vs. IPT/Ins)
Coping Style of Tx Included ES Focus ES Coping ES Fit (d) ES Fit (r) Vfit
Patients
D (Int-Ext) Both 1.63 0.75 0.35 0.076

D (Ext) Sym 0.63 0.30 0.053

D (Int-Ext) Both 1.16 1.64 0.63 0.124

I (Ext) Sym 0 .12 0 .6 8 0.37 0.18 0.091

I (Ext-Int) Both 0.18 0.18 0.09 0.048
D (Int-Ext) Both 0.81 0.38 0.003
D (Ext) Both 0 .6 0.29 0 .0 0 2

I (Ext) Both 0 .0 2 0.3 0.5 0.24 0.91

I (Ext) Both 0.13 0.13+ 0.06 0.14
D/I (Int-Ext) Both 1 .0 1 0.99 0.71 0.33 0 .12

I (Int) Both 0.92 0.71 0.33 0.087

D (Int) Both 0.94 0.94 0.17 0.08 0.015

D (Int) Ins 1.17 0.76 0.36 0 .0 0 2

D (Int-Ext) Both 0.61 0.29 0.045

I (Ext) Both 0.13 0.57 0.57+ 0.28 0.16

I (Ext) Both 0.44 0.43 0.43 0 .2 1 0.07
Table 3.2. Continued
Study Total N Design Type of Measure (Treatment Coping Style of Tx Included ES Focus ES Coping ES Fit (d) ES Fit (r) Vfit
Focus) Patients
Beutler et al. (2012) 258 RCT/MR I (CT/Sym vs. D/I (Ext-Int) Both 0.11 0.15 0.13 0.06 0.004
Psychodynamic, ES/Ins
Poulsen et al. (2014) 70 RCT I (PAT/Ins vs. CBT/Sym) I (Ext) Both 1.53 1.53 0 .6 0 .12

Total N Mean/ Weighted Avg Focus ES: Mean/ Weighted Avg Coping ES Mean Fit ES (Random r 95% CI
Effects Model)
1,947 0.60 0 .6 6 0.60 0.29 0.44 0.76 p < .0 0 1

Q:
115.84 p < .0 0 1

Notes. Design: RCT = randomized control trial, MR = multiple regression, NAT = naturalistic study, Q-E = quasi-experiment.
Measure Tx (Treatment) Focus = Either direct (designated as D) or indirect (designated as I). Indirect measures are based on the treatment model used and identified as either symptom
(Sym) or insight (Ins) focused; direct measures are based on an individual measure of the use of insight or symptom change procedures.
Direct measures: TOQ = Therapist Orientation Questionnaire, TPRS = Therapist Procedure Rating Scale. Indirect measures of Tx Focus are based on the model of treatment studied.
CT = cognitive therapy; EFT = focused expressive therapy; SSD = supportive self-directed therapy; CST = cognitive skills training; CBT = cognitive-behavioral therapy; ECBT = relation­
ship enhanced CBT; IPT = interpersonal therapy; Interact = interactive; FST = family system; NT = narrative therapy; PT = prescriptive therapy; PFP = panic focused psychodynamic;
ART = applied relaxation; STF = solution focused therapy; STD = short-term dynamic therapy; LTD = long-term dynamic therapy, ES = experiential systems; PAT = psychoanalytic therapy.
Coping style (CS) is measured either directly (designated as D) or indirectly (designed as I). Direct measures are an individual personality test; indirect measures are derived from the
diagnosis (see Table 3.1).
M ES (Mean Effect Size) (Focus) = The M ES attributable to the treatment focus variable combining all treatments; M ES (Coping) = The M ES attributable to the coping style variable
combining all treatments; M ES Fit = The mean difference between ES for “good” and “poor” fit, estimated in MR/Nat studies from correlational data. All ESs are expressed as d with the
exception of r . M r = mean effect size as correlation between treatment and total means combining all treatments; M V fit = mean differences between treatment and total variances for
“good” and “poor” fit.
+ An interaction effect for coping style and treatment on outcome was not available. Resulting score of fit was the result of difference of poor match from good match of treatment based
on coping style.
 73 Coping Style

Outcome Effect Sizes

The procedures for calculating the ways in which treatment outcome was identified
and measured in our meta-analyses are described here. We confined the analyses to
outcomes that were measured at the end of treatment except in those cases when only
more distal outcomes were reported. The outcome score was the mean d difference
for all of the treatments categorized as symptom focused and, separately, for those
categorized as insight focused (11 studies; Table 3.2, column 4).
In the case of treatments with multiple outcome measures, we followed the
suggestions of Borenstein and colleagues (2009, p. 234) and calculated the mean
standardized (d) score difference for each comparison. Then we pooled (averaged) the
difference scores across multiple outcome measures in a study to reach a single score
indicating pre-post differences for each insight- and symptom-focused treatment.
When two or more treatments with the same focus were in a study, their effects were
pooled via a weighted average using the inverse of the variance. When both pretreat­
ment and posttreatment scores were available, we calculated an effect size for each of
the treatments using the formula:

M 1- M 2
d=
SDpooled

When pretreatment (and often posttreatment) means were not reported in a study
(k = 8), we relied on the authors’ report of the individual treatment effect sizes as
expressed in a statistic indicating change (e.g., d, F, t). Weighted transformations to d
followed, as per Lipsey and Wilson (2001, pp. 172-206).
One study (Wilson et al., 2002) did not report either pretreatment scores or treat­
ment d scores. Rather, they reported the end of treatment scores only. In this case,
we estimated the intake or pretreatment scores for each treatment (CBT vs. IPT) by
first calculating two grand mean pretreatment score across all studies. One score was
expressed as the (weighted) mean pretest of all treatments representing an insight
focus, and the other expressed the pretest (weighted) mean pretest of all treatments
representing a symptom focus. We then calculated d (change) scores between IPT
(the insight-focused treatment) and the cross-study pretest mean score for all insight
treatments. Identically, we calculated a d difference score between the study CBT and
the cross-study, grand (weighted) mean of all symptom-focused treatments.
W ith these adjustments, we had d scores based on pretreatment and posttreatment
differences for 33 pooled or single treatments in all 18 studies (Table 3.2, column 6).

Treatment Focus
Using the rater system previously described, we classified all treatments into one of
two categories—insight focused or symptom focused. Among the 14 studies that used
indirect measures to assess therapy focus, names and descriptions of the therapies were
 74 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

used to classify them into insight or symptom focused. The procedure to classify each
therapy employed the three raters as described previously.
In all cases, indirect measures yielded a simple categorical distinction of either in­
sight or symptom focus, whereas direct measures produced a dimensional score that
assessed relative use of symptom- or insight-focused procedures. To make a classifi­
cation, the raters attended to the dominant procedures and guiding philosophy of the
psychotherapy system and then reached agreement on the categorical placement of
each therapy focus.
Both to assess the relative change between the treatment foci in each study and to
reduce the scores to one per study, we subtracted the pretreatment d score from the
posttreatment d score in all 18 studies. This difference score indicated which treatment
had done better and the magnitude of that advantage. That calculation also yielded a
single score for each study indicating the relative efficacy of insight focused compared
to symptom-focused treatments.
Whenever possible, we adopted the coding procedures used in the 2011 meta­
analysis to preserve consistency and comparability across reviews. We tried to uti­
lize the computation procedure that best fit the characteristics of the particular data
presented in each study.

Coping Style
Seven studies utilized a mixed inpatient or outpatient sample that could be reliably
divided into internalizing and externalizing subsamples using either direct or indi­
rect measures (see Table 3.2, column 8). Five of these studies (Beutler et al., 1993;
Beutler, Engle, et al., 1991; Beutler et al., 2012; Beutler, Moliero, et al., 2003; Johannsen
et al., 2009) utilized a direct measure of patient CS, whereas the remaining two studies
(Barber & Muenz, 1996; Calvert et al., 1998) used diagnosis as an indirect measure of
CS and then compared patients with internalizing and externalizing diagnoses.
Nine studies employed diagnostically homogeneous samples, without additional
assessment, that precluded them from being divided reliably into internalizing

Table 3 .3 . Intercorrelations am ong Internalizing and

E xternalizing C oping Style, Subjective D istress, and
R esistance in A lcohol Study“

Variable 1 2 3 4
1. Internalizer - -. 44 ** .48* -.2 6
2. Externalizer - -.0 3 .70*
3. Subjective - - .2 1

Distress
4. Resistance - - -

"Fisher et al. (1999).

“Correlations were significant at the .01 level (two-tailed).
 75 Coping Style

Table 3.4. Intercorrelations am ong Internalizing and

E xternalizing C oping Style, Subjective D istress, and Resistance
in Shyness Studya

Variable 1 2 3 4
1. Internalizer - .60* .78* .61*
2. Externalizer - - .60* .46*
3. Subjective Distress - - - .72*
4. Resistance - - -

aK im p a r a e t a l. ( 2 0 0 7 ) .

^ C o r r e la tio n s w e re s ig n if ic a n t a t th e .0 1 le v e l (t w o - t a ile d ) .

and externalizing groups. In these cases, the raters made a final classification (see
Table 3.1). Five of these studies (Litt et al., 1992; Longabaugh et al., 1994; Kadden et al.,
1989; Karno et al., 2003; Poulson et al., 2014) focused on externalizing patterns, pri­
marily substance abuse. Four of the studies (Kimpara et al., 2009; Knekt et al., 2008;
Leichsenring et al., 2009; Milrod et al., 2007) treated internalizing individuals with
mixed symptoms (e.g., social phobia, unipolar depression, obsessive thoughts).
Because these nine studies were each conducted on only a single group of patients
and, in some cases, with only a variation of a single treatment (e.g., Kimpara et al.,
2009), a complete test of CS effects could not be conducted. In these cases, an effect
size (d) was calculated for each treatment on a single group of patients, and we dis­
tinguished between those who had a “good fit” with treatment focus and those who
experienced a “poor fit.” While these studies did not permit a within study analysis of
interaction effects between CS and treatment focus, they did contribute to the overall
assessment of the fit or interaction when effect sizes were compared among collapsed
groups of externalizing and internalizing patients.
Only six studies (Barber & Muenz, 1996; Beutler, Engle, et al., 1991; Beutler et al.,
1993; Beutleret al., 2012; Calvert et al., 2003; Johannsen, 2009) included subsamples of
both internalizing and externalizing coping styles as well as treatments with two treat­
ment foci. All of these studies used a direct measure of CS, but not all used a direct
measure of therapy focus. For these studies, we conducted a complete 2 x 2 compar­
ison that included assessment of interaction effects.

Interaction Effects
Interaction effects were calculated as the weighted product of CS and treatment
focus in a standard meta-analysis (Lipsey & W ilson, 2001). Unfortunately, only 9
studies of the 18 reported both pre- and posttreatment test scores from which change
could be assessed, and only 7 studies reported main effects for both CS and therapy
focus (see Table 3.2, columns 7 and 8). Fortunately, however, all studies in our series
contained either pre- and posttreatment changes or the information necessary to
conduct regression analyses from which we extracted a change score associated with
each treatment.
 76 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

The mean effect size fit was expressed as multiplicative combination of the effects
of therapy focus and CS on d change scores. In this case, d change scores were also
weighted for variance and sample size. Thus, in the primary analysis, we entered in the
meta-analysis, mean d change scores for each treatment representing patient CS and
treatment focus. The analysis also calculated an interaction term, shown in column 9
of Table 3.2. This term is the weighted product of CS and therapy focus d change scores
and is taken as the index of effect size fit for each study.
For post hoc analyses, we calculated fit in a slightly different manner. We calculated
a “good fit” using a weighted mean change score occurring for all treatment pairs that
included internalizing patients x insight-focused therapy and externalizing patients x
symptom-focused therapy. To provide a contrast, we also calculated the efficacy of the
“poor fit”—that is, internalizing patients x symptom-focused therapy and externalizing
patients x insight-focused therapy.

Excluded Studies
To illustrate the complexities in our coding studies for patient CS and treatment focus,
we describe several studies that were omitted from our meta-analysis and explain the
reasoning that led us to exclude them from the data set.
One reason we excluded studies was the inability of the raters to agree on what pa­
tient groups were represented. That is, there were instances in which the three raters
could not agree on the type of CS of a study sample. Disagreement among raters was
most clearly seen in our efforts indirectly to cluster diagnostic samples into internalizing
versus externalizing categories. The classification of patients with both obsessions and
compulsions, with cyclothymic disorders, and with personality disorders presented
a significant challenge to reliability when determining a classification. For example,
the patients in one excluded study were Cluster C diagnosed patients (Berggraf et al.,
2014) with both internalizing and externalizing qualities (mixed coping styles) that
could not be easily disentangled. Though the study did analyze differential responses
to short-term psychodynamic (insight-oriented) and cognitive (symptom-focused)
psychotherapies on improvements of interpersonal problems, it proved to be im ­
possible to measure accurately a dominant CS. Two studies were excluded from our
meta-analysis due to their use of unreliable measures. One study compared cognitive
therapy and exposure therapy for hypochondriasis (Richtbeerg et al., 2017); however,
patients experiencing hypochondriasis proved to have a great deal of variability on
personality and CS indices. A second study assessed optimal outcomes in cognitive
therapy and interpersonal therapy for depressed individuals but was excluded due to
lack of variability in both patient coping styles and treatment focus (Hube et al., 2015).

Results
Data were analyzed with random effects meta-analyses using Wilson’s (2005) SPSS
macros with supplementary multivariate analyses. The analyses were based on 18
 77 Coping Style

studies and 1,947 patients, and we also compared the results from the 12 studies in the
2011 analysis to those obtained in the 6 studies added for the current analysis.
The new studies were similar to the old studies in several important ways. For ex­
ample, all of the new studies and a clear majority of the older ones were based on a
RCT design (Table 3.2, column 3). Both the older and newer studies had a similar dis­
tribution of studies favoring use of an indirect measure of therapy focus (76% vs. 83%;
Table 3.2, column 4).
On the other hand, nine (82%) of the old studies but only one (16%) of the new
studies used direct measures of patient CS (Table 3.2, column 5). A confounding fea­
ture was that all but two of those in the 2011 studies that used a direct measure of pa­
tient CS were from one laboratory. The original studies included 6 (50%) from a single
research group, and these studies were the only ones that specifically addressed the CS
x treatment focus interaction. In the newer studies, two (33%) specifically addressed
this effect, and only one of these (Barbar & Muenz, 1996) was from the research group
that dominated the 2011 studies. The importance of this difference was that only
studies that specifically addressed the CS x treatment focus used a methodology that
included two levels of both CS and focus.

Main and Interaction Effects for Total Sample

The main objective of this meta-analytic review was to study the interactive effect of
patient CS and treatment focus on psychotherapy outcomes. In an ideal study of the
interactions of CS (externalizers vs. internalizers) and treatment focus (insight vs.
symptom), at least two levels of each variable would be included. However, only 11
of the 18 studies in the current analysis specifically evaluated the fit of treatment to
the patient, and only 4 of these used a group comparison design (as opposed to a
regression-based design). The remaining studies were primarily concerned with de­
termining the efficacy of a particular psychotherapy and tested two treatments among
a diagnostically homogeneous sample of patients. These latter studies and the re­
gression designs frequently omitted reporting pretreatment scores on the dependent
variable. This omission limited our statistical comparisons. To address this problem,
we estimated the pre-post differences to capture a reliable interaction term from the
analyses reported by the authors, but we could not consistently estimate an effect size
(d) for each therapy focus or CS.
The meta-analysis produced a mean CS x therapy focus interaction effect size (d) of
.60 for all studies (standard error [SE] = .10; p < .001; 95% CI = 0.45-0.53; Table 3.2).
Such an effect size is considered to be medium to large (Cohen, 2008) and indicates
that about 23% of the variance among outcomes was a result of the interaction. A uni­
variate analysis on weighted scores proved to be significant as well (_F[1, 16] = 4.87; p <
.05). This result supports the effect of patient x treatment matching or fit in optimizing
outcomes; specifically, internalizing patients fare better in psychotherapy with
insight-focused therapy and externalizing patients fare better in symptom-focused
intervention.
 78 psy c h o th er a py rela tio n sh ips that w ork

The mean effect size for the interaction among the six new studies was .49. This
compared with the mean of the 2011 studies, .55. Thus the support for the matching
hypothesis continues but has decreased in newer studies. However, it is difficult to
make a firm conclusion given the variability in the manner in which the interaction
was investigated.

More Comparisons of Old and New Studies

Head-to-head comparisons of the main effects of CS and therapy focus on treatment
outcomes (change in d) between the new and the older studies proved problematic due
to the small number of studies. Only 4 of the original 12 studies reported a main ef­
fect for CS, and only 3 of the 6 new studies included this information, threatening the
usefulness of statistical comparisons. Likewise, eight studies produced information on
how therapy focus related to outcome in the older studies, and only four did so among
the newer studies.
Thus we resorted to group means (Table 3.2) and visual comparisons. New studies
produced a mean effect size associated with patient coping styles of (d) .47, while
old studies averaged .72 in terms of CS effects on outcome. The results favored the
use of symptom-focused strategies. The net influence of CS on outcome was .66 (see
Table 3.2).
Similar computations for the influence of therapy focus on outcome, surprisingly,
favored internalizing patients. The fact that internalizing patients did not make a good
fit with symptom-focused interventions, yet were favored if viewed alone, is a curiosity.
The oddity is extended by the further observation that old studies (n = 8) produced
a mean d of .87 as a change measure while new studies (n = 4) produced a mean
change of d of .21. Both of these values favored symptom-focused therapy. The com­
posite group earned a weighted therapy outcome score, based on focus, of (d) .60, also
favoring symptom focus. While the samples are too small for one to draw conclusions
about these differences, the curious mixed effects between CS and focus deserves fu­
ture attention by researchers.
Shifting for a moment to the results of the entire sample, the ds suggested that both
treatment focus (d = .60) and CS (d = .66) exert a direct effect on outcome. The overall
d of .60 favors the use of a symptom-focused treatment over an insight-oriented one in
these studies, and the d of .66 favors the prognosis of an internalizing individual over
an externalizing one.

EVIDENCE FOR CAUSALITY

We consider two lines of evidence to conclude that the match or fit of CS and therapy
focus causally leads to better treatment outcomes. First is the empirical evidence that
the interaction of patient and treatment qualities are moderately related both to good
and poorer clinical outcomes. The meta-analysis produced an effect size of d = .60 for
the interaction effect, supporting a causal impact or moderating effect. This medium
 79 Coping Style

to large effect size was based heavily on studies that used indirect measures of patient
and therapy factors. The degree of fit, therefore, is likely understated.
The second line of evidence is embodied in that 14 of the18 studies in the meta­
analysis were RCTs in which patients were randomly assigned to treatments, and
treatments were conducted independently of one another with many controls to limit
the effect of extraneous variables. RCTs aim to prove a causative link between patient
and treatment patterns, and the consistency of RCT findings in the form of interaction
effects in this research are consistent with their being moderators of change.
In sum, the evidence suggests a causal chain, but a final conclusion cannot be reli­
ably reached. We look forward to additional and well-controlled, prospective studies
to strengthen the conclusions.

LIMITATIONS OF THE RESEARCH

There are several limitations to this meta-analysis. First, the number of studies in the
meta-analysis is still relatively small. Second, most studies in the meta-analysis utilized
indirect measures, such as diagnostic categories (e.g., depression, substance use) as
proxies for client coping styles. Psychotherapy brand names were also used as proxies
for insight and symptom-centered foci. Ideally, each study would have included a di­
rect, individual measure of coping styles, such as M M PI-2 or STS/Innerlife, as well as
direct observational measures to quantify treatment focus (Beutler & Forrester, 2014).
Another limitation is that three-quarters of these studies took place in North
America. This leaves open the possibility that the findings are unique to the United
States or to English-speaking patients and therapists. W hile several studies in Asia and
South America have pointed to the generalizability of the findings, unfortunately, only
one of these studies included a direct assessment of treatment factors in optimizing
outcomes (Johannsen, 2009). These meta-analytic results require further research,
both in the United States and in other cultures or countries, using large sample sizes
with direct measures of coping styles and treatment focus. We consider several of these
limitations as we discuss diversity issues broadly.

DIVERSITY CONSIDERATIONS
Differences in non-Western and even non-English speaking Western cultures might
limit the generalizability of these meta-analytic results. Cross-cultural matching re­
search has been sparse and, when conducted, has rarely included culturally diverse US
samples. Cultural contexts have a great impact on how a client interacts with the world
and with psychotherapy (e.g., Beutler, 2009; Norcross, 2011; Sue, 2002). If countries
have values and traditions far removed from Western cultures, generalization of US
findings to these countries may prove difficult.
Coping styles may not mean the same thing abroad as they do in the United States.
For example, in many Eastern cultures, attribution styles are more likely to include
self-blame (internalizing CS) than in Western cultures (e.g., Kim, 2002; Kitayama
 80 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

et al., 2006); people in Eastern societies have historically tended to blame themselves
for mistakes and accord successes to others at a much higher rate than in the United
States.
Recent research on the distribution of CS scores suggest that some Eastern cultures,
such as the Japanese, Taiwan, Mainland China, and Korean, are beginning to adopt
both Western and Eastern values. As this transition proceeds, the meta-analytic results
and their clinical implications may become more easily generalized to Eastern cultures.
What can be said from the available research is that, at least among Spanish speakers
in Argentina and Spain, CS can be translated without losing reliability (Corbella et al.,
2003). In Argentina and Switzerland, matching patterns parallel what has been found
among US participants (Beutler, 2009; Beutler, Mohr, et al., 1991). In these countries,
patient CS seems to follow the same interactive parameters as is true in the United
States (Corbella et al., 2003). It has also been found that the patterns of influence apply
to men and women, but insufficient research has been conducted on clients of diverse
sexual orientations, religious commitments, and other intersecting dimensions of cul­
ture. Psychotherapy fit must consider individual differences in coping styles within
multiple cultures.

TRAINING IMPLICATIONS
Training is the key to turning research findings on treatment fit into clinical practices.
Systematic training on matching patient CS to therapists’ focus has been slow to de­
velop. Training programs seem to interpret patient-treatment fit as pertaining solely to
making treatment culturally sensitive via cultural adaptations.
Initial findings of training students to apply STS m atching dimensions have
been positive. STS principles, including CS matched to treatm ent focus, have been
shown to be more effective than usual training procedures (Holt et al., 2015).
A quasi-experim ental study used supervision as usual or STS assisted supervi­
sion among second- and third-year graduate student clinicians to examine the
outcomes o f their patients. The patients o f STS assisted supervisees achieved better
outcomes that the patients o f supervision as usual; the effect size difference was a
d of .67 (Stein, 2016).
STS training teaches participants to provide insight-oriented treatments to
internalizing patients and symptom-oriented treatments to externalizing clients. One
still emerging study has applied a pre-post analysis to treatment as it incorporates
this principle into an ongoing therapy. The results support the added value of treating
the patient x therapist fit as a moderator. The treatment program in this study pro­
ceeds in stages, with a principle inserted at various points along the way. The following
describes training designed to fit treatment focus to patient CS.
Training begins with a lecture and video demonstration of the two patient coping
styles—internalizing and externalizing. Then the student is introduced to a cloud-based
assessment procedure for determining coping styles. Again, with video demonstrations
 81 Coping Style

and lectures, students are taught the difference between symptom-focused treatments
and insight-focused treatments. These techniques are practiced with one or more
patients in the student’s caseload over a period of one to two weeks before the student
connects the treatment to the patient’s CS. During this time, students are given feed­
back about their proficiency with each type of treatment. The supervisor sets goals to
help each student improve. When judged proficient by their supervisor, students are
assigned to do an intake on a new case and to use the assessment for identifying the
patient’s CS. Students then provide a treatment plan that addresses the patient’s CS and
tailors the psychotherapy focus accordingly.

THERAPEUTIC PRACTICES
A client’s CS can guide therapists in applying treatments that produce optimal psycho­
therapeutic outcomes. Here we offer suggestions for therapeutic practice arising from
the meta-analytic research on coping styles.

♦ Assess patients’ coping styles in reviewing their life history and conducting intake
assessments.
♦ Develop a conceptual understanding of clients’ CS in stressful or aversive situations.
♦ Use symptom-focused treatments, such as behavioral or cognitive behavioral
psychotherapies, with externalizing patients.
♦ Use insight or relationship-oriented psychotherapies with internalizing patients.
♦ Develop competency in both symptom- and insight-focused treatments to optimally
match the needs of more clients.
♦ Be aware of client preferences and other transdiagnostic factors that can effectively
guide treatment selection.
♦ Tailor treatment to a given client following research leads.

REFERENCES
References m arked with an asterisk indicate studies included in the m eta-analysis.
*Barber, J. P, & Muenz, L. R. (1996). The role o f avoidance and obsessiveness in m atching
patients to cognitive and interpersonal psychotherapy: Em pirical findings from the
Treatm ent for Depression Collaborative Research Program. Journal o f Consulting and
Clinical Psychology, 64, 9 5 1 -9 5 8 . https://www.doi.org/10.1037/0022-006X.64.5.951
Berggraf, L., Ulvenes, P G., Oktedalen, T., Hoffart, A., Stiles, T., M cCullough, L., & Wampold,
B. E. (2014). Experience o f affects predicting sense o f self and others in short-term dy­
nam ic and cognitive therapy. Psychotherapy, 51, 2 4 6 -2 5 7 .
*Beutler, L. E. (2009). M aking science m atter in clinical practice: Redefining psychotherapy.
Clinical Psychology: Science and Practice, 16, 3 0 1 -3 1 7 .
*Beutler, L. E., Engle, D., Mohr, D., Daldrup, R. J., Bergan, J., M eredith, K., & Merry, W. (1991).
Predictors o f differential response to cognitive, experiential, and self-directed psychother­
apeutic techniques. Journal o f Consulting and Clinical Psychology, 59, 3 3 3 -3 4 0 .
 82 psy c h o th er a py r ela tio n sh ips that w ork

Beutler, L. E., & Forrester, B. (2014). W hat needs to change: M oving from research in ­
form ed practice to em pirically effective practice. Journal o f Psychotherapy Integration, 24,
1 6 8 -1 7 7 .
*Beutler, L. E., Forrester, B., G allagher-Thom pson, D., Thom pson, L., & Tom lins, J. B. (2012).
Com m on, specific, and treatm ent fit variables in psychotherapy outcome. Journal o f
Psychotherapy Integration, 22, 2 5 5 -2 8 1 .
Beutler, L. E., Harwood, T. M ., Kimpara, S., Verdirame, D., & Blau, K. (2011). Coping style. In J.
C. Norcross (Ed.), Psychotherapy relationship that work (2nd ed., pp. 3 3 6 -3 5 3 ). New York,
NY: O xford University Press.
Beutler, L. E., M achado, P. P. P., Engle, D., & M ohr, D. (1993). Differential patient x treatm ent
m aintenance aim ing cognitive, experiential, and self-directed psychotherapies. Journal o f
Psychotherapy Integration, 3, 15-3 1 .
*Beutler, L. E., Moleiro, C., Malik, M ., Harwood, T. M ., Rom anelli, R., Gallagher-Thom pson,
D . , & Thom pson, L. (2003). A com parison o f the Dodo, EST, and ATI indicators among
co-m orbid stimulant dependent, depressed patients. Clinical Psychology & Psychotherapy,
10, 6 9 -8 5 .
Beutler, L. E., & M oos, R. H. (2 0 0 3 ). C oping and coping styles in personality and treat­
m ent planning: Introd uction to th e special series. Journal o f Clinical Psychology, 59,
1 0 4 5 -1 0 4 8 .
Beutler, L. E., Mohr, D. C., Grawe, K., Engle, D. & M acDonald, R. (1991). Looking for differ­
ential effects: Cross-cultural predictors o f differential psychotherapy efficacy. Journal o f
Psychotherapy Integration, 1, 1 2 1-142.
Beutler, L. E., M oos, R. H., & Lane, G. (2003). Coping, treatm ent planning, and treatm ent out­
com e: Discussion. Journal o f Clinical Psychology, 5 9 , 1 1 51-1167.
Beutler, L. E., & W illiam s, O. B. (1999). Systematic treatment selection. Ventura, CA: C enter for
Behavioral Health Technology.
Beutler, L. E., W illiam s, O. B., & Norcross, J. C. (2009). Systematic Treatm ent Selection/
Innerlife [Software program ]. Palo Alto, CA. Retrieved from http://www.innerlife.com
Borenstein, M ., Hedges, L., Higgins, J., & Rothstein, H. (2009). Introduction to meta-analysis.
W est Sussex, England: Wiley.
Butcher, J. N. (1990). The MMPI-2 in psychological treatment. New York, NY: Oxford
University Press.
Butcher, J. N., Beutler, L. E., Harwood, T. M ., & Blau, K. (2011). The M M P I-2. In T. M.
Harwood, L. E. Beutler, & G. G roth-M arnat (Eds.), Integrative assessment o f adult person­
ality (3rd ed., pp. 1 5 2 -1 8 9 ). New York, NY: Guilford.
*Calvert, S. J., Beutler, L. E., & Crago, M. (1988). Psychotherapy outcomes as a function of
therapist-patient m atching on selected variables. Journal o f Social and Clinical Psychology,
6, 1 0 4 -1 1 7 .
Cohen, B. (2008). Explaining psychological statistics (3rd ed.). New York, NY: Wiley.
Corbella, S., Beutler, L. E., Fernandez-Alvarez, H., Botella, L., M alik, M. L., Lane, G ., &
Wagstaff, N. (2003). M easuring coping style and resistance am ong Spanish and Argentine
samples: Development o f the System atic Treatm ent Selection Self-Report (ST S-SR ) in
Spanish. Journal o f Clinical Psychology, 5 9 , 9 2 1 -9 3 2 .
Costa, P. T., & M cCrae, R. R. (1985). The NEO Personality Inventory manual. Odessa,
FL: Psychological Assessm ent Resources.
 83 Coping Style

Costa, P. T. Jr., & M cCrae, R. R. (1989). The NEO-PI/NEO-FFI manual supplement. Odessa,
FL: Psychological Assessm ent Resources.
Elkin, J., Shea, M. T., W atkin, J. T., Imber, S. D., Stsky, S. M ., Collins, J. F., . . . Partoff, M. B.
(1989). N IM H Treatm ent o f Depression Collaborative Research Program. G eneral effec­
tiveness o f treatments. Archives o f General Psychiatry, 4 6 , 9 7 1 -9 8 2 .
Eysenck, H. J. (1947). Dimensions o f personality. London, England: Routledge & Kegan Paul.
Fisher, D., Beutler, L. E., & W illiam s, O. B. (1999). M aking assessment relevant to treatm ent
planning: The ST S C linician Rating Form . Journal o f Clinical Psychology, 55, 8 2 5 -8 4 2 .
Freud, S. (1938). An outline o f psychoanalysis. Standard Edition, 2 3 , 1 3 9-207.
Freud, S. (1954). The origins o f psycho-analysis: Letters to Wilhelm Fliess, drafts and notes: 1887­
1902. New York, NY: Basic Books.
Harwood, T. M ., Beutler, L. E., W illiam s, O. B., & Stegm an, R. S. (2011). Identifying treatm ent­
relevant assessment: Systematic Treatm ent Selection. In T. M. Harwood, L. E. Beutler, &
G. G roth-M arnat (Eds.), Integrative assessment o f adult personality (3rd ed; pp. 6 1 -7 9 ).
New York, NY: Guilford.
Holt, H., Beutler, L. E., Kimpara, S., Macias, S., Haug, N. A., Shiloff, N., . . . Stein, M. (2015).
Evidence-based supervision: Tracking outcom e and teaching principles o f change in clin ­
ical supervision to bring science to integrative practice. Psychotherapy, 52, 1 8 5-189.
Hubers, M . J. H., Cohen, Z. D., Lem m ens, L. H. J. M ., Arntz, A., Peeters, F. P. M . L., Cuijers,
P., & DeRubeis, R. J. (2015). Predicting optim al outomes in cognitive therapy or inter­
personal psychotherapy for depressed individuals using the personalized advantage index
approach. PLoS One, 10, e0140771. https://www.doi.org/10.1371/journal.pone,0140771
*Kadden, R. M ., Cooney, N. L., Getter, H., & Litt, M. D. (1989). M atching alcoholics to coping
skills or interactional therapies: Posttreatm ent results. Journal o f Consulting and Clinical
Psychology, 57, 6 9 8 -7 0 4 .
Kagan, J. (1998). Galen’s prophecy: Temperament in human nature. New York, NY: Basic Books.
*K arno, M ., Beutler, L.E., & Harwood, T.M. (2003). Interactions between psychotherapy p ro­
cess and patient attributes that predict alcohol treatm ent effectiveness: A prelim inary re­
port. Addictive Behaviors, 2 7 (5 ), 7 7 9 -7 9 7 .
Kim , H. S. (2002). We talk, therefore we think? A cultural analysis o f the effect o f talking
on thinking. Journal o f Personality Social Psychology, 83, 8 2 8 -8 4 2 . https://www.doi.org/
10.1037/0022-3514.83.4.828
Kimpara, S., Henderson, L., & Beutler, L. E. (2007). An archival study o f the internalizing and
externalizing coping style and feelings o f shame and guilt among shy individuals in psycho­
therapy (D octoral dissertation). Palo Alto University, Palo Alto, CA.
Kitayama S., M esquita B., & Karasawa M . (2006). Cultural affordances and em otional ex­
perience: Socially engaging and disengaging em otions in Japan and the United States.
Journal o f Personality Social Psychology, 91, 8 9 0 -9 0 3 . https://www.doi.org/ 10.1037/
0 0 2 2 -3514.91.5.890
*K nekt, P , Lindfors, O., Hrakanen, T., Valikoski, M ., Virtala, E., & Laaksonen, M . A. (2008).
Random ized trial o f the effectiveness o f long-term and short-term psychodynam ic psy­
chotherapy and solution-focused therapy on psychiatric symptoms during a 3-year follow
up. Psychological Medicine, 3 8 , 6 8 9 -7 0 3 .
*Leichsenring, F., Salzer, S., Jaeger, U., Kachele, H., Kreische, R., Leweke, F., . . . Leibing, E.
(2009). Short-term psychodynam ic psychotherapy and cognitive-behavioral therapy in
 84 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

generalized anxiety disorder: A randomized, controlled trial. The American Journal o f

Psychiatry, 166, 8 7 5 -8 8 1 .
Lipsey, M . W., & W ilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: SAGE.
*Litt, M . D., Babor, T. F., D elBoca, F. K., Kadden, R. M ., & Cooney, N. L. (1992). Type o f alco­
holics: II. Application o f an em pirically derived typology to treatm ent matching. Archives
o f General Psychiatry, 49, 6 0 9 -6 1 4 .
*Longabaugh, R., Rubin, G .M ., Malloy, P , Beattie, M ., Clifford, P. R., & Noel, N. (1994).
D rinking outcom es o f alcohol abusers diagnosed as antisocial personality disorder.
Alcoholism: Clinical and Experimental Research, 18, 7 7 8 -7 8 5 .
M acCrae, R. R., Harwood, T. M ., & Kelly, S. I. (2011). The N EO Inventories. In T. M. Harwood,
L. E. Beutler, & G. G roth-M arnat (Eds.), Integrative assessment o f adult personality (3rd ed.,
pp. 2 5 2 -2 7 5 ). New York, NY: Guilford.
Malik, M . L., Beutler, L. E., G allagher-Thom pson, D., Thom pson, L., & Alim oham ed, S.
(2003). Are all cognitive therapies alike? A com parison o f cognitive and non-cognitive
therapy process and implications for the application o f em pirically supported treatm ents
(ESTs). Journal o f Consulting and Clinical Psychology, 71, 1 5 0-158.
*M ilrod, B., Leon, A .C., Busch, F., Rudden, M ., Schwalberg, M ., Clarkin, J., . . . Shear, M . K.
(2007). A randomized controlled clinical trial o f psychoanalytic psychotherapy for panic
disorder. The American Journal o f Psychiatry, 164, 2 6 5 -2 7 2 .
Norcross, J. C. (2011). Psychotherapy relationship that work (2nd ed.). New York, NY: Oxford
University Press.
Orlinsky, D. E. (2017). U nity and diversity am ong psychotherapies. In A. Consoli, L. E. Beutler,
& B. Bongar (Eds.), Comprehensive textbook o f psychotherapy: Theory and practice (2nd ed.,
pp. 1 1 -3 0 ). New York, NY: Oxford University Press.
*Poulsen, S., Lunn, S., Daniel, S. I. F., Folke, S., M athiesen, B. B., Katznelson, H., & Fairburn,
C. G. (2014). A randomized controlled trial o f psychoanalytic psychotherapy or cognitive-
behavioral therapy for bulim ia nervosa. The American Journal o f Psychiatry, 1 7 1 , 1 09-116.
Richtberg, S., Jakob, M ., Höfling, V., & Weck, F. (2017). Patient characteristics and patient
behavior as predictors o f outcom e in cognitive therapy and exposure therapy for hypo­
chondriasis. Journal o f Clinical Psychology, 73, 6 1 2 -6 2 5 . https://www.doi.org/10.1002/
jclp.22356
Shapiro, D. (1965). Neurotic styles. New York, NY: Basic Books.
*Stangier, U., Schram m , E., Heidenreich, T., Berger, M ., & Clark, D. M. (2011). Cognitive
therapy vs interpersonal psychotherapy in social anxiety disorder: A randomized
controlled trial. Archives o f General Psychiatry, 6 8 , 6 9 2 -7 0 0 .
Stein, M . (2016). The impact o f a common factors, principle-based supervisory approach on
treatment outcomes at a psychology training clinic (Unpublished doctoral dissertation). Palo
Alto University, Palo Alto, CA.
Strupp, H. H., & Binder, J. L. (1984). Psychotherapy in a new key. New York, NY: Basic Books.
Sue, D. (2002). M ulticultural therapy. In M . Hersen & W Sledge (Eds.), Encyclopedia o f psycho­
therapy (Vol. I, pp. 1 6 5 -1 7 3 ). Amsterdam , The Netherlands: Academic Press,
W ampold, B. E., & Imel, Z. E. (2015). The great psychotherapy debate: the evidence fo r what
makes psychotherapy work (2nd ed.). New York: Routledge.
 85 Coping Style

W ilson, D. B. (2005). M eta-analysis m acros for SAS, SPSS, and Stata [Com puter software].
Retrieved from http://mason.gmu.edu/~dwilsonb/ma.html
*W ilson, G. T., Fairburn, C. G., Agras, W. S., Walsh, B. T., & Kraemer, H. (2002). Cognitive-
behavioral therapy for bulim ia nervosa: T im e course and m echanism s o f change. Journal
o f Consulting and Clinical Psychology, 70, 2 6 7 -2 7 4 .
 4

C U L T U R A L A D A P T A T I O N S AND
MULTICULTURAL COM PETENCE

Alberto Soto, Timothy B. Smith, Derek Griner,

Melanie Domenech Rodnguez, and Guillermo Bernal

A client’s cultural experiences and background clearly impact the client-therapist re­
lationship, the selection of treatments, the formation of the alliance, and the outcome
of psychotherapy (Casas et al., 2016; Vasquez, 2007). Stated more explicitly, culture
is ubiquitous in the real world and in psychotherapy. It is crucial that mental health
treatments account for client culture (e.g., Yeh et al., 2011).
Clinicians also have their own cultural perspectives that impact the client and psy­
chotherapy; thus framing culture solely as a client characteristic oversimplifies the
presence of culture in the therapeutic encounter. In this chapter, we examine culture
both as a client characteristic and as a practitioner competence. Considering and
attending to culture improves the complex processes and multifaceted outcomes in­
herent in psychotherapy (La Roche & Christopher, 2008).
Over the past four decades, there have been major advances in improving aware­
ness of cultural factors in psychotherapy, with particular emphasis placed on tailoring
treatments. It is well beyond the scope of this chapter (or any one chapter or book) to
discuss all aspects of cultural experiences and backgrounds. For the purposes of this
chapter, we focus specifically on client race and ethnicity. Research has provided ev­
idence for the effectiveness of adapting treatments for racially and ethnically diverse
individuals, families, and groups (e.g., Benish et al., 2011; Griner & Smith, 2006; Smith
& Trimble, 2016).
These findings have examined the extent to which culturally adapted interventions,
as well as therapist cultural competencies, can be utilized to improve client retention
and treatment outcomes. The process of modifying interventions has been commonly
referred to as cultural adaptation . Therapist factors have most often been studied under
the broad umbrella of multicultural competencies. This chapter addresses both cultural
adaptations to treatments and therapist multicultural competence with clients from di­
verse racial and ethnic groups. Both are essential and provide a framework for how to
be with a client, as well as w hat to do to align psychotherapy with the client’s culture(s).

86
 87 Cultural Adaptations and Multicultural Competence

In this chapter, we consider definitions, measures, and examples of both cultural

adaptation and multicultural competence in psychotherapy. We report two meta­
analyses, one on the effects of cultural adaptations for clients from diverse racial/
ethnic backgrounds and one on multicultural competence, and then consider po­
tential moderators of their effects on treatment outcomes. Next, we consider evi­
dence for causality, patient contributions, and limitations of the extant research. The
chapter concludes with diversity considerations, training implications, and therapeutic
practices.

DEFINITIONS

Race and ethnicity are terms that have received extensive attention in the psychological
literature, yet precise and consensual definitions are debated and often the terms are
erroneously used interchangeably (Cokley, 2007; Helms & Talleyrand, 1997; Phinney,
1996). For the purpose of this chapter, we refer to race as the “characterization of a
group of people believed to share physical characteristics such as skin color, facial
features, and other hereditary traits” (Cokley, 2007, p. 225) and ethnicity as “a char­
acterization of a group of people who see themselves and are seen by others as having
a common ancestry, shared history, shared traditions, and shared cultural traits”
(Cokley, 2007, p. 225). Thus race and ethnicity differ in important aspects, with race
being a socially constructed category assigned to groups of people with shared physical
traits or characteristics (e.g., skin color), while ethnicity refers to acceptance of shared
group mores/practices stemming from an individual’s culture of origin, resulting in a
sense of belonging (American Psychological Association [APA], 2003).
Cultural adaptations of psychotherapies are defined as the systematic modification
of an intervention “to consider language, culture, and context in such a way that it
is compatible with the client’s cultural patterns, meanings, and values” (Bernal et al.,
2009, p. 362). Cultural adaptations modify treatment in a manner that reflects cultural
considerations (Cardemil, 2010a, 2010b; La Roche & Lustig, 2010), such as holistic/
spiritual conceptualizations of wellness among North American Indians and Alaskan
Natives (King et al., 2014). For instance, traditional psychotherapy does not typically
inquire regarding ancestral beliefs and connections, engage in cultural rituals, use
cultural metaphors and folktales, explore spiritual wellness, conduct treatment out­
doors, involve trusted family members, or extend sessions beyond an hour or two, yet
such adaptations may be suited to some North American Indian clients (Calabrese,
2008). Such treatment modifications may potentially decrease the fidelity of evidence-
based treatments, but systematic modifications made in collaboration with treatment
developers can protect the integrity of evidence-based treatments (Cardemil, 2010a;
Domenech Rodriguez & Bernal, 2012; Parra-Cardona et al., 2012).
Components of cultural adaptations can be categorized across broad domains, such
as treatment goals or methods. One model, the Ecological Validity Model (Bernal et al.,
1995), provides the following eight domains of psychotherapy that are amenable to
 88 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

cultural adaptations: language, persons, metaphors, content, concepts, goals, methods,

and context.
Language adaptations include oral and written adaptations to communication
(e.g., utilizing preferred/native language, code switching) and utilization of dialects or
culture-specific jargon. An adaptation to persons could include matching a client with
a therapist of the same ethnic or racial background. Metaphor adaptations includes
using cultural idioms to strengthen the understanding of specific components of
treatment and can also refer to the visual inclusion of metaphors, such as cultural art­
work in the therapy room. Content adaptations address cultural values, customs, and
traditions in treatment, such as the cultural values of familismo (importance of strong
family connections) or personalismo (emphasis on interpersonal courtesy, friendli­
ness, and personal relationships) when working with Latinx clients. Treatment goal
and method adaptations refer to aligning goals that are congruent with the culture of
the client (e.g., familial harmony rather than individual goals) and also the methods
or delivery of treatment, such as including family members in treatment or engaging
in culturally-sanctioned healing rituals. Finally, adaptations to context consider social
and political realities that can impact the process of therapy, such as systemic forms
of discrimination or concerns related to immigration status. Thus culturally adapted
therapy considers multiple factors not typically addressed in traditional psychotherapy
(Smith & Trimble, 2016).
Whereas cultural adaptations involve modifications to the treatment itself, multicul­
tural competence refers to the ability of the therapist to actively engage and work effec­
tively with diverse clientele (Chu et al., 2016; Domenech Rodríguez & Bernal, 2012).
The most common definition is based on foundational work that operationalized
competencies in a tripartite model (e.g., Arredondo et al., 1996; Sue, 1998) consisting
of cultural awareness, cultural knowledge, and cultural skills.
Cultural awareness refers to therapists’ ability to recognize both the cultural back­
ground of themselves and their clients, as well as being aware of their own perspectives,
assumptions, and biases that might influence the process of psychotherapy or their
ability to implement efficacious interventions across diverse clientele. Cultural know­
ledge refers to therapist understanding of specific cultural groups and their norms and
unique experiences (e.g., how the tenets of Islam and the traumatic experiences of
civil war impact a refugee family from Syria), as well as the ways in which histor­
ical forms of oppression (e.g., colonialism, unequal access to resources) continue to
impact the assumptions embedded in psychotherapy. Finally, cultural skills refer to
therapists’ ability to actively engage diverse clientele and to modify assessment or treat­
ment methods to better match the cultural needs of clients.
In addition, scholars have suggested two psychotherapy skills that culturally com­
petent therapists can use to provide effective treatment, namely dynamic sizing and
scientific mindedness (Sue, 1998). Dynamic sizing refers to psychotherapists’ ability to
understand the shared commonalities within a specific cultural group (e.g., Somali
communities) while still understanding the individuality and uniqueness pre­
sent in members of that group. Successful implementation of dynamic sizing allows
for the ability to attend to both cultural needs and individuality while avoiding
 89 Cultural Adaptations and Multicultural Competence

the perpetuation of stereotypes, which can easily occur when simple categorical
assumptions or beliefs are held. Scientific mindedness refers to psychotherapists’
ability to manage and test assumptions by gathering information, asking questions
based on hypotheses, and confirming or disconfirming hypotheses in the process of
treatment, rather than applying assumptions to clients. This perspective is not unique
to multicultural therapy, yet the attending to client culture requires the skill to confirm
and disconfirm hypotheses relevant to culture.
However defined, cultural competence has become part of ethical and profes­
sional standards of psychology (e.g., APA, 2017a, 2017b) and other mental health
professions (e.g., American Counseling Association, 2014; National Association of
School Psychologists, 2010). Indeed, multicultural competence is sought in all health­
care professions (e.g., Betancourt et al., 2003; US Department of Health and Human
Services, 2001).

MEASURES

Cultural Adaptations
In primary studies, researchers typically provide a brief narrative explanation of cul­
tural adaptations implemented in the study, without any quantification of the extent or
quality of the adaptations. However, scholars have distinguished between adaptations
developed prior to clinical implementation and created to target a broad population,
such as creating an intervention for African American clients, and local adaptations,
which are adaptations that are made just prior to or during the course of treatment
with a specific population, such as African American clients with diabetes in a par­
ticular community (Barrera et al., 2017). We are unaware of a specific measure that
assesses the degree of cultural adaptations of psychotherapy. Thus cultural adaptations
are quantified using systems of coding at the meta-analytic level.
Some reviews (e.g., Smith & Trimble, 2016) have quantified the extent of cultural
adaptations (Bernal et al., 1995). Other scholars have evaluated the extent of cultural
adaptations based on specific types of adaptations (e.g., language match, racial/ethnic
match), adaptation to explanatory model of illness (illness myth adaptation), adher­
ence to specific model of adaptation, or overall quality of adaptations (Benish et al.,
2011; Domenech Rodriguez & Bernal, 2012; Hall et al., 2016).

Cultural Competence
Several therapist self-report measures of therapist cultural competence exist. The four
measures used most often are the Multicultural Counseling Inventory (MCI; Sodowsky
et al., 1994), the Cross-Cultural Counseling Inventory-Revised (CCCI-R; LaFromboise
et al., 1991), the Multicultural Counseling Awareness Scale (MCAS; Ponterotto et al.,
2002), and the Multicultural Awareness Knowledge Skills Scale (Kim et al., 2003).
Reviews of these instruments have been largely favorable (e.g., Ponterotto et al.,
1994; Pope-Davis & Dings, 1994, 1995). One study (Ponterotto & Alexander,
 90 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

1996) found all four of the instruments to be sensitive to therapist change as a result
of training, although only the MCI and MCAS were found to be sensitive to increased
levels of multicultural experience. A review found all four instruments to be “easy to
administer and score, moderate in length (about 25 minutes to complete), and inex­
pensive” (Boyle & Springer, 2001, p. 65). A subsequent reliability generalization study
(Dunn et al., 2006) of the four measures, as well as five other rarely used measures of
cultural competence, found acceptable psychometric evidence of reliability across 89
studies; however, the authors pointed to inadequate evidence of divergent and con­
struct validity.
At the same time, these measures of therapist cultural competence manifest limi­
tations. First, participant responses are probably impacted by social desirability (e.g.,
Constantine & Ladany, 2000), although one review found that measures of social de­
sirability explained only 4% of the variance in measures of therapist cultural com­
petence (Dunn et al., 2006). Second, the factor structure of the instruments requires
additional explication (e.g., Constantine et al., 2002). Third, the self-report versions ac­
tually measure multicultural therapy self-efficacy, not demonstrated ability to counsel
diverse populations (Constantine et al., 2000, 2002). Fourth, the self-report measures
are unreliably related to observer-rated multicultural competence (e.g., Worthington
et al., 2000) or to clients’ evaluations of therapists’ cultural competence (Smith &
Trimble, 2016). Fifth, the high subscale intercorrelations and low concurrent validity
across measures of multicultural competence have “raised questions about the devel­
opment and definition of the constructs being measured” (Kocarek et al., 2001, p. 494).
Finally, the degree to which therapist self-reported cultural competence remains dis­
tinct from general therapist competence cannot be established using existing measure­
ment (Coleman, 1998; Smith & Trimble, 2016).
Scholars have also examined therapist variability among client outcomes between
diverse and White clients as a function of either therapists’ general competence or
cultural competence. One study demonstrated therapist variability in client outcomes
across all clients, yet some therapists evidenced systematic differences in outcomes
between diverse and White clients, supporting the notion that general competencies
and cultural competencies can be distinguished (Imel et al., 2011). An additional study
demonstrated that, although diverse clients and White clients did not differ in clinical
outcomes at posttreatment, some therapists were more effective, with this variability
being, in part, explained by client race or ethnicity (Hayes et al., 2014). Thus demon­
strable client improvement among diverse clientele is a strong indicator of therapist
cultural competence (Hayes et al., 2016).

CLINICAL EXAMPLES
“Nolan” is a 28-year-old African American male from Minnesota. He is in a committed
relationship and identifies as heterosexual. Spirituality is an important part of his life,
but he is not currently active with his church. Nolan was raised by his mother, who has
struggled with substance dependence, as has his younger brother. He presents to psy­
chotherapy with worsening depression after discovering that his brother and mother
 91 Cultural Adaptations and Multicultural Competence

were using drugs together, an experience that elicited deep feelings of betrayal and
isolation. In addition, Nolan states that the widely publicized shooting of Philando
Castile, and the resultant lack of legal conviction of the involved police officer, have
increased his feelings of depression and hopelessness.

I just feel so hopeless. I mean what am I supposed to do? I found out about my
brother and mom doing heroin together and that just sends me to this dark place.
I grew up with that crap all my life. I felt so alone growing up and I had to look out
for me and my brother. I learned to cook when my mom dropped us off at some
random crack house back when she was using, and everyone was too high to cook
or order us food. Now for them to be using together . . . it’s like he forgot about all
that crap she put us through. I feel alone, I feel scared just like I did when I was
a kid. Then you see this shit in the news about cops killing black people. It’s like
nobody gives a damn about you if your skin isn’t white. I don’t know whether to
feel angry or just feel more like giving up. I’ve tried therapy before . . . I don’t know
what can be done to make anything better.

As psychotherapy begins, his clinician wrestles with two aspects of Nolan’s experi­
ence that contribute significantly to his distress. The first is associated with the fear of
losing his family to substance dependence, a fear that has shaped Nolan’s past. Opiate
dependency is not unique to culture, but its impact on an individual from a more col-
lectivistic culture can exacerbate the feelings of loss. This feeling of familial loss or of
disconnection from one’s group can also parallel historical trauma associated with co­
lonialism and slavery (Cross, 1998; Eyerman, 2004; Okazaki et al., 2008). The second
aspect is Nolan’s experience as an African American man and how the recent police
shootings have had a negative impact on him.
As an initial step in adapting treatment, Nolan’s clinician examined the available
literature, consulted with colleagues with greater experience working with African
American men, asked for insights from African American community leaders, and
solicited feedback from Nolan regarding what aspects of his culture proved important
to him. Nolan expressed that he valued collectivism and struggled to fit in with others
who are not part of his cultural group. His therapist validated this cultural value and
encouraged Nolan to seek support from others in his life, such as his maternal grand­
mother, who Nolan reported has been a significant person to him throughout his life.
Nolan also indicated that, in the past, his spiritual faith had been exceptionally helpful
for him and discussed the possibility of once again being involved with his church.
Encouraging Nolan to utilize social support and his spiritual beliefs may appear to be
minor modifications, but these represent fundamental shifts from an individualistic
stance to a collectivist one.
The therapist also engaged in open discussion of police brutality, recent political
changes, and Black Lives Matter as they related to Nolan’s feelings of distress and isola­
tion. Topics such as police shootings, racism, poverty, and other forms of systemic op­
pression are social realities rarely broached in psychotherapy, but obviously influence
the daily lives of many diverse clients.
 92 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

During the course of treatment, both the context and content of therapy were
adapted to Nolan’s experience as an African American man. His therapist introduced
Afrocentric literature (e.g., Asante, 2009) in psychotherapy and asked Nolan whether
he would like to use an Afrocentric approach in their work together. After Nolan
expressed that he would, his therapist used concepts and principles associated with
NTU psychotherapy (Phillips, 1990), an Afrocentric approach to psychotherapy that
is spiritually based and founded in historic African and Afrocentric worldviews.
Using this culturally adapted model, the therapist helped Nolan explore his family
relationships, his spirituality, his connection to ancestors, and work toward the five
core NTU principles: harmony, balance, interconnectedness, cultural awareness, and
authenticity.
Multicultural competencies attend to the unique cultural backgrounds of clients
employing the requisite awareness, knowledge, and skills. In the following, we con­
sider three such competencies applied with Nolan, by his therapist. Lindsey, a 30-
year-old White American female who also grew up in Minnesota. She is empathically
attuned to Nolan’s mood as he discusses his concerns. Although she has worked with
diverse clientele, she is aware that she struggles to grasp central features of multicul­
tural competencies. The privileges proffered to her as a result of her racial background
and middle-class upbringing often bring a sense of guilt and shame as she engages in
psychotherapy with diverse clients. Lindsey is also aware that, growing up, the notions
of meritocracy and equality were presented as self-evident truths, yet her experiences
as a woman and working with diverse clients clash with those values. As she begins
treatment with Nolan, she thus understands that many of her worldviews and personal
beliefs conflict with Nolan’s experiences.
In terms of multicultural competence, Lindsey attempts to acknowledge potential
biases, blind-spots, or values that might interfere with her connection with Nolan.
In addition, she actively solicits feedback related to race, ethnicity, and culture. She
attends to personal cultural connections, remaining aware of how she has benefited
from White privilege (McIntosh, 1988) and how systemic forms of oppression have
negatively impacted her and others. Thus, as a clinician, Lindsey proactively identifies
aspects of her own identity and also of the treatment itself that might impact Nolan
and their work together.
As well in terms of multicultural knowledge, Lindsey understands the tendency of
psychotherapy to advance and perpetuate Eurocentric values. She is actively engaged
in social justice organizations, as well as professional organizations dedicated to mul­
ticultural services. In short, she is committed to ensuring that she remains knowledge­
able of others’ experiences and worldviews.
Lindsey recognizes that she has many limitations regarding her knowledge of
how the African American community reacted to the death of Philando Castile in
Minneapolis. She also recognizes that, while she has worked with African American
clients in the past that, there is much diversity within groups, and she wants to avoid
advancing stereotypical thinking or categorical assumptions. Lindsey starts to read
research on perceived racism, racial trauma, and culturally adapted treatments (e.g.,
Comas-Diaz, 2006; Smith et al., 2015). Lindsey also asks Nolan if he can help her
 93 Cultural Adaptations and Multicultural Competence

understand aspects of his culture that she may overlook. This is done not with the ex­
pectation that Nolan will teach her everything she needs to know (e.g., “Tell me . . .”),
but rather with the intention of creating an empathic and respectful dialogue (e.g.,
“Can you help me understand what I am missing?”).
Lindsey utilizes her cultural competency skills to effectively work with Nolan and
to incorporate culturally appropriate interventions. These overlap with other thera­
peutic skills, such as forming a strong alliance and repairing ruptures. In response to
Nolan’s initial statement about his experience hearing about Philando Castile, Lindsey
expressed:

Nolan, I hear your profound loneliness. While police officers kill Black men in
broad daylight and return to their lives without much commotion, the people you
love and who would understand your grief about Philando Castile’s death are not
there for you because they are caught up in their own problems, often related to
racism. Can you tell me more about your emotional reactions to the shooting?
(Brief discussion). Thank you for sharing your particular fears and also the im­
portance of feeling connected with the Black community. Those are essential to
our work together. I will do my best to understand, but would you be comfortable
letting me know when I am not “getting it”?

Lindsey’s multicultural competencies provide a framework for effectively attending

to Nolan’s cultural experiences. She communicates her personal awareness (while not
turning the therapeutic encounter into one that focuses on her needs), solicits more
information from Nolan and others about working with African American clients,
and addresses therapist and client differences. She thereby empathically responds and
engages in conversation about Nolan’s experiences with his family and with racism.
Her work demonstrates both cultural adaptations and cultural competence.

LANDMARK STUDIES

Cultural Adaptations
Many early studies of culturally adapted mental health treatments involved single
group, pre- to posttest designs that were subject to numerous sources of bias (Smith,
2010). The following landmark studies are exemplars that utilized rigorous research
designs, that applied treatments specific to populations that are historically overlooked,
such as refugees, and that compared interventions already accounting for primary cul­
tural consideration, such as active family involvement.
Rossello and Bernal (1999) used an experimental design to investigate the ef­
ficacy of a culturally adapted treatment for 71 adolescent Puerto Rican clients who
met diagnostic criteria for depression. Clients were randomly assigned to a waitlist
group, culturally modified cognitive-behavioral therapy (CBT), or culturally modi­
fied interpersonal psychotherapy (IP). Clients engaged in individual sessions for 12
weeks, with both CBT and IP adapted to incorporate Latinx values. Multiple outcome
 94 P S Y C H O T H E R A P Y R EL AT IO N S H IP S THAT WORK

measures were administered pretreatment, posttreatment, and at three months follow­

up. Clients in both CBT and IP showed significant increases in social adaptation and
self-esteem and also significant reductions in depressive symptoms relative to waitlist
participants. This landmark study demonstrated the acceptability of conducting
randomized controlled trials (RCTs) with cultural adaptations and also documented
their effectiveness over and above a control group.
Hinton and colleagues (2005) conducted a RCT to examine the efficacy of cul­
turally adapted CBT for trauma. The 40 patients were refugees who had survived
the Cambodian genocide (1975-1979) and met criteria for treatment-resistant
posttraumatic stress disorder (PTSD; including prior treatment and at least one year
of using an antidepressant medication). Using a repeated measure crossover design,
patients were stratified by gender and randomly assigned to an initial treatment
or delayed treatment condition (waitlist control). Over a 12-week period, active-
treatment participants received weekly individual therapy in which the therapist,
who was fluent in Cambodian, administered the culturally modified CBT. All patients
completed a number of measures relevant to PTSD that had been translated and back-
translated to ensure culturally appropriate language. Results indicated that those in­
volved in initial treatment benefitted significantly (a large effect size) compared to the
waitlist control. These results are particularly noteworthy given that these patients
had not responded to prior supportive therapy or medication regimen. This study
also proved to be consequential because of the comprehensive nature of the psycho­
logical treatment, the culture-sensitive outcome measures, and attention to refugee
patients.
Using multiple strategies to design a cultural adaptation, McCabe and colleagues
(2005) modified Parent-Child Interaction Therapy (PCIT) to be culturally appropriate
for Mexican American families. The researchers conducted focus groups with 24
Mexican American therapists and parents of youth with behavior disorders to design
the treatment. The resulting adapted treatment, Guiando A Niños Activos, was found
to be superior to treatment as usual in a RCT (McCabe & Yeh, 2009) but equivalent to
traditional PCIT on multiple indicators. This pivotal study raised the possibility that
cultural adaptations may prove superior to traditional interventions that already ac­
count for the primary cultural considerations, which in this instance was active family
involvement in treatment.

Cultural Competence
Although many publications emphasize the need for therapist multicultural compe­
tence, relatively few studies have been conducted with actual clients. The following
landmark studies involved actual clients and provided evidence regarding the validity
of cultural competencies, distinct from general therapy competencies. The studies also
evaluated components of cultural competencies, such as skills and knowledge.
Pope-Davis and colleagues (2002) reported a qualitative study using grounded
theory to understand clients’ perceptions o f therapists’ multicultural competence.
 95 Cultural Adaptations and Multicultural Competence

They interviewed 10 university student clients who had worked with a therapist
from a culture different from their own. The clients reported having discussed cul­
tural issues with their therapists when they felt the issues were important, as well
as when they perceived the therapist to be open to multicultural considerations.
Clients considered multicultural competence as pivotal in many instances, with
several clients reporting frustration when their therapist did not demonstrate such
competence. This study found that therapists’ multicultural competence impacted
clients’ disclosures, processes, and outcomes in psychotherapy and initiated other
investigations.
In another influential study, Constantine (2002) examined the relation between
client satisfaction and client perceptions of their counselors’ multicultural compe­
tence, after statistically accounting for general counseling competence. Data came
from 112 racial and ethnic minority college students in psychotherapy at their campus
counseling centers. After termination, the clients completed evaluations of the mental
health treatment, perceived counselor competence, perceived counselor multicultural
competence, and overall client satisfaction. Clients’ ratings of their counselors’ mul­
ticultural competence explained significant variance in client satisfaction, above and
beyond that explained by clients’ attitudes toward mental health treatment and general
counseling competence. This landmark study suggested that therapist multicultural
competence could be differentiated from general counseling competence when con­
sidering client satisfaction.
In another pivotal study, Constantine (2007) conducted focus groups on the impact
of microaggressions with 24 African American clients receiving treatment in a college
counseling center. Utilizing themes that emerged from these focus groups, Constantine
developed the Racial Microaggressions in Counseling Scale. This scale, along with
a number of other measures, was administered to 40 African American undergrad­
uate clients at termination of their therapy with European American practitioners.
A path analysis indicated that clients’ perceptions of racial microaggressions not only
decreased their satisfaction in counseling but also significantly weakened the working
alliance. Multicultural competence proved a more robust predictor of clients’ satisfac­
tion than general counseling competence.
In a landmark series of four investigations, Hook and colleagues (2013) focused at­
tention on cultural humility, including the therapists’ openness to cultural differences.
They developed a brief client-rated measure to assess therapists’ cultural humility and
used that measure in subsequent studies. Participants in three of the four studies were
college students from a variety of racial/ethnic backgrounds who had either attended
therapy in the past or were currently attending therapy. Participants in the fourth study
were Black individuals attending therapy recruited from a crowdsourcing website.
Results of the four studies provided psychometric evidence for their measure and also
indicated that clients rated therapist cultural humility as more important than other
therapist characteristics, including similarity, skills, knowledge, and experience. Client
perceptions of therapist cultural humility were also positively related with a strong
working alliance as well as therapeutic improvement.
 96 psy c h o th er a py rela tio n sh ips that w ork

RESULTS OF PREVIOUS META-ANALYSES

Cultural Adaptations
There have been numerous meta-analytic reviews of culturally adapted mental health
treatments. The first of these meta-analyses demonstrated the overall effectiveness
of these interventions (Griner & Smith, 2006). A later meta-analysis, specific to
interventions with youth (e.g., Huey & Polo, 2008), reported no difference between
adapted and nonadapted interventions. These findings deserve consideration given
the conclusion of null differences that differ from numerous other reviews. One con­
sideration is that cultural adaptations are equally as effective as treatment as usual;
alternately, the equivalency between interventions may have been due to the limited
information available about the interventions enacted or to the characteristics of the
youth sample, which was likely to be more highly acculturated to US values and norms
as opposed to older adults.
Several subsequent meta-analyses have found clear benefits to primarily adult
clients receiving culturally adapted treatments (e.g., Hall et al., 2016; Smith et al.,
2011; Smith & Trimble, 2016; van Loon et al., 2013). Meta-analytic reviews sup­
port the relative efficacy of culturally adapted interventions compared to preven­
tion studies (Hall et al., 2016), the importance o f focusing on specific rather than
broad adaptations (Sm ith et al., 2011), and the benefit of making more rather than
fewer cultural adaptations (Sm ith et al., 2011; Sm ith & Trimble, 2016). In addi­
tion, one review demonstrated the superior efficacy of adapted treatments even
when compared to bona fide interventions, as well as examining which types of
adaptations (such as language match) were more effective and establishing the im ­
portance of adapting the intervention to the explanatory model, or myth, of illness
(Benish et al., 2011).
There are numerous challenges in conducting and evaluating treatments that
have been culturally adapted. These complexities are reflected in conflicting results
regarding the efficacy o f modified interventions, both among individual studies as
well as at the meta-analytic level. Cultural adaptations vary greatly in terms of the
breadth and depth of the modifications, as well as the adherence to extant guidelines.
Additional complications can arise as a result of unclear rationale for modifications
and methodological concerns (e.g., small sample size, researcher allegiance, single
group designs). At the meta-analytic level, reviewers do not always specify the com ­
parison groups that are utilized in the studies (e.g., waitlist, treatment as usual), such
that results reported across the aggregate may not encapsulate the literature in a clear
manner.
An additional complication is that cultural adaptation often entails many factors in
psychotherapy beyond client characteristics. Given that scholars have suggested that
extra-therapeutic and relationship factors account for the majority of the variance of
therapeutic outcome (Lambert, 2013), cultural adaptation may further promote effi­
cacy by attending to both process and relational variables associated with outcome
(e.g., Norcross & Wampold, 2011).
 97 Cultural Adaptations and Multicultural Competence

Cultural Competence
To date, there have been two meta-analytic reviews on multicultural competencies in
psychotherapy. The first review examined the association between client-rated multi­
cultural competencies and psychotherapy process variables (e.g., working alliance, real
relationship, session depth), in addition to client outcomes (Tao et al., 2015). In this re­
view, 18 studies met inclusion criteria, with the clients being primarily women (68%),
averaging 24 years of age, with an average of 80% of clients being racially or ethnically
diverse. All studies included utilized correlational research designs, with client-reported
therapist multicultural competencies being correlated with the dependent variable
(process variables or client outcomes). Effect sizes were characterized by substantial
heterogeneity, with moderate to large effect sizes reported across variables: therapeutic
alliance, r = .61 (k = 16); client satisfaction, r = .72 (k = 5); and general counseling
competencies, r = .62 (k = 7). The average effect size of multicultural competencies
on improved outcomes was moderate, r = .29 (k = 7). This review demonstrated the
overlap between multicultural competencies and several important processes, as well
as with improved client outcomes. This study also provided support for the overlap be­
tween general counseling competencies and multicultural competencies, a distinction
that proves difficult for clinicians, clients, and researchers to make.
The second meta-analysis examined the association of multicultural competencies
across client experiences and outcomes (Smith & Trimble, 2016). Sixteen studies were
identified as meeting inclusion criteria, containing data on 2,025 clients. All studies
included both client and therapist-self reported measures of therapist cultural compe­
tency. The average age for clients in this review was 28 years, with 64% being female and
over 90% being racially or ethnically diverse participants. Across 10 studies the cor­
relation of therapist multicultural competence (both client and therapist self-reports)
with client perceptions of the therapist was r = .50. Across four studies evaluating
client participation in treatment (e.g., premature termination vs. completion), the re­
ported correlation between client and therapist-reported multicultural competencies
and client participation was r = .26. When examining outcome, the association be­
tween client and therapist-reported multicultural competencies and improved client
outcomes was small, r = .16.
This second meta-analysis identified a significant amount of heterogeneity across
studies that was not attributable to publication bias but rather moderating variables.
Specifically, the source of the multicultural competency rating (therapist self-rating
vs. client-rated) accounted for a significant amount of this heterogeneity. Therapist
self-ratings of their own multicultural competencies did not appear to be significantly
associated with client outcomes, but client-rated measures were robust predictors
of all client outcomes. This study demonstrated the importance of multicultural
competencies from the perspective of the client.
In the following sections, we build on these previous meta-analyses and pro­
vide updated reviews on both multicultural competencies and culturally adapted
interventions. The results of the two separate meta-analyses furnish empirical support
for clinicians tailoring their interventions to culture.
 98 p sy c h o th er a py rela tio n sh ips that w ork

METHODS OF THE META-ANALYTIC REVIEWS

Inclusion and Exclusion Criteria

For the first meta-analysis of cultural adaptations, we only included studies that used
experimental and quasi-experimental research designs that evaluated mental health
treatments that were culturally adapted based on considerations for people of color in
the United States or Canada. For the second meta-analysis of cultural competencies,
we included studies that examined the correlation between therapist multicultural
competencies and client outcomes (i.e., clinical improvement and treatment utilization
or retention). Thus our two meta-analyses involved studies with fundamentally different
research designs; the meta-analysis of cultural adaptations included only experimental
or quasi-experimental studies, while the meta-analysis of multicultural competencies in­
cluded cross-sectional and longitudinal evaluations of the association among variables.
The specific criteria for inclusion were as follows: (a) studies were of people of color
with less than 10% non-Latinx Whites or European Americans; (b) data were clin­
ical in nature, excluding analogue data; (c) primary outcomes were psychological in
nature (e.g., depression, obsessive-compulsive disorder, aggression), excluding sub­
stance abuse treatment and outcomes; (d) studies were in English or Spanish and were
conducted in the United States or Canada; and (e) effect sizes could be extracted for the
effectiveness or efficacy of culturally adapted interventions (for the first meta-analysis)
or for the relation between multicultural competencies and client outcomes (for the
second m eta-analysis).

Information Sources and Search Strategies

To locate relevant published and unpublished studies through May 2017, we searched
the following electronic databases: PsycINFO, EMBASE, Academic Search Premier,
and ProQuest Digital Dissertations. To obtain as many studies as possible, we used
dozens of synonymous search words and phrases that were cross-referenced with
one another. For instance, root phrases such as racial/ethnic/cultur* specific/sensi-
tive/adapted were crossed with a list of word roots and phrases relevant to mental
health treatments: client, counsel, treatment, clinic, therapy, psychotherapy, mental
health treatment/service, and so forth using the Boolean OR to include all terms. As
an additional step, we searched the references of all articles meeting preliminary in­
clusion criteria (e.g., containing data) to identify additional studies. Team members
further searched for articles citing the retrieved studies. Solicitations for unpublished
manuscripts were posted on professional listservs.

Data Collection Processes

Our research team, consisting of two university faculty members, two graduate
students, and five undergraduate students, received extensive training on coding
procedures. They coded manuscripts that met our inclusion criteria in pairs to enhance
 99 Cultural Adaptations and Multicultural Competence

the accuracy of coding decisions and data entry. Each article was coded by two sepa­
rate pairs of coders (four total coders per article). The interrater agreement of initial
coding decisions for the first meta-analysis on cultural adaptations averaged .95 using
intraclass correlation coefficients for continuous variables and .70 using Cohen’s kappa
for categorical variables. For the second meta-analysis on multicultural competence,
the coefficients were .89 and .74, respectively. Discrepancies were resolved through
discussion and mutual agreement.
In addition to effect size data, coders extracted information about possible
moderating variables, including participant characteristics (e.g., age, gender, race) and
study characteristics (e.g., study design, sample size). The coding of the meta-analysis
on cultural adaptations included the following participant and study moderating
variables: publication status; year published; sample type (normal community
members, clinical population, at-risk); client mean age; client percentage female; ra-
cial/ethnic percentage for both clients and clinicians; treatment modality (individual,
group, family, combined); treatment type (psychotherapy, prevention program); re­
search design; type of comparison group (no intervention, alternative intervention,
bona fide treatment); racial/ethnic match with therapist; sample size for both treat­
ment and comparison group; type of dependent measure (e.g., global mental health,
symptom specific outcome); source of outcome rating (patient, therapist, external);
and time of outcome.
Additionally, the following unique treatment cultural adaptations were coded for
each article: matching native language therapists with clients; utilizing cultural values/
concepts/examples; multicultural training for therapists; external services provided
beyond psychological intervention (e.g., transportation, child care, home visits); con­
sultation with experts/stakeholders/family; language translation of materials or meas­
ures; reading accommodations; modification of décor to align with diverse clientele;
outreach programming; referrals to outside resources; intervention theoretical basis
described; modification of instrumentation to be culturally congruent; utilization of
metaphors; cultural conceptualization of presenting concern; identifying treatment
goals based on clients’ desires; treatment methods that align with culture; consid­
eration of broader social, economic, political realities; and total number of cultural
adaptations.
The meta-analysis on multicultural competencies coded for the following
moderating variables: manuscript publication status; year of study; percent female
of clients; mean age of clients; client ethnicity; source of multicultural competency
rating (client, therapist, external); and reliability coefficient of multicultural compe­
tence scale. The moderator variables for culturally adapted treatments and multicul­
tural competencies are included in Tables 4.1 to 4.4.

Data Analyses
We used the Meta-Analysis Calculator software (Lyons & Morris, 2017) to transform
statistical estimates derived from a variety of metrics (e.g., t, F, and p values) to Cohen’s
Table 4.1. C haracteristics o f 99 Studies o f Culturally Adapted
Treatm ents

Characteristic M Number o f Studies (k)

Year of report 2003
Before 1980 0
1 9 8 0 -1 9 8 9 7
1 9 9 0 -1 9 9 9 26
2 0 0 0 -2 0 0 9 41
2 0 1 0 -2 0 1 7 25
Publication status
Published 82
Unpublished 17
Research design
Q uasi-experim ental 23
Experim ental 76
Sample type
Com m unity mem bers 12
At-risk clients 51
Clinical clients 36
Sample size 140
<50 40
5 0 -9 9 34
1 0 0 -1 9 9 14
2 0 0 -3 9 9 6
4 0 0 -9 9 9 2
>1,000 3
Age of clients“ 26.8
Children (<13 yrs.) 22
Adolescents (1 3 -1 8 yrs.) 21
Young adults (1 9 -2 9 yrs.) 9
M iddle-aged adults (3 0 -5 5 yrs.) 37
Senior adults (>56 yrs.) 4
Gender of clients (% Female) 62.7
Race of clients’1 (%)
A frican Am erican 15
Asian A m erican 39
Hispanic/Latinx Am erican 40
Native A m erican Indian 5
O ther 1

Note. Not all variables sum to the total number of studies due to missing data.
aAverage age category of participants within studies; not all participants in
a study were necessarily included in the category. bThe racial composition
of participants across all studies, calculated by multiplying the number of
participants within studies by the percentage of participants from each racial
group and dividing that product by the total number of participants
 10 1 Cultural Adaptations and Multicultural Competence

d or Pearson’s r. For the first meta-analysis of cultural adaptations, positive effect sizes
indicated superior efficacy of adapted treatments while negative effect sizes indicated
that the nonadapted interventions were superior. For the second meta-analysis of mul­
ticultural competencies, positive effect sizes indicated improved client outcomes asso­
ciated with therapist multicultural cultural competence, whereas a negative effect sizes
indicated poorer client outcomes were associated with multicultural competence. The
correlational data for this meta-analysis were analyzed after Fisher’s z transformation.
For both meta-analyses, effect sizes were aggregated within studies, with each study
contributing one effect size to the overall analyses. When studies contained multiple
effect sizes (e.g., depression and anxiety outcomes), the multiple effect sizes were
averaged, weighted by the standard error or number of participants. Data analyses
were conducted in STATA using random effects models.

META-ANALYSIS OF CULTURAL ADAPTATIONS

TO TREATMENTS

The Studies
We located 99 studies evaluating a cultural adaptation to a mental health treatment
that met our inclusion criteria, including 21 additional studies beyond those re­
ported in a prior meta-analysis (Smith & Trimble, 2016). Most of the 99 studies were
conducted over the past 20 years, and the vast majority of these studies were published
(Table 4.1). Clients largely represented either children/adolescents or middle-aged
adults, which is reflective of a trend in the literature toward treatments that are ei­
ther prevention oriented (treating at-risk groups) or treatment oriented (examining
groups in community mental health clinics). We included both kinds of treatments in
our meta-analysis because prior meta-analyses have included both, and we analyzed
the differences.
Overall, the studies contained data from 13,813 individuals, with a median of 60
participants. Asian American (39%) and Hispanic/Latinx American (40%) clients were
the most commonly evaluated racial and ethnic groups (Table 4.1). African Americans
and Native American Indian clients were evaluated less frequently and were present in
only 15% and 5% of the studies, respectively.
The types of cultural adaptations reported in individual studies varied substan­
tially, both in terms of the number of adaptations and in terms of the specific types
of adaptations. Some studies clearly followed existing guidelines available in the
professional literature (e.g., Bernal et al., 1995), whereas others provided minimal
descriptions and/or had restricted adaptations that focused on only a few aspects of
culture. Across all studies:

♦ 75% provided treatment in the clients’ preferred language when other than English.
♦ 75% included explicit mention of cultural content/values in treatment.
♦ 55% matched clients with therapists of similar ethnic/racial backgrounds.
♦ 52% addressed clients’ broader contexts (e.g., experiences of racism, employment).
 102 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

♦ 49% modified the methods of delivering psychotherapy (e.g., including cultural

rituals) beyond the other cultural considerations listed here.
♦ 48% developed the cultural adaptations through consultation with individuals from
the culture.
♦ 42% utilized metaphors from client cultures.
♦ 42% adhered to the clients conceptualization of the presenting problem.
♦ 28% reported that mental health staff had received training in the cultural
adaptations.
♦ 16% modified the wording of outcome measures to be culturally appropriate.
♦ 14% explicitly solicited culturally congruent outcome goals from the client.

On average, studies incorporated, or reported utilizing, 3.8 out of the 8 components

proposed by Bernal and colleagues (1995). Only 18 studies (18%) involved more than
five of the eight components. Thus researchers typically made several attempts to cul­
turally adapt treatments; a minority made comprehensive adaptations.
In 76 studies using an experimental design, the effect size represented an estimate
of the effectiveness of culturally adapted treatments, with positive values reflecting the
superiority of the adapted treatment. For 23 studies using quasi-experimental designs
(non-random group composition) and the 45 studies using no-treatment controls
(clients on a waiting list), the magnitude of the effect size estimates was impacted by
factors other than the culturally adapted nature of the intervention. As such, these
studies warranted separate analyses. Given that the nature of the services provided to
at-risk populations, as well as client outcomes and rates of change likely differ rela­
tive to clinical populations, we also considered it essential to distinguish results from
studies using mental health treatments with clinical populations from results of studies
involving prevention-oriented interventions for at-risk populations.

Overall Results
Across all 99 studies examining a culturally adapted mental health treatment (13,813
total participants), the random effects weighted effect size was d = .50 (standard error
[SE] = .039, 95% confidence interval [CI] = 0.42-0.58, p < .0001). The heterogeneity of
the findings was high (I2 = 71.4, 95% CI = 6 5 -7 7 ; Q(98) = 342.1, p < .0001), indicating
that the results tended to be inconsistent across studies (see Figure 4.1).
Across the 21 studies added since the last meta-analysis (Smith & Trimble, 2016),
the random effects weighted effect size was d = .57 (SE = .073, 95% CI = 0.41-0.73, p <
.001), with moderate heterogeneity (I2 = 59.7, 95% CI = 34 -7 5 ; Q(20) = 47.2, p < .001).
These effect sizes did not differ significantly (Q(98) = 1.1, p = .30) from those reported
previously. The 21 new studies had more female clients than the previously reported
studies (78% vs. 58%), but otherwise the new studies did not differ (p > .05) on any
of the variables coded, listed previously. Given the similarity of the 21 new studies
with the previously reported studies, all subsequent analyses were conducted with the
combined data set.
 103 Cultural Adaptations and Multicultural Competence

- 2 - 1 0 1 2 3

E f f e c t S iz e (d )

• S tu d ie s ■ p < 1% ■ 1% < p < 5 % ■ 5 % < p < 10% p > 10%

figure 4 .1 C o n t o u r - e n h a n c e d f u n n e l p lo t o f e f f e c t siz e s ( C o h e n ’s d) b y s ta n d a r d e r r o r f o r 9 9 s tu d ie s

o f c u ltu r a lly a d a p te d t r e a tm e n t s . T h is g ra p h sh o w s th e d is t r ib u t io n o f e f f e c t siz e s a s a f u n c t io n o f th e

n u m b e r o f p a r tic ip a n ts i n th e stu d y (o p e r a t io n a liz e d as s ta n d a r d e r r o r ) .

Publication Bias
When findings are obtained that are nonsignificant (null findings), those results are
less likely to be published or may not exist and are therefore unable to be located in
a literature search and included in the meta-analysis. This inability to locate unpub­
lished studies can result in publication bias. In this meta-analytic review, publication
bias appeared to influence the overall findings. Figure 4.1 provides a visual indicator
and shows that studies with negative or null results were not located in our literature
search (few studies appear in the bottom left). The data in Figure 4.1 were asymmetric,
with a notable dearth of studies with few participants that did not achieve statistically
significant results. The distribution strongly suggests publication bias in the available
literature.
In addition, Egger’s regression test (an estimate of effect size asymmetry) was sta­
tistically significant (p < .001), providing further evidence of publication bias. Begg’s
test for small-study effects, based on the rank correlation between the effect size and
its standard error, also reached statistical significance (p < .001). In our analyses,
published studies had higher average effect sizes than unpublished studies (d = .54 vs.
d = .31, Q = 4.2, p = .04). In addition, the trim and fill method (Duval & Tweedie,
2000) identified 25 “missing” studies in the distribution. When those hypothetically
missing data were accounted for, the resulting omnibus effect size would be reduced to
 104 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

d = .35 (95% CI = 0.27-0.43). Thus the results presented in the previous section prob­
ably represent liberal estimates likely impacted by publication bias.

Moderators
Given the heterogeneity of the overall findings, we conducted a series of random­
effects weighted analyses of variance and meta-regressions to examine variables that
could have accounted for the variance in effect size estimates. We first examined study,
treatment, and participant characteristics separately (Table 4.2). As an additional step,
we included these variables simultaneously in a meta-regression model (Table 4.3).

Table 4.2. Moderator Analyses for the Studies of Cultural Adaptations to Mental Health
Treatments
Variable Q p k d 95% CI
Study Characteristics
Research Design 9.1 .003
Quasi-experimental 23 .31 [.16, .45]
Experimental 76 .56 [.48, .65]
Control Group Type 0.5 .47
No treatment (e.g., waiting list) 45 .53 [.41, .64]
Treatment as usual 53 .47 [.37, .57]
Outcome Measuresa 15.0 .002
General/multidimensional 18 .23 [.06, .41]
wellbeing
General/multidimensional mental 12 .26 [.04, .48]
health symptoms
Specific mental health symptoms 43 .55 [.43, .67]
(e.g., depression)
Specific behavioral outcomes (e.g., 46 .58 [.46, .70]
aggression)
Treatment Characteristics
Type of Treatment b 6.3 .01
Mental health treatments 29 .41 [.31, .52]
Prevention programs targeting 12 .68 [.50, .86]
problem behaviors
Treatment Modality 1.2 .75
Individual 18 .48 [.30, .67]
Group 46 .48 [.36, .59]
Family 14 .63 [.38, .87]
Combined individual/group 11 .50 [.27, .73]
Adaptations for Client Language c 4.7 .03
No 17 .35 [.16, .54]
 105 Cultural Adaptations and Multicultural Competence

Table 4.2. C ontinued

Variable Q P k d 95% CI
Yes 52 .59 [.48, .70]
Measurement Language Translationc 18.2 <.0001
No 23 .28 [.15, .42]
Yes 46 .66 [.55, .76]
Goals Based on Cultural Values 4.9 .03
No 85 .46 [.38, .54]
Yes 14 .69 [.50, .88]
Use of Cultural Metaphors 5.3 .02
No 57 .43 [.33, .52]
Yes 42 .61 [.49, .73]
Adaptations to Treatment Methods 4.3 .04
No 51 .42 [.32, .53]
Yes 48 .58 [.47, .69]
Participant Characteristics
Gender o f Client 0.3 .57
Female 22 .53 [.35, .70]
Male 6 .42 [.08, .75]
Ethnicity o f Client 7.6 .06
A frican Am erican 23 .50 [.33, .67]
Hispanic/Latinx A m erican 44 .52 [.41, .64]
Asian Am erican 18 .67 [.47, .88]
Native Am erican 5 .14 [-.18, .46]
Clients All Same Ethnicity 6.3 .01
No 9 .29 [.05, .54]
Yes 67 .63 [.53, .74]
Sample Type 4.0 .04
Community/at-risk samples 63 .44 [.35, .54]
C linical populations 36 .61 [.48, .74]
Treatment by Age Grouping 9.1 .003
A t-risk samples under age 18 33 .36 [.23, .49]
C linical samples over age 24 20 .70 [.52, .88]

Note. Q = Q-value for variance between groups. p = significance of between group differences.
k = number of studies. d = Cohen’s d, random effects weighted effect size. CI = confidence interval.
a This analysis compared all outcome measures within studies. b Culturally adapted treatments
compared with treatment as usual. c Among populations likely speaking English as a second language.

Moderation by Study Characteristics

A s seen in T able 4 .2 , th e 76 stud ies in w h ich p a rticip a n ts w ere ra n d o m ly assign ed
to tre a tm e n t co n d itio n s (i.e., e x p e rim e n ta l d esig n s) averaged sig n ifica n tly m o re effec­
tive resu lts th a n th e 23 stud ies u sin g n o n -ra n d o m a ssig n m en t o f clien ts to tre a tm e n t
co n d itio n s (d = .56 vs. d = .3 1 , Q (9 8 ) = 9 .1 , p = .0 0 3 ). T h is fin d in g is u n ex p ected , as
 106 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

T able 4 .3 . Random Effects Regression Weights fo r Study Characteristics Associated

with Effect Sizes from Studies o f Cultural Adaptations to Mental Health Treatments
Variable R2 ? P
35.4***
Estim ate o f publication bias a .256 .002
Assessments o f specific symptoms b .246 .003
Random assignm ent to treatm ent type .214 .01
Sum o f cultural adaptations c .175 .04
Average client age .177 .09
C linical sample vs. at-risk or com m unity sample .150 .13
Treatm ent for problem atic behavior d .142 .11
Year o f study publication .083 .37
Client racial hom ogeneity within studies e .070 .41

"Inverse of the number of participants in the study. bContrast of symptom-specific measures of mental
health and prosocial behaviors with multidimensional assessments or global well-being/mental health.
cSum of the eight indicators of the ecological model (Bernal et al., 1995). dContrast of treatments for
problematic behaviors with treatments for social and mental health. 'Contrast of studies in which all
participants were of the same race with studies in which participants’ race varied.
*** p < .0001.

experimental designs typically account for multiple potential confounds (e.g., impact
of self-selection on treatment effectiveness) and typically result in more conservative
effect sizes, compared to designs in which confounds are uncontrolled. The findings
of the 45 studies comparing outcomes of treatment groups to those of no-treatment
control groups (i.e., clients on a waiting list; d = .53) yielded similar findings to the
53 studies comparing outcomes in the experimental group to outcomes of clients re­
ceiving some kind of treatment (e.g., treatment as usual; d = .47, p > .10). It would be
expected that studies comparing adapted treatments to no treatment would result in
larger effect sizes than those comparing the adapted condition to another intervention,
but the difference observed was small, an average of d = .06.
Results also varied significantly by the type of outcome assessment (Q = 15.0,
p = .002). As indicated in Table 4.2, comparisons were made across four kinds of out­
come measures. Studies evaluating general/multidimensional well-being averaged
d = .23, and studies evaluating general/multidimensional mental health symptoms
averaged d = .26. However, studies evaluating specific mental health symptoms (e.g.,
depression) averaged d = .55, and studies evaluating specific behavioral outcomes (e.g.,
aggression) averaged d = .58. These two sets of evaluations, general versus specific,
yielded quite different results.
We also found that effect size differences differed as a function of the year of study
publication. The random effects weighted correlation of study year with effect size was
.20 (p = .02). All of the studies with effect sizes greater than 1.50 appeared during the
years 2004 to 2014. There were no studies prior to the year 1999 that reported effect sizes
 107 Cultural Adaptations and Multicultural Competence

larger than 1.0. This trend in the data could, in part, be explained by improvements to
the culturally adapted treatments: We observed a moderate correlation of .24 (p = .04)
between an estimate of adaptation quality and year of study publication. However, we
were also concerned that the observed trend over time could potentially be the result
of publication bias. After the first meta-analysis demonstrating the effectiveness of cul­
tural adaptations appeared (Griner & Smith, 2006), it is possible that journal editors
became less likely to publish studies that did not result in significant findings or that
they were more likely to publish studies with favorable results even with a questionably
low number of participants. Because the number of participants in a study can have
notable impact on statistical significance (and the likelihood of research publication),
we correlated the inverse of both the number of participants and the standard error,
as estimates of publication bias (Peters et al., 2006), with the year of study publication.
The resulting correlations were very small (< .06 absolute value), suggesting that the
number of participants (an indicator associated with publication bias) did not account
for the trend for larger effect sizes in the recent literature.

Moderation by Treatment Characteristics

After restricting analyses to studies using a comparison group that received a treat­
ment (rather than a no-treatment control), mental health treatments were less effec­
tive (d = .41) than programs targeting problem behaviors such as aggression (d = .68;
Q = 6.3, p =.01). However, average effect sizes were similar across individual, group,
and family treatments (p > .10).
Systematic differences in effect sizes were found across the kinds of cultural
adaptations made. Specifically, treatment adaptations that focused on language
appeared to be most effective. That is, treatments that were conducted in the preferred
language of the client (when the client did not speak English as a native language) were
much more effective than those that did not explicitly report conducting treatment
in the preferred language of the client (d = .59 vs. d = .35, Q = 4.7, p = .03). Similarly,
studies that provided written assessments translated to clients’ preferred language had
much larger effect sizes than those that did not (d = .66 vs. d = .28, Q = 18.2, p < .001).
When clinical goals were explicitly based on client cultural values, effect sizes were
much larger than when no such information was provided (d = .69 vs. d = .46, Q = 4.9,
p = .03). Treatments involving cultural metaphors were more effective than those
that did not mention that particular cultural adaptation (d = .61 vs. d = .43, Q = 5.3,
p = .02). Additionally, when the cultural adaptations explicitly modified the treatment
methods (e.g., performing cultural rituals, consulting with family members as part of
individual therapy), those studies yielded larger effect sizes than studies not explicitly
modifying standard treatment methods (d = .58 vs. d = .42, Q = 4.3, p = .04). Overall,
the more cultural adaptations reported by study authors, the larger the corresponding
effect size (r = .21, p = .01).
 108 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Moderation by Participant Characteristics

Effect sizes were not moderated by participant gender. Studies involving only men had
equivalent effect sizes to those involving only women. Furthermore, the percentage
of women within studies was not significantly correlated with the study effect size.
Similarly, the percentage of clients and the percentage of therapists from various racial
groups were unrelated to effect size magnitude. However, experimental studies that
contained all participants of the same race had larger effect sizes than studies in which
clients of various races who received the same culturally adapted treatment (d = .63 vs.
d = .29, Q = 6.3, p = .01).
Clinical samples (d = .61) obtained stronger benefits from treatment than commu­
nity and at-risk samples (d = .44, Q = 4.0, p = .04). Even after restricting our analyses to
experimental designs, this same pattern remained. Twenty-seven treatments for clin­
ical populations yielded an effect size of d = .80, and 49 interventions for community
and at-risk samples yielded an average effect size of d = .48 (Q = 8.7, p =.003). Thus the
type of client sample moderated the overall findings.
Effect sizes were also moderated by participant age. The random effects weighted
correlation between participant average age and study effect size was .25 (p =.004).
Further examination of the associated scatterplot confirmed that studies with
participants older than 40 years yielded larger effect sizes. Previously we found that
there was an association with age distribution and the clinical nature of samples.
Given that the vast majority of prevention-oriented studies focused on children and
adolescents and that most clinical studies involved adults, we conducted an additional
analysis contrasting clinical studies with participants over age 24 with prevention-
oriented studies with participants under age 18. As anticipated, experimental studies
with at-risk children and adolescents tended to be much less effective (d = .36) than
those with adult clinical samples (d = .70; Q = 9.1, p = .003).

Meta-Regression of Moderating Variables

The previous analyses in this section identified several variables that explained sys­
tematic variance in effect sizes. Some of these variables, such as participant age and
clinical status, clearly overlapped. We therefore sought to ascertain which of the several
moderating variables would remain predictors of treatment effectiveness in the pres­
ence of the others.
A meta-regression including only the 93 studies reporting participant age explained
35.4% of the variance in effect sizes (p < .001), with the results depicted in Table 4.3.
All variables except two contributed at least 1% of variance to the model; year of
study publication and racial homogeneity of clients (studies with all clients from the
same racial or ethnic group contrasted with clients from multiple groups) were weak
predictors of effect size when considered simultaneously with the other variables. The
inverse of study sample size, an indirect estimate of publication bias (Peters et al.,
2006), remained the strongest predictor, explaining 6.6% of effect size variance. The
use of specific versus general assessments and random assignment to treatments also
 109 Cultural Adaptations and Multicultural Competence

explained substantial amounts of variance (6.1% and 4.6%, respectively). In the pres­
ence of the other variables, the number of reported cultural adaptations remained a
statistically significant predictor; the efficacy of a treatment increased as more cultural
adaptations were incorporated. Although client age and at-risk versus clinical status
were both associated with effect size magnitude at the univariate level, neither one
reached statistical significance when considered simultaneously.
A separate meta-regression was conducted to examine which, if any, particular
cultural adaptation was more predictive of positive client outcomes than another.
Indicators of the eight components of the ecological validity model (Bernal et al.,
1995) explained 14.5% of the variance in effect sizes (p < .01). The two types of cul­
tural adaptations that remained statistically significant in the presence of the others
were (a) explicitly basing treatment on the client’s goals, informed by cultural values
(standardized beta = 0.24, p = .01) and (b) providing treatment in clients’ preferred
language (standardized beta = 0.18, p < .05).

m e t a - a n a l y s is o f t h e r a p i s t
m u l t ic u l t u r a l c o m p e t e n c e

The Studies
We identified 15 studies containing data on 2,640 clients’ experiences in mental health
treatments as a function of their therapists’ level of multicultural competence. Nine
studies (60%) involved clients receiving individual psychotherapy and seven studies
evaluated clients participating in various modalities (individual and group treatments).
Table 4.4 summarizes information on study and participant characteristics. The
vast majority of studies appeared since the year 2010, including many unpublished
doctoral dissertations. All studies used convenience samples, with the majority using
university/college students as participants. Studies averaged 176 clients, with one study
having more than 500 participants. Most often clients were either African American or
Hispanic/Latinx American, with limited research investigating other racial or ethnic
groups. Nine studies involved cross- sectional (correlational) data, and six studies
evaluated changes over time.
In this meta-analysis, two outcomes were considered: client participation in treat­
ment (premature termination vs. completion), evaluated in four studies, and client
outcomes, evaluated in 11 studies. We analyzed these two distinct outcomes separately
in the present meta-analysis.

Overall Results
Across four studies that evaluated the level of client participation in treatment, the
overall value was r = .26 (95% CI = .05-.44, p = .02; d = .54). Across 11 studies that
evaluated client outcomes, the value was r = .24 (95% CI = .10-.37, p < .001; d = .50).
High heterogeneity characterized the findings the two types of studies
analyzed: studies evaluating client participation in treatment (I2 = 80.9, 95%
 110 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

T able 4 .4 . C h aracteristics o f 15 C orrelational Studies o f Therapist M ulticultural

C om petence and C lient E xperiences in Treatm ent

Characteristic M Number of Studies (k) %

Year o f R ep o rt 2010
Before 2000 1 7
2 0 0 0 -2 0 0 9 3 20
2 0 1 0 -2 0 1 7 11 73
P u b lica tio n Status
Published 9 60
Unpublished 6 40
Sam ple Type
Outpatient m ental health clinic 3 23
University/college counseling center 8 62
University/c ollege students (nonclinical) 1 7
At-risk com m unity m embers 1 7
Sam ple Size 176
<50 0 0
5 0 -9 9 3 20
101-199 9 60
2 0 0 -3 9 9 2 13
4 0 0 -9 9 9 1 7
>1,000 0 0
A ge o f C lien ts“ 28.8
Children (<13 yrs.) 1 10
Adolescents (1 3 -18 yrs.) 0 0
Young adults (1 9 -2 9 yrs.) 6 60
M iddle-aged adults (3 0 -5 5 yrs.) 3 30
Senior adults (>56 yrs.) 0 0
G end er o f clien ts (% Fem ale) 66.6
R ace o f clien tsb (%)
A frican Am erican 41
Asian Am erican 14
Hispanic/Latinx Am erican 32
Native A m erican Indian 1
O ther 7

Note. Not all variables sum to the total number of studies due to missing data.
aAverage age category of participants within studies; not all participants in a study were necessarily
included in the category. bThe racial composition of participants across all studies, calculated by
multiplying the number of participants within studies by the percentage of participants from each ra­
cial group and dividing that product by the total number of participants.

C I = 5 0 - 9 3 ; Q (3 ) = 1 5 .7 , p = .0 0 1 ) and stud ies ev alu atin g clie n t o u tco m es ( 12 = 8 9 .6 ,

9 5 % C I = 8 3 - 9 3 ; Q (1 0 ) = 9 6 .5 , p < .0 0 0 1 ). In co n siste n t fin d in g s ch a ra cteriz ed th is
m eta -a n a ly sis, re su ltin g in th e a sso cia tio n b e tw een th erap ist m u lticu ltu ra l co m p e te n ce
and clie n t e x p e rie n ces in tre a tm e n t b e in g h ig h ly v ariab le acro ss stud ies. T h is m ad e it
 11 1 Cultural Adaptations and Multicultural Competence

difficult to interpret the averages reported here. We therefore sought explanations for
the observed variability.

Moderators
As seen in Figure 4.2, the effect sizes were unevenly distributed. Specifically, studies
tended either to cluster around r = 0 (indicative of no effect) or to be statistically sig­
nificant (located beyond the shaded regions to the right in Figure 4.2), with few studies
filling in the space between those extremes. It was therefore necessary to examine the
data for systematic differences that could account for this unusual distribution.
As shown in Table 4.5, the eight effect sizes from studies measuring client perceptions
of therapists’ multicultural competence with client outcomes (both participation and
clinical improvement) averaged r = .38 (d = .82), whereas the five effect sizes from
studies using a measure of multicultural competence completed by the therapist
averaged r = .06 (d = .12). This difference explained a remarkable 48.2% of the vari­
ance in effect sizes in a random effects weighted regression model that controlled for
the type of outcome evaluated. Thus this model accounted for the disparate findings
shown in Figure 4.2.
In Figure 4.2, all but one of the nonsignificant effect sizes (those in the center of
the graph) reflect studies that relied on therapist self-report of cultural competencies,
suggesting no relation between cultural competencies and client outcomes. In the same
figure, all but one of the statistically significant studies (to the right of the shaded lines)

• Studies ■ p < 1% ■ 1% < p < 5% ■ 5% < p < 10% p > 10%

f i gu re 4.2 Contour-enhanced funnel plot of effect sizes (Pearson r) by standard error for 15
correlational studies of therapist multicultural competence and client treatment outcomes. This
graph shows the distribution of effect sizes as a function of the number of participants in the study
(operationalized as standard error). The results are highly scattered.
 112 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HA T WO RK

Table 4 .5 . M oderator A nalyses for th e Studies o f M ulticultural C om petence

Variable Q b p k r 95% CI
B etw een Stu dy D ifferences
Source o f Competence Evaluation 12.5 .0004
Therapist (self-report) 5 .06 [ -.0 8 ,.2 0 ]
Clients 8 .38 [.27, .48]
Published 0.5 .47
No 6 .19 [.00, .37]
Yes 9 .28 [.13, .41]
Clients All Same Ethnicity 0.2 .67
No 4 .29 [.06, .52]
Yes 11 .23 [.10, .37]
C o rrela tio n s w ith E ffect Size
Year o f Study .29 .25 15
Average Participant Age -.1 5 .67 10
G ender (% females) -.11 .68 14
% A frican Am ericans .16 .53 15
% Asian Americans - .0 7 .79 15
% Hispanic/Latinx Am ericans -.2 9 .25 15
M ulticultural com petence measure .49 .04 13
reliability

Note. Q = Q-value for variance between groups. b = Standardized coefficient for the association with
effect size. p = significance of between group differences. k = number of studies. r = random effects
weighted correlation coefficient. CI = confidence inteveral.

utilized client-rated measures of cultural competencies, demonstrating a positive re­

lation between cultural competencies and outcomes. This provides a visual represen­
tation that the source of the cultural competency rating influenced the distribution of
the effect sizes.
One additional variable appeared to moderate the overall findings. The reliability
coefficient of the measure of therapist multicultural competence was strongly associ­
ated with the effect size obtained in the study, r = .49 (p = .04). Reliability coefficients
(Cronbach’s alpha) for measures completed by clients averaged .94, whereas measures
completed by therapists averaged .78 (p < .001). However, a subsequent regression
model accounting for the source of the multicultural competence rating (therapist or
client) demonstrated that the source of the rating was a more robust predictor than
the reliability coefficient of the competence scale. The source of the rater’s evaluation
of multicultural competence accounted for 43.7% of the variance in effect sizes, and
the corresponding reliability coefficient now accounted for only .02% of the variance.
No other study or participant characteristic moderated the overall results. Thus
shared rater variance explained the findings obtained within studies, even after ac­
counting for the different types of outcomes measured (client participation in treat­
ment and client clinical outcomes), with the added possible confound of shared
 113 Cultural Adaptations and Multicultural Competence

measurement variance (similar content between the CCCI-R or Cultural Humility

Scale and clients’ ratings of other therapists’ attributes) likely influencing the findings
of the studies measuring client perceptions of therapists.

Publication Bias
We also examined the unusual distribution of effect sizes (Figure 4.2) for the possi­
bility of publication bias influencing the results. In the case of this meta-analysis, the
notable gap between statistically significant studies (on the right) and nonsignificant
studies (in the center) suggested several “missing” studies (i.e., studies conducted
but unpublished with values that would fill in the missing spaces in the existing dis­
tribution). Nevertheless, across all studies and controlling for the type of outcome
evaluated by studies, published studies yielded results of about the same magnitude as
unpublished studies (p > .20 when testing for differences), suggesting that the results
of the meta-analysis were not attributable to publication status. Furthermore, two
analyses using the formulas provided by Egger and by Begg failed to reach statis­
tical significance, and the trim and fill method failed to identify any “missing” studies
when conducted on the overall data and when applied separately for the two types of
study outcomes. No evidence of publication bias was found when these analyses were
repeated separately for therapist-rated and client-rated data (accounting for the data
distribution). Thus we concluded that the unusual distribution of data (Figure 4.2)
was attributable to the source of data evaluation and not to publication bias.

EVIDENCE FOR CAUSALITY

All studies included in the meta-analysis of cultural adaptations to mental health
treatments involved experimental or quasi-experimental designs. Effect sizes
obtained from the 76 experimental designs were significantly larger than those
obtained from the 23 quasi-experimental designs. This finding provides consider­
able evidence for causality. More specifically, culturally adapted treatments showed
superior positive outcomes. The greater the number of adaptations, the better the
outcomes. Also, treatments provided in the client’s preferred language, goals based
on the client cultural values, and using cultural metaphors were associated with
improved outcomes.
The fact that an indirect estimate of publication bias (inverse sample size) remained
the strongest independent predictor of effect size magnitude remains a major hurdle to
establishing causality. For instance, we cannot determine the extent to which journal
reviewers/editors accepted manuscripts showing strong benefits of cultural adaptations,
such that the overall field sees evidence for those adaptations more often than they en­
counter small findings, null findings, or even negative findings. We estimated that at
least 25 additional studies should have been included in the meta-analysis but were not
located, possibly because they remained unpublished and were thus difficult to locate.
To establish solid evidence of causality, we urge researchers and journal reviewers to
 114 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

attend to factors that could contribute to publication bias, including the publication of
nonsignificant findings obtained from quality research.
The studies evaluating multicultural competence have not yet provided evidence
of causality. The vast majority of the studies included in the meta-analysis contained
correlational data. Only six studies examined changes over time. No studies involved
group comparisons or experimental designs. Longitudinal research and experimenta­
tion will be needed if the field is to move beyond conjecture to causal inferences.
The meta-analysis specific to multicultural competencies was characterized by a
high degree of heterogeneity. Such heterogeneity makes it difficult to draw conclusions
from the overall results. However, when examined, this variability was explained by
the source of the rating of the therapists’ multicultural competencies. Therapist self­
ratings of their multicultural competencies do not correspond to client experiences in
psychotherapy. Therapist self-reports appear to lack validity, potentially due to system­
atic bias from therapist self-evaluations.
Client ratings of therapist multicultural competencies proved more robust
predictors of positive client outcomes. However, client ratings introduce the potential
confound of attenuation. Specifically, it may be possible that a halo effect operates in
which clients who generally like their therapists, or who generally experience therapy
positively, are more likely to rate their therapist positively across many domains. Thus,
without attending to client attenuating factors, it is difficult to make a conclusive state­
ment regarding the extent to which clients can distinguish multicultural competencies
from their own positive perspectives of the therapist.
Client participation and client outcomes were associated with therapist multicul­
tural competence, when rated by the client. Therapist multicultural competencies
accounted for nearly 5.8% of the observed variance in client improvements and nearly
6.8% of client participation (e.g., premature termination vs. completion).
An important limitation to causality involves the validity of therapist self-report
measures. In our meta-analysis, therapist self-report measures failed to demonstrate
criterion validity: they were uncorrelated with clients’ outcomes. The difference be­
tween studies using client-reported measures and those using therapist-reported
measures was so large, explaining 48% of the variance in effect sizes, that one begins
to question to what extent therapist self-report measures are useful in clinical settings.
If self-report measures remain problematic in more targeted evaluations of criterion
validity, then the field may benefit from examining therapist variability among client
outcomes (Hayes et al., 2016) rather than relying on therapist self-report.

LIMITATIONS OF THE RESEARCH

Cultural adaptations to mental health treatments typically prove more effective than
traditional treatments when used with clients of color; however, multiple limitations
characterize the current research. One limitation is that even though the data in
our present meta-analysis indicate that treatments are most effective when adapted
to multiple features of a client’s cultural background, many studies provided lim­
ited descriptions of the adaptations enacted. This precludes systematic reviews from
 115 Cultural Adaptations and Multicultural Competence

examining the quality or specificity of adaptations implemented. When quantifying

adaptations using the Ecological Validity Model (Bernal et al., 1995), few studies (18%)
implemented five or more of the eight broad components covered by the model. Only
about half reported modifying treatment methods (49%), consulting with individuals
from the culture when developing the treatment (48%), or relying on the client’s con­
ceptualization of the presenting problem (42%). Only 14% explicitly solicited cultur­
ally congruent outcome goals from the client, even though that particular practice had
been identified as a key aspect of cultural adaptations in the first meta-analysis on the
topic (Griner & Smith, 2006). Thus, while cultural adaptations are more effective than
nonadapted interventions, it is still not clear which specific adaptations are more con­
sequential than others. Clinical practice and future scholarship will benefit from more
detailed descriptions of specific adaptations utilized (Domenech Rodríguez & Bernal,
2012; Sidani et al., 2017).
The current research literature does not adequately or consistently account for
several diverse groups. Most of the research on cultural adaptations in the North
American literature involves Asian Americans and Hispanic/Latinx Americans. Arab
Americans, Native Americans, African Americans, and Polynesian Americans remain
underrepresented. Therefore, while generalizations can be made about the effective­
ness of culturally adapted interventions with Asian Americans and Hispanic/Latinx
Americans, more research is needed with other groups.
Now that research has consistently documented that cultural adaptations work, fu­
ture research can evaluate the reasons why they prove more effective than traditional
treatments for clients of color. Why are some culturally adapted treatments very effec­
tive while others are equivalent to control group conditions (see Figure 4.1)? Which
components of cultural adaptations are the most consequential: culturally adapting
the explanatory model of the presenting concern, enhancing language congruence
through cultural idioms, or providing a culturally cohesive approach to therapy,
adapting methods and goals? Cultural adaptations work, but we need more specific
information about their underlying mechanisms (Smith, 2010).
Many limitations characterize the meta-analysis on multicultural competencies.
A major limitation is the paucity of studies that met inclusion criteria, with only 15
studies identified across two different outcomes. The empirical and clinical foundations
of psychotherapy will likely benefit from renewed interest and scholarship on this topic.
An additional limitation related to methodology surrounds the use of clinicians’
self-report measures, which do not correlate with clients’ outcomes. If the field cannot
rely on therapist self-report, whether due to social desirability, overconfidence,
or other reasons, then we must ascertain how to best measure and assess multicul­
tural competencies. Whether through direct observation, client reports, responses
to vignettes, or alternative methods, it is essential that scholars produce empirical
support.
Client reports of therapists’ cultural competence also introduce several confounding
factors. The first is that many studies have thus far relied on retrospective cross­
sectional design methodology, with clients often being asked to simultaneously re­
flect on psychotherapy processes, including cultural competencies, in one survey. Such
 116 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

an approach introduces halo effects such that clients cannot distinguish meaningful
nuances of psychotherapy processes. Clients may also be more likely to positively rate
therapists when therapy is effective, or globally rate them negatively when it is not, fur­
ther obfuscating the extent to which client self-reports are influenced by halo effects.
Future research on multicultural competence can specify whether clients can distin­
guish multicultural competencies from global therapist skills.

DIVERSITY CONSIDERATIONS

A large body of literature has concluded that client attributes must be considered
when planning and implementing mental health treatments (e.g., Casas et al., 2016).
Our meta-analytic review evaluated only treatment adaptations based on client race,
culture, and ethnicity; numerous other client characteristics (e.g., religious affilia­
tion, sexual orientation, socioeconomic status, age, gender) also deserve attention.
Traditional approaches to psychotherapy have been adapted to account for several of
those characteristics, as seen throughout this volume.
Individuals espouse multiple identities at once; researchers and practitioners can
account for the intersections of human diversity (Rosenthal, 2016; Smith & Draper,
2004), such as how gender roles affect well-being among individuals with disabilities.
Every client possesses many cultural attributes, and these attributes intersect to form
complex identities.
A focus on only one client attribute, even an attribute as consequential as race, eth­
nicity, or the broader construct of culture, could unwittingly minimize the relevance
of other salient client factors, such as gender or spirituality, on improved outcomes
(e.g., Cabral & Smith, 2011). We are particularly concerned that widespread attention
to certain aspects of diversity, such as culture, has overshadowed the need to adapt
mental health treatments for people living in poverty (e.g., Smith & Brewster, 2015).
Economic advantage is a primary source of privilege. Discrimination against people
in poverty (Lott, 2002) compounds adverse circumstances, which include differential
access to quality mental health services. Multiple treatment adaptations, including cul­
ture and socioeconomic status, will be required to increase rates of mental health treat­
ment utilization and retention among people in poverty.
We believe it likely that treatments simultaneously accounting for multiple client
identities could prove even more effective than those targeting only one client at­
tribute. That is, the more a treatment is tailored to match the precise characteristics of
a client, the more likely that client will engage in treatment, remain in treatment, and
experience improvement as a result of treatment (Beutler et al., 2012; Castonguay &
Beutler, 2006; Smith & Trimble, 2016). For instance, effective spiritual and religious
adaptations to psychotherapy (Gonsalves et al., 2015; Martinez et al., 2007; Smith et al.,
2007) can be combined with other considerations to develop an effective treatment
adapted to the sexual, spiritual, and family identities of a Latinx adolescent (Duarté-
Vélez et al., 2010). Although clinicians cannot account simultaneously for every client
attribute, they can be aware of intersecting identities, consider how those identities
operate in therapy, and adapt treatment to the needs of an individual client.
 117 Cultural Adaptations and Multicultural Competence

TRAINING IMPLICATIONS

Research has shown that multicultural training can improve therapist multicultural
competence (Smith et al., 2006). However, major limitations of that research include
responses confounded with social desirability, inaccurate trainee self-evaluations, and
clinical nonsignificance, in which change on a few questionnaire items does not equate
to meaningful changes in real-world practice (Smith & Trimble, 2016). Clinicians-in-
training may exaggerate self-reported competency (Constantine & Ladany, 2000) and
have difficulty applying specific multicultural skills (Constantine et al., 2000).
Our meta-analytic results confirm a major problem: Therapists’ self-ratings of mul­
ticultural competence were not associated with client outcomes. Given this finding,
we need to go beyond simply recommending that therapists and trainees receive more
training in multicultural competencies. The key considerations are how to design, im ­
plement, and improve multicultural training that results in demonstrable skills when
working with diverse clients.
Our first recommendation for enhancing clinician multicultural competence
through training is to attune them to client cultural experiences . Traditional multi­
cultural training aims to raise awareness and knowledge of others’ experiences and
worldviews. However, care must be taken to avoid perpetuating stereotypes and cat­
egorical thinking about cultural complexities; therapists must engage in hypothesis
testing and dynamic sizing (Sue, 1998). If trainees learn cultural generalities about a
given group, say Haitian immigrants, they may come away with just enough informa­
tion to develop a heuristic that paradoxically reduces their active curiosity about the
lived experiences of their next Haitian immigrant client. Optimally, training will foster
explicit openness to all aspects of clients’ experiences, bolstered by a healthy dose of
cultural humility (Hook et al., 2013). Training for cultural humility can allow for cul­
tural learning to occur across a lifetime (Smith, 2004).
Second, multicultural training can account for the discrepancy between therapist
and client perceptions of multicultural competence by requiring trainees to solicit
client feedback about cultural considerations. Soliciting feedback during treatment
is essential to bridging the gap between client and therapist perceptions of cultural
considerations.
Third, multicultural training can shift from didactic instruction to research-based
m ethods f o r skill developm ent , such as modeling by a skilled individual followed by
observed rehearsal by the trainee (Beidas et al., 2014). Experiential learning is essen­
tial. For instance, community engagement and service/leadership can provide trainees
with more opportunities to internalize multicultural competencies. Developing multi­
cultural competencies extends well beyond an academic or intellectual pursuit (Smith,
2004; Smith et al., 2015); instructors need to utilize learning methods that enable be­
havioral and worldview change.
Fourth, multicultural training can provide research-supported content , rather than
perpetuating ideas prevalent in the multicultural literature not based on research
(Smith & Trimble, 2016). Just as psychotherapy has changed over time based on re­
search findings, such as those reported in this book and its prior editions, cultural
 118 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

adaptations to treatments and therapist multicultural competence will benefit from a

greater reliance on data. The findings from this meta-analysis that cultural adaptations
to mental health treatments are more effective with adult clients than adolescents (see
also Huey et al., 2008), for instance, can be translated into practice by closely consid­
ering clients’ level of acculturation.
Fifth, multicultural training should require skill dem onstration and ongoing evalua­
tion . Trainees can be asked to demonstrate multicultural competencies both in class­
room settings and in clinical settings. For instance, students can engage in role-plays
with the instructor in class and with others, video recorded outside of class. They can
respond to case scenarios and seek out opportunities to demonstrate a skill, such as
repairing of an alliance rupture over racial tensions. It is essential that clinicians-in-
training learn to walk the walk rather than simply talk the talk.
Sixth, because treatments were most effective when conducted in the preferred lan­
guage of the client, training programs can purposefully recruit multilingual candidates .
At the very least, training programs in mental health can maintain lists of professional
language interpreters and require trainee proficiency in working with them (Frandsen
et al., 2018).

THERAPEUTIC PRACTICES
Cultural adaptations to mental health treatments typically prove more effective than
treatment as usual with clients of color in North America. Among clients of color, their
perceptions of a therapist’s ability to successfully work with culture is moderately as­
sociated with their outcomes in treatment, yet these findings are susceptible to strong
halo effects, wherein clients globally view therapists positively when therapy goes well.
Therapists’ self-rated multicultural competencies have essentially nothing to do with
the clients’ outcomes. Nevertheless, individual therapist variability of client outcomes
between diverse and White clients suggests the importance of cultural competencies
as they pertain to improved outcomes among clients of color (Hayes et al., 2016; Hayes
et al., 2014; Imel et al., 2011). Given these findings, we urge the field to consider the
following research-supported practices:

♦ Assess regularly patients’ racial and ethnic backgrounds and their salient worldviews
and race-related experiences. One cannot adapt treatments to what one does
not know.
♦ Improve client outcomes by aligning psychological treatment with clients’ cultural
backgrounds. Seek “cultural fit” when planning treatments.
♦ Incorporate multiple cultural adaptations, as they tend to be more effective than
treatments with only a few cultural adaptations. However, the specific procedures
taken to align therapy with client culture may matter less than the fact that therapists
attempt to make the alignment by using several methods (Smith, 2010).
♦ Consider particularly cultural adaptations for adults, as they tend to prove more
efficacious than with children and adolescents. This repeatedly demonstrated
disparity is likely a function of level of acculturation (i.e., integration with North
 119 Cultural Adaptations and Multicultural Competence

American society vs. maintaining ancestral cultural worldviews). Therapists should

attend to how client age and acculturation interact with potential cultural adaptations
to treatments.
♦ Conduct psychotherapy in the client’s preferred language in most cases. This research
finding confirms the stance taken by the American Psychological Association
(1993): “Psychologists interact in the language requested by the client and, if this
is not feasible, make an appropriate referral . . . If this is not possible, psychologists
offer the client a translator with cultural knowledge and an appropriate professional
background” (p. 147).
♦ Adapt group therapy to a specific group, rather than a culturally mixed group
of clients. The more culturally specific the treatment, the more effective it will
probably be.
♦ Handle cultural issues for clients of color in a sensitive and humble way. How
psychotherapists do so is strongly related with those clients’ treatment outcomes.
♦ Assume that therapist’s own self-evaluations regarding multicultural competence do
not align with the perspectives of clients of color. To overcome that gap, therapists
can actively seek to learn each client’s perspectives.

REFERENCES
References m arked with an asterisk indicate studies included in the m eta-analysis (single as­
terisk = cultural adaptations; double asterisks = m ulticultural com petence).
*A legría, M ., Ludman, E., Kafali, E. N., Lapatin, S., Vila, D., Shrout, P. E., & Bauer, A. M.
(2014). Effectiveness o f the Engagement and Counseling for Latinos (ECLA ) intervention
in low -incom e Latinos. Medical Care, 52(11), 9 8 9 -9 9 7 .
A m erican Counseling Association. (2014). ACA code o f ethics. Alexandria, VA: Author.
Retrieved from http://www.counseling.org/
A m erican Psychological Association. (1993). Guidelines for providers o f psychological services
to ethnic, linguistic, and culturally diverse populations. American Psychologist, 4 8 , 4 5 -4 8 .
A m erican Psychological Association. (2003). Guidelines on m ulticultural education, training,
research, practice, and organizational change for psychologists. American Psychologist, 58,
3 7 7 -4 0 2 .
A m erican Psychological Association. (2017a). Ethical principles o f psychologists and code o f
conduct. W ashington, DC: Author. Retrieved from http://www.apa.org/ethics/code/
A m erican Psychological Association. (2017b). Multicultural guidelines: An ecological approach
to context, identity, and intersectionality. W ashington, DC: Author. Retrieved from http://
www.apa.org/about/policy/multicultural-guidelines.pdf
Arredondo, P., Toporek, R., Brown, S. P., Jones, J., Locke, D. C., Sanchez, J., & Stadler, H.
(1996). O perationalization o f the M ulticultural Counseling Com petencies. Journal o f
Multicultural Counseling and Development, 2 4 (1 ), 4 2 -7 8 . https://www.doi.org/10.1002/
j.2 1 6 1 -1 9 1 2 .1 9 9 6 .tb 0 0 2 8 8 .x
Asante, M. K. (2009). Afrocentricity. Retrieved from http://www.asante.net/articles/1/
afrocentricity/
*Banks, R., Hogue, A., Tim erlake, T., & Liddle, H. (1998). An A frocentric approach to group
social skills training with inner-city A frican A m erican adolescents. Journal o f Negro
Education, 6 5(4), 4 1 4 -4 2 3 . https://www.doi.org/10.2307/2967144
 120 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

*Banks, S. R. (1998). The impact o f social support and active coping on enhancing mental health
and employability outcomes among African-Americans and Latinos with disabilities: A com­
munity based group intervention (D octoral dissertation). George W ashington University,
W ashington, DC.
Barrera, M ., Berkel, C., & Castro, F. G. (2017). Directions for the advancem ent o f cultur­
ally adapted preventive interventions: Local adaptations, engagement, and sustainability.
Prevention Science, 18(6), 6 4 0 -6 4 8 .
**Bath je, G. J. (2010). Multicultural competence and the process and outcome o f counseling
(D octoral dissertation). University o f G eorgia, Athens.
*Bedoya, C. A., Traeger, L., Trinh, N. H. T., Chang, T. E., Brill, C. D., Hails, K., . . . Yeung, A.
(2014). Im pact o f a culturally focused psychiatric consultation on depressive symptoms
am ong Latinos in prim ary care. Psychiatric Services, 65(10), 1 2 56-1262.
*Beeber, L. S., Holditch-Davis, D., Perreira, K., A Schwartz, T., Lewis, V , Blanchard, H., . . .
G oldm an, B. D. (2010). Short-term in-hom e intervention reduces depressive symptoms
in Early Head Start Latina m others o f infants and toddlers. Research in Nursing & Health,
3 3(1), 6 0 -7 6 .
Beidas, R. S., Cross, W , & Dorsey, S. (2014). Show me, don’t tell me: Behavioral rehearsal as a
training and analogue fidelity tool. Cognitive and Behavioral Practice, 2 1(1), 1 -1 1 . https://
www.doi.org/10.1016/j.cbpra.2013.04.002
*Belgrave, F. Z. (2002). Relational theory and cultural enhancem ent interventions for African
A m erican adolescent girls. Public Health Reports, 117(1), s76-s81.
*Belgrave, F. Z., Chase-Vaughn, G ., Gray, F., Dixon-A ddison, J., & Cherry, V. R. (2000). The
effectiveness o f a culture and gender specific intervention for increasing resiliency among
A frican A m erican pre-adolescent females. Journal o f Black Psychology, 26, 1 23-147.
https://www.doi.org/10.1177/0095798400026002001
Benish, S. G., Q uintana, S., & W ampold, B. E. (2011). Culturally adapted psychotherapy
and the legitim acy o f myth: A direct-com parison m eta-analysis. Journal o f Counseling
Psychology, 58(3), 2 7 9 -2 8 9 . https://www.doi.org/10.1037/a0023626
Bernal, G., Bonilla, J., & Bellido, C. (1995). Ecological validity and cultural sensitivity for
outcom e research: Issues for the cultural adaptation and development o f psychosocial
treatm ents with Hispanics. Journal o f Abnormal Child Psychology, 2 3(1), 6 7 -8 2 . https://
www.doi.org/10.1007/bf01447045
Bernal, G ., Jim énez-Chafey, M . I., & D om enech Rodriguez, M . M. (2009). Cultural ad­
aptation o f treatm ents: A resource for considering culture in evidence-based practice.
Professional Psychology: Research and Practice, 4 0 (4 ), 3 6 1 -3 6 8 . https://www.doi.org/
10.103 7/a0016401
*B ernstein, K., Park, S. Y., Hahm, S., Lee, Y. N., Seo, J. Y., & Nokes, K. M . (2015). Efficacy of
a culturally tailored therapeutic intervention program for com m unity dwelling depressed
Korean A m erican women: A non-random ized quasi-experim ental design study. Archives
o f Psychiatric Nursing, 30(1), 19-2 6 .
Betancourt, J. R., G reen, A. R., Carrillo, J. E., & Ananeh-Firem pong, A. (2003). D efining cul­
tural com petence: A practical fram ework for addressing racial/ethnic disparities in health
and health care. Public Health Reports, 118(4), 2 9 3 -3 0 2 . https://www.doi.org/10.1093/phr/
118.4.293
Beutler, L. E., Forrester, B., G allagher-Thom pson, D., Thom pson, L., & Tom lins, J. B. (2012).
Com m on, specific, and treatm ent fit variables in psychotherapy outcome. Journal o f
Psychotherapy Integration, 2 2(3), 2 5 5 -2 8 1 .
 12 1 Cultural Adaptations and Multicultural Competence

*B otvin, G. J., Schinke, S. P., Epstein, J. A., & Diaz, T. (1994). Effectiveness o f culturally fo ­
cused and generic skills training approaches to alcohol and drug abuse prevention am ong
m inority youths. Psychology o f Addictive Behavior, 8 (2), 1 1 6 -1 2 7 . https://www.doi.org/
10.1037//0893-164x.8.2.116
Boyle, D. P., & Springer, A. (2001). Toward a cultural com petence measure for social work with
specific populations. Journal o f Ethnic and Cultural Diversity in Social Work, 9 (3 -4 ), 5 3 -7 1 .
https://www.doi.org/10.1300/J051v09n03_03
*Cabiya, J. J., Cotto, L. P , Gonzalez, K., Cestero, J. S., Taboas, A. M ., & Sayers, S. (2008).
Effectiveness of a cognitive-behavioral intervention for Puerto Rican children.
Interamerican Journal o f Psychology, 4 2 (002), 195-202.
Cabral, R. R., & Sm ith, T. B. (2011). Racial/ethnic m atching o f clients and therapists in m ental
health services: A m eta-analytic review o f preferences, perceptions, and outcomes. Journal
o f Counseling Psychology, 58(4), 5 3 7 -5 5 4 . https://www.doi.org/10.1037/a0025266
Calabrese, J. D. (2008). C linical paradigm clashes: E thnocentric and political barriers to Native
A m erican efforts at self-healing. Ethos, 36(3), 3 3 4 -3 5 3 .
Cardem il, E. V. (2010a). Cultural adaptations to em pirically supported treatm ents: A research
agenda. The Scientific Review o f Mental Health Practice, 7(2), 8 -2 1 .
Cardem il, E. V. (2010b). The com plexity o f culture: Do we em brace the challenge or avoid it?
The Scientific Review o f Mental Health Practice, 7(2), 4 1 -4 7 .
*Cardem il, E. V , Reivich, K. J., Beevers, C. G., Seligman, M . E. P, & James, J. (2007). The
prevention o f depressive symptoms in low -incom e, m inority children: Two-year follow­
up. Behaviour Research and Therapy, 4 5(2), 3 1 3 -3 2 7 . https://www.doi.org/10.1016/
j.b rat.2006.03.010
*C arter, M. M ., Sbrocco, T., G ore, K. L., M arin, N. W , & Lewis, E. L. (2003). C ognitive-
behavioral group therapy versus a w ait-list control in the treatm ent o f A frican A m erican
women with panic disorder. Cognitive Therapy and Research, 27 (5), 5 0 5 -5 1 8 .
Casas, J. M ., Suzuki, L. A., Alexander, C. M ., & Jackson, M. A. (Eds.). (2016). Handbook o f
multicultural counseling. Thousand Oaks, CA: SAGE.
Castonguay, L. G ., & Beutler, L. E. (2006). Principles o f therapeutic change: A task force on
participants, relationships, and technique factors. Journal o f Clinical Psychology, 62(6),
6 3 1 -6 3 8 .
*C astro-O livo, S. M. (2014). Prom oting social-em otional learning in adolescent Latino
ELLs: A study o f the culturally adapted Strong Teens program . School Psychology Quarterly,
29 (4), 5 6 7 -5 7 7 .
*Cherry, V. R., Belgrave, F. Z., Jones, W , K ennon, D. K., Gray, F. S., & Phillips, F. (1998).
N TU : An A frocentric approach to substance abuse prevention am ong A frican A m erican
youth. The Journal o f Primary Prevention, 18, 3 1 9 -3 3 9 .
**C hristensen, C. (2001). Therapist cultural sensitivity and premature termination rates with
ethnic minority adolescents (Unpublished doctoral dissertation). University o f Akron,
A kron, OH.
Chu, J., Leino, A., Pflum, S., & Sue, S. (2016). A m odel for the theoretical basis o f cultural
com petency to guide psychotherapy. Professional Psychology: Research and Practice, 47(1),
1 8 -2 9 . https://www.doi.org/10.1037/pro0000055
Cokley, K. (2007). C ritical issues in the m easurem ent o f ethnic and racial identity: A refer­
endum on the stat o f the field. Journal o f Counseling Psychology, 54(3), 2 2 4 -2 3 4 .
Colem an, H. L. K. (1998). G eneral and m ulticultural counseling com petency: Apples and
oranges? Journal o f Multicultural Counseling and Development, 2 6 , 1 47-156.
 122 P S Y C H O T H E R A P Y R EL AT IO N S H IP S THAT WORK

Com as-D íaz, L. (2006). Cultural variation in the therapeutic relationship. In C. D. G oodheart,
A. E. Kazdin, & R. J. Sternberg (Eds.), Evidence-based psychotherapy: Where practice and
research meet (pp. 8 1 -1 0 5 ). W ashington, DC: A m erican Psychological Association.
Constantine, M . G. (2000). Social desirability attitudes, sex, and affective and cognitive em ­
pathy as predictors o f self- reported m ulticultural counseling com petence. Counseling
Psychologist, 28, 8 5 7 -8 7 2 .
**C onstantine, M . G. (2002). Predictors o f satisfaction with counseling: Racial and ethnic
m inority clients' attitudes toward counseling and ratings o f their counselors' general and
m ulticultural counseling com petence. Journal o f Counseling Psychology, 4 9(2), 2 5 5 -2 6 3 .
**C onstantine, M. G. (2007). Racial m icroagressions against A frican A m erican clients in
cross-racial counseling relationships. Journal o f Counseling Psychology, 54(1), 1 -1 6 .
Constantine, M. G ., G loria, A. M ., & Ladany, N. (2002). The factor structure underlying
three self-report m ulticultural counseling com petence scales. Cultural Diversity & Ethnic
Minority Psychology, 8, 3 3 4 -3 4 5 .
Constantine, M. G., & Ladany, N. (2000). Self-report m ulticultural counseling com petence
scales: Their relation to social desirability attitudes and m ulticultural case conceptualiza­
tion ability. Journal o f Counseling Psychology, 4 7 , 1 5 5-164.
*C osantino, G., Malgady, R. G., & Primavera, L. H. (2009). Congruence between cultur­
ally com petent treatm ent and cultural needs o f older Latinos. Journal o f Consulting and
Clinical Psychology, 77(5), 9 4 1 -9 4 9 .
*Crespo, M. M . (2006). Effects o f culturally specific dynamically oriented group art therapy with
immigrant Latinas (Unpublished doctoral dissertation). California Institute o f Integral
Studies, San Francisco, CA.
Cross, W J. (1998). Black psychological functioning and the legacy o f slavery: Myths and
realities. In Y. D anieli (Ed.), International handbook o f multigenerational legacies o f trauma
(pp. 3 8 7 -4 0 0 ). New York, NY: Plenum.
**C um m ins, D. M. (2004). Multicultural competence and levels o f effectiveness in adven­
ture based counseling (Unpublished doctoral dissertation). University o f West Virginia,
Morgantown.
*D ai, Y., Zhang, S., Yamamoto, J., Ao, M ., Belin, T. R., Cheung, F., & Hifum i, S. S. (1999).
Cognitive behavioral therapy o f m in or depressive symptoms in elderly Chinese
Am ericans: A pilot study. Community Mental Health Journal, 35(6), 5 3 7 -5 4 2 .
*Davey, M. (2013). A culturally adapted fam ily intervention for A frican A m erican families
coping with parental cancer: outcom es o f a pilot study. Psycho-Oncology Journal, 2 2(7),
1 5 7 2 -1 5 8 0 .
**D avis, D. E., DeBlaere, C., Brubaker, K., Owen, J., Jordan, T. A., Hook, J. N., & Van Tongeren,
D. R. (2016). M icroaggressions and perceptions o f cultural hum ility in counseling. Journal
o f Counseling & Development, 94 (4), 4 8 3 -4 9 3 .
**D illon, F. R., O dera, L., Fons-Scheyd, A., Sheu, H. B., Ebersole, R. C., & Spanierm an, L.
B. (2016). A dyadic study o f m ulticultural counseling com petence. Journal o f Counseling
Psychology, 63(1), 5 7 -6 6 .
D om enech Rodríguez, M. M ., & Bernal, G. (2012). Frameworks, models, and guidelines
for cultural adaptation. In G. Bernal, & M . M. D om enech Rodríguez (Eds.), Cultural
adaptations: Tools fo r evidence-based practice with diverse populations (pp. 2 3 -4 4 ).
W ashington, DC: A m erican Psychological Association.
Duarté-Vélez, Y., Bernal, G., & Bonilla, K. (2010). Culturally adapted cognitive-behavior
therapy: Integrating sexual, spiritual, and family identities in an evidence-based treatm ent
o f a depressed Latino adolescent. Journal o f Clinical Psychology, 6 6 (8), 8 9 5 -9 0 6 .
 123 Cultural Adaptations and Multicultural Competence

D unn, T., Sm ith, T. B., & Montoya, J. (2006). M ulticultural com petency instrum entation: A
review and analysis o f reliability generalization. Journal o f Counseling and Development,
84, 4 7 1 -4 8 2 .
Duval, S., & Tweedie, R. (2000). Trim and fill: A simple funnel-plot-based m ethod o f testing
and adjusting for publication bias in m eta-analysis. Biometrics, 56(2), 4 5 5 -4 6 3 .
*E ll, K., Katon, W , Xie, B., Lee, P. J., Kapetanovic, S., Guterm an, J., & Chou, C. P. (2010).
Collaborative care m anagem ent o f m ajor depression am ong low -incom e, predom inantly
Hispanic subjects with diabetes. Diabetes Care, 33(4), 7 0 6 -7 1 3 .
Eyerm an, R. (2004). The past in the present: Culture and the transm ission o f memory. Acta
Sociologica, 47 (2), 1 5 9 -1 6 9 . https://www.doi.org/10.1177/0001699304043853
*Falconer, J. (2002). The effectiveness o f a culturally relevant eating disorder prevention interven­
tion with African American college women (Unpublished doctoral dissertation). University
o f M issouri, Columbia.
*Feldm an, J. M ., M atte, L., Interian, A., Lehrer, P. M ., Lu, S. E., Scheckner, B. & Shim , C.
(2 0 1 6 ). Psychological treatm ent o f com orbid asthm a and panic disorder in Latino
adults: Results from a random ized controlled trial. Behaviour Research and Therapy, 87,
1 4 2 -1 5 4 .
*Flaskerud, J. H., & Akutsu, P D. (1993). Significant influence o f participation in ethnic-
specific program s on clinical diagnosis for Asian Am ericans. Psychological Reports, 72,
122 8 -1 2 3 0 .
*Flaskerud, J. H., & Hu, L. (1994). Participation in and outcom e o f treatm ent for m ajor depres­
sion am ong low incom e A sian-Am ericans. Psychiatry Review, 53, 2 8 9 -3 0 0 .
Frandsen, C., Sm ith, T. B., Griner, D., & Beecher, M . (2018). Beyond English only: Addressing
language interpretation in multicultural psychology training. M anuscript submitted for pub­
lication, Brigham Young University, Provo, UT.
**Fuertes, J. N. (2002). Clients' ratings o f counselor m ulticultural competency. Cultural
Diversity and Ethnic Minority Psychology, 8 (3), 2 1 4 -2 2 3 .
**Fuertes, J. N., Stracuzzi, T. I., Bennett, J., Scheinholtz, J., Mislowack, A., Hersh, M ., &
Cheng, D. (2006). Therapist m ulticultural com petency: A study o f therapy dyads.
Psychotherapy: Theory , Research , Practice, Training, 43 (4), 4 8 0 -4 9 0 .
*Fung, M. P, & Fox, R. A. (2014). The culturally-adapted Early Pathways program for young
Latino children in poverty: A randomized controlled trial. Journal o f Latina/o Psychology,
2 (3 ), 1 3 1 -1 4 5 .
*G allagher-Thom pson, D., Arean, P, Rivera, P , & Thom pson, L. W (2001). Psychoeducational
intervention to reduce distress in Hispanic fam ily caregivers: Results o f a pilot study.
Clinical Gerontologist, 2 3 (1 -2 ), 1 7-32.
**G am st, G., Dana, R. H., Meyers, L. S., D er-K arabetian, A., & Guarino, A. J. (2009). An anal­
ysis o f the m ulticultural assessm ent intervention process model. International Journal o f
Culture and Mental Health, 2 (1 ), 5 1 -6 4 .
*Garza, Y., & Bratton, S. C. (2005). School-based child-centered play therapy with Hispanic
children: O utcom es and cultural considerations. International Journal o f Play Therapy,
14(1), 5 1 -7 9 .
**G illispe, J. F. (1998). Construct validation o f the multicultural counseling inventory: The con­
tribution o f client satisfaction (Unpublished doctoral dissertation). Fielding Institute, Santa
Barbara, CA.
*G insburg, G. S., & Drake, K. L. (2002). School-based treatm ent for anxious A frican-
A m erican adolescents: A controlled pilot study. Journal o f the American Academy o f Child
& Adolescent Psychiatry, 4 1(7), 7 6 8 -7 7 5 .
 12 4 PS Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Gonsalves, J. P., Lucchetti, G., M enezes, P. R., & Vallada, H. (2015). Religious and spiritual
interventions in m ental health care: A systematic review and m eta-analysis o f randomized
controlled clinical trials. Psychological Medicine, 4 5(14), 2 9 3 7 -2 9 4 9 .
**G onzalez, J. (2015). Client outcome: An exploratory investigation o f multicultural competence
and the working alliance (D octoral dissertation). University o f Central Florida, Orlando.
*G onzalez, M. (2003). The effectiveness o f a culturally based social work activity group pin
the promotion o f ethnic identity and adaptive behaviors: A study o f Puerto Rican children
(D octoral dissertation). Catholic University o f Am erica, W ashington, DC.
Griner, D., & Sm ith, T. B. (2006). Culturally adapted m ental health intervention: A m eta­
analytic review. Psychotherapy: Theory, Research, Practice, Training, 43(4), 5 3 1 -5 4 8 .
*Grodnitzky, G. R. (1993). Hero modeling versus non-hero modeling as interventions fo r Puerto-
Rican and Anglo adolescents exhibiting behavior problems (Unpublished doctoral disserta­
tion). Hofstra University, Hempstead, NY.
*G uinn, B., & V incent, V. (2002). A health intervention on Latina spiritual well-being
constructs: An evaluation. Hispanic Journal o f Behavioral Sciences, 2 4(3), 3 7 9 -3 9 1 .
*Gutierrez, L. M ., & Ortega, R. (1991). Developing m ethods to empower Latinos: The im por­
tance o f groups. Social Work with Groups, 14(2), 2 3 -4 3 .
Hall, G. C. N., Ibaraki, A. Y., Huang, E. R., M arti, C. N., & Stice, E. (2016). A meta-analysis of
cultural adaptations o f psychological interventions. Behavior Therapy, 4 7(6), 9 9 3 -1 0 0 3 4 .
https://www.doi.org/10.1016fj.beth.2016.09.005
*H am m ond, W. R., & Yung, B. R. (1991). Preventing violence in at-risk A frican-A m erican
youth. Journal o f Health Care fo r the Poor and Underserved, 2 (3 ), 3 5 9 -3 7 3 .
Hayes, J. A., McAleavey, A. A., Castonguay, L. G., & Locke, B. D. (2016). Psychotherapists’
outcom es with W hite and racial/ethnic clients: First the good news. Journal o f Counseling
Psychology, 63(3), 2 6 1 -2 6 8 .
Hayes, J. A., Owen, J., & Bieschke, K. J. (2014). Therapist differences in symptom change with
racial/ethnic m inority clients. Psychotherapy, 5 2 , 3 0 8 -3 1 4 .
Helms, J. E., & Talleyrand, R. M. (1997). Race is not ethnicity. American Psychologist, 52(11),
1 2 4 6 -1 2 4 7 .
**H ijioka, S. (2012). Effects o f cultural com petency on etiology belief congruence and therapy
process am ong Asian A m erican clients. Dissertation Abstracts International, 72, 4984.
*H inton, D. E., Chhean, D., Pich, V , Safren S. A., Hofmann, S. G., & Pollack, M. H. (2005). A
randomized controlled trial o f cognitive-behavior therapy for Cam bodian refugees with
treatm ent-resistant PTSD and panic attacks: A cross-over design. Journal o f Traumatic
Stress, 18(6), 6 1 7 -6 2 9 .
*H inton, D. E., H ofm ann, S. G., Pollack, M. H., & O tto, M . W. (2009). M echanism s o f efficacy
o f C B T for Cam bodian refugees with PTSD : Im provement in em otion regulation and or­
thostatic blood pressure response. CNS Neuroscience & Therapeutics, 15(3), 2 5 5 -2 6 3 .
*Hinton, D. E., Hofmann, S. G., Rivera, E., Otto, M. W , & Pollack, M. H. (2011). Culturally adapted
C B T (CA -CBT) for Latino women with treatment-resistant PTSD: A pilot study comparing
CA -CBT to applied muscle relaxation. Behaviour Research and Therapy, 4 9 , 275-280.
*H inton, D. E., Pham, T., Tran, M ., Safren, S. A., Otto, M. W., & Pollack, M . H. (2004). C B T
for Vietnam ese refugees with treatm ent-resistant PTSD and panic attacks: A pilot study.
Journal o f Traumatic Stress, 17(5), 4 2 9 -4 3 3 .
*Hogue, A., Liddle, H. A., Becker, D., & Johnson-Leckrone, J. (2002). Fam ily-based prevention
counseling for high-risk young adolescents: Im mediate outcomes. Journal o f Community
Psychology, 30(1), 1 -2 2 .
 125 Cultural Adaptations and Multicultural Competence

**H ook, J. N., Davis, D. E., Owen, J., W orthington Jr, E. L., & Utsey, S. O. (2013). Cultural hu­
m ility: M easuring openness to culturally diverse clients. Journal o f Counseling Psychology,
6 0(3), 3 5 3 -3 6 6 .
*Huey, S. J., & Pan, D. (2006). Culture-responsive one-session treatm ent for phobic Asian
Am ericans: A pilot study. Psychotherapy: Theory, Research, Practice, Training, 43(4),
5 4 9 -5 5 4 .
Huey, S. J., & Polo, A. J. (2008). Evidence-based psychosocial treatm ents for ethnic m inority
youth. Journal o f Clinical Child & Adolescent Psychology, 37(1), 2 6 2 -3 0 1 . https://www.doi.
org/10.1080/153 74410701820174
*Huey, W. C., & Rank, R. C. (1984). Effects o f counselor and peer-led group assertive training
on Black adolescent aggression. Journal o f Counseling Psychology, 32 (1), 9 5 -9 8 .
*Hwang, W. C., Myers, H. F., Chiu, E., M ak, E., Butner, J. E., Fujim oto, K., & M iranda, J.
(2015). Culturally adapted cognitive-behavioral therapy for Chinese Am ericans with de­
pression: A randomized controlled trial. Psychiatric Services, 66(10), 1 0 35-1042.
Imel, Z. E., Baldwin, S., Atkins, D. C., Owen, J., Baardseth, T., & W ampold, B. E. (2011). Racial/
ethnic disparities in therapist effectiveness: A conceptualization and initial study o f cu l­
tural com petence. Journal o f Counseling Psychology, 58, 2 9 0 -2 9 8 .
*Jackson, P. A. (1997). The effect o f exposure to culturally relevant/historically based material on
level o f frustration tolerance, level o f depression, and mediation o f anger in African-American
young males (Unpublished doctoral dissertation). California School o f Professional
Psychology, Alameda.
*Jesse, D. E., Gaynes, B. N., Feldhousen, E. B., Newton, E. R., Bunch, S., & Hollon, S. D. (2015).
Perform ance o f a culturally tailored cognitive-behavioral intervention integrated in a
public health setting to reduce risk o f antepartum depression: A randomized controlled
trial. Journal o f Midwifery & Women’s Health, 60(5), 5 7 8 -5 9 2 .
*Johnson, D. L., & Breckenridge, J. N. (1982). The Houston Parent-Child Development Center
and the prim ary prevention o f behavior problem s in young children. American Journal o f
Community Psychology, 20(3), 3 0 5 -3 1 6 .
*Jones, L. V. (2008). Preventing depression: Culturally relevant group work with Black women.
Research on Social Work Practice, 28(6), 6 2 6 -6 3 4 .
*Jones, L. V., A hn, S., & Chan, K. T. (2016). Expanding the psychological wellness threshold
for Black college women: An exam ination o f the Claim ing Your C onnections intervention.
Research on Social Work Practice, 2 6(4), 3 9 9 -4 1 1 .
*Jones, L.V., & W arner, L. A. (2011). Evaluating culturally responsive group work with Black
women. Research on Social Work Practice, 22 (6), 7 3 7 -7 4 6 .
*Kaslow, N. J., Leiner, A. S., Revier, S., Jackson, E., Bethea, K., Bhaju, J., . . . Thom pson, M.
(2010). Suicidal, abused A frican A m erican women's response to a culturally inform ed in ­
tervention. Journal o f Consulting and Clinical Psychology, 78(4), 4 4 9 -4 5 8 .
Kim , B. S. K., Cartw right, B. Y., Asay, P. A., & DAndrea, M . J. (2003). A revision o f the
M ulticultural Awareness, Knowledge, and Skills Survey-C ounselor Edition. Measurement
and Evaluation in Counseling and Development, 3 6 , 1 6 1-180.
*K im , B. S. K., Om izo, M. M ., & D'Andrea, M . J. (1998). The effects o f culturally consonant
group counseling on the self-esteem and internal locus o f control orientation am ong
Native A m erican adolescents. Journal o f Specialists in Group Work, 2 3(2), 1 45-163.
*K ing, B. S. (1999). The effect o f a cultural-based life skills curriculum on American Indian ad­
olescent self-esteem and locus o f control (D octoral dissertation). University o f Arkansas,
Fayetteville.
 126 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

King, J., Trimble, J. E., M orse, G. S., & Thomas, L. R. (2014). North A m erican Indian and
Alaska Native spirituality and psychotherapy. In P. S. Richards (Ed.), Handbook o f psy­
chotherapy and religious diversity (2nd ed., pp. 4 5 1 -4 7 2 ). W ashington, DC: Am erican
Psychological Association.
Kocarek, C. E., Talbot, D. M ., Batka, J. C., & A nderson, M. Z. (2001). Reliability and validity
o f three m easures o f m ulticultural competency. Journal o f Counseling and Development,
79, 4 8 6 -4 9 6 .
*K ohn, L. P., O den, T., M unoz, R. E , Robinson, A., & Leavitt, D. (2002). Adapted cognitive be­
havioral group therapy for depressed low -incom e A frican A m erican women. Community
Mental Health Journal, 38, 4 9 7 -5 0 4 .
*Kopelowicz, A., Zarate, R., Sm ith, V. G ., M intz, J., & Liberm an, R. P. (2003). Disease m an­
agem ent in Latinos with schizophrenia: A family-assisted, skills training approach.
Schizophrenia Bulletin, 2 9(2), 2 1 1 -2 2 7 .
*Kung, W. W., Tseng, Y. F., Wang, Y., Hsu, P. C., & Chen, D. (2012). Pilot study o f ethni­
cally sensitive family psychoeducation for Chinese-A m erican patients with schizophrenia.
Social Work in Mental Health, 10(5), 3 8 4 -4 0 8 .
LaFrom boise, T. D., Colem an, H. L., & Hernandez, A. (1991). Developm ent and factor structure
o f the Cross-Cultural Counseling Inventory-Revised. Professional Psychology: Research
and Practice, 22 , 3 8 0 -3 8 8 .
*LaFrom boise, T., & Howard-Pitney, B. (1995). The Zuni life skills development cu rric­
ulum: Description and evaluation o f a suicide prevention program. Journal o f Counseling
Psychology, 4 2(4), 4 7 9 -4 8 6 .
*Lagom asino, I. T., Dw ight-Johnson, M ., Green, J. M ., Tang, L., Zhang, L., Duan, N., &
M iranda, J. (2016). Effectiveness o f collaborative care for depression in public-sector p ri­
m ary care clinics serving Latinos. Psychiatric Services, 68(4), 3 5 3 -3 5 9 .
Lam bert, M. J. (2013). Bergin and Garfield’s handbook o f psychotherapy and behavior change.
H oboken, NJ: Wiley.
La Roche, M ., & Christopher, M. S. (2008). Culture and em pirically supported treatments: On
the road to a collision? Culture & Psychology, 14(3), 3 3 3 -3 5 6 .
La Roche, M . J., & Lustig, K. (2010). Cultural adaptations: U npacking the m eaning o f culture.
Scientific Review o f Mental Health Practice, 7 (2), 2 6 -3 0 .
**Larrison, C. R., Schoppelrey, S. L., Hack-Ritzo, S., & Korr, W. S. (2011). Clinician factors
related to outcom e differences between Black and W hite patients at CM H Cs. Psychiatric
Services, 6 2(5), 5 2 5 -5 3 1 .
*Lau, A. S., Fung, J. J., Ho, L. Y., Liu, L. L., & Gudino, O. G. (2011). Parent training with high­
risk im m igrant Chinese families: A pilot group randomized trial yielding practice-based
evidence. Behavior Therapy, 4 2 , 4 1 3 -4 2 6 .
*Lau, A., & Zane, N. (2000). Exam ining the effects o f ethnic-specific services: An analysis of
cost-utilization and treatm ent outcom e for Asian A m erican clients. Journal o f Community
Psychology, 28 (1), 6 3 -7 7 .
*Le, H. N., Perry, D. F., & Stuart, E. A. (2011). Random ized controlled trial o f a preventive in ­
tervention for perinatal depression in high-risk Latinas. Journal o f Consulting and Clinical
Psychology, 79(2), 1 3 5-141.
**L i, L. C., & Kim , B. S. K. (2004). Effects o f counseling style and client adherence to Asian
cultural values on counseling process with Asian A m erican college students. Journal o f
Counseling Psychology, 51(2), 1 5 8-167.
 12 7 Cultural Adaptations and Multicultural Competence

*Lopez-Appelo, S. G. (2007). Use o f biopsychosocial services, cultural formulation and cul­

turally sensitive service integration in public sector mental health: Impact on outcomes fo r
Hispanic adult outpatients (Unpublished doctoral dissertation). Seattle Pacific University,
Seattle, WA.
Lott, B. (2002). Cognitive and behavioral distancing from the poor. American Psychologist,
5 7(2), 1 0 0 -1 1 0 .
Lyons, L. C., & M orris, W A. (2017). The meta-analysis calculator. Retrieved from http://www.
lyonsmorris.com/ma1/index.cfm
*Malgady, R. G ., Rogler, L. H., & Constantino, G. (1990a). Culturally sensitive psychotherapy
for Puerto Rican children and adolescents: A program o f treatm ent outcom e research.
Journal o f Consulting and Clinical Psychology, 58(6), 7 0 4 -7 1 2 .
*Malgady, R. G., Rogler, L. H., & Costantino, G. (1990b). Hero/heroine m odeling for Puerto
Rican adolescents: A preventive m ental health intervention. Journal o f Counseling
Psychology, 58 , 4 6 9 -4 7 4 .
*M artinez, C. R. Jr., & Eddy, J. M. (2005). Effects o f culturally adapted parent m anagem ent
training on Latino youth behavioral health outcomes. Journal o f Consulting and Clinical
Psychology, 7 3(4), 8 4 1 -8 5 1 .
M artinez, J. S., Sm ith, T. B., & Barlow, S. H. (2007). Spiritual interventions in psycho­
therapy: Evaluations by highly religious clients. Journal o f Clinical Psychology, 63(10),
9 4 3 -9 6 0 .
*M atos, M ., Bauermeister, J. J., & Bernal, G. (2009). Parent-child interaction therapy for
Puerto R ican preschool children with A D H D and behavior problems: A pilot study. Family
Process, 4 8(2), 2 3 2 -2 5 2 .
*M ausbach, B. T., Bucardo, J., M cKibbin, C .L., Cardenas, V., & Barrio, C. (2008). Evaluation
o f a culturally tailored skills intervention for Latinos with persistent psychotic disorders.
American Journal o f Psychiatric Rehabilitation, 11(1), 6 1 -7 5 .
*M cC abe, K., & Yeh, M. (2009). Parent-child interaction therapy for M exican Am ericans: A
randomized clinical trial. Journal o f Clinical Child & Adolescent Psychology, 38(5), 7 5 3 -7 5 9 .
M cCabe, K. Yeh, M ., Garland, A. F., Lau, A. S., & Chavez, G. (2005). The GANA program : A
tailoring approach to adapting parent child interaction therapy for M exican Americans.
Education and Treatment o f Children, 2 8(2), 1 1 1-129.
**M cC a n n , N. B. (2006). Client perceptions o f multicultural counseling competence and
their impact on effective treatment: An exploration o f White counselor/African American
client dyads (U npublished d octoral dissertation). Sam Houston State University,
Huntsville, T X .
M cIntosh, P (1988). White privilege and male privilege: A personal account o f coming to
see correspondences through work in Women’s Studies. W orking Paper 189. Wellesley,
M A: W ellesley College C enter for Research on Women.
**M enapace, B. (1997). Correlates o f effectiveness fo r White psychotherapists working with
African American clients: Multicultural competence, clinical experience, racial consciousness,
multicutural training, and social experience (Unpublished doctoral dissertation). California
School o f Professional Psychology, Alameda.
National A ssociation o f School Psychologists. (2010). Principles fo r professional ethics.
Retrieved from https://www.nasponline.org/
*M ickens-English, P (1996). The efficacy o f an Afrocentric/holistic group psychotherapy ap­
proach fo r black women (D octoral dissertation). Kent State University, Kent, OH.
 128 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

*M iller, L. D., Laye-Gindhu, A., Bennett, J. L., Liu, Y., Gold, S., M arch, J. S., . . . W aechtler, V E.
(2011). An effectiveness study o f a culturally enriched school-based C B T anxiety preven­
tion program . Journal o f Clinical Child & Adolescent Psychology, 4 0(4), 6 1 8 -6 2 9 .
*M iranda, J., Duan, N., Sherbourne, C., Schoenbaum , M., Lagomasino, I., Jackson-Triche, M ., &
Wells, K. B. (2003). Improving care for m inorities: Can quality improvement interventions
improve care and outcom es for depressed m inorities? Results o f a randomized, controlled
trial. Health Services Research, 38(2), 6 1 3 -6 3 0 .
*M okuau, N., Braun, K. L., Wong, L. K., Higuchi, P., & Gotay, C. C. (2008). Development
o f a family intervention for native Hawaiian women with cancer: A pilot study. National
Association o f Social Workers. 53(1), 9 -1 9 .
*M oran, J. R. (1999). Preventing alcohol use am ong urban A m erican Indian youth: The seventh
generation program. Journal o f Human Behavior in the Social Environment, 2 (1 -2 ) , 5 1 -6 7 .
*M yers, H. F., Alvy, K. T., A rrington, A., Richardson, M. A., M arigna, M ., Huff, R., . . .
Newcomb, M.D. (1992). The im pact o f a parent training program on inner-city A frican-
A m erican families. Journal o f Community Psychology, 2 0 , 1 3 2-147.
Norcross, J. C., & W ampold, B. E. (2011). W hat works for whom: Tailoring psychotherapy to
the person. Journal o f Clinical Psychology, 67(2), 1 27-132.
*O choa, M . L. (1981). Group counseling Chicana troubled youth: An exploratory group coun­
seling project (D octoral dissertation). University o f Massachusetts, Amherst.
Okazaki, S., David, E. J. R., & Abelm ann, N. (2008). Colonialism and psychology o f culture.
Social and Personality Psychology Compass, 2(1), 9 0 -1 0 6 .
*O tto, M . W., Hinton, D., Korbly, N. B., Chea, A., Ba, P., Gershuny, B. S., & Pollack, M . H.
(2003). Treatm ent o f pharm acotherapy-refractory posttraum atic stress disorder among
Cam bodian refugees: A pilot study o f com bination treatm ent with cognitive-behavior
therapy vs sertraline alone. Behaviour Research and Therapy, 4 1 , 1 271-1276.
**O w en, J., Leach, M . M ., Rodolfa, E., & W ampold, B. (2011). Client and therapist variability
in clients' perceptions o f their therapists' multicultural com petencies. Journal o f Counseling
Psychology, 58(1), 1 -9 .
**O w en, J., Tao, K. W , Drinane, J. M ., Hook, J., Davis, D. E., & Kune, N. F. (2016). Client
perceptions o f therapists’ m ulticultural orientation: Cultural (missed) opportunities and
cultural humility. Professional Psychology: Research and Practice, 47(1), 3 0 -3 7 .
**O w en, J. J., Tao, K., Leach, M . M ., & Rodolfa, E. (2011). Clients' perceptions o f their
psychotherapists' m ulticultural orientation. Psychotherapy, 48 (3), 2 7 4 -2 8 2 .
*Pan, D., Huey, S. J. Jr., & Heflin, L. H. (2017). Ethnic differences in response to directive
vs. non-directive brief intervention for subsyndromal depression. Psychotherapy Research,
2 9 (2 ), 1 8 6 -1 9 7 .
*Pan, D., Huey, S. R., & Hernandez, D. (2011). Culturally adapted versus standard exposure
treatm ent for phobic Asian Am ericans: Treatm ent efficacy, m oderators, and predictors.
Cultural Diversity and Ethnic Minority Psychology, 17(1), 11-2 2 .
Parra Cardona, J. R., D om enech Rodríguez, M ., Forgatch, M ., Sullivan, C., Bybee, D., Holtrop,
K., & Bernal, G. (2012). Culturally adapting an evidence-based parenting intervention for
Latino imm igrants: The need to integrate fidelity and cultural relevance. Family Process,
5 1(1), 5 6 -7 2 .
*Patterson, T. L., Bucardo, J., M cKibbin, C. L., M ausbach, B. T., M oore, D., Barrio, C., . . . Jeste,
D. V. (2005). Development and pilot testing o f a new psychosocial intervention for older
Latinos with chronic psychosis. Schizophrenia Bulletin, 31(4), 9 2 2 -9 3 0 .
 129 Cultural Adaptations and Multicultural Competence

*Penedo, F. J., Traeger, L., Dahn, J., M olton, I., Gonzalez, J. S., Schneiderm an, N., & A ntoni, M.
H. (2007). Cognitive behavioral stress m anagem ent intervention improves quality o f life
in Spanish m onolingual Hispanic m en treated for localized prostate cancer: Results o f a
randomized controlled trial. International Journal o f Behavioral Medicine, 14(3), 1 6 4-172.
Peters, J. L., Sutton, A. J., Jones, D. R., Abrams, K. R., & Rushton, L. (2006). Com parison o f two
m ethods to detect publication bias in m eta-analysis. JAMA, 2 9 5 (6 ), 6 7 6 -6 8 0 .
**Peterson, M. G. (2013). Therapist cultural intelligence as a moderator o f working alliance
and outcome in multicultural counseling (Unpublished doctoral dissertation). University of
M innesota, M inneapolis.
Phillips, F. B. (1990). N TU psychotherapy: A n A frocentric approach. Journal o f Black
Psychology, 17(1), 5 5 -7 4 .
Phinney, J. S. (1996). W hen we talk about A m erican ethnic groups, what do we mean?.
American Psychologist, 51(9), 9 1 8 -9 2 7 .
Ponterotto, J. G. (1996). Evaluating and selecting research instrum ents. In F. T. L. Leong & J. T.
Austin (Eds.), The psychology research handbook: A guide fo r graduate students and research
assistants (pp. 7 3 -8 4 ). Thousand Oaks, CA: SAGE.
Ponterotto, J. G., & Alexander, C. M. (1996). Assessing the m ulticultural com petence of
counselors and clinicians. In L. A. Suzuki, P. J. Meller, & J. G. Ponterotto (Eds.), Handbook
o f multicultural assessment (pp. 6 5 1 -6 7 2 ). San Francisco, CA: Jossey-Bass.
Ponterotto, J. G ., G retchen, D., Utsey, S. O., Rieger, B. P., & Austin, R. (2002). A revision of
the M ulticultural Counseling Awareness Scale. Journal o f Multicultural Counseling and
Development, 30(3), 1 5 3-180.
Ponterotto, J. G., Rieger, B. P., Barrett, A., & Sparks, R. (1994). Assessing m ulticultural coun­
seling com petence: A review o f instrum entation. Journal o f Counseling & Development,
7 2 ,3 1 6 -3 2 2 .
Pope-Davis, D. B., & Dings, J. G. (1994). An em pirical com parison o f two self-report m ulticul­
tural counseling com petency inventories. Measurement and Evaluation in Counseling and
Development, 27 , 9 3 -1 0 2 .
Pope-Davis, D. B., & Dings, J. G. (1995). The assessment o f m ulticultural counseling
com petencies. In J. G. Ponterotto, J. M . Casas, L. Suzuki, & C. A lexander (Eds.), Handbook
o f multicultural counseling (pp. 2 8 7 -3 1 1 ). Thousand Oaks, CA: SAGE.
Pope-Davis, D. B., Toporek, R. L., O rtega-Villalobos, L., Ligiéro, D. P., Brittan-Pow ell, C. S.,
Liu, W M ., . . . Liang, C. T. H. (2002). Client perspectives o f m ulticultural counseling co m ­
petence: A qualitative exam ination. The Counseling Psychologist, 30(3), 3 5 5 -3 9 3 . https://
www.doi.org/10.1177/0011000002303001
*Ram irez, S. Z., Jain, S., Flores-Torres, L. L., Perez, R., & Carlson, R. (2009). The effects o f
cuento therapy on reading achievem ent and psychological outcom es o f M exican-A m erican
students. Professional School Counseling, 12(3), 2 5 3 -2 6 2 .
*R obinson, W L., Case, M. H., W hipple, C. R., G ooden, A. S., Lopez-Tamayo, R., Lambert, S.
F., & Jason, L. A. (2016). Culturally grounded stress reduction and suicide prevention for
A frican A m erican adolescents. Practice Innovations, 1(2), 1 1 7-128.
Rosenthal, L. (2016). Incorporating intersectionality into psychology: An opportunity to p ro­
m ote social justice and equity. American Psychologist, 71(6), 4 7 4 -4 8 5 .
*Rosselló, J., & Bernal, G. (1999). The efficacy o f cognitive-behavioral and interpersonal
treatm ents for depression in Puerto Rican adolescents. Journal o f Consulting and Clinical
Psychology, 6 7(5), 7 3 4 -7 4 5 .
 130 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

*Rossello, J., Bernal, G., & Rivera-M edina, C. (2008). Individual and group C B T and IP T
for Puerto Rican adolescents with depressive symptoms. Cultural Diversity and Ethnic
Minority Psychology, 14(3), 2 3 4 -2 4 5 .
*Rowland, M . D., Halliday-Boykins, C. A., Henggeler, S. W., Cunningham , P. B., Lee, T. G.,
Kruesi, M . J. P., & Shapiro, S. B. (2005). A randomized trial o f m ultisystemic therapy with
Hawaii’s Felix Class youths. Journal o f Emotional and Behavioral Disorders, 13(1), 13-2 3 .
*Royce, D. (1998). M entoring high-risk m inority youth: Evaluation o f the Brothers project.
Adolescence, 33 (1 29), 1 4 5-158.
*Santisteban, D. A., Coatsworth, J. D., Perez-Vidal, A., Kurtines, W. M ., Schwartz, S. J.,
LaPerriere, A., & Szapocznik, J. (2003). Efficacy o f brief strategic family therapy in
m odifying Hispanic adolescent behavior problem s and substance use. Journal o f Family
Psychology, 17(1), 1 2 1-133.
**Sarm iento, I. A. (2012). Perspectives on the therapeutic process: An examination o f cultural
competence and the working alliance among college students in treatment (D octoral disser­
tation). Clark University, Worcester, MA.
*Schwarz, D. A. (1989). The effect o f a Spanish pre-therapy orientation videotape on Puerto
Rican clients’ knowledge about psychotherapy, improvement in therapy, attendance patterns
and satisfaction with services (Unpublished doctoral dissertation). Temple University,
Philadelphia, PA.
*Shin, S. K. (2004). Effects o f culturally relevant psychoeducation for Korean Am erican
families o f persons with chronic m ental illness. Research on Social Work Practice, 14(4),
2 3 1 -2 3 9 .
*Shin, S. K., & Lukens, E. P. (2002). Effects o f psychoeducation for Korean Am ericans with
chronic m ental illness. Psychiatric Services, 53(9), 1 1 25-1131.
Sidani, S., Ibrahim , S., Lok, J., Fan, L., Fox, M ., & Guruge, S. (2017). An integrated strategy for
the cultural adaptation o f evidence-based interventions. Health, 9, 7 3 8 -7 5 5 .
*Sim oni, J. M ., W iebe, J. S., Sauceda, J. A., Huh, D., Sanchez, G., Longoria, V., . . . Safren, S.
A. (2013). A prelim inary R C T o f C B T -A D for adherence and depression am ong H IV ­
positive Latinos on the U S-M exico border: The Nuevo D ia study. AIDS and Behavior,
17(8), 2 8 1 6 -2 8 2 9 .
Sm ith, L., & Brewster, M . E. (2015). Counseling in the context o f poverty. In P. B. Pedersen, W.
J. Lonner, J. G. Draguns, J. E. Trimble, & M. R. Scharron-del Rio (Eds.), Counseling across
cultures (7th ed., pp. 3 6 1 -3 8 2 ). Thousand Oaks, CA: SAGE.
Sm ith, T. B. (2010). Culturally congruent practices in counseling and psychotherapy: A re­
view o f research. In J. G. Ponterotto, J. M. Casas, L. A. Suzuki, & C. M. A lexander (Eds.),
Handbook o f multicultural counseling (3rd ed., pp. 4 3 9 -4 5 0 ). Thousand Oaks, CA: SAGE.
Sm ith, T. B. (Ed.). (2004). Practicing multiculturalism: Affirming diversity in counseling and
psychology. Boston, M A: Allyn & Bacon.
Sm ith, T. B., Bartz, J., & Richards, P. S. (2007). Outcom es o f religious and spiritual adaptations
to psychotherapy: A m eta-analytic review. Psychotherapy Research, 17(6), 6 4 3 -6 5 5 .
Sm ith, T. B., Constantine, M . G ., D unn, T., Dinehart, J., & Montoya, J. A. (2006). M ulticultural
education in the m ental health professions: A m eta-analytic review. Journal o f Counseling
Psychology, 5 3 , 1 3 2-145.
Sm ith, T. B., D om enech Rodriguez, M. D., & Bernal, G. (2011). Culture. Journal o f Clinical
Psychology, 67(2), 1 6 6-175.
Sm ith, T. B., & Draper, M . (2004). Understanding individuals in their context: A relational
perspective o f m ulticultural counseling and psychotherapy. In T.B. Sm ith (Ed.), Practicing
 131 Cultural Adaptations and Multicultural Competence

multiculturalism: Affirming diversity in counseling and psychology (pp. 3 1 3 -3 2 3 ). Boston,

M A: Allyn & Bacon.
Sm ith, T. B., Soto, A., Griner, D., & Trimble, J. E. (2015). Multicultural counseling foundations: A
synthesis o f research findings on selected topics. In P. B. Pedersen, J. Draguns, W Lonner,
J. Trimble, & M . Scharrón del Río (Eds.), Counseling across cultures (7th ed., pp. 7 5 -9 6 ).
Thousand O aks, CA: SAGE.
Sm ith, T. B., & Trimble, J. E. (2016). Foundations o f multicultural psychology: Research to in­
form effective practice. W ashington, DC: A m erican Psychological Association.
Sodowsky, G. R., Taffe, R. C., Gutkin, T. B., & W ise, S. L. (1994). Developm ent o f the
M ulticultural Counseling Inventory: A self-report m easure o f m ulticultural competencies.
Journal o f Counseling Psychology, 4 1 , 1 3 7-148.
Sue, S. (1998). In search o f cultural com petence in psychotherapy and counseling. American
Psychologist, 5 3(4), 4 4 0 -4 4 8 .
*Szapocznik, J., Rio, A., Perez-Vidal, A., Kurtines, W , Hervis, O., & Santisteban, D. (1986).
Bicultural Effectiveness Training (B E T ): An experim ental test o f an intervention m o ­
dality for families experiencing intergenerational/intercultural conflict. Hispanic Journal
o f Behavioral Sciences, 8(4), 3 0 3 -3 3 0 .
*Szapocznik, J., Santisteban, D., Rio, A., Perez-Vidal, A., Santisteban, D., & Kurtines, W. M.
(1989). Family Effectiveness Training: An intervention to prevent drug abuse and problem
behaviors in Hispanic adolescents. Hispanic Journal o f Behavioral Sciences, 11(1), 4 -2 7 .
Tao, K. W , Owen, J., Pace, B. T., & Imel, Z. E. (2015). A m eta-analysis o f multicultural
com petencies and psychotherapy process and outcom e. Journal o f Counseling Psychology,
62, 3 3 7 -3 5 0 .
*Telles, C., Karno, M ., M intz, J., Paz, G., Arias, M ., Tucker, D., & Lopez, S. (1995). Im m igrant
families coping with schizophrenia: Behavioural fam ily intervention v. case m anagem ent
with a low- incom e Spanish-speaking population. British Journal o f Psychiatry, 167(4),
4 7 3 -4 7 9 .
*Tim berlake, T. L. (2000). A comprehensive approach to social skills training with urban
African American adolescents (Unpublished doctoral dissertation). Temple University,
Philadelphia, PA.
**T im b o, D. M. (2016). African American clients' perceptions o f therapists' multicultural coun­
seling competencies (D octoral dissertation). Carlow University, Pittsburgh, PA.
U S D epartm ent o f Health and Human Services. (2001). National standards fo r culturally and
linguistically appropriate services in health care: Final report. W ashington, DC: Author.
Vasquez, M. J. (2007). Cultural difference and the therapeutic alliance: An evidence-based
analysis. American Psychologist, 62(8), 8 7 8 -8 8 5 .
van Loon, A., van Schaik, A., Dekker, J., & Beekm an, A. (2013). Bridging the gap for ethnic
m inority adult outpatients with depression and anxiety disorders by culturally adapted
treatm ents. Journal o f Affective Disorders, 1 4 7 (1 -3 ), 9 -1 6 .
^Villalobos, G. (2009). The mediating effect o f acculturation on the effectiveness o f culturally
adapted cognitive behavioral therapy with Mexican Americans suffering from depression
(Unpublished doctoral dissertation). University o f Texas, Austin.
*V illarreal, C. E. (2008). School-based child parent relationship therapy (CPRT) with Hispanic
parents (Unpublished doctoral dissertation). Regent University, Virginia Beach, VA.
**W ard, E. C. (2002). Cultural competence, cultural mistrust, working alliance and racial and
ethnic minority clients’ experience o f counseling: A mixed method study (Unpublished d oc­
toral dissertation). University o f W isconsin, Madison.
 132 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

*W iniarski, M .G ., Beckett, E., & Salcedo, J. (2005). Outcom es o f an inner-city H IV m ental

health program m e integrated with prim ary care and emphasizing cultural responsiveness.
AIDS Care, 17(6), 7 4 7 -7 5 6 .
W orthington, R. L., Mobley, M ., Franks, R. P , & Tan, J. A. (2000). M ulticultural counseling
com petencies: Verbal content, counselor attributions, and social desirability. Journal o f
Counseling Psychology, 47, 4 6 0 -4 6 8 .
Yeh, C. J., Parham, T. A., Gallardo, M. E., & Trimble, J. E. (2011). Culturally adaptive counseling
skills: Demonstrations o f evidence-based practices. Thousand Oaks, CA: SAGE.
 5

GENDER ID EN TITY

Stephanie L. Budge and Bonnie Moradi

The social meanings that are tied to gender permeate societies around the world and
often dictate how humans feel, think, and act (Gilbert & Scher, 2009). Gender is de­
fined by systems of power that shape and are shaped by gender norms and hierarchies
that intersect with other axes of power (such as race, sexual orientation, and class).
These norms and hierarchies disadvantage transgender people relative to cisgender
people and disadvantage women relative to men (e.g., Bond & Allen, 2016; Moradi
& Yoder, 2011; Serano, 2016). This conceptualization of gender as power and social
context contrasts with how gender is typically operationalized in psychotherapy re­
search. In psychotherapy research, gender is often treated as a binary demographic
variable whereby women/female clients are compared with men/male clients, and all
clients are assumed to be cisgender. In this narrow approach, clients with transgender
or nonbinary gender identities are rendered invisible and gender as a dimension of
power, including oppression and privilege, is not examined.
In its infancy, psychotherapy was dominated by gendered theories, such as Freudian
theories focusing on gender (specifically in relation to one’s parents) as the etiology
of most presenting concerns (Brown, 1994; Burman et al., 1998). In more modern
theories of change and psychotherapy, it is less clear what role gender plays in the ef­
fectiveness of psychotherapy. It might be reasonable to hypothesize that gender as a
demographic variable does not play a large role in the overall process of change and
effectiveness of treatments, given the evidence on common factors (Wampold & Imel,
2015) and meta-analyses indicating a lack of gender demographic differences across
a multitude of psychosocial domains (Hyde, 2005). However, researchers have pos­
ited that gender may shape treatment outcomes due to differential rates of certain
diagnoses (e.g., Felmingham & Bryant, 2012; Ogrodniczuk, 2006) or may impact the
process of how specific types of interventions are communicated—for example, how
therapists self-disclose or embed gender stereotypes within explanations for treatment
(Haddock & Lyness, 2002; Heru et al., 2006), or how long clients remain in therapy
(Cottone et al., 2002; Swift et al., 2013). Despite these speculations, empirical studies

133
 134 PS Y C H O T H E R A P Y R EL AT IO N S H IP S THAT WORK

have yielded mixed results as to whether gender impacts such psychotherapy processes
and outcomes (Clarkin & Levy, 2004; Ogrodniczuk, 2006; Owen et al., 2009).
Gender, conceptualized or measured as a demographic variable, may not yield sta­
tistically reliable or meaningful differences in psychotherapy process and outcome
research. Nevertheless, attending to gendered social contexts and systems of power,
including gendered exposure to risk factors that promote psychological symptoma­
tology, remains important. In this chapter, we focus on psychotherapy outcomes for
treatments that address gender, including transgender populations. Specifically, we re­
view the research evidence on the effectiveness of adapting or tailoring psychotherapy
to clients’ gender identity, which we conceptualize more as a system of power than as
an identity variable in psychotherapy. We begin by providing definitions and meas­
ures of gender identity, clinical examples, and landmark studies. We then summarize
our search for studies that either (a) conducted randomized controlled psychotherapy
trials focusing on transgender clients or (b) compared the outcomes of gender-focused
psychotherapies (for all individuals, regardless of gender identity) with another bona
fide psychotherapy. The search indicated there were not enough studies that fit the
search criteria; thus meta-analyses were not conducted. However, we summarize a
content analysis (k = 108) with studies focusing on psychotherapy with transgender
individuals. The chapter concludes by highlighting the limitations of the research base,
the training implications, and the clinical practices that focus on gender, including
gendered systems of power, oppression, and privilege.

DEFINITIONS
Feminist scholars (e.g., Bem, 1993, West & Zimmerman, 1987) distinguish sex from
gender. Sex reflects the biological and anatomical characteristics used to assign people
at birth to sex categories (e.g., male, female, intersex). Gender is the social meaning or
collection of characteristics prescribed to sex categories in a given society or culture.
Gender identity reflects one’s sense of self and identification (e.g., gender nonbinary,
genderqueer, male/man, female/woman). Gender expression reflects the variety of ways
in which people communicate their gender and gender identity in a given sociocul­
tural context (e.g., hair, clothes, voice).
Transgender is an umbrella term that captures a variety of people whose gender
identity is different from their assigned sex at birth. In addition, nonbinary gender
identities describe individuals who identify outside of the man/woman binary or
along a spectrum between man/woman. Cisgender is a term that describes individuals
whose gender identity is the same as the sex they were assigned at birth. Cisnormativity
describes the collection of individual and systemic biased assumptions that all people
are and should be cisgender and binary (man/woman) and, by extension, that there is
something wrong with people that do not fit within this standard.
Transphobia is defined as prejudicial attitudes, behaviors, and systems that deni­
grate transgender individuals; it can span from subtle forms (microaggression) to overt
forms (violence). Internalized transphobia is the internalization of such prejudice, such
that societal bias is turned inward toward the self and results in shame, guilt, and
 135 Gender Identity

internalized stigmatization. Androcentrism is centering men’s (most often cisgender

men’s) experiences as the normative, best, and focal experiences in defining human
experience, in turn marginalizing the experiences of anyone who does not identify as
a cisgender man as a suboptimal deviation of the norm. Relatedly, sexism describes
prejudicial attitudes, behaviors, and systems that define women and femininity in re­
strictive ways, oppress women (transgender inclusive) relative to men, and are rooted
in the notion that women are inferior to men.

MEASURES
As noted in Chapter 9, sex and sexual orientation variables should be assessed sep­
arately from gender identity, with an understanding that there can be some overlap
in certain constructs. The GenlUSS Group (2014) recommends separately assessing
individuals’ assigned sex at birth (male or female) and whether they are also intersex.
These variables, in turn, can be assessed separately from gender identity and gender
expression. Gender expression can be assessed continuously to include masculine and
feminine continua.
In the largest survey to assess a range of gender identities in the United States
(James et al., 2016), researchers asked individuals to respond to the following ques­
tion: “If you had to choose only one of the following terms, which best describes your
current gender identity? (Please choose only one answer)” and offered the following
options: crossdresser, man, woman, trans woman (M TF), trans man (FTM ), and
nonbinary/genderqueer. Participants were then instructed to fill out a text box to de­
scribe their gender identities. This open-ended description allows participants to have
agency over how researchers quantify their gender identity (rather than researchers
categorizing participants’ gender identities) and also enables participants to describe
their gender identity label(s) for research purposes.
Beyond assessing gender as identity and demographic variables, a few measures as­
sess feminist psychotherapy behaviors. For example, the Therapy with Women Scale
(Robinson & Worell, 1991) is a 40-item measure that attempts to differentiate feminist
therapy behaviors from other types of therapy behaviors, specifically by assessing how
the therapist and client share power in the therapeutic relationship (“I establish an
egalitarian relationship with my client”) and address affirming women in therapy (“I
support and value my female clients’ relationships with other women”). The Feminist
Couple Therapy Scale (McGeorge et al., 2009) is a 38-item measure that assesses “the
degree to which couple and family therapists agreed with principles and utilized
practices associated with promoting equality in heterosexual couple relationships”
(p. 198).
A third measure is the Feminist Family Therapist Behavior Checklist (Chaney &
Piercy, 1988), and its subsequent versions, the Feminist Therapy Behaviors (Juntunen
et al., 1994) and the Feminist Therapy Behaviors-Revised (FTB-R; Moradi et al., 2000).
These measures are applicable across psychotherapy formats (e.g., individual, family)
and assess use of feminist therapy behaviors with clients of all genders. Consistent
with a focus on systems of power, an empirical study using the FTB-R found that the
 13 6 PS Y C H O T H E R A P Y R EL AT IO N S H IP S THAT WORK

strongly identified feminist therapists were differentiated from other therapists by en­
gaging in more FTB-R behaviors that reflected attention to systems of oppression (e.g.,
sexism, racism, heterosexism) and socialization (Moradi et al., 2000).
Despite the availability of these clinical measures and their value in addressing
gender as social context and system of power, psychotherapy researchers typically
assess and use gender as a dichotomous moderating variable (man/woman) to de­
termine if there are gender demographic differences in outcomes. These binary
operationalizations also mean that the inclusion or exclusion of transgender people in
these studies remains unknown. Transgender people may be included in these studies
but not categorized by their affirmed gender, they may be excluded by researchers, or
they may opt out from participation. These limitations suggest error in measurement
of gender as a demographic variable within psychotherapy studies and meta-analyses
that impede interpretation of the results of gender comparisons and the generaliza-
bility of these results to transgender people.
Because it is not typical for researchers to go beyond assessing gender as a demo­
graphic variable in psychotherapy research, many measures have not been validated or
adapted for use with people with a diversity of gender identities. Budge and colleagues
(2017) provide recommendations for psychotherapy researchers interested in using
measures that have been validated/adapted for participants with specific gender
identities. These recommendations span information for how to include commonly
used measures in psychotherapy research (such as the Beck Depression Inventory),
how to include measures that focus on minority stress, and reminders to critically eval­
uate measures that focus on gender, as they could be outdated.

CLINICAL EXAMPLES
Feminist psychotherapy is a constructivist approach that incorporates principles of
person-centered therapies into technically eclectic treatment methods, while simul­
taneously focusing on the politics of clinical practice, critically addressing gender in
psychotherapy, and infusing the client’s social location into all interventions (Brown,
2006). As such, feminist therapy is applicable to clients of all genders. A core aspect
of feminist therapy also includes the power relationship between the therapist and the
client (Worrell & Remer, 2002).
Gender aware therapy (GAT) is another approach that infuses feminist princi­
ples into therapy (Good et al., 1990). GAT employs five overarching principles: (a)
ensuring that gender is integral to the treatment plan, (b) using social context to un­
derstand a client’s presenting concern, (c) infusing activism into the therapeutic pro­
cess, (d) working toward a collaborative therapeutic relationship, and (e) respecting
clients’ freedom to choose how they choose to enact gender roles. Although feminist
therapy and GAT are applicable to clients of all genders, explicit articulation of using
these approaches with transgender people remains limited.
In influential analogue research (Salierno, 2000), four clinical case examples were
provided to 97 participants to determine if the type of therapy (feminist vs. cognitive
behavioral) and client gender were related to the perception of effective outcome. The
 137 Gender Identity

vignette here focuses on a cisgender male therapist conducting feminist therapy with
a cisgender male client (from Salierno, 2000, p. 129). The case history indicates that
James (the client) has been seeing the therapist (Mr. Jones) for about three months.
The presenting concern for James is primarily related to difficulties in his marriage and
ensuing depression.

d r . jo n e s: It sounds like career is very important to you.

ja m e s : Yes, I think career is, especially for a man. You can’t really feel like a man un­
less you make your mark in your career.
d r . jo n e s: So a man who stays at home to raise children while his wife works is
not a real man. I think we could come up with a list of qualities that makes a man
without even mentioning work, like . . . someone who is caring, strong, honest,
and giving.
ja m e s : Well, maybe it’s not so much about what I think, but we all have to live in the
real world. It’s about what people expect of you.
d r . jo n e s: You know what we’ve discussed about this, just because society has un­
fair expectations about what a woman should be like or what a man should be like
doesn’t mean we have to buy into it. We have to decide these things for ourselves.
ja m e s : I guess if it were totally up to me, I would rather come home earlier and spend
more time with my family.
d r . jo n e s: Perhaps then, that’s what we should be working on.
ja m e s : You mean how to be able to not work so late all of the time. That would be
difficult, because the boss is so unreasonable and intimidating. I’ll just get yelled at
and embarrassed in front of the whole company like the last person who tried to
stand up to the boss.
d r . jo n e s: Tell me, what kind of expectations do you think that working late will set
up in your boss’s mind?
ja m e s : I guess I’m showing that working late means I care about the job.
d r . jo n e s: So if you ever try to stop working late . . .
ja m e s : They’ll think I don’t care anymore. So you’re telling me that I’ll be stuck
working late forever unless I do something about it. I’m sure not looking forward to
explaining this to my boss.
d r . jo n e s: It’s always difficult to confront someone who has power over you.
Whenever there is such a difference in power in a relationship, it puts the less pow­
erful person at a disadvantage and that person can easily be taken advantage of,
just like the employees in your company. It’s just like what we talk about in terms of
differences between men and women. Women have less power, and they are paid
less for the same work than men. But these are things we can do to help you deal
with this situation. The first thing I want to discuss with you are some assertiveness
training techniques. . . .

In this excerpt, the therapist addresses gender role expectations and discusses
power with the client, which are central aspects of feminist therapy. Several additional
points in this excerpt warrant critical analysis. First, the statement about a man who
 138 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

stays home and his wife who works outside the home communicates an implicit het­
eronormative assumption (i.e., men are heterosexual and have wives) that should be
avoided. Instead, a more exploratory and open-ended response could be used to help
the client deepen his understanding of social norms versus his own values, for ex­
ample, “Can you tell me more about what it means to feel like a man and what it means
for you personally to feel good about yourself? I wonder how these things might be
similar and different.”
A second point of caution about this excerpt is the therapist emphasizes individual
agency without sufficient attention to the consequences of that agency in gender and
class systems of power. For example, a realistic consequence of assertiveness in this
example might be that the client loses his job. Thus, while the therapeutic intervention
may include assertiveness training, there must also be deep empathy, full exploration
of costs and benefits, and additional strategies to help mitigate potential costs of the
client’s assertiveness.
A recent case study described the process and outcome of psychotherapy with a
client seeking a letter for hormones. In this exchange (from Budge, 2015, p. 289), the
therapist (Stephanie) and the client (Liz, a pseudonym) are discussing the possible
rupture on the relationship due to the therapist’s gatekeeping role:

St e p h a n ie : Lia, I’d like to talk to you about some of the information that I need to put
in the letter. The university health services requires that transgender patients receive
a diagnosis of gender dysphoria to receive hormones—how do you feel about this?
l ia : Um, well, I guess you need to do what you need to do.
St e p h a n ie : I thought it would be important to bring this up, because I want you
to know that I do not believe that being trans means you have a mental disorder.
Instead, in the letter, I will indicate that you meet criteria for the diagnosis, since you
have given me information that fits enough of the criteria to provide this diagnosis.
l ia : No, I get it. It makes me pretty mad to think that someone would think that
I have some type of mental disorder, but I understand that it is part of the process.
St e p h a n ie : I definitely understand how that could make you mad. How do you feel
about working with me after I have given you this diagnosis as part of the letter?
l ia : Oh, it’s okay—I know that you are writing the letter to help me get hormones and
it helps to know how you see the diagnosis. I’m not mad at you, more mad that it
even exists in the first place.

This part of the exchange illustrates key principles of feminist therapy, including
acknowledging and critically analyzing the underlying system of power in psycho­
therapy, in this case, the necessity of a pathologizing diagnosis to receive care and
the potential impact of this on the client and on the client-therapist dyad. The next
exchange (Budge, 2015, pp. 289-290) examines the process of the client reading the
letter.

St e p h a n ie : So, what was it like to read your letter?

l ia : It was weird.
 139 Gender Identity

St e p h a n i e : Weird.
lia : Yeah. I’ve never actually seen a document that was written about me that uses my
name—Lia— or female pronouns [appears a little tearful]. It feels good . . . what’s the
word? Empowering. But I’m having a lot of feelings I’m not sure how to talk about.
More than good, but also weird. Maybe a little overwhelming.
St e p h a n i e : That makes sense—a lot of trans people tell me they feel the same way
when they read their letters for the first time. How are you feeling about our rela­
tionship after reading through the letter?
lia : Oh, I feel good. I feel like you really advocated for me and that you understood
the things I told you. It was funny to read the letter because I could see the things
I told you written down. I feel like you really get it.

In one of the final exchanges (p. 292), the therapist and client discuss the client’s per­
ception of how therapy interacted with her taking hormones:

St e p h a n i e : How do you feel like taking hormones has impacted your mood?
lia : Greatly. I used to be all types of crazy and wild and my moods would just pop up.
During the first month or so of hormones it was worse, but now I just feel like, since
I’m living my life as a woman, everything is just chill now. It just feels right.
St e p h a n i e : How much of that do you attribute to the hormones?
lia : A lot. Because I can see my progress through hormones, like getting boobs.
St e p h a n i e : How do you think therapy interacted with the hormones?
lia : Well, it helped me notice how different my moods actually were and how I’m
much more open now than I was before. I attribute that to hormones, because I do
not think I would have done it if I hadn’t started hormones. I do not know where
I would be . . .
St e p h a n i e : Where do you think you would be?
lia : Well, I would still be that quiet do-not-want-to-talk-about-anything type of
person.
St e p h a n i e : So not wanting to open up to others or show who you are?
lia : Yeah.

Here, the therapist and the client were able to connect the client’s gender transition
process to her ability to be a more authentic individual with others in her life. Although
the client attributes much of her change to external factors (e.g., hormones), with the
assistance of her therapist, she is able to internalize that she is able to be more open
with others as a result of engaging in psychotherapy.

LANDMARK STUDIES

The majority of psychotherapy research on gender identity focuses on gender differences

or if a general treatment was effective for cisgender men or women. However, there
was one pioneering study that determined if gendered content within an intervention
would contribute to treatment differences. Greenfield and colleagues (2014) conducted
 140 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

a randomized clinical trial on cisgender women’s responsiveness to group therapy for

substance use disorders, Women’s Recovery Group (W RG). Fifty-two women were
randomized to WRG, 48 women were randomized to mixed-gender drug counseling,
and 58 men were assigned to mixed-gender drug counseling. The W RG group was “a
12-session, structured relapse-prevention group therapy that utilizes a cognitive behav­
ioral approach and includes gender-specific content and single-gender group composi­
tion” (p. 246). The gender-specific content in the group primarily focused on violence
and abuse, relationships, caretaking, and women’s health. The control group was also a
12-week session that focused on substance-related psycheducation and did not address
gender specifically. Both groups showed improvement, and there was not a significant
difference between the groups in the amount of improvement (as noted by a reduction
in substance use).
In one of the first documented studies that focused on transgender individuals’ psy­
chotherapy experiences, Rachlin (2002) asked participants to complete a survey about
why they sought therapy and their experiences of treatment. A total of 93 participants
(70 trans men and 23 trans women) completed the survey. Participants could report
experiences for both a current therapist and one former therapist (if applicable). When
reporting on recent therapy, participants indicated the top five reasons for seeking
therapy: to secure a letter for hormones or surgery (58%), to help making decisions
regarding gender transition (54%), to reduce emotional discomfort (49%), to improve
relationships (47%), and to enhance personal growth (46%). These contrast with former
therapy experiences, in which reducing emotional discomfort (67%), enhancing per­
sonal growth (58%), and improving relationships (57%) were endorsed as the primary
reasons for attending therapy (19% indicated seeking therapy previously for letters
and 21% indicated seeking therapy previously for help regarding gender transition).
Rachlin analyzed correlations of participants’ perceptions of therapist expertise re­
garding transgender issues with participants’ perceptions of outcomes. Therapist ex­
pertise was correalted positively with clients’ progress in gender exploration/transition
(r = .60) and satisfaction with treatment (r = .47) and it was correlated negatively with
harm from therapy (r = -.30).
Rachlin’s (2002) landmark study was a springboard for other researchers to under­
stand the experiences of transgender individuals in psychotherapy. For example, one
set of researchers (Mizock & Lundquist, 2016) conducted a qualitative study with 45
transgender and gender-nonconforming (TGNC) individuals to determine the psy­
chotherapy missteps that occur. In this grounded theory study, the authors noted that
there were eight therapist missteps reported by transgender clients:

♦ educational burdening (therapists expecting their clients to educate them on TGNC

issues)
♦ gender inflation (therapists exaggerating the importance of gender identity on topics
where gender identity was not the primary theme)
♦ gender narrowing (therapists relying on stereotyped notions of gender to guide
interventions)
 14 1 Gender Identity

♦ gender avoidance (therapists actively avoiding topics that were gender focused; one of
the clients in this study reported that the therapist had “ruled out” that gender was an
issue for the client, and thus it was not explored)
♦ gender generalizing (therapists assuming that “all TGNC people are the same”)
(p. 151)
♦ gender repairing and gender pathologizing (therapists treated the clients’ gender
identity as something that needed to be “fixed” and overly pathologized gender
identity)
♦ gatekeeping (therapists determining whether hormones or surgery were indicated,
which clients saw as a barrier to gender-affirming care)

Although there is little research that addresses how affirmative psychotherapies

work, some research has begun to compile evidence-based practices with transgender
individuals. A content analysis of literature focusing on transgender people between
2000 and 2012 revealed an increase in transgender-focused research (Moradi et al.,
2016). The content analysis found that the top four topics of research were psycho­
logical symptoms and disorders, identity development and coming out, medical tran­
sition procedures, and gender identity disorder and gender dysphoria. Most of the
research centered on mental health concerns, and only a small number of studies
broached psychotherapy process or principles with transgender individuals. At the
same time, the clear trend was that research on psychotherapy with transgender
people is advancing.

RESULTS OF PREVIOUS META-ANALYSES

Most psychotherapy meta-analyses that include gender as a m oderating variable
operationalize it as a binary demographic variable, comparing women/female
clients with men/male clients; these studies report m inim al to no differences in
psychotherapy outcomes between these groups (Bowman et al., 2001; Felmingham
& Bryant, 2012; Swift et al., 2013; Swift & Greenberg, 2012; Tao et al., 2015;
Weisz et al., 1995). This pattern is consistent with the gender sim ilarities hypo­
thesis and broader empirical evidence that there is more within group variability
than between-group differences between women and men on m ost psychosocial
variables (Hyde, 2005).
However, a fundamental limitation of psychotherapy research on gender is the
reliance on three im plicit assumptions: (a) that gender is intrapersonal, catego­
rical, and fixed rather than contextual, continuous, or dynamic; (b) that there are
only two gender categories com prising women and men; and (c) that all women
and men are cisgender. These assumptions lim it the generalizability of psycho­
therapy studies on gender to transgender people as well as to many other people
who do not fit the rigid gender binaries imposed in dominant operationalizations
of gender.
 142 PS Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Current Study
We could not locate any prior meta-analyses or systematic reviews on the effectiveness
of psychotherapy for transgender clients or gender-focused psychological treatments.
Moreover, a recent systematic review of randomized controlled trials (RCTs) of be­
havioral and psychological interventions for anxiety and depression (Heck et al.,
2017) revealed that no studies reported on gender categories beyond female/woman
or male/man. Thus we aimed to conduct two meta-analyses of (a) the outcomes of
psychotherapy with transgender people and (b) the outcomes of gender-focused
psychotherapies.

META-ANALYTIC REVIEW
To identify research relevant to the present analysis for the first meta-analysis, we
considered studies that either (a) compared psychotherapy outcomes between trans­
gender and cisgender people or (b) compared the outcomes of gender-focused
psychotherapies (for individuals with any gender identity) with a bona fide psycho­
therapy. To determine the final databases we would use in our search, we conducted
keyword searches in Academic Search Premier, PsycINFO, Pubmed, and ISI Web
of Science databases. We determined that the most comprehensive search was via
ProQuest’s PsycINFO.
For the first meta-analysis, relevant searches were identified by using keyword
searches that focused on psychotherapy trials and lesbian, gay, bixesual, transgender,
and queer (LGBTQ+) populations; we used keywords to capture LGBQ as well as
transgender populations because studies often collapse across these groups. In addi­
tion, we distributed a call for unpublished data to the professional listserves of the
American Psychological Association’s Divisions 12, 29, 17, 44, 35, 49, 51 and POWR-L
(feminist psychology list). We received one email in response that provided a literature
review of some data that were thought to fit our inclusion criteria. This call yielded no
unpublished psychotherapy outcome studies about transgender people.

Inclusion and Exclusion Criteria

To be included in this meta-analysis, multiple criteria needed to be met, including

(a) English-language publications from 1990 or later; (b) two treatment conditions,
with at least one bona fide psychotherapy condition and an additional treatment con­
dition where participants engaged in some type of intervention (i.e., not a waitlist);
(c) treatment conditions that either compared transgender and cisgender individuals
or had a focus on treatments adapted/designed for transgender individuals; and
(d) statistics necessary to calculate effect sizes. Ideally, we were also hoping to have
an inclusion criterion that focused primarily on trans-affirmative psychotherapy;
however, given that the search yielded no comparison studies specific to transgender
individuals, we were open to including studies that were broader in scope.
 143 Gender Identity

Results of Search Focused on Transgender Studies

As noted in Figure 5.1, the initial search on June 1, 2017, for psychotherapy trials and
LGBTQ+ studies resulted in k = 2,257. After all duplicates were deleted, the search
resulted in 2,189 studies. All abstracts were downloaded and coded into the following
categories: (a) meets inclusion for meta-analysis, (b) addresses psychotherapy and
LGBTQ+ people without data, (c) includes data on psychotherapy with LGBTQ+
people, (d) addresses non-psychotherapy interventions with LGBTQ+ people, and
(e) discard. We also included an extra coding category to identify publications that
specifically addressed transgender people or issues. The results focusing on LGBQ+
populations are reported in Chapter 9 of this volume. For the search focusing on
gender-affirming and trans-affirming and psychotherapy trials, there were no studies
that met the inclusion criteria.

figure 5 .1 PRISMA flow diagram.

 144 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Results of 10 Identified Studies

Because there were no psychotherapy studies that fit the criteria for this meta-analysis,
we conducted a content analysis and coded articles into categories (see Figure 5.1). Of
the 10 articles that provided empirical data on psychotherapy, 8 were focused on trans­
gender individuals’ experiences of psychotherapy, 1 was focused on therapist com­
petence with and attitudes towards transgender clients, and 1 was a content analysis
focused on family therapy journal articles published on transgender issues. Table 5.1
provides detailed information about the purpose and main findings of each of these
10 empirical studies.
The results of these 10 studies converge in several themes or patterns. First, it is ap­
parent from the results that clients experience anxiety and uncertainly when they dis­
cuss gender or deep emotions with their psychotherapists. Second, results from several
of the studies indicate that clients uniformly reject Diagnostic and Statistical Manual o f
Mental Disorders/International Classification of Diseases diagnoses focused on gender
identity. Third, participants indicate the value and need for a strong working relation­
ship, characterized by empathy, validation, and safety; they also indicate this is im ­
portant in general, as well as specifically to assist with transition. Fourth, participants
indicate the importance of having a knowledgeable and affirming psychotherapists
andacknowledge that without training psychotherapists can err by ignoring gender
too much or focusing on gender when it is not relevant. Fifth, psychological treatments
may prove beneficial, but clients indicate harm from some psychotherapy experiences;
for example, in one study, 90% described transaffirmative and healing experiences
but 100% described negative first therapy experiences. Finally, there is a strong con­
sensus among clients, practitioners, and researchers that psychotherapy competence
and using a transaffirmative approach is essential in how therapy is conducted with
transgender clients.

Results of Search Focused on Gender-Focused Studies

For the second meta-analysis, we centered the search on gender-focused therapies.
Again, we conducted a search via ProQuest’s PsycINFO by using keywords to capture
two primary topics: (a) psychotherapy trials comparing two bona fide psychotherapy
treatments and (b) treatments focused on gender or systems of power as they relate to
gender (e.g., feminist therapy, gender-aware therapy). The search was conducted on
July 31, 2017. The search yielded 331 articles that had been published between 1974
and 2017. There were no articles that matched the specific criteria for the meta-analysis.
Our searches revealed a paucity of research on the outcomes of psychotherapy with
transgender clients, the outcomes of transgender affirmative psychotherapies, and the
outcomes of gender-focused therapies. There were no articles that fit the criteria to
conduct a meta-analysis on these topics. Unfortunately, the lack of studies that eval­
uate treatments that attend to gender beyond a demographic variable and empower
individuals from a gendered perspective (e.g., transgender-affirmative therapy, femi­
nist therapy) is unsurprising. We know that cisgender women and transgender people
Table 5.1. Empirical Articles on Psychotherapy with Transgender Clients
Authors Method Sample Purpose
Applegarth & Nuttall Qualitative N =6 Explore transgender
(2016) individuals’ experiences o f
psychotherapy

Bess & Stabb (2009) Qualitative N =7 Explore transgender

clients’ perspectives of
the therapeutic alliance
and satisfaction with
psychotherapy
Main Findings
1. D escribed fear as a barrier in therapy, m ainly due to anxiety about how to
talk about gender or about talking about deep em otions in psychotherapy
2. D escribed the im portance o f a good working alliance but also described
com plicated feelings about their therapists (e.g., feeling deeply connected
but also managing therapy boundaries)
3. Indicated personal growth related to gender (clients learned new gender
identity labels and gained a greater sense o f humanity)
4. Psychotherapy helped participants think long term about how to handle
problem s when therapy concluded
1. Rejected a DSM diagnosis
2. Reported that m any therapists provided explicit transition-related support
3. Appreciated a lack o f rigidity in gender expectations
4. Stated that the prim ary aspects o f therapeutic alliance were support and
em pathy
5. Indicated m ixed feelings about the Standards o f Care
6 . All felt that their therapists helped with self-acceptance, self-definition,
validation, and norm alization
7. Found it helpful to have therapists who identified as women
8 . Found group therapy to be “particularly beneficial”
9. Two found therapy not to be helpful (e.g., lack o f therapist com petence,
therapist expression o f hostility)
{continued.)
Table 5.1. Continued
Authors Method Sample Purpose Main Findings
Blum er et al. (2012) Quantitative K=9 C ond uct a content analysis o f 1. O f 10,739 articles reviewed from journals that publish family therapy
(content fam ily therapy journals to content, only 9 (.0008% ) focused on transgender issues.
analysis) determ ine focus on TG N C 2. O f the 9 articles, 5 focused on therapy with transgender clients, 2 focused
clients on fam ily-of-origin issues, 1 focused on transgender identity, and 1
focused on “other”
E ld e r(2016) Qualitative N = 10 G ain understanding 1 .90% described transaffirmative and healing experiences
o f older transgender 2. All described negative first therapy experiences
individuals’ perspectives of 3. All said they have seen improvement over the years in how therapy is
psychotherapy conducted with TG N C clients
4. All provided recom m endations for therapists, including becom ing m ore
educated, better ground rules in groups, m ore affordable rates, and offering
gender-specific spaces for groups
Hunt (2014) M ixed m ethod N = 74 (quant.) Explore transgender clients’ 1. Found counseling through their general practitioners (48% ), the Internet
N = 5 (qual.) experiences o f seeking and (24% ), personal recom m endation (18% ), or transgender com m unity (16% )
experiencing therapy 2. 74% o f survey participants received therapy two or m ore tim es, with 75%
o f the sample attending 1 to 12 sessions
3. M ost clients sought therapy when com ing out or seeking m edical
interventions (only 8% and 9% sought therapy “post-transition” or
“surgery”)
4. M ost clients reported feeling accepted by their therapist and trusted them ,
but the num ber who felt understood by their therapist was lower
5. Som e clients thought that therapists ignored gender too m uch; some
clients thought that therapists focused on gender when that was not the
prim ary concern
Kanam ori & Quantitative N =95
=9 5 D eterm ine therapists’ and
Cornelius-W hite therapists-in-trainings’
(2017) attitudes toward
transgender clients

M athy (2011) Quantitative N =73 TG N C D eterm ine how transgender

clients clients com pare to
cisgender/heterosexual
and cisgender/LGBQ+
individuals on seeking
psychotherapy
1. Exam ined three attitudes toward transgender people: interpersonal
com fort (how m uch em pathy and interpersonal closeness one feels toward
transgender people), sex/gender beliefs (e.g., connection between m ental
illness and transgender identity), and hum an value (transgender people
should be treated with dignity and treated sim ilarly to cisgender people)
2. Means for all three attitudes were all above 5 (somewhat agree) on a 7-
point scale, indicating positive attitudes toward transgender clients
2. Group m ean for women was higher than the group m ean for m en on all
three attitudes
3. Group m ean for LG BQ + individuals was higher than the group m ean for
heterosexual participants on all three attitudes
4. Therapists who reported m ore personal contact with transgender clients
reported less com fort with and less hum an value o f transgender clients
5. Therapists who reported m ore training in transgender issues reported less
hum an value o f T G N C clients
1. Transgender individuals reported greater use o f psychiatric m edication
and psychotherapy than cisgender/heterosexual or lesbian/gay com parison
groups

( continued)
Table 5.1. Continued
Authors M ethod Sample Purpose Main Findings
M izock & Lundquist Qualitative N = 45 Explore therapist missteps 1. Seven them es o f therapist missteps emerged: (a) relying on clients to
(2016) from the perspectives of educate them about transgender issues, (b) overfocusing on gender,
transgender clients (c) stereotyping gender, (d) avoiding topics related to gender, (e) acting as
though gender identity needs to be fixed, (f) pathologizing gender identity,
and (g) controlling therapy as a gatekeeper
Yuksel et al. (2000) M ixed methods N = 25 (for Provide demographic 1. Depression was the m ost frequent diagnosis
the therapy inform ation and outcomes 2. Over the course o f the three years, four participants (16% ) dropped out o f
portion o f from group psychotherapy therapy
the study) with transgender women 3. Themes from the group were (a) clients experienced relationship
difficulties with fam ilies, work, and partner; (b) the group facilitated a way
to m eet other transgender individuals; (c) clients felt they could share their
difficulties with others who understood them and receive support; (d) the
group was established as a “self-help resource”; and (e) the group facilitated
m aking decisions about gender affirm ing treatm ents
R achlin (2002) Quantitative N = 93 Q uantify transgender clients’ 1. Two prim ary reasons emerged for seeking therapy: personal growth and
experiences with therapists help with gender-affirm ing treatm ents and gender identity processes
(e.g., num ber o f therapists, 2 . Clients in treatm ent longer indicated that they were in therapy for personal
reasons for seeking therapy) growth, rather than to discuss gender identity
3. Therapist expertise with gender was correlated positively with rapport and
correlated negatively with the length o f therapy
4. 15% o f the sample reported that they were actively harm ed by their
therapist

Note. TGNC = transgender and gender-nonconforming; LGBQ+ = lesbian, gay, bisexual, and queer; DSM = Diagnostic and Statistical M anual o f Mental Disorders.
 149 Gender Identity

seek therapy and are diagnosed with some mental health concerns at higher rates than
cisgender men (e.g., Addis & Mahalik, 2003; Kessler et al., 2012; Salk et al., 2017),
but therapies that focus on systemic gender oppression that might explain these
disparities have not been evaluated scientifically. This omission continues the cycle
that ignores how systems of power create disparities and lend to internalized sexism
and internalized transphobia, which, in turn, lead to more mental health concerns
(Meyer, 2003; Testa et al., 2015).
Despite these discouraging findings that precluded meta-analyses from being
conducted, the meta-analytic search identified articles that addressed psychotherapy
with transgender individuals. It is encouraging that 60 articles were published that de­
scribe psychotherapy techniques and processes with transgender clients. A previous
analysis of all literature published about transgender people or issues from 2000 to
2012 (Moradi et al., 2016) found only 21empirical articles on counseling process and
24 empirical articles on counseling techniques.
One particular strength of the content analysis in the current study is that we were
able to examine empirical studies that emerged from the search (see Table 5.1). After
exploring the content of these 10 articles, we are still left with the same question that pi­
oneering psychotherapy researchers asked: Does psychotherapy work for transgender
clients, and, assuming it does, how does it work? We know that psychotherapy works
(Wampold & Imel, 2015) and to some extent how it works with cisgender people, but
we have no data on the efficacy and effectiveness of psychotherapy with transgender
people.

LIMITATIONS OF THE RESEARCH

The first obvious limitation, as noted, is that there is not enough research to deter­
mine the outcomes of gender-focused psychotherapy. Thus, we call for research on
processes and outcomes of psychotherapy, specifically focusing on gender dynamics
that can exist in psychotherapy. Specifically, there is a large gap in outcome research
with transgender clients. The content analysis from the current study determined that
most research in psychotherapy with transgender clients is retrospective; if clients
were currently in psychotherapy, they were not asked about changes or processes that
were specific to how psychotherapy was assisting them to grow, adapt, and change.
Researchers should conduct psychotherapy trials that focus on how treatments work,
specifically focusing on gender dynamics (e.g., systems of power related to gender) and
transgender individuals.
A second limitation is that we do not understand how stressors specific to
marginalized populations interact with psychotherapy processes and outcomes. W E
recommend that psychotherapy researchers answer targeted questions about specific
concerns for transgender populations. For example, as minority stress appears to be
one of the primary reasons for transgender individuals’ high rates of mental health
concerns, targeted psychotherapies for minority stress could be compared with other
bona fide treatments to determine if minority stressors should be targeted specifically
or if improvement occurs due to changes in common factors.
 150 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

Third, feminist therapists have been vocal in critiquing research methods that are
highly controlled and unrepresentative of clinical reality (Brown, 2006). We recom­
mend that researchers track psychotherapy outcomes with all clients in naturalistic
settings and that therapists administer outcome measures to their clients. The benefit
of publishing longitudinal data from naturalistic settings is that such findings will be
more generalizable to how therapy is conducted with clients in the “real world” and
also allow researchers to determine if there are meaningful differences related to health
disparities based on gender identity (see Eubanks-Carter et al., 2012).
Finally, a limitation from many of the studies we reviewed is that the therapeutic
relationship is either ignored or barely mentioned. We recommend that all psycho­
therapy studies focusing on gender identity measure the complex therapeutic relation­
ship. A recurrent theme from the studies focusing on transgender clients’ experiences
concerned their relationship with the therapist. Their alliance is particularly impor­
tant to explore because many medical providers and clients may believe (errone­
ously) that psychotherapy is mandated for gender-affirming treatments (Budge, 2015).
Researchers have hypothesized that the working alliance between transgender clients
and their therapists is impacted negatively by an assumption (by both parties) that
therapy is mandated (Budge et al., 2017).

DIVERSITY CONSIDERATIONS
There is a diversity of gender identities, and gender identity is inherently connected with
power dynamics, oppression, and privilege. As such, all of the considerations discussed
in this chapter thus far are “diversity considerations.” As well, there are implications for
how gender identities relate to other systems of power. For example, psychotherapy
in the United States with an incarcerated White cisgender man compared to therapy
conducted with an incarcerated African American cisgender woman will likely involve
distinct conversations that focus on race, gender, and other systems of power within
the United States (Mulay et al., 2017).

TRAINING IMPLICATIONS
One of the most important areas for advancing training is to promote therapists’ com­
petence in understanding power, gender dynamics, and gender-affirmative care. We
recommend that all training and continuing education programs examine how power
impacts psychotherapy and subsequently provide specific training regarding gender.
In particular, clinical training should be offered on transgender populations. Four
studies (Elder, 2016; Hunt, 2014; Mizock & Lundquist, 2016; Rachlin, 2002) noted sub­
stantial negative experiences that resulted from psychotherapy for transgender and
gender-diverse clients. These findings reinforce the need for quality training.
There are very few training programs focus on systems of power in psycho­
therapy. One training program that attends to these issues is the Feminist Psychology
Institute, which is sponsored by the Society for the Psychology of Women/American
 151 Gender Identity

Psychological Association Division 35. The Feminist Psychology Institute (https://fem-

inism.org) provides continuing education on feminist therapy, trauma therapy based
on a gendered analysis, and barriers to mental health care for diverse women. The Jean
Baker Miller Training Institute (jbmti.org) also provides training from a gendered lens
that addresses relational-cultural therapy and the dynamics of dominance and subor­
dination in human relationship.
Another training program that specifically addresses transgender clients is the
World Professional Association for Transgender Health’s Global Education Initiative,
which focuses on training mental health professionals’ in foundational knowledge on
transgender issues. These two-day multidisciplinary courses discuss evidence-based
practices and case discussions. However, this training may prove time or cost prohib­
itive. We also recommend free training resources provided by the Fenway Institute
(fenwayhealth.org/the-fenway-institute); the institute has multiple one-hour training
webinars about transgender individuals and mental health.

THERAPEUTIC PRACTICES

Based on the literature review and content analysis, we offer suggestions for improving
psychotherapy outcomes by attending to gender-based issues of power, oppression,
and privilege.

♦ Address gender dynamics and gender topics explicitly in session. Gender can, and
should, be addressed as a system of power in clients’ lives and in the psychotherapy
process. The clinical challenge is to know when and how to bring gender into
the therapeutic space. Therapists should follow feminist therapy principles of
collaborative gender analysis to integrate gender in psychotherapy based on accurate
information (rather than stereotypes) and with a spirit of curiosity and collaborative
exploration.
♦ Privilege clients’ experiences and avoid assumptions . The error of making assumptions
based on both gender and sexual orientation was evident in a therapist’s account
of seeing a cisgender man who identified as heterosexual and came to treatment
wanting to address his pornography usage (Walters & Spengler, 2016). The
therapist (Walters) never asked about the gender identity of the individuals in the
pornography his client was watching because he made assumptions from his client’s
disclosures and gender presentation. However, once he determined that the client
was watching cisgender men in pornography, the content of the sessions changed
and focused on masculinity and shame, which ultimately led to a better outcome.
♦ Check in with clients to determine if the clinician missed the mark or is putting “too
much” emphasis on gender. Transgender clients want their therapists to bring up
gender, but not when it is stereotyped or not relevant to the content (Mizock &
Lundquist, 2016). Here we make the same recommendation as before; it appears
from the qualitative studies that more damage can be done from not bringing up
gender than bringing it up when it is not relevant.
 15 2 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

♦ Consider transgender affirmative methods and methods that focu s on systems o f pow er
in psychotherapy. A benefit of affirmative psychotherapies from a gendered lens is
that they are applicable to all individuals, since all individuals have a relationship
to gender identity. Information gleaned about the effectiveness of affirmative
psychotherapy could benefit all clients. Although there is a paucity of evidence from
RCTs, other research evidence from the perspectives of transgender clients supports
the use of affirmative psychotherapy (e.g., Bess & Stabb, 2009; Elder, 2016).
♦ Use a social justice fram ew ork fo r interventions and systems. Feminist therapies and
transgender affirming therapies suggest a social justice orientation to psychotherapy.
Because gender-based oppression is not internal to the client, addressing it requires
pushing beyond traditional conceptualization of psychotherapy to include advocacy
and activism. Therapists can assist clients with coping mechanisms, but in the
broader context of continuing oppression, it is unlikely that coping alone will be
sufficient. Therapists can advocate for and with the client without compromising
confidentiality, for example, by intervening with institutions, finding lists of
transgender affirming housing and groups, serving on local committees, working
within the system to ensure better policies and trainings, and advocating in political
structures for equal rights for people with all gender identities. Collaborating with
clients on such efforts can also empower clients and serve as a social justice-focused
therapeutic intervention.
♦ Stay inform ed on gendered language. Frequently, mental health practitioners worry
that the information is changing “too fast to keep up with the language.” When we
hear this concern from colleagues, we often wonder what is behind this fear. One
hypothesis is that therapists are afraid of making mistakes and use the defense of “I
don’t have time” or “It happens too fast to keep up with it” to cover their worries.
We recommend that practitioners search blogs, social media pages, and websites
that post gender-specific definitions and terms about once every six months to
remain updated on language and gender. This may take approximately 15 minutes
to complete, which should be feasible for all practitioners. It is relatively quick and
painless to keep up with the changing language and nuanced understanding of
gender as it relates to psychotherapy.
♦ Use an intersectional approach when focu sin g on gen der an d gen der identity. We
draw attention to the importance of understanding that people who use the
same gender identity labels are not homogenous. Clients should be approached
with humility and curiosity and understood with their full humanity, salient
experiences, and identities, . This humble approach should be paired with therapist
understanding of the sociopolitical climate for transgender individuals. Clients’
experiences should be understood within the context of intersecting systems of
power around gender, race/ethnicity, sexual orientation, and other inequalities
(Hook et al., 2013).

REFERENCES
Addis, M . E., & M ahalik, J. R. (2003). M en, masculinity, and the contexts o f help seeking.
American Psychologist, 58, 5 -1 4 .
 153 Gender Identity

Applegarth, G., & Nuttall, J. (2016). The lived experience o f transgender people o f talking
therapies. International Journal o f Transgenderism, 17, 6 6 -7 2 .
Bem , S. L. (1993). The lenses o f gender: Transforming the debate on sexual inequality. New
Haven, CT: Yale University Press.
Bess, J., & Stabb, S. (2009). The experiences o f transgendered persons in psychotherapy: Voices
and recom m endations. Journal o f Mental Health Counseling, 31, 2 6 4 -2 8 2 .
Blumer, M . L., G reen, M. S., Knowles, S. J., & W illiam s, A. (2012). Shedding light on thirteen
years o f darkness: C ontent analysis o f articles pertaining to transgender issues in marriage/
couple and family therapy journals. Journal o f Marital and Family Therapy, 38, 2 4 4 -2 5 6 .
Bond, M . A., & Allen, C. T. (2016). Beyond difference: G ender as a quality o f social settings.
In T.-A. Roberts, N. Curtin, L. E. Duncan, & L. M . C ortin a (Eds.), Feminist perspectives
on building a better psychological science o f gender (pp. 2 3 1 -2 5 4 ). New York, NY: Springer
International.
Bowman, D., Scogin, F., Floyd, M ., & M cKendree-Sm ith, N. (2001). Psychotherapy length of
stay and outcome. Psychotherapy: Theory, Research, Practice, Training, 38, 1 4 2-148.
Brown, L. S. (1994). Subversive dialogues: Theory infeminist therapy. New York, NY: Basic Books.
Brown, L. S. (2006). Still subversive after all o f these years: The relevance o f fem inist therapy in
the age o f evidence-based practice. Psychology o f Women Quarterly, 3 0 , 15-2 4 .
Budge, S. L. (2015). Psychotherapists as gatekeepers: An evidence-based case study highlighting
the role and process o f letter w riting for transgender clients. Psychotherapy, 52, 2 8 7 -2 9 7 .
Budge, S. L., Israel, T., & M errill, C. R. (2017). Improving the lives o f sexual and gender
m inorities: The prom ise o f psychotherapy research. Journal o f Counseling Psychology, 64,
3 7 6 -3 8 4 .
Burm an, E., Gowrisunkur, J., & Sangha, K. (1998). Conceptualizing cultural and gendered
identities in psychological therapies. The European Journal o f Psychotherapy, Counselling
& Health, 1, 2 3 1 -2 5 5 .
Clarkin, J. F., & Levy, K. N. (2004). The influence o f client variables on psychotherapy. In M . J.
Lam bert (Ed.), Bergin and Garfield’s handbook o f psychotherapy (pp. 1 9 4 -2 2 6 ). New York,
NY: Wiley.
Cottone, J. G., Drucker, P., & Javier, R. A. (2002). G ender differences in psychotherapy
dyads: Changes in psychological symptoms and responsiveness to treatm ent during
3 m onths o f therapy. Psychotherapy: Theory, Research, Practice, Training, 39, 2 9 7 -3 0 8 .
Elder, A. B. (2016). Experiences o f older transgender and gender nonconform ing adults in
psychotherapy: A qualitative study. Psychology o f Sexual Orientation and Gender Diversity,
3 ,1 8 0 -1 8 6 .
Eubanks-Carter, C., G orm an, B. S., & Muran, J. C. (2012). Quantitative naturalistic methods
for detecting change points in psychotherapy research: A n illustration with alliance
ruptures. Psychotherapy Research, 2 2 , 6 2 1 -6 3 7 .
Felmingham, K. L., & Bryant, R. A. (2012). G ender differences in the m aintenance o f response
to cognitive behavior therapy for posttraum atic stress disorder. Journal o f Consulting and
Clinical Psychology, 80 , 1 96-200.
Flores, A. R., Herm an, J. L., Gates, G. J., & Brown, T. N. T. (2016). How m any adults identify as
transgender in the United States? Retrieved from http://williamsinstitute.law.ucla.edu/wp-
content/uploads/H ow -M any-Adults-Identify-as-Transgender-in-the-United-States.pdf
G enIU SS Group (2014). Best practices for asking questions to identify transgender and other
gender m inority respondents on population-based surveys. Retrieved January 7, 2019
from https://williamsinstitute.law.ucla.edu/research/census-lgbt-demographics-studies/
geniuss-report-sept-2014/.
 154 P S Y C H O T H E R A P Y R E L A T IO N S H IP S THAT WORK

G ilbert, L. A., & Scher, M. (2009). Gender and sex in counseling and psychotherapy. Eugene,
O R: W ipf and Stock.
G ood, G. E., G ilbert, L. A., & Scher, M . (1990). G ender aware therapy: A synthesis o f fem inist
therapy and knowledge about gender. Journal o f Counseling & Development, 68, 3 7 6 -3 8 0 .
Greenfield, S. F., Sugarman, D. E., Freid, C. M ., Bailey, G. L., Crisafulli, M. A., Kaufman, J. S., . . .
Fitzm aurice, G. M. (2014). Group therapy for women with substance use disorders: Results
from the W omen’s Recovery Group Study. Drug and Alcohol Dependence, 142, 2 4 5 -2 5 3 .
Haddock, S. A., & Lyness, K. P. (2002). Three aspects o f the therapeutic conversation in couples
therapy: Does gender make a difference? Journal o f Couple & Relationship Therapy, 1, 5 -2 3 .
Heck, N. C., M irabito, L. A., LeMaire, K., Livingston, N. A., & Flentje, A. (2017). Om itted
data in randomized controlled trials for anxiety and depression: A systematic review of
the inclusion o f sexual orientation and gender identity. Journal o f Consulting and Clinical
Psychology, 85, 7 2 -7 6 .
Heru, A. M ., Strong, D., Price, M ., & Recupero, P. R. (2006). Self-disclosure in psychotherapy
supervisors: G ender differences. American Journal o f Psychotherapy, 60, 3 2 3 -3 3 4 .
Hook, J. N., Davis, D. E., Owen, J., W orthington, E. L. Jr., & Utsey, S. O. (2013). Cultural hu­
mility: M easuring openness to culturally diverse clients. Journal o f Counseling Psychology,
60, 3 5 3 -3 6 6 .
Hunt, J. (2014). An initial study o f transgender people’s experiences o f seeking and receiving
counselling or psychotherapy in the UK. Counselling and Psychotherapy Research , 14,
2 8 8 -2 9 6 .
Hyde, J. S. (2005). The gender sim ilarities hypothesis. American Psychologist, 60, 5 8 1 -5 9 2 .
James, S. E., Herm an, J. L., Rankin, S., Keisling, M ., M ottet, L., & Ana, M. (2016). The report
o f the 2015 U.S. transgender survey. W ashington, DC: National C enter for Transgender
Equality.
Kanam ori, Y., & Cornelius-W hite, J. H. (2017). Counselors’ and counseling students’ attitudes
toward transgender persons. Journal o f LGBT Issues in Counseling, 11, 3 6 -5 1 .
Kessler, R. C., Petukhova, M ., Sam pson, N. A., Zaslavsky, A. M ., & W ittchen, H. U. (2012).
Twelve-month and lifetim e prevalence and lifetim e m orbid risk o f anxiety and m ood
disorders in the United States. International Journal o f Methods in Psychiatric Research,
2 1 , 1 6 9 -1 8 4 .
Lev, A. I. (2004). Transgender emergence: Therapeutic guidelines fo r working fo r gender-variant
people and theirfamilies. New York, NY: Routledge.
M cGeorge, C. R., Carlson, T. S., & G uttorm son, H. (2009). The fem inist couple therapy
scale: A measure o f therapists’ ability to prom ote equality in couple relationships. Journal
o f Feminist Family Therapy, 2 1(3), 1 98-215.
Meyer, I. H. (2003). Prejudice, social stress, and m ental health in lesbian, gay, and bisexual
populations: Conceptual issues and research evidence. Psychological Bulletin, 129, 6 7 4 -6 9 7 .
M izock, L., & Lundquist, C. (2016). Missteps in psychotherapy with transgender
clients: Prom oting gender sensitivity in counseling and psychological practice. Psychology
o f Sexual Orientation and Gender Diversity, 3 , 1 4 8-155.
M oradi, B., Fischer, A. R., Hill, M. S., Jom e, L. M ., & Blum, S. A. (2000). Does “fem inist” plus
“therapist” equal “fem inist therapist”?: An em pirical investigation o f the link between self­
labeling and behaviors. Psychology o f Women Quarterly, 2 4 (4 ), 2 8 5 -2 9 6 .
 155 Gender Identity

M oradi, B., Tebbe, E. A., Brewster, M. E., Budge, S. L., Lenzen, A., Ege, E., . . . Hiner, D.
L. (2016). A content analysis o f literature on trans people and issues: 2 0 0 2 -2 0 1 2 . The
Counseling Psychologist, 44, 9 6 0 -9 9 5 .
M oradi, B., & Yoder, J. D. (2011). The psychology o f women. In E. M . A ltm aier & J.-I. C.
Hansen (Eds.), The Oxford handbook o f counseling psychology (pp. 3 4 6 -3 7 4 ). New York,
NY: O xford University Press.
Mulay, A. L., Kelly, E., & Cain, N. M. (2017). Psychodynamic treatm ent o f the crim inal of­
fender: M aking the case for longer-term treatm ent in a longer-term setting. Psychodynamic
Psychiatry, 4 5 , 1 4 3-173.
Ogrodniczuk, J. S. (2006). M en, women, and their outcom e in psychotherapy. Psychotherapy
Research, 16, 4 5 3 -4 6 2 .
Owen, J., W ong, Y. J., & Rodolfa, E. (2009). Em pirical search for psychotherapists’ gender co m ­
petence in psychotherapy. Psychotherapy: Theory, Research, Practice, Training, 4 6 , 4 4 8 -4 5 8 .
Rachlin, K. (2002). Transgender individuals’ experiences o f psychotherapy. International
Journal o f Transgenderism, 6, 1 -19.
Robinson, D., & Worell, J. (1991). The Therapy with Women Scale (TWS). Unpublished m anu­
script, University o f Kentucky, Lexington.
Salierno, E. F. (2000). The effect o f theoretical orientation and patient gender on perceptions o f
therapy process and outcome: An analogue investigation o f feminist therapy (Unpublished
doctoral dissertation). Fairleigh Dickinson University, Teaneck, NJ.
Salk, R. H., Hyde, J. S., & Abramson, L. Y. (2017). G ender differences in depression in rep­
resentative national samples: M eta-analyses o f diagnoses and symptoms. Psychological
Bulletin, 143(8), 7 8 3 -8 2 2 .
Serano, J. (2016). Whipping girl: A transsexual woman on sexism and the scapegoating o f fem i­
ninity. New York, NY: Hachette.
Swift, J. K., Callahan, J. L., Ivanovic, M ., & Kom iniak, N. (2013). Further exam ination o f the
psychotherapy preference effect: A m eta- regression analysis. Journal o f Psychotherapy
Integration, 2 3 , 1 3 4-145.
Swift, J. K., & Greenberg, R. P. (2012). Premature discontinuation in adult psychotherapy: A
m eta-analysis. Journal o f Consulting and Clinical Psychology, 80, 5 4 7 -5 5 9 .
Tao, K. W , Owen, J., Pace, B. T., & Imel, Z. E. (2015). A m eta-analysis o f multicultural
com petencies and psychotherapy process and outcom e. Journal o f Counseling Psychology,
6 2(3), 3 3 7 -3 5 0 . https://www.doi.org/10.1037/cou0000086
Tebbe, E. A., & Moradi, B. (2016). Suicide risk in trans populations: An application o f m inority
stress theory. Journal o f Counseling Psychology, 63, 5 2 0 -5 3 3 .
Testa, R. J., Habarth, J., Peta, J., Balsam, K., & Bockting, W (2015). Development o f the G ender
M inority Stress and Resilience Measure. Psychology o f Sexual Orientation and Gender
Diversity, 2 , 6 5 -7 7 .
W alters, N. T., & Spengler, P. M. (2016). Clinical errors and therapist discom fort with client
disclosure o f troublesom e pornography use: Im plications for clinical practice and error
reduction. Psychotherapy, 53, 3 5 4 -3 5 9 .
W ampold, B. E., & Imel, Z. E. (2015). The great psychotherapy debate: The evidence fo r what
makes psychotherapy work. New York, NY: Routledge.
 15 6 P S Y C H O T H E R A P Y R EL AT IO N S H IP S THAT WORK

Weisz, J. R., Weiss, B., Han, S. S., Granger, D. A., & M orton, T. (1995). Effects o f psychotherapy
with children and adolescents revisited: A m eta-analysis o f treatm ent outcom e studies.
Psychological Bulletin, 117, 4 5 0 -4 6 8 .
W est, C., & Zim m erm an, D. H. (1987). Doing gender. Gender & Society, 1 , 1 2 5-151.
W orell, J., & Remer, P. (2002). Feminist perspectives in therapy: Empowering diverse women.
New York, NY: Wiley.
Yiiksel, §., Kulaksizoglu, I. B., Tiirksoy, N., & §ahin, D. (2000). Group psychotherapy with
fem ale-to-m ale transsexuals in Turkey. Archives o f Sexual Behavior, 29, 2 7 9 -2 9 0 .
 6

PREFERENCES

Joshua K. Swift, Jennifer L. Callahan,

Mick Cooper, and Susannah R. Parkin

Over the past couple of decades, the field has seen an increasing recognition of the
crucial role that client preferences can play in psychotherapy. For example, in 2006 the
American Psychological Association (APA) defined evidence-based practice in psy­
chology as “the integration of the best available research with clinical expertise in the
context of patient characteristics, culture, and preferences” (p. 273). As another ex­
ample, a quick Google Scholar search using the term “preferences in psychotherapy”
yields almost 100,000 citations between the years 1970 and 2016. A review of the
number of citations by year reveals a clear curvilinear trend with upward growth, es­
pecially over the past 10 years (see Figure 6.1).
In this chapter, we provide the results of a meta-analysis that demonstrates that ac­
commodating client preference in psychotherapy is associated with fewer treatment
dropouts and more positive treatment outcomes. In this chapter, we also define the
construct of client preferences, describe how they are measured and studied in the
field, and provide a clinical example. We conclude with limitations of the research,
diversity considerations, training implications, and practice recommendations for
incorporating patient preferences into psychotherapy.

DEFINITIONS

Client preferences in psychotherapy are defined as the specific conditions and activ­
ities that clients want in their treatment experience (Swift et al., 2011). Preferences
have generally been grouped into three broad categories (Swift et al., 2011). First,
activity preferences refer to the activities that clients hope they and their therapists
will engage in throughout the course of psychotherapy. For instance, one client may
hope that his therapist will use confrontations as they work together. Another client
may wish that her therapist will not assign homework between treatment sessions.
Activity preference can also include desires about the format of the intervention
(e.g., cognitive-behavioral therapy offered individually or in a group). This type of

157
 158 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

Year

figure 6.1 Google Scholar results for “preferences in psychotherapy” by year.

preference has previously been referred to as role preferences; however, we refer to it

as activity preferences to be more inclusive of the setting and format preferences that
are also included in this category.
Second, client preferences about the therapist refer to the type of practitioner
with whom clients would like to work. Often these preferences are based on demo­
graphics; however, research has indicated that clients frequently report even stronger
preferences for a therapist’s personality characteristics (Swift & Callahan, 2010; Swift
et al., 2015). Client preferences about their therapists has historically been referred to
as therapist preferences. However, this type of client preference can be distinguished
from preferences held by therapists, an area where future research is needed. Thus,
throughout this chapter, we refer to this type of client preference as preferences about
the therapist.
Last, treatment preferences include clients’ desires regarding the type of interven­
tion that is used in their care. Most often studied are preferences for psychotherapy
versus medication; however, preferences can also exist between different types of psy­
chotherapy (e.g., client-centered psychotherapy, cognitive-behavioral therapy, psycho­
dynamic psychotherapy) and between psychotherapy and other interventions (e.g.,
self-help books, peer-support groups; McLeod, 2012; Tompkins et al., 2017).
Although research has shown that they are related, preferences can be distinguished
from client expectations in psychotherapy (Tompkins et al., 2017). Expectations rep­
resent what clients believe will occur in psychotherapy, whereas preferences represent
their desires. For example, a female client who belongs to an ethnic minority group
may very much desire to work with a female therapist who also belongs to the same
ethnic minority group; however, given where she lives and the resources around her,
she may not expect that could happen. As another example, a male client with an al­
cohol use disorder may hope that his new therapist is one who will be nonjudgmental
and accepting; however, given his experiences with previous authority figures, he may
expect that his new therapist will be critical of his drinking behavior.
 159 Preferences

Perhaps even more closely related to preferences are clients’ prognostic beliefs
about treatment success. Even though prognostic beliefs undoubtedly inform clients’
preferences for psychotherapy, specific preferences may also be based on other
factors, such as motivation to change, personality characteristics, and past treatment
experiences. An example of the difference between these two constructs can be seen
in a client with social phobia who prefers to attend individual treatment sessions even
though he expects that a group treatment would prove most effective for addressing
his problems.

MEASURES

Multiple ways of measuring patient preferences have been utilized in research and in
practice. Perhaps the most popular measure has been to directly ask patients what
condition they would prefer to receive: for example, asking patients if they would
prefer medication, psychotherapy, or a combination of the two (e.g., Kocsis et al.,
2009) or asking patients if they would prefer a male or a female therapist (e.g., Zlotnick
et al., 1998).
In a variation of this type of measure, a few studies have provided patients with
descriptions and/or demonstrations of their options prior to asking them to state a
preference. For example, some researchers have played video or audio recordings of
therapists providing descriptions of themselves and their treatment approaches and
then asked patients to indicate which therapist they would prefer to work with (e.g.,
Devine & Fernald, 1973; Manthei et al., 1982). Other researchers have had clients
briefly discuss therapy options with a psychotherapist or physician prior to being asked
to state a preference for one treatment or another (e.g., Adamson et al., 2005; Calsyn
et al., 2000).
In contrast to directly asking patients to state their preferences, some researchers
have employed questionnaires or rating scales that assess preferences as well as their
degree or strength. Assessing preference strength is of value because one might ex­
pect that stronger preferences, compared with weaker preferences, would exert greater
influence on treatment outcomes. For example, researchers have not only invited de­
pressed patients to indicate if they preferred interpersonal psychotherapy or pharma­
cotherapy but also asked them to rate on a 5-point Likert-type scale how strongly they
wanted their preferred treatment (e.g., Raue et al., 2009).
In research settings, investigators have also used delay-discounting methods to as­
sess psychotherapy preference strength. Delay discounting is a method that has tra­
ditionally been used in the field of economics to measure individuals’ preferences for
smaller immediate rewards compared to larger delayed ones. For example, individuals
could be asked if they prefer to receive $10 today or $20 in one week. Depending on
the initial choice, the delayed reward can be raised or lowered to find the value that
individuals place on time over money. With slight variations, this paradigm has also
been used to assess preference values. For example, Swift and Callahan (2010) asked
participants how much they were willing to sacrifice in terms of treatment efficacy to
receive a therapist who they could develop a positive working relationship with. In that
 160 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

study, participants were initially asked if they would prefer a treatment that had been
found to work for about 70% of clients in recent clinical trials but was delivered by a
therapist who was difficult for them to relate to or a treatment that had been found to
work for about 10% of clients in recent clinical trials but was delivered by a therapist
who was easy for them to relate to. Although most clients picked the 70% option at
first, on average they switched to the less effective treatment once it reached 30%—
thus indicating they were willing to sacrifice 40% in treatment efficacy to ensure that
positive relationship. Swift and colleagues (2015) used the same method with racial/
ethnic variables and found that participants had much stronger preferences for ther­
apist multicultural training and use of culturally adapted treatment techniques than
they did for simple racial/ethnic matching with their therapists. More recently, Boswell
and colleagues (2018) used this method to examine preferences for clinician perfor­
mance information.
Within the research and clinical literature, three patient preference measures
have been developed and validated: the Psychotherapy Preferences and Experiences
Questionnaire (PEX; Sandell et al., 2011), the Preference for College Counselling
Inventory (PCCI; Hatchett, 2015), and the Cooper-Norcross Inventory of Preferences
(C-NIP; Cooper & Norcross, 2016). In addition, two standardized qualitative strategies
for assessing patient preferences have been established: The Treatment Preference
Interview (Vollmer et al., 2009) and the use of decision aids (Health Foundation, 2014).

Psychotherapy Preferences and Experiences Questionnaire

The PEX, previously titled the Treatment Preferences and Experiences questionnaire
(Levy Berg et al., 2008), is a 25-item measure that asks respondents to rate, on a 6-point
Likert-type scale, the extent to which they believe a range of therapist and client activ­
ities would be helpful for them (e.g., “getting active support”). The items are grouped
according to five subscales, supported by factor analysis: Outward Orientation (di­
rective, concrete, and problem-solving therapist activities); Inward Orientation (re­
flective and insight-oriented client and therapist activities); Support (encouraging,
warm, and friendly therapist activities and characteristics); Catharsis (emotionally ex­
pressive client and therapist activities); and Defensiveness (avoidant and emotionally
suppressive client activities). The PEX subscales have satisfactory internal consistency
(Cronbach’s a = .78-.86), with evidence of concurrent (Sandell et al., 2011) and predic­
tive (Levy Berg et al., 2008) validity.

Preference for College Counselling Inventory

The PCCI is a 90-item questionnaire designed for assessing patients’ preferences about
potential therapists and psychotherapy roles (Hatchett, 2015). It was designed for use
within a college counseling context, but its items are probably relevant to a broad range
of psychotherapy settings. The first part of the PCCI includes seven open questions
about preferences for particular therapist characteristics, such as therapist gender
 16 1 Preferences

and sexual orientation. The second part consists of 32 items focusing on preferences
for therapist characteristics and activities. Respondents are provided with the stem
“I would prefer a therapist . . and then asked to rate items such as “. . . who will be
caring and kind,” and “. . . who is intelligent.” Each of these items are rated on a 1 (not
true) to 5 (definitely true) scale. The third part, which uses the same scoring scheme as
the second part, consists of 28 items focusing on preferences for client activities, for
instance “I would like to experience my feelings more intensely.”
Principal components analysis identifies three components for Part 2 of the
measure: Therapist Expertise (a well-trained and competent therapist), Therapist
Warmth (a kind and understanding therapist), and Therapist Directiveness (an ac­
tive, goal-oriented, and structured therapist). For the third part of the PCCI, a sepa­
rate principal components analysis identified two further components: Task-Oriented
Activities (technical, problem- solving activities) and Experiential/Insight-Oriented
Activities (developing self-understanding and emotional experiencing). Each of these
five subscales have been shown to have high levels of internal consistency (Cronbach’s
a = .8 9 - .92; Hatchett, 2015). Adequate levels of test-retest reliability have also been
identified for four of the subscales (r = .73-.90), with weaker test-retest reliability for
the Therapist Expertise subscale (r = .50; Hatchett, 2015).

Cooper-Norcross Inventory of Preferences

The C-NIP, based in part on the Therapy Personalisation Form (Bowen & Cooper,
2012), is an activity preference measure designed for clinical use either at assessment
or as part of ongoing therapy (Cooper & Norcross, 2016). The first part of the inven­
tory presents clients with the stem “I would like my therapist to . . . ,” and then 18
semantic differential items in which they indicate their preferences from 3 (a “strong
preference” for one end of the item) to - 3 (a “strong preference” for the other end of
the item). Examples of items are “Focus on specific goals”— “Not focus on specific
goals” and “Focus mainly on my thoughts”—“Focus mainly on my feelings.” Based
on principal components analyses, the items are grouped into four scales with cut
points for strong preferences in both directions: Therapist-Directiveness versus Client-
Directiveness (internal consistency a = .84), Emotional Intensity versus Emotional
Reserve (a = .67), Past Orientation versus Present Orientation (a = .73), and Warm
Support versus Focused Challenge (a = .60). In the unscored second part of the inven­
tory, patients are presented with 11 open-ended questions regarding preferences about
the therapist, activities, and treatment.

Treatment Preference Interview

The Treatment Preference Interview (Vollmer et al., 2009; Table 6.1) is a semi­
structured, discussion-based tool that assesses patients’ preferences about the ther­
apist, psychotherapy roles and behaviors, and type of treatment. In the first part of
the interview, patients are asked about previous episodes of therapy and what they
 16 2 psy c h o th er a py rela tio n sh ips that w ork

Table 6 .1 . Treatm ent P reference Interview

Preference Factor Question Content and Examples

Therapist’s Strong preferences for counselor’s: gender, age, ethnicity or race,

characteristics language, sexual orientation, religion, or other?
Activity preferences P rior therapy or experience being helped: W hat was m ost helpful?
W hat was the worst a therapist could do?
Preferences for the counselor’s approach: Preference for a
therapist who takes charge, is active/talkative and expressive/
warm, or client taking charge and the therapist is m ore quiet
and reserved?
Preferences for treatm ent modality: Individual, couple, group, or
family sessions?
Preferences for therapy tasks: Try new things between sessions,
reading self-help books, watching self-help m ovies, going on­
line for inform ation
Type o f therapy Beliefs about the causes o f the problem : W ill o f God, unlucky
experiences, biological m ake-up, unm et em otional needs,
unrealistic expectations, relationship conflicts, lack o f self­
knowledge, life style, or lack o f will power?
Preferences for type o f therapy: Solution-Focused, Cognitive-
Behavioral, or Psychodynamic therapy? (Therapy descriptions
were also provided including typical goals, therapist-client
relationship, and tasks.)
Preferences for who decides about the type o f therapy: Client
makes the decision, client and therapist collaborate, or therapist
makes the decision?

might have found helpful or hindering. They are then asked to rate their preferences
about the therapist, treatment modality, psychotherapy roles and behaviors, and their
beliefs about the causes of their problems. The final section presents patients with a
description of different treatments and asks them to rate their preferences for each,
as well as whether they would prefer that they, or their therapist, decide on their
treatment.

Decision Aids
Decision aids are a class of healthcare tools developed to help patients identify and
articulate their treatment preferences (Health Foundation, 2014). These tools can be
used to provide prospective patients with information about the available treatments
for their particular problems. The aids also discuss the likely impact and the pros and
cons of each intervention. To date, these decision aids have primarily focused on phys­
ical health conditions. However, decision aids for mental health problems are now
 163 Preferences

available, both as written pamphlets (see BM J Group, 2015a, for an example) and as
web-based resources (see BMJ Group, 2015b, for an example).

CLINICAL EXAMPLE

The following case example demonstrates how client preferences can be addressed in
psychotherapy. This example shows how client preferences can be incorporated into
the initial treatment planning as well as ongoing therapy decision-making using struc­
tured methods.
“Ayo,” a 24-year-old student of African descent, was referred to his college counseling
service because he was experiencing low mood and struggling with his course work.
At assessment, Ayo’s score on the Patient Health Questionnaire-9 (Spitzer, Kroenke,
& Williams, 1999) indicated severe depression, and Ayo reported experiencing in­
tense anxiety and sadness since his childhood. When Ayo was about eight, his mother
started to experience manic episodes. This led to a hostile and complex separation
between Ayo’s parents—with Ayo trying to protect his father—and also thrown the
family into poverty. Many days, said Ayo, he went without any lunch. Ayo was badly
bullied at school. He had a limp that he was teased about, and he went through his
school years isolated and as an outsider. At assessment, Ayo also indicated that there
had been some traumatic event in his past, but he declined to say more. Ayo lived with
a partner who, for the first time, he experienced as loving and caring but felt that he
was “holding back” from him and becoming increasingly withdrawn. A major cause
of his depression, added Ayo, was the chronic pain that he experienced in his lower
back and leg. Partly because of this, Ayo would sleep most days until late afternoon,
and, with little time left in the day, he was finding it impossible to keep up with his
course work.
Ayo did not want to set specific goals for his therapy; however, his scores on the C-
NIP (Cooper & Norcross, 2016) at assessment indicated a strong preference for Past
Orientation and for Focused Challenge. He also showed a tendency toward wanting
Therapist Directiveness, scoring high on items such as “I would like the therapist . . . to
take a lead in therapy.”
To accommodate these preferences, his psychotherapist—an integrative psycholo­
gist with a predominantly person-centered leaning—strove to adopt a relatively active
and probing psychotherapeutic stance. He encouraged Ayo to talk about his “negli­
gent” and “disinterested” father and a romantic relationship that Ayo had in his early
20s, which he had experienced as very controlling and damaging.
For the psychotherapist, these early sessions seemed to be going well. His per­
ception was that Ayo was engaging fully in the therapeutic work and was exploring
past difficulties at a level of emotional depth. However, Ayo’s responses on the feed­
back measures used at the clinic indicated otherwise. On the post-session Session
Effectiveness Scale (Elliott, 2000), for instance, Ayo reported making only “a little
progress,” and he rated his first session at 3 on a 1 (very poor) to 7 (perfect) scale.
At Session 4, Ayo confirmed these experiences. He said that he “really didn’t know”
how he felt about the psychotherapy to date and that he was “not sure whether it had
 16 4 psy c h o th er a py r ela tio n sh ips that w ork

been helpful or not.” Ayo said that he felt he had a “bubble around him,” which made
it “difficult to let others in.” He was not sure whether his psychotherapist understood
him. On completing the C-NIP at this second time point, Ayo again indicated that his
preference was for a Focused Challenge way of working. This time, he also showed a
stronger preference toward Therapist Directiveness. This was explored further in the
following exchange.

t : And are there ways that I can be more helpful? I mean, I saw that in one of the
forms you were doing there [the C-NIP], about, wanting structure, does it feel too
unstructured or . . . ?
c : Slightly too unstructured. I kind of, don’t know what to talk about. So prompting’s
probably good.

The feedback from Ayo prompted the psychotherapist to increase his directiveness.
He invited Ayo to talk more about his previous, controlling relationship; and when
Ayo said that he just could not say—“it feels like a security screen coming down in a
bank”—the therapist encouraged him to stay with it and to try and “open things up” if
he could. Ayo did so and began to talk about the way that this partner would twist what
had happened to Ayo in his past, saying that things were Ayo’s fault. This led the psy­
chotherapist, again adopting a more directive stance, to encourage Ayo to say some­
thing about what this past event was. Ayo hesitated but, with the psychotherapist’s
persistence, began to talk about a painful episode of sexual abuse in his early teens.
At the end of this session, Ayo and his therapist reviewed how the session:

t : I just wanted to ask— I mean—you’ve been really open. And I guess I’ve been
pushing you a little bit more. Has that felt okay, or . . .?
c : Yeah . . . Yeah . . . I find it easier to talk about, I think, with you pushing me, rather
than just being left to speak.
t : Okay, okay, I guess it’s really important for me in that—that you can say, you know
“I’m not going to talk about that,” “I don’t want” . . . It’s really important that you
feel safe not to.
c : Things come out better when coaxed, I think, I just have that sort of brain (laughs).

Ayo confirmed this in his post-session feedback form. He wrote that what had been
particularly helpful in the session was that he “Spoke loosely about some things I did
not feel capable of speaking about at the start of the session.” In terms of what the ther­
apist had done that had felt helpful, he wrote, “Gently encouraged with his perceptions
and questions me to talk about things I didn’t feel capable of—find it easier to talk
when prodded.”
This example highlights, first, the benefits of assessing clients’ preferences
throughout the ongoing work, not just during initial assessment, and second, that the
psychotherapist’s “intuitive” sense of how treatment is proceeding can be an inaccu­
rate representation of progress. Hence, the use of a structured method proved valuable
in accessing a truer picture of how the client was experiencing therapy. This was also
 165 Preferences

Ayo’s view. At post-therapy he rated the C-NIP as “very helpful," saying “because it
meant that I got to make the decisions on paper, rather than telling a person" By this,
Ayo meant that it was easier for him to be more honest on a form rather than in speech
because he was less concerned about hurting the other person’s feelings. Ayo also said
that he valued the C-NIP because, “there are a lot of questions and they’re all very spe­
cific, which is great because they’re things I probably wouldn’t have thought of"

LANDMARK STUDIES
Landmark Study 1

Devine and Fernald’s (1973) investigation of receiving preferred, nonpreferred, or ran­

domly allocated group treatments for snake phobia is the earliest empirical study of
preference effects to be included in our meta-analysis. It is an example of a study that
randomized clients to match/choice and mismatch/no-choice conditions.
The study involved 48 participants who had an extreme fear of snakes, 32 of whom
were first shown a 40-minute video that described and illustrated four approaches
to treating snake phobia. These treatments were systematic desensitization, an “en­
counter approach” (which explored personal feelings), a rational-emotive technique
focusing on irrational thoughts, and a modeling/behavioral rehearsal approach. These
32 participants were then asked to indicate their liking for each approach on a 5-point
scale and were then allocated to a treatment for which they either expressed a strong
liking or disliking. The remaining 16 participants, who had not seen the video, were
then randomly allocated to one of the four treatments. Hence, each therapy group had
12 participants: 4 of whom had expressed a liking for that treatment, 4 of whom had
expressed a disliking for that treatment, and 4 of whom had been randomly allocated
to it. Each therapy group met for two one-hour sessions. At one-week follow-up, the
16 participants who had received their preferred intervention showed significantly less
fear of snakes than those participants who had received a nonpreferred therapy or had
been randomly assigned.
A major limitation of the study, as noted by the authors, was that therapist effects
were confounded with treatment effects. That is, as each treatment was presented and
implemented by a specific therapist, it is not clear whether the outcomes were due to
activity preference effects or preferences for the therapists. In addition, there is no evi­
dence from the article that the experimenter, who performed the behavioral ratings at
post-therapy, was blind to participants’ allocation. Hence, the outcomes may have been
due to biases in the experimenter’s ratings or to other experimenter effects.

Landmark Study 2

Elkin et al’s (1999) analysis of the relation between patients’ attitudes toward treatment
and their early engagement and outcomes came from the National Institute of Mental
Health’s Treatment of Depression Collaborative Research Program (Elkin et al., 1989).
It is an example of a preference study that analyzes data from clients randomized to
 166 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

different treatments. This multisite trial was once described as “the most methodo­
logically sophisticated study ever done” (Duncan et al., 2004, p. 26). It compared the
effectiveness of cognitive-behavior therapy (CBT), interpersonal psychotherapy (IPT),
imipramine plus clinical management, and placebo plus clinical management across
250 clients suffering from major depressive disorder.
Patients’ preferences for treatment were inferred based on their predilections for
each treatment. Predilections were defined as “their beliefs about the origins of their
distress and their expectations about what will be helpful to them” (Elkin et al., 1999,
p. 438). Based on responses to several items from a predilection questionnaire, patients
were classed as having a preference for a particular treatment if they scored above the
midpoint on the scale that corresponds to that treatment and if this score was at least
1 point higher than their scores for the two other scales. This resulted in 43 patients
with a predilection for CBT, 4 patients with a predilection for IPT, and 24 patients with
a predilection for medication. In addition, the authors identified 28 patients as having
a predilection for psychotherapy of either type—these were patients who scored above
the midpoints on both the CBT and IPT scales, at a level more than 1 point higher than
on the medication scale.
Elkin and colleagues (1999) compared early engagement outcomes for patients
who received a treatment congruent with their predilections against those who re­
ceived a treatment that was noncongruent with their predilections. Patients in the con­
gruent group consisted of those with a CBT, IPT, or psychotherapy predilection who
were assigned to either CBT or IPT (n = 32) and those with a medication predilection
assigned to imipramine plus clinical management (n = 8). The noncongruent sample
consisted of 42 patients. Four measures of early engagement were used: early attri-
tion/dropout; patients’ ratings of their therapists’ empathy, warmth, and genuineness;
observers’ ratings of the clients’ contribution to the therapeutic alliance; and depressive
symptoms.
Consistent with predictions, patients receiving their preferred treatment were sig­
nificantly less likely to show attrition, with 9 of the 11 early terminators (81.8%) coming
from the noncongruent group. Follow-up analysis of these patients’ explanation for
dropout indicated that, for at least six of the nine (66.7%), a major determinant had
been “dissatisfaction with treatment or desire for another treatment” (Elkin et al., 1999,
p. 445). In addition, patients in the congruent group had significantly higher ratings of
the therapist-provided therapeutic conditions and significantly higher ratings on the
patient’s contribution to the therapeutic alliance. After controlling for initial severity
of depression, preference congruence was found to explain 6% and 14% of these two
outcomes, respectively.
Elkin et al.’s (1999) study provides evidence that patients with a predilection for
medication or psychotherapy will engage more with that respective treatment. In par­
ticular, the differences in dropout between congruent and noncongruent groups was
quite striking (5% and 21.4%, respectively). A major limitation of this study was the
lack of evidence of reliability or validity for the predilection measure. Further, it is un­
clear whether scores on the predilection measure purely represent preferences, given
 16 7 Preferences

that the questions overlap with outcome expectations. As indicated, while patients who
expect or otherwise believe that certain treatments will help them (and attribute their
problems to related causes) are likely to prefer those treatments, it cannot be assumed
that this will automatically be the case.

Potential Landmark Study-

Qualitative research into clients’ experiences of expressing preferences in psycho­
therapy and having those preferences honored or not has the potential to become
a fertile area of psychotherapy research, with significant implications for practice.
However, to date, there is no published qualitative research examining the effects of
preference accommodation—let alone landmark studies. Here, therefore, we report
on a study that is in progress that illustrates some of the methods that may be used for
such research.
Gibson et al. (in preparation) examined clients’ experiences of shared decision­
making in therapy. Shared decision-making is a “process in which clinicians and
patients work together to select tests, treatments, management, or support packages,
based on clinical evidence and patients’ informed preferences” (Coulter & Collins,
2011, p. vii). It is conceptually and clinically close to preference accommodation.
The specific study aims were to identify (a) clients’ perceptions about whether
clinical decisions are shared, (b) the elements of the therapeutic process that facili­
tate shared decision-making, and (c) the impact that shared decision-making has on
clients’ experiences in psychotherapy. This study focused on shared decision-making
in pluralistic therapy (Cooper & Dryden, 2016; Cooper & McLeod, 2011), which is an
integrative approach that emphasizes eliciting and accommodating clients’ preferences.
In this treatment, therapists strive to understand and accommodate client preferences
through both structured clinical tools (such as the C-NIP) and unstructured clinical
conversations.
To examine clients’ experiences of shared decision-making and preference ac­
commodation, the study focused on initial assessment sessions. These sessions were
recorded and the primary researcher extracted segments in the recording in which
clinical decisions could be identified: for instance, the goals for therapy had been set,
clinical methods had been established, or the treatment contract had been agreed
upon. Within one week of this initial assessment, a researcher met with each client
to conduct an Interpersonal Process Recall interview (Elliott, 1986; Kagan, 1973).
In this process of stimulated recall, clients were played the selected audio segments
and asked—through structured interview questions—about their experiences of the
decision-making process: whether they felt it was shared, what had facilitated sharing
in the process, and what the impact of the sharing was. For triangulation purposes, a
second interview was subsequently conducted with each client after his or her fourth
psychotherapy session, and a similar series of questions were asked. Qualitative
responses for each question will be analyzed using an adapted grounded theory ap­
proach (Rennie et al., 1988).
 168 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

RESULTS OF PREVIOUS META-ANALYSES

Several meta-analyses have been conducted examining the influence of preference ac­
commodation on psychotherapy outcomes. In 2011, in the first edition of this book,
Glass et al. (2001) conducted a review identifying 4 studies that compared activity
preferences to treatment outcomes, 10 studies for treatment preferences, and 1 study
of preferences about the therapist. Using a box count method, 3 of the 15 reviewed
studies did find a significant relation between preference matching and treatment
outcomes; however the remaining 12 found mixed or null results. Although this re­
view concluded with the recommendation for therapists to pay close attention to their
clients’ preferences, it also indicated that the empirical research to that point had not
provided a strong justification for doing so.
The first meta-analysis examining the psychotherapy preference effect was
conducted just under a decade ago (Swift & Callahan, 2009). Researchers found 26
studies that examined the impact of preference matching on treatment dropout or
outcome. The researchers found an average an odds ratio (OR) of 0.58 for the rela­
tionship between preference accommodation and dropout. This effect indicates that
clients whose preferences were matched were between a third and half as likely to
prematurely terminate from psychotherapy compared to clients whose preferences
were not matched. They also found an average correlation of .15 for the relation be­
tween preference matching and treatment outcome. This effect indicates that clients
whose preferences were matched had a 58% chance of showing greater improvement
compared to preference mismatched clients. Although important, the results of this
first meta-analysis primarily focused on treatment preferences.
Shortly after, two more meta-analyses were conducted that expanded on those
results. The first meta-analysis (Swift et al., 2011), in the second edition of this book,
pooled data from 35 studies to examine the impact of client preferences, again on both
treatment dropout and treatment outcomes. Very similar effects were found—OR of .59
for the relationship between preference accommodation and dropout and a Cohen’s d
of .31 for the association with treatment outcomes. In examining potential moderators,
the researchers found that preference type, type of outcome measurement, and timing
of outcome measurement did not have a significant impact on the psychotherapy pref­
erence effect; however, treatment options, diagnosis, and study design did. Specifically,
the preference accommodation effect was stronger when the treatment options were
psychotherapy versus medication, compared to two types of psychotherapy; prefer­
ence accommodation was most important for clients with anxiety, depression, or a
substance use problem; and partially randomized preference trials and studies where
clients were allocated to choice conditions tended to show smaller effects compared to
studies where clients were allocated to treatment conditions.
The second meta-analysis used the same data set to examine the relationship be­
tween several additional client (age, ethnicity, education level, gender, relationship
status) and treatment (duration) variables and the value of preference accommodation
in psychotherapy (Swift et al., 2013). The authors found that the preference outcome
 169 Preferences

effect was not moderated by any of these variables, suggesting that preference accom­
modation is equally important across client types and treatment durations.
The most recent meta-analysis of preference accommodation in psychotherapy
was published in 2014 (Lindhiem et al., 2014). This meta-analysis built on previous
ones by also examining the influence of preferences on treatment satisfaction and
the therapeutic alliance and by including studies of treatments for physical health
problems; however, it missed several studies that were identified by the previous
reviews (even with broader inclusion criteria, only 32 trials were identified). Based
on the authors’ included studies, this meta-analysis found a Cohen’s d of 0.34 for the
relationship between preference accommodation and treatment satisfaction (similar
in magnitude to previous meta-analyses) but a d of only .15 for treatment outcome.
The authors also found a slightly smaller relation between preference accommodation
and treatment dropout (OR = 1.37 in favor of mismatched being more likely to drop
out prematurely).

META-ANALYTIC REVIEW

The purpose of the current meta-analysis was to examine whether client preference ac­
commodation in psychotherapy is associated with fewer treatment dropouts and more
positive treatment outcomes. For some of the included studies, preference accom­
modation occurred through matching clients to a preferred treatment or therapeutic
condition. For other studies, preference accommodation occurred by allowing clients
to directly choose their therapist, treatment behaviors, and/or interventions. We first
present the overall findings regarding the impact of client preference accommodation
on both dropout and outcome. We then provide the results regarding moderators and
covariates of psychotherapy preference.

Inclusion Criteria
Our goal was to include all studies that were published in the English language that
quantitatively examined the impact of preference accommodation on treatment
dropout or other treatment outcomes. To be included, studies had to either assess
preferences directly and then make a comparison between preference match and non­
match conditions or include a comparison of conditions where clients were placed into
choice/no-choice conditions. Studies were excluded if they used a nonclinical sample,
if the treatments were for a nonclinical problem, and if they did not involve at least
one treatment condition that was an in-person psychological intervention. We also
excluded studies of family therapy, couple therapy, and treatments for children and
adolescents. This decision was made because in those treatments it is more difficult to
determine whose preferences actually inform decisions that may impact dropout and
outcomes (e.g., a child client whose preferences are not matched may still complete
treatment if his or her parents’ preferences are met; in couple therapy one partner’s
preferences may be matched while the other’s preferences are not matched). Where we
 17 0 psy c h o th er a py r ela tio n sh ips that w ork

identified multiple articles that used the same data, only the study with the most com­
plete data related to preference matching was included in the meta-analysis.

Search Strategy
Two independent coders reviewed each potential article to determine if it met the in­
clusion criteria for this meta-analysis. Each potential article that was identified through
the three search strategies was reviewed by both coders at the abstract and title level.
I f either coder believed that based on the abstract and title that the article could po­
tentially fit the inclusion criteria, then the article was reviewed independently by both
coders at the full text level. Figure 6.2 presents a flow chart of the search strategies and
results.
Our first search strategy was a review of all articles that had been included in one (or
more) of the six previous meta-analyses/reviews on psychotherapy preference (Glass
et al., 2001; Lindhiem et al., 2014; Rosen, 1967; Swift & Callahan, 2009; Swift et al.,
2011; Swift et al., 2013). We also reviewed the abstracts and titles of all articles listed
in PsycINFO (k = 181) and PubMed (k = 64) that cited one of these previous reviews.
This resulted in 68 articles passed to the full text review.
Our second search strategy included term searches conducted in PsycINFO and
PubMed. In PsycINFO the terms preference OR choice, AND therapy OR psycho­
therapy OR treatment OR therapist OR counselor OR role OR therapeutic alliance, AND
matching OR outcome resulted in 9,924 citations that were reviewed at the abstract and
title level. In PubMed the terms preference OR choice, AND psychotherapy OR therapist
OR counselor OR therapeutic alliance, AND matching OR outcome resulted in 1,405
citations that were reviewed at the title and abstract level. This resulted in 67 additional
studies that passed to full text review.
Our third and final search strategy was a root and branch search in PsycINFO of all
of the identified potential articles after the abstract and title review. This included 417
articles that were in the reference lists of the studies that were identified for potential
inclusion and 2,475 articles that cited one of the studies that were identified for poten­
tial inclusion. After a review of the titles and abstracts of these articles, 14 were passed
on for full text review.
The full text review was performed on 149 articles identified as potential relevant
through the three search strategies, plus 2 more articles that were recommended by
researchers with expertise in studying client preferences. We removed 98 of the po­
tential articles after the full text review because 25 of them did not include choice or
preference conditions, 14 were not quantitative studies, 14 did not address a clinical
problem, 10 did not include psychotherapy as a treatment condition, 8 were from the
same data set as an included article, 8 did not include at least one in-person psycholog­
ical intervention, 6 did not assess treatment outcomes or psychotherapy dropout, 4 did
not report data for preference match/non-match conditions, 3 had child participants,
2 had preference match/non-match conditions that also differed completely in the
treatments that were provided, and 1 was not available in English. The full text of three
 17 1 Preferences

f ig u r e 6 . 2 Flow chart of the search.

additional studies that were passed on for full text review could not be obtained, even
after contact with the study authors. The full text review resulted in 53 studies that
met all inclusion criteria to be included in this meta-analysis.

Study Coding
Each included study was coded by two independent coders. These reviewers coded the
study year, the study design (correlational, partially randomized preference trial, ran­
domization to choice/no-choice conditions, randomization to treatment conditions,
 17 2 P S Y C H O T H E R A P Y R EL AT IO N S H IP S THAT WORK

other), the preference type (activity, treatment, therapist), the treatment options (psy­
chotherapy vs. psychotherapy, psychotherapy vs. medication, psychotherapy vs. other),
the duration of the psychological intervention, the client presenting problem (anxiety,
depression, behavioral health problems, psychotic disorders, substance use problems,
other), client average age, percentage of the sample that was female, percentage of the
sample that was White, non-Hispanic, and average years of patient education. The two
coders independently agreed on 94.9% of the ratings. For the disagreements, the two
coders jointly reviewed the article to discuss their ratings and then jointly made a final
coding decision.

Methodological Decisions
Since the purpose of this study was to compare dropout rates and outcomes between
preference match/choice and preference non-match/no-choice groups, two effect
sizes were calculated for each study. First, an OR that represents the likelihood of
a mismatched/no-choice client dropping out over a matched/choice client drop­
ping out was calculated for each study that included the necessary data. In these
calculations, an OR greater than 1 indicates that mismatched/no-choice clients were
more likely to drop out prematurely, an OR less than 1 indicates that matched/choice
clients were more likely to drop out prematurely, and an OR of 1 indicates that clients
in both groups were equally likely to drop out. Second, a Cohen’s d, which represents
differences in the degree of improvement between matched/choice and mismatched/
no-choice groups, was calculated for each study that included the necessary data.
Several studies included data for more than one outcome variable. In these cases, an
average d was calculated for each study based on all of the outcome data that was re­
ported within that study.
After calculating individual study effect sizes, an overall weighted OR and an overall
weighted d was calculated averaging across all studies using a random effects model.
A one-study removed analysis was conducted to examine whether outlier studies were
having a large impact on the overall findings. A Q-statistic was used to test for hetero­
geneity and an I2 statistic was calculated to examine the degree of heterogeneity in the
study effect sizes. Last, a fail-safe N was calculated to examine the robustness of the
overall results against missing studies.
Several study, preference, and client characteristics were then tested as potential
moderators. Categorical moderators were tested using a Q-statistic and a mixed-effects
model. With this type of moderator analysis, a significant between group Q-statistic
indicates a higher than chance level of between-group differences in effect sizes. Post
hoc pairwise comparisons were conducted for each significant categorical moder­
ator. Continuous moderators were tested using meta-regression techniques. Using a
random-effects model, a regression coefficient was calculated representing the linear
relationship between the continuous moderator and the study effect size. For the pref­
erence dropout effect, the predictor variables were regressed onto the log ORs for the
studies. For the preference outcome effect, the predictor variables were regressed onto
the studies’ d values.
 173 Preferences

Effect on Dropout
Twenty-eight of the 53 included studies reported data on treatment dropout. These
28 studies included data from 3,237 clients. Figure 6.3 shows a forest plot of the OR
effect sizes for each study and the weighted average effect size. The overall preference
effect on psychotherapy dropout was significant, OR = 1.79, 95% confidence interval
(CI) = 1.44-2.22, p < .001. This indicates that clients whose preferences were not
matched or who were not given a choice of their treatment conditions were 1.79 times
more likely to prematurely terminate than clients who were matched to their prefer­
ence or who were given a choice of their conditions. The one study removed analyses
resulted in ORs ranging from 1.71 to 1.89, indicating that the results were not highly
influenced by any single study. Calculation of the fail-safe N indicated that 300 un­
published studies with nonsignificant results would be necessary to reduce this result
to a nonsignificant level. Significant heterogeneity in the study ORs was not found,
Q(27) = 34.76, p = .15, I2 = 22.33.

Effect on Outcome
Fifty-one of the 53 included studies reported data on treatment outcomes, including
data from 16,269 client participants. Figure 6.4 shows a forest plot of the outcome
effects sizes for each study and the weighted average effect size. The overall preference

Study Effect (O . R .)
Bakker et al., 2000 1.19
Carey et al., 2013 3.15
Dunlop et al., 2012 1.94
Dunlop et al., 2017 2.20
Elkin et al., 1999 5.18
Ersner-Hershfield et al., 1979 2.95
Fuller, 1988 2.23
Hegerl et al., 2010 1.77
Kerns et al., 2014 0.86
Kludt, 1999 1.74
Kocsis et al., 2009 1.24
Kwan et al., 2010 16.29
Le et al., 2014 1.91
Leykin et al., 2007 1.61
Macias et al., 2005 4.34
Manthei et al., 1982 2.17
McKay et al., 1995 1.32
McKay et al., 1998 0.89
Mergl et al., 2010 1.20
Moradveisi et al., 2014 7.31
Proctor & Rosen, 1981 2.81
Raue et al., 2009 21.34
Renjilian et al., 2001 0.70
Rokke et al., 1999 12.00
Sterling et al., 1997 1.20
Van et al., 2009 1.59
Wheaton et al., 2016 3.71
Zlotnick et al., 1998 1.59
Weighted Average 1.79
0.00 1.00 2.00 3.00 4.00 5.00 6.00

f ig u r e 6 . 3 Forest plot of the preference dropout effect (odds ratios with 95% confidence interval
bars).
 174 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

Study Effect (d)

Adamson et al., 2005 0.11
Al-Otaiba et al., 2008 0.25
Atkinson et al., 1991 0.12
Bakker et al., 2000 0.28
Berg et al., 2008 0.90
Brown et al., 2002 0.45
Calsyn et al., 2000 0.18
Carey et al., 2013 0.12
Chilvers et al., 2001 0.22
Cooper et al., 2017 0.61
Cooper, 1980 Study 1 0.69
Cooper, 1980 Study 2 0.54
Devine & Fernald, 1973 1.19
Dowrick et al., 2011 0.78
Dunlop et al., 2012 -0.07
Dunlop et al., 2017 0.24
Dyck & Spinhoven, 1997 0.22
Elkin et al., 1999 0.56
Fuller, 1988 -0.16
Gossop et al., 1986 0.51
Gum et al., 2006 0.08
Hamann et al., 2007 0.05
Handelzalts & Keinan, 2010 0.30
Hegerl et al., 2010 0.15
Iacoviello et al., 2007 1.15
Kadish, 1999 0.43
Kay-Lambkin et al., 2012 0.21
Kerns et al., 2014 0.01
Kludt, 1999 0.16
Kocsis et al., 2009 0.70
Kwan et al., 2010 0.50
Le et al., 2014 0.08
Leykin et al., 2007 0.18
Lin et al., 2005 0.27
Macias et al., 2005 0.06
Manthei et al., 1982 0.23
McKay et al., 1995 0.03
McKay et al., 1998 -0.01
Mergl et al., 2010 0.33
Moradveisi et al., 2014 0.25
Raue et al., 2009 -0.18
Renjilian et al., 2001 -0.14
Rokke et al., 1999 0.37
Sidani et al., 2015 0.27
Sterling et al., 1997 0.07
Van et al., 2009 0.36
Wallach, 1988 0.49
Ward et al., 2000 -0.04
Wheaton et al., 2016 0.37
Williams et al., 2016 0.74
Zlotnick et al., 1998 0.36
Weighted Average 0.28
-0.40

f ig u r e 6 . 4 Forest plot of the preference outcome effect (standardized d values with 95% confidence
interval bars).

effect on psychotherapy outcome was significant, d = 0.28, 95% CI = .17-.38, p < .001.
This effect size indicates a small but meaningful difference in outcomes in favor of
clients who are given their preferred psychotherapy. The one study removed analyses
resulted in d s ranging from 0.23 to 0.29, indicating that the outcome results were
 175 Preferences

similarly not highly influenced by any single study. Calculation of the fail-safe N indi­
cated that 4,177 unpublished studies with nonsignificant results would be necessary to
reduce the outcome effect size to a nonsignificant level. Contrasting the results from
the dropout analyses, significant heterogeneity in the outcome effect sizes (ds) between
studies was found, Q(50) = 657.3, p < .001, I2 = 92.39.

Review of Results
Based on the results from 28 studies and over 3,000 clients, preference accommodation
does have an influence on rates of premature termination. Specifically, clients whose
preferences are not matched in psychotherapy are almost twice as likely to drop out
prematurely. This result appears to be both precise (based on the confidence intervals)
and robust (based on the results of the one study removed analyses and calculation of
the fail-safe N).
Based on the results from 51 studies and over 16,000 clients, we can confidently say
that preference accommodation also has an impact on treatment outcomes. The overall
preference outcome effect that was found in this meta-analysis was small, d = 0.28, but
significant and clinically meaningful. A high degree of precision and robustness was
also observed for the preference outcome effect; however, the studies were found to be
highly heterogeneous (I2 = 92.39) in the magnitude of the effects that were reported.

MODERATORS AND COVARIATES

Study Characteristics
We investigated as potential moderators of the preference-dropout and preference-
outcome associations for five characteristics of the study: research design, publication
date, treatment duration, outcome measure, and timing of outcome measurement.
Four study designs were identified in the literature: correlational studies, partially
randomized preference trials, studies that randomized clients to choice/no-choice
conditions, and studies that randomized clients to treatment conditions. The average
dropout percentages did not significantly differ, QB(2) = 4.68, p = .10. However, signif­
icant differences in the preference outcome effect were found between study designs,
Qb(3) = 13.76, p = .003. Specifically, a significantly higher preference outcome ef­
fect was found in studies that randomized clients according to treatment conditions
(k = 23, d = 0.36, 95% Cl = .2 7 - .46) compared to partially randomized preference
trials (k = 7, d = 0.13, 95% Cl = .0 0 4 - .263), QB(1) = 7.642, p = .006. Studies that
randomized clients according to treatment conditions also reported a significantly
larger preference effect than studies that randomized clients to choice/no-choice
conditions (k = 17, d = 0.14, 95% Cl = .04-.23), QB(1) = 10.90, p = .001. Although the
largest preference outcome effect was observed in the correlational studies (d = 0.46,
95% Cl = - . 14-1.06), the effect was not significantly different from that seen in the
other designs due to decreased power to detect differences (only two studies used a
correlational design).
 17 6 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

The studies included in this meta-analysis ranged in year from 1973 to 2017 (see
Figure 6.4 for a breakdown of number of studies by year). There was not a significant
association between year of publication and the dropout OR effect sizes, B = -0.001,
95% CI= -.0 2 -.0 2 , z = 0.12, p = .90, R2 . = .00. There was also not a significant asso­
ciation between year of publication and the outcome d effect sizes, B = -0.001, 95% CI
= -.0 1 -.0 1 , z = 0.14, p = .89, R2 , = .00.
Meta-regression analyses were conducted comparing the study treatment durations
to the dropout and outcome effects. A significant relation between the preference
dropout effect and treatment duration was not found, B = -0.02, 95% CI = -.0 6 -.0 2 ,
z = 1.14, p = .25, R2 , = .05. There was also no evidence for a significant relation be­
tween the preference dropout effect and treatment duration, B = 0.001, 95% CI = -.0 1 ­
.01, z = 0.16, p = .87, R2 , = .00.
A total of 179 outcomes were reported across the 51 studies that reported outcomes
between preference match/choice and preference mismatch/no-choice groups. Each of
those outcomes were coded as a measure of a psychotherapy process variable (e.g., the
therapeutic alliance), client satisfaction with treatment, or a treatment outcome vari­
able (e.g., scores on a behavioral/symptom measure, frequency of heavy drinking). The
preference outcome effect was found to differ significantly depending on which type
of outcome was assessed, QB(2) = 13.60, p = .001. For assessments of process variables
(k = 15), the preference effect was d = 0.51, 95% CI = .28-.75. For assessments of client
satisfaction (k = 10), the difference between preference matched/choice and preference
mismatched/no-choice clients was only d = 0.03, 95% CI = -.0 9 -.1 7 . For actual treat­
ment outcomes (k = 154), the preference effect was d = 0.23, 95% CI = .16-.29. Post
hoc pairwise comparisons indicated significant differences in the preference effects for
process variables compared to outcome variables, QB(1) = 5.37, p = .02, between the
process variables and ratings of satisfaction for psychotherapy, QB(1) = 12.19, p < .001,
and between the outcome variables and ratings of satisfaction, QB(1) = 6.64, p = .01.
In summary, preference matched/choice clients express similar levels of satisfaction
with their treatments as preference mismatched/no-choice clients; however, prefer­
ence matched clients are more likely to experience positive process experiences and
treatment outcomes while in psychotherapy.
We also coded the timing of assessment for each outcome variable as either
midtreatment or posttreatment. The average effect size for midtreatment measurements
(k = 16) was d = 0.44, 95% CI = .2 7 - .61, was higher than the average effect size for
posttreatment measurements (k = 163), and was d = 0.25, 95% CI = .19-.31. Although
both effect sizes were significant, there was also a significant difference between them,
Qb(1) = 5.73, p = .02.

Preference Characteristics
We also examined as potential moderators two characteristics of preferences: type of
preference and treatment options. See Figure 6.5 for a breakdown of preference studies
by preference type and year. Neither the preference dropout effect, QB(2) = 3.39,
 177 Preferences

9 r

Treatment preferences □ Activity preferences

Preference about the therapist □ Mixed preferences

f ig u r e 6. 5 Number of preference effect studies identified by year and preference type.

p = .18, nor the preference outcome effect, QB(2) = 2.00, p = .37, differed depending
on the type of preference—activity, therapist, or treatment. Treatment options (psy­
chotherapy vs. psychotherapy, psychotherapy vs. medication, psychotherapy vs. other
type of treatment) was also not a significant moderator of either dropout, QB(2) = 1.06,
p = .59, or outcome, QB(2) = 0.51, p = .78.

Client Characteristics
Preference accommodation may have a larger or smaller effect depending on the
client. We tested the client predictors of problem being treated, average age, percentage
of the sample that was female, percentage of the sample that was White, non-Hispanic,
and average years of education to determine if they moderated the overall preference
dropout and outcome effects that were found through the main analyses.
Problem type (depression, behavioral health problems, substance abuse) was not
a significant moderator of treatment dropout, QB(2) = 4.71, p = .10; however, it was
a significant moderator of the preference outcome effect, QB(4) = 15.57, p = .004. The
highest outcome effects were observed in treatments for anxiety ( k = 7, OR = .49,
95% CI = .2 7 - .71), followed by depression ( k = 20, OR = .30, 95% CI = .1 9 - .42),
psychotic disorders ( k = 3, OR = .14, 95% CI = .0 1 - .26), substance use problems
( k = 10, OR = 0.12, 95% CI = .04-.21), and lastly behavioral health problems ( k = 5,
OR = .07, 95% CI = -.1 4 -.2 8 ). Post hoc pairwise comparisons indicated that the pref­
erence outcome effect was not significantly different between anxiety and depres­
sion, Qb(1) = 2.15, p = .14. However, the preference outcome effect was significantly
lower in psychotic disorders than both anxiety, QB(1) = 7.34, p = .007, and depression,
Qb(1) = 3.79, p = .05. The preference outcome effect was also significantly lower in sub­
stance use disorders compared to both anxiety, QB(1) = 9.20, p = .002, and depression,
Qb(1) = 6.54, p = .01. Similarly, the preference outcome effect was significantly lower
 17 8 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

in the treatment of behavioral health concerns compared to the treatment of anxiety,

Qb(1) = 7.41, p = .007, or depression, QB (1) = 3.87, p = .05. In summary, preference ac­
commodation appears to have a larger impact on treatment outcomes for clients who
are seeking treatment for an anxiety disorder or for depression compared to clients
who are seeking treatment for behavioral health concerns, psychotic disorders, or sub­
stance use problems.
Age was not found to moderate the preference dropout effect, k = 25, B = .005, 95%
CI = -.0 3 -.0 4 , z = 0.25, p = .80, R 2 . = .002, nor the preference outcome effect,
k = 45, B = -.002, 95% CI = -.0 1 -.0 0 , z = 0.49, p = .62, R 2 . = .00. Similarly, neither
the preference dropout effect, k = 25, B = .60, 95% CI = -.2 0 -1 .4 1 , z = 1.46, p = .14,
R2 , = .08, nor the outcome effect, k = 47, B = .17, 95% CI = - 0 .2 4 - .59, z = 0.82,
p = .41, R 2analog = .03, were moderated by client gender. Percentage of the sample who
was White, non-Hispanic did not moderate dropout rates, k = 19, B = .76, 95% CI
= -.1 0 -1 .6 1 , z = 1.73, p = .08, R 2 , = .15, or outcomes, k = 31, B = .36, 95% CI = -
.09-.81, z = 1.58, p = .11, R 2 analog
, = .13. Likewise, neither the preference dropout effect,
k = 10, B = .17, 95% CI = -.1 0 -.4 4 , z = 1.21, p = .23, R 2analog
, = .14, nor the outcome ef-
fect, k = 17, B = .03, 95% CI = -.0 3 -.1 0 , z = 1.02, p = .31, R 2analog
, = .05, was moderated
by client years of education.

Review of Moderator Results

In summary, the heterogeneity in the preference outcome effect was partially
explained by study design, with higher effects being seen in correlational studies
and studies that randomized clients to treatment conditions. In addition, preference
accommodation had a larger impact on treatment process and treatment outcomes
and a much smaller impact on clients’ ratings of treatment satisfaction. Also,
preferences had a larger impact on midtreatment measures of outcome compared
to posttreatment measures. Finally, the preference outcome effect was significantly
stronger for treatments o f anxiety and depression compared to treatments for psy­
chotic disorders, substance use problems, and behavioral health problems. The
preference outcome effect was not related to the year of publication, the treatment
duration, the type of preference, the treatment options, or the client’s age, gender,
ethnicity, or years of education.

LIMITATIONS OF THE RESEARCH

The lim itations of the present meta-analyses include that they were English-
only studies; there was a preponderance of evidence com ing from W estern, de­
veloped countries; and many studies examining the effect of patient preferences
stemmed from post hoc analyses in which preferences were not a central focus
o f the study. The number of available studies that allow for more fine-grained
analyses examining the association of specific patient characteristics and/or types
o f preferences is small. In addition, the finding of between-study heterogeneity in
 1 79 Preferences

the preferences-outcom es association strongly suggests a need for more research.

In particular, future research is strongly encouraged that measures the strength of
specific preferences and/or examines the possibility that preference strength may
vary according to the phase of treatment. Finally, although new measures of pref­
erence have appeared in the literature, psychom etric study o f those measures has
been limited and the extent to which they will be used clinically or in research is
not yet known.

DIVERSITY CONSIDERATIONS

Diversity constitutes an important component of patient preferences. For example,

ethnic minorities may not seek needed psychotherapy due to concerns that their
preferences will not be accommodated (e.g., González et al., 2010; Smith et al., 2011;
Sue & Zane, 2009). Unfortunately, in this meta-analysis we were only able to group
patients into two groups, minority versus nonminority, which probably overlooks vital
race or ethnicity effects. Based on our analyses, the results indicate that it will prove
equally effective to accommodate preferences for all clients, regardless of the racial or
ethnic background. We strongly encourage researchers to report race and ethnicity
data with greater clarity in future studies.
A large body of research has sought to identify the preferences that various ra­
cial and ethnic groups hold for psychotherapy. Much of this research has focused on
clients’ preferences for a therapist whose race and/or ethnicity matches their own.
Reviews of this research have found a moderately strong effect size (d = 0.61) for such
a match (Cabral & Smith, 2011). Other studies have indicated that racial and ethnic
minority clients may value other therapist characteristic (e.g., attitudes, values) more
strongly than demographic match (Atkinson et al., 1998; Bennett & BigFoot-Sipes,
1991; Stewart et al., 2013). For example, ethnic minority clients have been found
to value the therapists’ multicultural competence and a willingness to use cultural
adaptations over twice as strongly as they value racial or ethnic matching (Swift et al.,
2015). Thus therapists may best meet the preferences of individuals with different
ethnic backgrounds through openness and a willingness to work within the client’s
cultural framework.
A smaller body of research has tested for preference differences based on gender.
The results of this meta-analysis suggest that it is equally effective to accommodate
preferences for both male and female clients; however, previous research has indi­
cated that males and females may hold different preferences for psychotherapy. For
example, in one survey, males preferred more stereotypical feminine sex-role traits in
their therapists while females preferred more stereotypical masculine sex-role traits
(DeGeorge et al., 2013). Further research is needed examining gender differences in
psychotherapy preferences.
Very little research has examined preferences for psychotherapy based on client
sexual orientation, socioeconomic status, disability, or religion. These are all impor­
tant areas for future study.
 180 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

TRAINING IMPLICATIONS

Evidence-based practice necessitates attending carefully to patient characteristics,

culture, and preferences (APA, 2006). One simple method for providing training
opportunities in preference accommodation is to structure clinical training so that
each student works with a variety of patients in diverse settings. W ith variability,
students are more likely to gain experience in recognizing and working with several
different preferences that a diverse group of clients may hold.
Given the impact of preference accommodation on premature termination and
treatment outcomes, clinical programs can help trainees to develop competences in
this area. Competences might include actively inviting clients to express their opinions
regarding their preferred activities, therapist characteristics, and treatments; effec­
tively using patient preference scales; and impartially communicating to patients a
range of treatment options. We also recommend training students to develop higher­
order “meta-competences” in this area. These might include the capacity to respond
to clients when their preferences may be more difficult to accommodate, balance pref­
erence accommodation against a psychologically informed assessment of what the
client needs, and recognize the individuals client’s preferences regarding the treatment
decision-making process (some clients may prefer that the therapist make all of the
treatment decisions). We recommend preference training be formally programmed
into a training syllabus, rather than scheduled on an informal and ad hoc basis.
Supervisors are encouraged to bring diversity considerations and patient
preferences into discussion regularly and guide trainees in determining when, how, or
if they should implement preference accommodations. In that context, we recommend
supervisors teach trainees shared decision-making marked by several components
(Charles et al., 2003); information must be shared across the dyad, who discuss
preferences as a precondition to any treatment decision and then attain consensus on
treatment implementations.

THERAPEUTIC PRACTICES

Based on the research, we conclude that client preferences exert a positive influence on
therapy dropout and treatment outcomes. Given this significant preference effect, we
offer the following:

♦ Assess clients’ preferences in each of the three domains: activity preferences, therapist
preferences, and treatment type preferences.
♦ Assess strong patient preferences within each domain. Activity preferences can
include the therapists “style” (e.g., more or less directive), therapy tasks, the goals of
treatment, and number/length/frequency of sessions. Preferences about the therapist
can include his or her gender, race/ethnicity, sexuality orientation, religion, and
personality characteristics.
♦ Seek to overcome barriers that might prevent clients from expressing their
preferences, such as paucity of information about therapy options, lack of trust in the
therapist, or low readiness to change.
 18 1 Preferences

♦ Invite clients to express their preferences, particularly in areas where they may fear
offending or upsetting the therapist.
♦ Provide clients with an understanding of the different options available. The
information provided can be informed by the evidence available, but therapists are
recommended to hold a position of “decisional equipoise.”
♦ Convey an attitude of nonjudgmental acceptance to whatever choices clients make,
unless ethically contraindicated.
♦ Pay particular attention to any strong preferences expressed by clients, including
strong disliking for particular tasks or therapist characteristics.
♦ Inform patients of the available options prior to the start of treatment. Assessment
sessions can be used to develop a deeper understanding of clients’ preferences.
During the initial session it may be particularly useful to ask clients about previous
experiences of therapy and what they found helpful or unhelpful.
♦ Consider using measures or structured interviews to comprehensively assess clients’
preferences, both initially and on an ongoing basis. When using measures, raw
responses are most useful when followed up through dialogue and further elicitation
of views.
♦ Address client preferences throughout the therapy process. Clients may change their
preferences after starting treatment, or they may believe that their preferences are not
being addressed despite therapists’ attempts to do so. Schedule periods of review in
session.
♦ Share therapist belief when a client’s treatment preferences are not in the client’s best
interest, so that treatment decisions can still be made collaboratively.
♦ Tailor the degree of dialogue about—and accommodation to—clients’ preferences to
the individual client. It is not recommended that clients be pressured into expressing
and discussing preferences if they cannot, or do not want to, do so.

REFERENCES
References m arked with an asterisk indicate studies included in the m eta-analysis.
*A dam son, S. J., Sellm an, J. D., & Dore, G. M . (2005). Therapy preference and treatm ent out­
com e in clients with m ild to m oderate alcohol dependence. Drug and Alcohol Review, 24,
2 0 9 -2 1 6 . https://www.doi.org/10.1080/09595230500167502
*A l-O taiba, Z., W orden, B. L., McCrady, B. S., & Epstein, E. E. (2008). A ccounting for self­
selected drinking goals in the assessm ent o f treatm ent outcom e. Psychology of Addictive
Behaviors, 22, 4 3 9 -4 4 3 . https://www.doi.org/10.1037/0893-164X.223.439
A m erican Psychological A ssociation Presidential Task Force on Evidence-Based Practice.
(2006). Evidence-based practice in psychology. American Psychologist, 61, 2 7 1 -2 8 5 .
https://www.doi.org/10.1037/0003-066X.61.4.271
*A tkinson, D., W orthington, R. L., Dana, D. M ., & G ood, G. E. (1991). Etiology beliefs,
preferences for counseling orientations, and counseling effectiveness. Journal o f Counseling
Psychology, 38, 2 5 8 -2 6 4 . https://www.doi.org/10.1037/0022-0167.383.258
*A tkinson, D. R., Wampold, B. E., Lowe, S. M ., Matthews, L., & A hn, H. (1998). Asian
A m erican preferences for counselor characteristics: Application o f the Bradley-Terry-Luce
m odel to paired com parison data. The Counseling Psychologist, 26, 1 0 1 -1 2 3 . https://www.
doi. org/10.1177/0011000098261006
 18 2 p sy c h o th er a py rela tio n sh ips that w ork

*Bakker, A., Spinhoven, P., Van Balkom , A. J. L. M ., Vleugel, L., & Van Dyck, R. (2000).
Cognitive therapy by allocation versus cognitive therapy by preference in the treatm ent
o f panic disorder. Psychotherapy and Psychosomatics, 69, 2 4 0 -2 4 3 . https://www.doi.org/
10.1159/000012402
Bennett, S. K., & BigFoot-Sipes, D. S. (1991). A m erican Indian and W hite college student
preferences for counselor characteristics. Journal o f Counseling Psychology, 38, 4 4 0 -4 4 5 .
https://www.doi.org/10.1037/0022-0167.38.4.440
B M J Group. (2015a). Depression decision aid. Retrieved from http://sdm.rightcare.nhs.uk/
pda/depression/
B M J Group. (2015b). Shared decision making: Deciding what to do about depression.
Retrieved from sdm.rightcare.nhs.uk/shared-decision-making-sheets/depression/
Boswell, J. F., Constantino, M. J., Oswald, J. M ., Bugatti, M ., Goodw in, B., & Yucel, R. (2018).
M ental health care consum ers’ relative valuing o f clinician perform ance inform ation.
Journal o f Consulting and Clinical Psychology, 86(4), 3 0 1 -3 0 8 .
Bowen, M ., & Cooper, M . (2012). Developm ent o f a client feedback tool: A qualitative
study o f therapists’ experiences o f using the Therapy Personalisation Forms. European
Journal of Psychotherapy and Counselling, 14, 4 7 -6 2 . https://www.doi.org/10.1080/
13642537.2012.652392
*Brow n, T. G ., Seraganian, P , Tremblay, J., & Annis, H. (2002). M atching substance abuse af­
tercare treatm ents to client characteristics. Addictive Behaviors, 27, 5 8 5 -6 0 4 . https://www.
doi.org/10.1016/ S0306-4603(01)00195-2
Cabral, R. R., & Sm ith, T. B. (2011). Racial/ethnic m atching o f clients and therapists in m ental
health services: A m eta-analytic review o f preferences, perceptions, and outcomes. Journal
o f Counseling Psychology, 58, 5 3 7 -5 5 4 . https://www.doi.org/10.1037/a0025266
*Calsyn, R. J., W inter, J. P , & M orse, G. A. (2000). Do consum ers who have a choice o f treat­
m ent have better outcomes? Community Mental Health Journal, 36, 1 4 9 -1 6 0 . https://www.
doi.org/10.1023/A: 1001890210218
Charles, C. A., W helan, T., Gafni, A., W illan, A., & Farrell, S. (2003). Shared treatm ent deci­
sion m aking: W hat does it m ean to physicians? Journal o f Clinical Oncology, 21, 9 3 2 -9 3 6 .
https://www.doi.org/10.1200/JC0.2003.05.057
*Chilvers, C., Dewey, M ., Fielding, K., G retton, V., Millwer, P, Palmer, B., . . . Harrison, G.
(2001). Antidepressant drugs and generic counseling for treatm ent o f m ajor depression in
prim ary care: Random ized trial with patient preference arms. British Medical Journal, 322,
1 -5 . https://www.doi.org/10.1136/bmj.322.7289.772
*C ooper, J. (1980). Reducing fears and increasing assertiveness: The role o f dissonance reduc­
tion. Journal of Experimental Social Psychology, 16, 1 9 9 -2 1 3 . https://www.doi.org/10.1016/
0 0 2 2 -1 0 3 1 (8 0 )9 0 0 6 4 -5
Cooper, M ., & Dryden, W. (2016). The handbook o f pluralistic counselling and psychotherapy.
London, England: SAGE.
Cooper, M ., & M cLeod, J. (2011). Pluralistic counselling and psychotherapy. London
England: SAGE.
*C ooper, M ., Messow, C., M cConnachie, A., Freire, E., Elliott, R., Heard, D., . . . M orrison,
J. (2018). Patient preference as a predictor o f outcomes in a pilot trial o f person-centered
counselling versus low -intensity cognitive behavioural therapy for persistent sub-threshold
and m ild depression. Counselling Psychology Quarterly, 31(4), 4 6 0 -4 7 6 . https://www.doi.
org/10.1080/ 09515070.2017.1329708
 183 Preferences

Cooper, M ., & Norcross, J. C. (2016). A brief, m ultidim ensional measure o f clients’ therapy
preferences: The Cooper-N orcross Inventory o f Preferences (C -N IP ). International
Journal o f Clinical and Health Psychology, 16(1), 8 7 -9 8 . https://www.doi.org/ 10.1016/
j.ijch p .2015.08.003
Making shared decision-making a reality: No decision about
Coulter, A., & Collins, A. (2011).
me, without me. London, England: The King’s Fund.
DeGeorge, J., Constantino, M . J., Greenberg, R. P., Swift, J. K., & Sm ith-H ansen, L. (2013).
Sex differences in college students’ preferences for an ideal psychotherapist. Professional
Psychology: Research and Practice, 44(1), 2 9 -3 6 . https://www.doi.org/10.1037/a0029299
*D evine, D. A., & Fernald, P. S. (1973). O utcom e effects o f receiving a preferred, randomly
assigned, or nonpreferred therapy. Journal o f Consulting and Clinical Psychology, 41, 104­
107. https://www.doi.org/10.1037/h0035617
*D ow rick, C., Flach, C., Leese, M ., Chatwin, J., M orriss, R., Peveler, R., . . . TH R EA D Study
Group. (2011). Estim ating probability o f sustained recovery from m ild to m oderate de­
pression in prim ary care: Evidence from the T H R EA D study. Psychological Medicine, 41,
1 4 1 -1 5 0 . https://www.doi.org/10.1017/S0033291710000437
The heroic client: A revolutionary way to
Duncan, B. L., Miller, S. D., & Sparks, J. A. (2004).
improve effectiveness through client-directed, outcome-informed therapy. San Francisco,
CA: Jossey-Bass.
*Dunlop, B. D., Kelley, M. E., Aponte-Rivera, V., Mletzko-Crowe, T., Kinkead, B., Ritchie, J. C., . . .
Mayberg, H. S. (2017). Effects ofpatient preferences on outcomes in the Predictors ofRemission
in Depression to Individual and Combined Treatments (PReD IC T) study. American Journal
o f Psychiatry, 174, 546-556. https://www.doi.org/10.1176/appi.ajp.2016.16050517
*D unlop, B. W , Kelley, M . E., Mletzko, T. C., Velasquez, C. M ., Craighead, W E., & Mayberg,
H. S. (2012). Depression beliefs, treatm ent preference, and outcom es in a randomized trial
for m ajor depressive disorder. Journal of Psychiatric Research, 46, 3 7 5 -3 8 1 . https://www.
doi.org/10.1016/j.jpsychires.2011.11.003
*D yck, V R., & Spinhoven, P. (1997). Does preference for type o f treatm ent matter?
A study o f exposure in vivo with or without hypnosis in the treatm ent o f panic disorder
with agoraphobia. Behavior Modification, 21, 1 7 2-186. https://www.doi.org/10.1177/
01454455970212003
Elkin, I., Shea, M . T., Watkins, J. T., Imber, S. D., Sotsky, S. M ., Collins, J. F., . . . Parloff, M.
B. (1989). National Institute o f M ental Health Treatm ent o f Depression Collaborative
Research Program: G eneral effectiveness o f treatm ents. Archives of General Psychiatry, 46,
9 7 1 -9 8 2 . https://www.doi.org/10.1001/archpsyc.1989.01810110013002
*E lkin , I., Yamaguchi, J. L., Arnkoff, D. B., Glass, C. R., Sotsky, S. M ., & Krupnick, J. L. (1999).
“Patient-treatm ent fit” and early engagement in therapy. Psychotherapy Research, 9, 4 3 7 ­
451. https://www.doi.org/10.1093/ptr/9.4.437
Elliott, R. U. (1986). Interpersonal process recall (IPR ) as a psychotherapy process research
m ethod. In L. S. G reenberg & W M. P insof (Eds.), The psychotherapeutic process: A re­
search handbook (pp. 5 0 3 -5 2 7 ). New York, NY: Guilford.
Elliott, R. (2000). The session effectiveness scale. Unpublished questionnaire. University of
Toledo. Toledo, OH.
*Ersner-H ershfield, S., Abramowitz, S. I., & Baren, J. (1979). Incentive effects o f choosing
a therapist. Journal of Clinical Psychology, 35, 4 0 4 -4 0 6 . https://www.doi.org/10.1002/
1 0 9 7 -4 6 7 9 (1 9 7 904)35:2< 404::A ID -JC L P 2270350235> 3.0.C 0;2-0
 18 4 p sy c h o th er a py rela tio n sh ips that w ork

*Fuller, T. C. (1988). The role o f patient preference for treatm ent type in the m odification of
weight loss behavior. Dissertation Abstracts International, 49, 2932.
Glass, C. R., Arnkoff, D. B., & Shapiro, S. J. (2001). Expectations and preferences.
Psychotherapy: Theory, Research, Practice, Training, 48, 4 5 5 -4 6 1 . https://www.doi.org/
10.1037/0033-3204.38.4.455
González, H. M ., Vega, W A., W illiam s, D. R., Tarraf, W , West, B. T., & Neighbors, H. W
(2010). Depression care in the U nited States: Too little for too few. Archives o f General
Psychiatry, 67, 3 7 -4 6 . https://www.doi.org/10.1001/archgenpsychiatry.2009.168
*G ossop, M ., Johns, A., & G reen, L. (1986). Opiate withdrawal: Inpatient versus outpatient
program m es and preferred versus random assignm ent to treatm ent. British Medical
Journal, 293, 1 0 3 -1 0 4 . https://www.doi.org/10.! 136/bmj.293.6539.103
*G um , A. M ., Arean, P. A., Hunkeler, E., Tang, L., Katon, M ., H itchcock, P., . . . Unutzer, J.
(2006). Depression treatm ent preferences in older prim ary care patients. The Gerontologist,
46 , 1 4 -2 2 .
*H am ann, J., Cohen, R., Leucht, S., Busch, R., & Kissling, W (2007). Shared decision m aking
and long-term outcom e in schizophrenia treatm ent. The Journal o f Clinical Psychiatry, 68,
9 9 2 -9 9 7 . https://www.doi.org/10.4088/JCP.v68n0703
*Handelzalts, J. E., & Keinan, G. (2010). The effect o f choice between test anxiety treatm ent
options on treatm ent outcomes. Psychotherapy Research, 20, 1 0 0 -1 1 2 . https://www.doi.
org/10.1080/1050330903121106
Hatchett, G. T. (2015). Development ofthe Preferences for College Counseling Inventory. Journal
o f College Counseling, 18, 3 7 -4 8 . https://www.doi.org/ 10.1002/j.2161-1882.2015.00067.x
Health Foundation. (2014). Person-centred care: From ideas to action. London, England: Author.
*Hegerl, U., Hautzinger, M ., Mergl, R., K ohnen, R., Schutze, M ., Scheunem ann, W , . . . Henkel,
V (2010). Effects o f pharm acotherapy and psychotherapy in depressed prim ary-care
patients: A randomized, controlled trial including a patients’ choice arm . International
Journal of Neuropsychopharmacology, 13, 3 1 -4 4 . https://www.doi.org/10.1017/
S 1461145709000224
*Iacoviello, B. M ., M cCarthy, K. S., Barrett, M . S., Rynn, M ., Gallop, R., & Barber, J. P (2007).
Treatm ent preferences affect the therapeutic alliance: Im plications for randomized
controlled trials. Journal o f Consulting and Clinical Psychology, 75, 1 9 4 -1 9 8 . https://www.
doi.org/10.1037/ 0022-006X .75.1.194
*Kadish, D. A. (1999). Psychological m indedness and psychotherapy orientation preference
as predictors o f treatm ent outcom e for social phobia. Dissertation Abstracts International,
60, 832.
Kagan, N. (1973). Can technology help us toward reliability in influencing hum an interaction?
Educational Technology, 13, 4 4 -5 1 .
*Kay-Lam bkin, F. J., Baker, A. L., Kelly, B. J., & Lewin, T. J. (2012). It’s worth a try: The treatm ent
experiences o f rural and urban participants in a randomized controlled trial o f com puter­
ized psychological treatm ent for com orbid depression and alcohol/other drug use. Journal
o f Dual Diagnosis, 8, 2 6 2 -2 7 6 . https://www.doi.org/10.1080/15504263.2012.723315
*K erns, R. D., Burns, J. W , Shulm an, M ., Jensen, M . P , Nielson, W , Czlapinski, R., . . .
Rosenberger, P (2014). C an we improve cognitive-behavioral therapy for chronic back
pain treatm ent engagem ent and adherence? A controlled trial o f tailored versus standard
therapy. Health Psychology, 33, 9 3 8 -9 4 7 . https://www.doi.org/10.1037/a0034406
*Kludt, C. J., & Perlmuter, L. (1999). Effects o f control and m otivation on treatm ent outcome.
Journal o f Psychoactive Drugs, 31, 4 0 5 -4 1 4 .
 185 Preferences

*K ocsis, J. H., Leon, A. C., Markowitz, J. C., Manber, R., Arnow, B., Klein, D. N., & Thase,
M . E. (2009). Patient preference as a m oderator o f outcom e for chronic form s o f m ajor
depressive disorder treated with Nefazodone, Cognitive Behavioral Analysis System of
Psychotherapy, or their com bination. Journal o f Clinical Psychiatry, 70, 3 5 4 -3 6 1 . https://
www.doi.org/10.4088/JCP.08m04371
*Kwan, B. M ., Dim idjian, S., & Rizvi, S. L. (2010). Treatm ent preference, engagement, and
clinical improvem ent in pharm acotherapy versus psychotherapy for depression. Behaviour
Research and Therapy, 48, 7 9 9 -8 0 4 . https://www.doi.org/10.1016/j.brat.2010.04.003
*Le, Q. A., Doctor, J. N., Zoellner, L. A., & Feeny, N. C. (2014). Cost-effectiveness o f prolonged
exposure therapy versus pharm acotherapy and treatm ent choice in posttraum atic stress
disorder (the Optim izing PTSD Treatm ent Trial): A doubly randomized preference trial.
Journal of Clinical Psychiatry, 73, 2 2 2 -2 3 0 . https://www.doi.org/10.4088/JCPT3m06719
*Levy Berg, A. L., Sandahl, C., & Clinton, D. (2008). The relationship o f treatm ent
preferences and experiences to outcom e in generalized anxiety disorder. Psychology and
Psychotherapy: Theory, Research, and Practice, 81, 2 4 7 -2 5 9 . https://www.doi.org/10.1348/
147608308X 297113
*Leykin, Y., DeRubeis, J., Gallop, R., A msterdam , J. D., Shelton, R. C., & Hollon, S. D. (2007).
The relation o f patients’ treatm ent preferences to outcom e in a randomized clinical trial.
Behavior Therapy, 38, 2 0 9 -2 1 7 . https://www.doi.org/10.1016/j.beth.2006.08.002
*L in , P., Cam pbell, D. G ., Chaney, E. F., Liu, C., Heagerty, P., Felker, B. L., & H edrick,
S. C. (2 0 0 5 ). The influence o f patient preference on depression treatm ent in p ri­
m ary care. Annals o f Behavioral Medicine, 30, 1 6 4 -1 7 3 . https://www.doi.org/10.1207/
s 1532479 6 a b m 3002_9
Lindhiem , O., Bennett, C. B., Trentacosta, C. J., & McLear, C. (2014). Client preferences af­
fect treatm ent satisfaction, com pletion, and clinical outcome: A meta-analysis. Clinical
Psychology Review, 34, 5 0 6 -5 1 7 . https://www.doi.org/10.1016/jxpr.2014.06.002
*M acias, C., Barreira, P., Hargreaves, W., Bickm an, L., Fisher, W., & A ronson, E. (2005).
Im pact o f referral source and study applicants’ preference for randomly assigned service
on research enrollm ent, service engagement, and evaluative outcomes. American Journal
o f Psychiatry, 162, 7 8 1 -7 8 7 . https://www.doi.org/10.1176/appi.ajp. 162.4.781
*M anthei, R. J., Vitalo, R. L., & Ivey, A. E. (1988). The effect o f client choice o f therapist on
therapy outcome. Community Mental Health Journal, 18, 2 2 0 -2 2 9 . https://www.doi.org/
10.1007/BF00754338
*M cKay, J. R., Alterman, A. I., M cLellan, A. T., Boardm an, C. R., Mulvaney, F. D., & O ’Brien,
C. P. (1998). Random versus nonrandom assignm ent in the evaluation o f treatm ent for c o ­
caine abusers. Journal o f Consulting and Clinical Psychology, 66, 6 9 7 -7 0 1 . https://www.doi.
org/10.1037/0022-006X .66.4.697
*M cKay, J. R., Alterman, A. I., M cLellan, A. T., Snider, E. C., & O ’Brien, C. P. (1995). Effect
o f random versus nonrandom assignment in a com parison o f inpatient and day hospital
rehabilitation for male alcoholics. Journal o f Consulting and Clinical Psychology, 63, 7 0 -7 8 .
https://www.doi.org/10.1037/0022-006X.63.1.70
M cLeod, J. (2012). W hat do clients want from therapy? A practice-friendly review o f research
into client preferences. European Journal o f Psychotherapy and Counselling, 14, 19-3 2 .
https://www.doi.org/10.1080/13642537.2012.652390
*M ergl, R., Henkel, V , Allgaier, A., Kramer, D., Hautzinger, M ., K ohnen, R., . . . Hegerl, U.
(2011). Are treatm ent preferences relevant in response to serotonergic antidepressants
and cognitive-behavioral therapy in depressed prim ary care patients? Results from
 186 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

a randomized controlled trial including a patients’ choice arm . Psychotherapy and

Psychosomatics, 80, 3 9 -4 7 . d oi:10.1159/000318772
”Proctor, E. K., & Rosen, A. (1981). Expectations and preferences for counselor race and their
relation to interm ediate treatm ent outcomes. Journal o f Counseling Psychology, 28, 4 0 -4 6 .
https://www.doi.org/10.1037/0022-0167.28.1.40
*Raue, P. J., Schulberg, H. C., Heo, M ., Klim stra, S., & Bruce, M . L. (2009). Patients’ depression
treatm ent preferences and initiation, adherence, and outcom e: A randomized prim ary care
study. Psychiatric Services, 60, 3 3 7 -3 4 3 . https://www.doi.org/10.! 176/appi.ps.60.3.337
*R enjilian, D. A., Nezu, A. M ., Sherm er, R. L., Perri, M. G ., McKelvey, W. G., & Anton, S. D.
(2001). Individual versus group therapy for obesity: Effects o f m atching participants to
their treatm ent preferences. Journal o f Consulting and Clinical Psychology, 69, 7 1 7 -7 2 1 .
https://www.doi.org/10.1037/ 0022-006X .69.4.717
Rennie, D. L., Phillips, J. R., & Quartaro, G. K. (1988). Grounded theory: A prom ising ap­
proach to conceptualization in psychology? Canadian Psychology, 29, 1 3 9 -1 5 0 . https://
www.doi.org/10.1037/h0079765
”Rokke, P. D., Tomhave, J. A., & Jocic, Z. (1999). The role o f client choice and target selection
in self-m anagem ent therapy for depression in older adults. Psychology and Aging, 14, 155­
169. https://www.doi.org/10.1037/0882-7974.14.1.155
Rosen, A. (1967). Client preferences: An overview o f the literature.The Personnel and Guidance
Journal, 45, 7 8 5 -7 8 9 . https://www.doi.org/10.1002/j.2164-4918.1967.tb04797.x
Sandell, R., Clinton, D., Frovenholt, J., & Bragesjo, M. (2011). Credibility clusters,
preferences, and helpfulness beliefs for specific form s o f psychotherapy. Psychology and
Psychotherapy: Theory, Research and Practice, 84, 4 2 5 -4 4 1 . https://www.doi.org/ 10.1111/
j.2 0 4 4 -8 3 4 1 .2 0 1 0 .02010.x
”Sidani, S., Epstein, D. R., Bootzin, R. R., M iranda, J., & Cousins, J. (2015). The contribution
o f treatm ent allocation m ethod to outcom es in intervention research. Canadian Journal of
Nursing Research, 47, 6 2 -8 0 .
Sm ith, T. B., Rodriguez, M. M. D., & Bernal, G. (2011). Culture. Journal o f Clinical Psychology,
67, 1 6 6 -1 7 5 . https://www.doi.org/10.1002/jclp.20757
Spitzer, R. L., Kroenke, K., & W illiam s, J. B. (1999). Validation and utility o f aself-report ver­
sion o f P R IM E-M D : The PH Q prim ary care study. JAMA, 282, 1 7 37-1744.
”"Sterling, R. C., Gottheil, E., Glassm an, S. D., W einstein, S. P., & Serota, R. D. (1997).
Patient treatm ent choice and compliance: Data from a substance abuse treatm ent pro­
gram. The American Journal on Addictions, 6, 1 6 8 -1 7 6 . https://www.doi.org/10.3109/
10550499709137028
Stewart, T. J., Swift, J. K., Freitas-M urrell, B. N., & W hipple, J. L. (2013). Preferences for
m ental health treatm ent options am ong Alaska Native college students. American Indian
and Alaska Native Mental Health Research, 20(3), 5 9 -7 8 . https://www.doi.org/10.5820/
aian.2003.2013.59
Sue, S., & Zane, N. (2009). The role o f culture and cultural techniques in psychotherapy: A
critique and reformulation. Asian American Journal o f Psychology, 1(Suppl.), 3 -1 4 . https://
www.doi.org/10.1037/1948-1985.S. 1.3
Swift, J. K., & Callahan, J. L. (2009). The im pact o f client treatm ent preferences on outcom e: A
m eta-analysis. Journal o f Clinical Psychology, 65, 3 6 8 -3 8 1 . https://www.doi.org/10.1002/
jclp.20553
 187 Preferences

Swift, J. K., & Callahan, J. L. (2010). A com parison o f client preferences for intervention em ­
pirical support versus com m on therapy variables. Journal o f Clinical Psychology, 66, 1217—
1231. https://www.doi.org/10.1002/jclp.20720
Swift, J. K., Callahan, J. L., Ivanovic, M ., & Kom iniak, N. (2013). Further exam ination o f the
psychotherapy preference effect: A m eta- regression analysis. Journal o f Psychotherapy
Integration, 23, 1 3 4 -1 4 5 . https://www.doi.org/10.1037/a0031423
Swift, J. K., Callahan, J. L., Tompkins, K. A., Connor, D. R., & Dunn, R. (2015). A delay­
discounting measure o f preference for racial/ethnic m atching in psychotherapy.
Psychotherapy, 52, 3 1 5 -3 2 0 . https://www.doi.org/10.1037/pst0000019
Swift, J. K., Callahan, J. L., & Vollmer, B. M. (2011). Preferences. Journal o f Clinical Psychology,
67, 1 5 5 -1 6 5 . https://www.doi.org/10.1002/jclp.20759
Tom pkins, K. A., Swift, J. K., Rousm aniere, T. G., & W hipple, J. L. (2017). The relationship
between clients’ depression etiological beliefs and psychotherapy orientation preferences,
expectations, and credibility beliefs. Psychotherapy, 54, 2 0 1 -2 0 6 . https://www.doi.org/
1 0 .1037/pst0000070
*Van, H. L., Dekker, J., Koelen, J., Kool, S., Aalst, G. V , Hendriksen, M ., . . . Schoevers, R.
(2009). Patient preference com pared with random allocation in short-term psychodynamic
supportive psychotherapy with indicated addition o f pharm acotherapy for depression.
Psychotherapy Research, 19, 2 0 5 -2 1 2 . https://www.doi.org/10.1080/10503300802702097
Vollmer, B., Grote, J., Lange, R., & Walker, C. (2009). A therapy preferences inter­
view: Empowering clients by offering choices. Psychotherapy Bulletin, 44, 3 3 -3 7 .
*W allach, H. S. (1988). Clients’ expectations and results o f psychological therapy for
dysmenorrheal. Dissertation Abstracts International, 49, 1961.
*W ard, E., King, M ., Lloyd, M ., Bower, P., Sibbald, B., Farrelly, S., . . . Addington-Hall, J. (2000).
Random ized controlled trial o f non-directive counseling, cognitive-behavior therapy, and
usual general practitioner care for patients with depression I: Clinical effectiveness. British
Medical Journal, 321, 13 8 3 -1 3 8 8 . https://www.doi.org/10.1136/bmj.321.7273.1383
*W heaton, M. G., Carpenter, J. K., Kalanthroff, E., Foa, E. B., & Sim pson, H. B. (2016).
Augmenting SRIs for obsessive-compulsive disorder: Patient preference for Risperidone
does not lim it effectiveness o f exposure and ritual prevention. Psychotherapy and
Psychosomatics, 85, 3 1 4 -3 1 6 . https://www.doi.org/10.1159/000445356
*W illiam s, R., Farquharson, L., Palmer, L., Bassett, P., Clarke, J., Clark, D. M ., & Crawford, M.
J. (2016). Patient preference in psychological treatm ent and associations with self-reported
outcom e: National cross-sectional survey in England and Wales. BMC Psychiatry, 16(4),
1 -8 . https://www.doi.org/10.1186/s12888-015-0702-8
*Z lotnick, C., Elkin, I., & Shea, M. T. (1998). Does the gender o f a patient or the gender o f a
therapist affect the treatm ent o f patients with m ajor depression?Journal of Consulting and
Clinical Psychology, 66, 6 5 5 -6 5 9 . https://www.doi.org/10.1037/0022-006X.66.4.655
 7

REACTANCE LEVEL

Christopher J. Edwards, Larry E. Beutler, and Kathleen Someah

Patients who willingly enter psychotherapy typically do so because of a desire to

change. At the same time, that desired change can prove a difficult and frightening
process and is often met with what clinicians refer to as resistance. The clinicians is
then left with the thorny question: “Why is it that one who wants to change then resists
doing so when offered the opportunity?” The role of the clinician is to serve as a fa­
cilitator for reconciling these contradictory inclinations of the patient and to produce
change and growth in the process. Psychological change in psychotherapy has long
been considered to result from the persuasiveness and compatibility of the therapist
when addressing an ambivalent patient (Strong & Matross, 1973). Many of these per­
suasive forces are mightily tested when the therapist addresses patient resistance.
Resistance in the context of psychotherapy implies a fundamental apprehension
and aversion to change (Firestone, 2015). To a clinician, resistance is an attribute of
the person and indicates, as the term suggests, a pulling back and digging in to prevent
change from occurring. It is usually met with an interpretation or other confrontation.
Within the frame of social psychology, it is useful to differentiate between resistance
to change and “reactance” to change. Reactance is an extreme form of resistance and
indicates not only the lack of an inclination to change but an oppositional reaction to
the persuader (Brehm & Brehm, 1981). Moreover, its presence reflects an interpersonal
process that is best met by a reduction in persuasive demand and an analysis of the lis-
tener/patient’s immediate fears and anticipated consequences of the behavior. Brehm
and Brehm convincingly argue that the central motivational theme that occurs when
one is reactant is the fear of losing one’s independence.
In this chapter, we accept the distinction between resistance and reactance and view
the two terms as a reflection of two points along a continuum of avoidance. Reactance,
even more than the broader term resistance, is our point of focus since it appears
strongly to be both activated and deactivated by a psychotherapist. Reactance decreases
when the therapist is able to avoid challenging or threatening the recipient/patient’s
fear of losing some aspect of personal freedom. And, conversely, reactance may be acti­
vated if the therapist is too confrontive or too uninvolved. The effective therapist, from
this perspective, understands that any patient may directly reassert his or her freedom
through oppositional behavior within the therapy room or via premature termination.

188
 189 Reactance Level

Likewise, the patient who declines to engage in therapeutic tasks or homework, or

ignores the therapist’s reflection by interrupting him or her, is engaged in the common
task of avoiding the loss of freedom.
Borrowing from social psychology scholars, reactance and, to a lesser extent, resist­
ance, is responsive to the moderating effects of therapist directiveness. That is, its destruc­
tive impact on psychotherapy outcome can be modified by how much confrontation and
direction the therapist chooses to provide (Beutler et al., 2001). Without the moderating
influence of therapist nondirectiveness and nonconfrontation, the patient’s resistance and
reactance is thought to correlate negatively to treatment outcome (Hara et al., 2015).
This chapter begins by providing definitions, measures, and clinical case examples
of reactance levels in psychotherapy. Subsequently, previous and current meta-analytic
reviews of resistance are presented. We then review evidence of causation via directive
and nonconfrontational interventions and inspect the potential limitations of the re­
search evidence. We conclude with diversity considerations, training implications, and
therapeutic practices based on this research evidence.

DEFINITIONS

Resistance/Reactance Level
Resistance is the tendency of an individual patient to avoid making the changes
advocated by the therapist. As we have noted, reactance is an extreme example wherein
the patient not only resists but changes in a direction away from that advocated by the
therapist. In a practical sense, a therapist intuitively knows that a resistant patient may
not carry out homework or may behave in ways that ensure the maintenance of the
symptoms. In contrast, a reactant patient may do the homework but even the simplest
assignments will be wrong. Or the reactant patient may suddenly have more symptoms
and be highly distressed about things the therapist says or recommends.
While the foregoing paragraph defines resistance and reactance, those definitions
are incomplete without considering the nature of the relationship that exists between
the therapist and the patient. Resistance in psychotherapy may be a consequence of
negative interpersonal conflicts between the therapist and the patient (Mitchell, 2013).
The clinician plays an integral role in the formation of resistance through the use of
demanding an authority-based directive. In contrast, the therapist may relieve reac­
tance by the noting and changing communication patterns that subsequently initiate
reactivity on the part of the patient, thereby obstructing treatment. Strong and Matross
(1973) emphasize that resistance emerges through the therapist’s request for change,
not the behavior change itself. This is an important relation for a therapist to remember.
Thus therapist beliefs influence the way that different interventions are employed
and thereby determine the degree to which the patient will accept these interventions.
For instance, cognitive-behavioral therapists describe resistance as behavioral non­
compliance and correspondingly seek to apply a behavioral correction to the meanings
held by the patient and applied within the circumstances that co-occur with its expres­
sion. In contrast, psychoanalytic clinicians explain resistance largely as an expression of
 190 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

internal conflicts and transference, which are to be expunged. Oppositional reactions

occur as a function of the severity of the early learning experiences and the inten­
sity of the resulting conflict (Messer, 2002). Early experiences with authorities are
thought to instill a distrust of any authority and to provoke resistance as an automatic
response to psychotherapy by virtue of this automatic (or unconscious) distrust. The
recommended response is to interpret the resistance and encourage the patient to rec­
ognize his or her unconscious contributions to therapy’s failure.
It is also important to distinguish between state- and trait-like resistance. In most
instances, reactance is understood to be a state-like quality. It is a response to a partic­
ular intervention by the therapist. Resistance, on the other hand, is a more enduring and
trait-like set of behaviors. The measurement of resistant traits rely on the measurement of
repetitive and varied protective behaviors across situations. One who is highly resistant
will be strongly inclined to be reactant. Therefore, resistance has the characteristics of a
continuing response to both the feeling of threat and the activation of reactance. Both
state and trait actions reflect a direct expression of the desire to avoid being controlled.
Whether state or trait, the psychotherapy objectives when dealing with reactance
are the same—to make an environment in which resistance and reactance are not nec­
essary. While all psychotherapy theories have articulated a variety of techniques with
which to deal with the “problem” of resistance, they seldom consider the possibility
offered by Brehm and Brehm (1981) that resistance in psychotherapy may mark the
failure of the therapist to present an environment that does not challenge the particular
patient’s fear of losing freedom.

Therapist Directiveness
Directiveness is defined as the degree to which the therapist uses suggestions,
interpretations, and assignments to guide the patient’s movement through psycho­
therapy. Directiveness may apply both to the use of techniques within the therapy ses­
sion and to the use of homework outside of the session. We have already described the
role of clinician’s directiveness in making a request or setting topics and leads for the
session. We have noted that directiveness may be informed either by direct observations
or by the therapist’s theoretical orientation and that, frequently, cognitive and behav­
ioral therapies are infused with a large number of directive interventions. Homework
assignments are particularly widely used in these therapies and are inherently directive
in nature. Within sessions, instructions and questions are directive interventions that
are used to keep the patient focused and to teach concepts. These techniques contrast
with the use of reflection, self-disclosure, and even many interpretations in most dy­
namic and experiential interventions.

MEASURES

Patient resistance can either be caused by situational (state) or temperamental (trait)

variables. Likewise, their measurement can be either direct or indirect. In the same
way, therapist directiveness can either be provoked by the situation or hinged on the
 19 1 Reactance Level

therapist’s preferred model of treatment. Measurement, accordingly, can be direct or in­

direct. Although direct measures both of patient resistance and therapist directiveness
are obviously superior, indirect measures are usually applied to one or both of these
dimensions because direct measures are not available within a particular study.

Level of Measurement
Linking therapist’s behavior (level of directiveness) to the reduction of patient resist­
ance is the nucleus of emerging research that fits the intervention to the patient’s re­
active expression. The motivation that produces resistance from this perspective, and
the key to its treatment, has been described as “a state of mind aroused by threat to
one’s perceived legitimate freedom, motivating the individual to restore the thwarted
freedom (Brehm & Brehm, 1981, p. 4).
Few reliable observational tests of patient resistance or of therapist directiveness
exist. These weaknesses have not changed since 2011. Moreover, many (or most) rele­
vant studies of the fit of patient and therapist were not intentionally aimed at this phe­
nomenon and did not use direct measures of how resistant the patients were, making
the use of indirect measures necessary. In the case of patients, these indirect measures
have usually taken the form of diagnoses under the assumption that people with cer­
tain disorders (e.g., paranoid personality, antisocial personality, etc.) are imbued with
a proclivity to be resistant.
In kind, indirect measures of therapist directiveness usually derive from the par­
ticular therapy model that is applied by the therapist under the assumption that
treatments differ in part by virtue of the amount of directiveness embodied in the
theory used (e.g., motivational enhancement therapy vs. behavior therapy). But these
concessions require that patient and therapist level measures are sacrificed and both
reactance and directiveness must be inferred from patient diagnoses or the treatment
model used.
While the use of such indirect measures as diagnosis and treatment type are reli­
able (Beutler et al, 2011), they frequently produce results that understate the degree
to which treatment “fit” induces therapeutic change. While it follows from an under­
standing of the basis of diagnosis and therapeutic intervention that an intervention
like cognitive-behavior therapy (CBT) is more directive than an intervention like
interpersonal therapy or motivational enhancement therapy, such measures mistak­
enly assign the same rating to all therapists or patients within a given treatment or
diagnostic group.
W hile it is logical that individuals with paranoid personality disorder are likely
to be more resistant than groups o f patients with unipolar depression, the lack of
within-group variation in scores reduces the sensitivity of the measure. Thus it is im ­
portant in reviewing the literature to specify when direct observations are being used
versus the articulation of indirect inferences that are based on group-level indicators
(e.g., diagnosis and therapeutic model). W hen identifying the effects of a patient or
therapist quality, one must remember that indirect measures produce understated
results.
 19 2 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

In the following, we separately review direct and indirect measures that are com­
monly used in reactance research. While preference is given to the use of direct meas­
ures of trait-like constructs throughout, we highlight both the direct and the indirect
measures and differentiate between state and trait measures.

Direct Measures of Reactance

Research on reactance has proceeded along two different branches. The best studies
rely on individual-level assessment and are comprised of “direct” measures. That is,
the measurements are made at the level of the person and specifically identify the fre­
quency or intensity of patient reactance. The other measurement method is the use of a
group-wise indicator of reactance and thus often involves the use of indirect measures.
The most used direct measures of patient resistance in studies comprising this re­
view are those that look at compliance with homework or in-session behavior. Less
direct measures, but ones that are nonetheless aimed at individual behavior, are
drawn from a personality test or an interview where a complex score for resistance is
computed. Most of these measures focus on a particular point in time or a particular
setting; they are “state” measures rather than “trait” measures. State measures may tell
us about a specific point in time but do not work if we want to use reactance to plan a
therapeutic approach over a period of time. There, trait measures generally prove more
useful.
Direct measures of patient resistance may usefully be identified from an anchored
rating sheet, one that includes a series of statements that describe a continuum of op­
position. Instructions that ask a patient to rate the current moment or event is a state
measure, while a trait measure usually asks one to rate the “past few” weeks or ses­
sions. This kind of measure (depending in part upon the number of items included)
yields a relatively continuous score that is preferred over single-ltem scales among
psychotherapy researchers because of its relative sensitivity. Such measures also allow
for coding patient resistance from their behaviors, such as missed appointments or re­
peated failures to complete homework assignments (Aviram & Westra, 2011; Karno &
Longabaugh, 2005a, 2005b; Westra et al., 2009).
The measurement of resistance, of course, also can be measured nominally (i.e.,
categorically). Two such categorical measures, the Adapted Client Resistance Code
(Westra et al., 2009) and the Client Resistance Code (CRC; Chamberlain et al., 1984),
have been used in studies of patient resistance. The CRC delineates 11 categories of
resistance behavior. Patients are identified by which category best fits them. Resistance
is identified as one of the 11 varieties, all of which depict behaviors that interfere
with the direction of psychotherapy. Ratings are made by the therapist based upon
verbalizations from the patient. Thus, although these measures are clearly direct, some
of the classifications require an inference.
The CRC considers resistance as a set of behaviors embedded in an interpersonal
process between the patient and clinician (Hara et al., 2015). A revised version of the
CRC measure relies on a similar definition of resistance but involves an altered coding
process to improve its reliability and validity via providing a rating of variability on a
 193 Reactance Level

single, global resistance score. Since the responses from which scores are earned all
address the patient’s current activity within psychotherapy, trait aspects of the measure
are lost. Interpretations of any findings are confined to the singular event of psycho­
therapy with this particular therapist.
The most widely used self-report measures of resistance in psychotherapy gauge
current situations— states rather than traits. The Patient Resistance Inventory (PRI;
Dowd et al., 1991) and Therapeutic Reactance Scale (TRS; Dowd et al., 1991) are re­
lated instruments, both of which can be used to measure resistance over time and ei­
ther within a specific course of treatment or with psychotherapy generally. These two
patient self-report measures are similar in design but differ in the response alternatives
that are available and the characteristics they purport to reveal.
The PRI consists of 28 items completed by the patient. The PRI uses a yes/no format
and produces one reliable score indicating level of reactance in the therapy session or
to psychotherapy more generally. In contrast, each item of the TRS is rated on a 4-point
Likert-type scale, anchored from strongly agree to strongly disagree, in which higher
scores suggest greater levels of resistance (Buboltz et al., 2003). This measure produces
a total score and two subscores. The Total score indicates overall level of receptivity to
psychotherapy. The TRS subscores differentiate between verbal reactance and behav­
ioral reactance. Items that load on the Verbal reactance scale scores are reflective of
such qualities as verbal negativity and argumentativeness. Behavioral reactance reflects
the presence of oppositional behavior. Elevations on this latter scale include questions
such as “I have a strong desire to maintain personal freedom” and “I find that I often
have to question authority.” In both cases, the items address patients’ current actions
and impulses not only within their current psychotherapy but in their daily life. Thus
trait aspects of reactance is obscured and confounded with more situational events in
this scale.
Some measures of trait-like resistance are available in the scales of omnibus per­
sonality tests. Among the 11 studies identified in the 2011 analysis (Beutler et al.,
2011), four utilized scales drawn from one of versions of the Minnesota Multiphasic
Personality Inventory versions: MMPI-1 (Hathaway & McKinley, 1940) and M M PI-2
(Butcher, 1999). Two general scales and three content subscales possess content con­
sistent with the trait of resistance (Butcher et al., 2011). These are Pd (psychopathic de­
viate), Pa (paranoia), TRT (treatment readiness), CYN (cynicism), and ANG (anger).
The general scales are empirically derived and tap global personality characteristics
and are represented in clinical populations whereas content scales are derived from ap­
parent similarity of items. The general scales are the usual scales that are administered
and graphed when the M M PI-2 is used in clinical practice. Combinations of the fore­
going scores were successfully used (e.g., Beutler, Engle, et al., 1991; Beutler et al., 1993;
Karno et al., 2002) in several studies of resistance levels.
STS/Innerlife (Beutler et al., 2008) is a cloud-based psychological assessment (www.
innerlife.com) that consists of 173 items. STS/Innerlife produces an intake narrative
and recommended treatment plan along with graphic and narrative information on six
global symptom measures (depression, anxiety, somatic complaints, chemical abuse,
thought disorder, and risk of self-harm) and up to 16 symptom domain scales that
 194 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

achieve the clinical range. In addition, several continuous scales yield patient trait
qualities, including reactance level. Innerlife measures patient resistance as an en­
during, cross-situational trait, the only instrument listed in this chapter that explicitly
does so. STS/Innerlife scales earn high to moderate reliabilities (mean of alphas = .85;
Kimpara, Regner, et al., 2015) and requires from 15 to 35 minutes to complete on an
iPad, iPhone, or computer.

Indirect Measures of Reactance

When direct measures of an individual patient’s reactance level are unavailable, then
an indirect measure of group membership can be used. Such a measure may identify
groups by their shared diagnosis and the associated inference that a particular level
of resistance characterizes all the patients in that diagnostic group. Since the method
relies on similarity among all group members, individual variations are ignored.
The most frequent research design keeps the diagnosis constant across patients and
then compares two or more treatments using a randomized controlled trial (RCT). In
this case, the focus of the investigator is on the effects of a particular brand of psycho­
therapy; the level of fit and the role of reactance are afterthoughts. No individual-level
measurement of personal traits like resistance is likely to be included and, frequently,
neither are individual-level measures of directiveness. The patient’s diagnosis is used to
infer patient reactance and one’s therapeutic school is used to infer level of directiveness
(e.g., Clarkin et al., 2007; Gregory et al., 2008).
For example, suppose investigators conduct an RCT in which two brands of
psychotherapy—cognitive therapy (CT) and motivational enhancement (M E)— are
applied to clients with a diagnosis of alcohol dependence. The investigators find that
the two treatments do not produce distinguishing outcomes. If they want to further
investigate the hypothesis regarding a poor fit between patient reactance levels and
therapist directiveness, they are forced to do so retrospectively and most often use an
indirect measure of both the major variables. They might take therapy type as a proxy
variable for directiveness— surely this would be justifiable via the distinctive theories
of these two approaches. They might also take the diagnosis itself—alcoholics are as
a group reactant to treatment. They now have two levels of directiveness (high, low)
with a highly reactant group. The investigators might propose that with this highly
reactant population, ME would be a better intervention than CT. And they probably
would be right, but they have lost much measurement sensitivity both by using indi­
rect or proxy measures rather than direct ones and by the omission of a comparison
patient type.

Measures of Therapist Directiveness

Measurement of therapist directiveness, as noted with respect to patient resistance, is
ideally direct rather than inferred, and individual assessments are preferred over those
that are group based. Such measures are more powerful than those that infer level of
directiveness from a treatment brand.
 195 Reactance Level

Indirect measures of therapist directiveness, based on the type of psychotherapy,

are relatively easy to apply and as noted previously are frequently the only ones that
are available to investigators. Certain treatment orientations are considered more di­
rective than others. For instance, prior research has used CBT as a measure of high
directiveness and motivational interviewing (MI) as low directiveness (Karno &
Longabaugh, 2005a; 2005b). Traditionally, CBT entails that the therapist adopts an
active, structured stance. For instance, a CBT therapist may guide a patient through
interventions such as cognitive restructuring and incorporate the use of homework
to be completed between sessions. In contrast, MI involves rolling with the patient’s
resistance and working with the patient where he or she is in terms of motivation for
change.
One of the few direct measures of therapist directiveness is the Therapy Process
Rating Scale (Fisher et al., 1995), which is rated by trained external observers. The
revision of the scale permits the therapist or supervisor to make ratings as well as an
external rater. The measure yields five scores: (a) therapist directiveness, (b) facilitating
change through insight/awareness, (c) focus on symptom change, (d) therapist skillful­
ness, and (e) facilitating change through the induction of arousal.

CLINICAL EXAMPLES
“Kate,” a 28-year-old, Caucasian, heterosexual female presented to psychotherapy after
losing custody of her child in a divorce action. Kate described an extensive history
of substance use, as well as a series of traumatic experiences resulting from an abu­
sive marriage and subsequent homelessness. On intake, Kate was administered the
STS/Innerlife (Beutler et al., 2008), which reveled elevations in depression and anx­
iety on the global scales. She also scored in the clinical range on measures of social
anxiety, family-related distress, and posttraumatic stress disorder on more narrowly
defined scales. On treatment planning scales, she was very high on the Resistance/
Reactance scale.
Although Kate was compliant with treatment at the onset of psychotherapy, her
high reactance became increasingly apparent as her treatment progressed. She missed
several appointments with her therapist, often blaming public transportation or last­
minute meetings for her tardiness or absence. When the therapist provided homework
assignments, Kate found ample reasons to not complete them, earning high in-session
scores on reactance.
Her therapist’s behavior was typically followed by an oppositional act on Kate’s part.
When the therapist leaned forward, Kate consistently moved backward in her seat.
As her history unfolded, the persistence and generalizability of this pattern became
apparent and suggested the presence of a vacillating attachment style or process and
probably a lack of readiness for accepting directive and guided change. She also lived
a life that was fraught with efforts to disengage from and avoid any hint of someone’s
directiveness or control over her, portending the presence of poor interpersonal en­
gagement that probably was linked to her high resistance. Although Kate voiced an
interest in engaging in psychotherapy and changing her behaviors, she demonstrated
 19 6 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

otherwise when the therapist took a more directive stance, such as assigning home­
work and requesting that Kate attend community support groups.
In response to Kate’s high reactance, the therapist decreased her level of directiveness.
Rather than encouraging the patient to engage in exposure methods for her trauma,
the therapist made statements that supported and even prescribed her withdrawal,
such as “I don’t want you to expose yourself to any situation or person that could be
reminiscent of traumatic experience.” The practitioner aimed to build a trusting alli­
ance with Kate by acknowledging and occasionally advocating for avoidance as a par­
adoxical intervention.
With regard to Kate’s substance use, the therapist adopted a relatively nondirective,
more collaborative approach to change. Sessions incorporated MI techniques that
sometimes met her resistance with a prescription for the symptom or a directive to
avoid changing until she becomes “ready.” These injunctions were intended to meet
Kate where she was in readiness for behavioral change. Practicing in this manner both
allowed Kate to assume more control of her behavior and allowed her resistance to
sufficiently subside such that psychotherapy eventually proved successful. Had Kate’s
high reactance met with high therapist directiveness, we fear that she would have
terminated treatment prematurely.

LANDMARK STUDIES
Many studies have explored the treatment implications of reactance in psychotherapy.
The three studies reported here are identified as landmark because of the impact of
their findings on subsequent research and clinical practice.

Study 1
The study by Beutler and colleagues (1991) was the first RCT to test the moderating
role played by therapist directiveness among patients who varied in trait-l ike reac­
tance. The investigators randomized depressed patients to three treatments—CT,
experiential/existential therapy (FEP), and a self-directed treatment (S/SD)—which
were designed to vary in the level of therapist directiveness prescribed. All treatments
were standardized through the use of manuals. Therapists were assigned to treatment
only after they had achieved a stable and meaningful level of adherence/compliance
with one of the manuals. The licensed therapists, all of whom were trained to specific
manuals of CT and FEP, attended case conferences in which they were supervised in
the use of a manual of their choice in psychotherapy.
The S/SD treatment was comprised of bibliotherapy and regular telephone support
by a graduate student trainee. S/SD was designed to represent a low level of therapist
direction. The therapist encouraged clients to select books from a list provided and to
both read the books and to identify ways that they were helped by this reading. The
therapist did not provide direction for the patient beyond this point and never insisted
that the patient comply with the treatment.
 197 Reactance Level

Patient trait-like resistance was assessed via a pretest administration of the MMPI
and the use of the TRT subscale, a scale denoting negative attitudes toward treat­
ment and health providers (Butcher et al., 2011). Treatment outcome was measured
using the Beck Depression Inventory, the Brief Symptom Inventory, and the Hamilton
Rating Scale for Depression. The investigators also tracked patient retention (or, nega­
tively stated, dropout) in the three psychotherapies.
The research design allowed investigators to test the comparative effects of high
(C T), medium (FEP), and low directive (S/SD) therapies among patients who varied
widely on reactance. On depression reduction, the results found modest effects (d = .34)
favoring therapist directiveness as a general contributor to change. Results of an anal­
ysis for the number of sessions completed (intended treatment length was 24 ses­
sions) mirrored that based on symptom reduction. Specifically, therapist directiveness
enhanced symptom outcomes to a moderate degree and improved patient retention.
The largest symptom decrease and retention rates were obtained as a function of
an inverse fit between patient resistance and therapist directiveness. Patients did best
when assigned to a therapist who used an intervention that inversely fit the patient’s
pretherapy level of resistance. In other words, patients with a high propensity to re­
sist therapist directives responded best to therapists who seldom became directive or
confrontive. A good response in this case consisted both of symptom reduction and
treatment retention. In contrast, patients with a low propensity to be defensive and
resistant responded best to therapists who used directive and confrontive techniques.
A “good fit,” defined as an inverse relation between therapist directiveness and patient’s
resistance traits, improved treatment outcomes above the average levels of those who
were not well matched to their therapists.

Study 2
A follow-up study (Beutler et al. 1991) confirmed the durability and even the in­
crease of this moderating effect 12 months after the end of treatment. Patients were
recontacted and administered outcome measures again. At follow-up, the size of the
effect attributable to the fit between patient resistance and therapist directiveness had
increased. The mean effect size (ES) for the variable of “fit” had grown to d = 1.4 while
the size of the direct effects attributable to therapist directiveness and patient resist­
ance, alone, had not changed.

Study 3
Karno and Longabaugh (2005a) examined the interactions of clinician directiveness
with patient anger and reactance among 139 alcohol-abusing patients who received
treatment as part of Project MATCH (Matching Alcoholism Treatments to Client
Heterogeneity), a large-scale, national study conducted by the National Institute on
Alcohol Abuse and Alcoholism. Patient trait reactance was assessed via observer
ratings of treatment sessions using the Reactance scale drawn from the Systematic
 19 8 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Treatment Selection-Clinician Rating Form (Fisher et al., 1995; Malik et al., 2003).
Clinician directiveness was measured directly by observers using the Directiveness
subscale of the Therapy Process Rating Scale (Fisher et al., 1995).
Findings from this study revealed an interaction between directiveness and anger
(an aspect of resistance), as well as directiveness and reactance in predicting patients’
alcohol use posttreatment. Specifically, patients with high reactance demonstrated
comparatively and significantly worse drinking outcomes when treated with an inter­
vention that emphasized structure and confrontation when compared to a less con­
frontational, paradoxical intervention.
The major significance of this study was its confirmation of psychotherapy-induced
reactance. The result was an iatrogenic deterioration of some high reactant patients
when treated with a highly directive therapy. It also confirmed that the moderating
effects of treatment directiveness among high-reactance patients was ameliorated
among similar patients receiving low-directive therapy.

RESULTS OF PREVIOUS META-ANALYSES

The landmark studies reviewed here were included in a previous meta-analysis

published in the second edition of this book (Norcross, 2011). This section is devoted
to reviewing the methods and findings of that earlier meta-analysis. To be included in
that 2011 meta-analysis of patient reactance x therapist directiveness interaction, the
authors identified six methodological criteria to rate the quality of the research. Each
study included

1. A breadth of reliably applied psychotherapies and trained therapists to ensure

variance in level of directiveness.
2. A similarly wide range of moderately impaired patients.
3. Individualized direct measures patient resistance or therapy and therapist
directiveness.
4. Random assignment of patients within and among treatments.
5. Reliable monitoring both of treatment variability/consistency on dimensions of
therapist directiveness and of patient resistance.
6. Objective and uniform outcome measurement that includes an analysis of fit
between patient resistance and therapy/therapist directiveness.

The 2011 meta-analysis automatically included studies that met five or six of these
rating criteria. Prospective studies uniformly embodied all six criteria, but post hoc
studies, which almost always were aimed at other objectives, usually did not conform
to criterion 3 and often to criterion 5 as well. These two criteria emphasized the role of
individual, participant assessment of the directiveness of the treatment received and
the resistance of the patient who received it.
The 2011 meta-analysis included 11 studies: 5 studies met all six criteria and an­
other 5 studies met five of the six criteria. One additional study met four of the criteria
and had special strengths to merit its inclusion.
 199 Reactance Level

That meta-analysis focused on the fit between patient resistance and therapy
directiveness. It tested the hypothesis that high patient resistance is best treated by low
directiveness and vice versa. The results supported the contention that nondirective
interventions are more effective among patients expressing high levels of reactance/
resistance. The ES (d ) averaged .82 when weighted for the number of participants in
the study. Thus the authors concluded that the fit between patient resistance and thera­
pist (or therapy-associated) directiveness meaningfully improves treatment outcomes.
At the same time, ESs varied widely among individual studies, from .14 to 1.40. The
authors tentatively ruled out cultural differences as a major contributor to this wide
variation but did implicate measurement factors. Specifically, the mean ES in studies
that used indirect measures of either therapy directiveness or patient resistance earned
a d of .43 while those that used one or more direct measures yielded a mean d of 1.16.
Two ideas about reactance are clear in this research: (a) both reactance and resist­
ance, more generally, result in avoidance, and (b) the avoidance takes place in response
to a controlling social environment. One of these findings is that therapists must look
to themselves, not only to patients, as a source of reactance. And if patients evince low
levels of both resistance (traits) and reactance (states), directive and confrontational
interventions may yield good results. A psychotherapist can, by controlling the level of
directiveness, create an environment of freedom and choice that can temper the client’s
reactant tendencies.

META-ANALYTIC REVIEW

The primary aim of the current meta-analysis was to investigate a causal moderating
influence of patient reactance on psychotherapy outcome. Specifically, we sought to
test the hypothesis that high-reactant patients would benefit more in relatively low-
directive treatments whereas low-reactant patients would benefit more from relatively
more directive treatments.

Literature Search
To build on the research from the previous meta-analysis, we undertook an extensive
literature review. Relevant research studies were included in the meta-analysis if they
satisfied the following criteria: (a) investigated psychotherapy outcome with actual
patients and psychotherapists, (b) used a quantifiable measure of clinically relevant
outcomes, (c) employed an RCT or modified RCT design with a sample size of 10 or
more, (d) provided the numerical data needed to calculate ESs that indicate the mag­
nitude of effect on treatment exerted by the match of therapist/therapy directiveness
and patient resistance, and (e) were published in a scientifically recognized and peer-
refereed English-language journal.
While not identical with the criteria used to select studies in the 2011 meta-analysis,
the criteria overlap and encompass the previously used criteria. In a practical sense,
by insisting on the use of an RCT design, the inclusion criteria for this meta-analysis
and those for the 2011 study are similar. Both searches used either direct (e.g., the
 200 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

use of an individual, reliable measure completed by a clinician or patient) or indirect

(e.g., assignment based on group membership) measurement of the targeted variables.
Moreover, both the earlier and the current analyses allowed selection into the study
even if the investigators used one proxy indicator of patient reactance or treatment
directiveness when studies did not include direct measures.
When reviewing the studies, the first question addressed by the coder was whether
the study investigated real psychotherapy with real patients. Second, the code deter­
mined whether each of the three constructs (outcome, resistance, directiveness) em­
ployed individual measurement. The third question addressed whether any proxy
variables could be identified that would substitute for the absence of an individual-
based assessment of these three variables. If so, the calculation was provided and the
range and reliability of the proxy variable was noted. In most cases, the proxy variables
were categorical in nature (e.g., diagnosis or treatment brand).
Seventeen new studies were identified in a preliminary but thorough literature
search of indexed listings, abstracting services, and primary journals using multiple
search terms (see Figure 7.1). Web-based databases such as PsycArticles, PsycBOOKS,
Psychology and Behavioral Sciences Collection, PsycINFO, MEDLINE, Google Scholar,
and others were used to search. We used a variety of synonymous terms, such as: “re­
sistance” and/or “reactance,” “psychotherapy,” “treatment outcomes,” “directiveness,”
and “difficult patients.” We specifically searched for studies that analyzed the effect of
matching patient resistance and therapy/therapist directiveness
Fourteen of the 17 new studies did not meet the inclusion criteria for the current
meta-analysis. Most often this was because the study lacked a quantitative measure for
level of directivenes (n = 3; 19%) or lacked a quantifiable measure of resistance (n = 4;
44%). One study of the 17 (6%) was excluded (e.g., Aviram & Westra, 2011) because
it was an additional analysis of an earlier published study that was already included
in the meta-analysis (Westra et al., 2009). One study used reactance itself as the de­
pendent measure rather than as a measure of mental health status or symptoms. The
other omitted studies did not permit the assessment of the fit between reactance and
directiveness.
The two studies that survived the screening used various methodologies and meas­
ures and warrant more detailed description here. Westra and colleagues (2009) used a
within-subjects design with two treatments and one group of reactant patients. Patients
were judged to be highly reactant at entry to the study and were placed in MI groups
prior to receiving CBT. Reactance was measured throughout therapy by the Client
Motivation for Psychotherapy Scale, Client Resistance Code, and homework comple­
tion. Resistance was measured over time as a dependent variable alongside a measure
of symptom change. For our analysis, we coded MI as a low-directive therapy and CBT
as a high-directive therapy. The results indicated that reactance decreased as a result
of MI and that lower reactance significantly mediated better treatment outcomes, per
pre-posttreatment scores on the Penn State Worry Questionnaire.
Another study by the same group (Aviram et al., 2016) also illustrates a successful
effort to unravel the moderating effects of treatment fit. These authors explored the hy­
pothesis that therapists who used MI techniques (less directive) during periods of high
 201 Reactance Level

figure 7 .1 PRISMA flow diagram.

* T w o e f f e c t s iz e s w e r e c o m p u t e d f r o m o n e s t u d y b e c a u s e d i f f e r e n t p a t i e n t s a m p le s w e r e u s e d ( K a r n o &

L o n g a b a u g h , 2 0 0 4 ).

conflict (high reactance in our coding) would have better outcomes than alternatively
treatment-matched groups. Outcomes were assessed by worry scores and overall reac­
tance scores, but we focused only on the Penn State Worry Questionnaire scores. Level
of directiveness was indexed in this study via a categorical variable—type of therapy.
However, the authors supplanted the group-level measure with the Motivational
Interviewing Treatment Integrity Scale, an individual measure of directiveness within
each therapist-patient pair. Reactance was measured by the CRC and “three visual
analog scales” (Aviram et al., 2016, p. 785). The results of this study provided evidence
for the hypothesis that more reactant patients benefitted more from less structured
treatments. Among patients whose reactance levels increased when confronting inter­
personal disagreements, low-directive interventions proved more effective than high-
directive interventions.

Effect Size Coding

The final sample for the current meta-analysis consisted of 11 studies from the 2011
meta-analysis and the 2 new studies found in the literature search, for a total of 13
 202 PSY C H O TH E R A PY R ELA TIO N SH IPS THAT WORK

studies. Three ESs were extracted from each of the articles, when possible: (a) the mean
ES of patient resistance/reactance on treatment outcome, (b) the mean ES of the level
of therapy/treatment directiveness (i.e., more directive vs. less directive), and (c) the
overall effect of matching therapeutic directiveness to patient reactance level.
A unit-free ES, d, was calculated by determining the difference in means between
groups and dividing the results by the pooled standard deviation (Borenstein et al.,
2009, p. 26).

M 1- M 2
d=
SDpooled

When means were not available, the ES (d) was estimated directly from signifi­
cance tests (i.e., t, F, X 2) according to the procedures/equations indicated by Lipsey
and Wilson (2001, pp. 172-206). The signs of the ESs in all cases were changed when
necessary to ensure that positive signs indicated support for the hypothesis. Multiple
outcomes within studies were aggregated using the method of Borenstein et al. (2009).

Calculating Mean ES

To determine the overall effect of fit between patient reactance and therapist
directiveness, we employed a random-effect meta-analysis using Wilson’s (2005) SPSS
macros. Similar to the previous meta-analysis, interaction effects were examined by
eliciting ESs as the product of the patient and therapist variables, using all studies that
related to a “fit” of reactance and directiveness. The resulting ESs from each study were
used to calculate a mean ES across studies by weighting each study by the inverse of
its variance. As well, ESs were calculated for the main effects of both reactance and
directiveness. The numbers of studies reporting ESs for directiveness (k = 6) alone and
for resistance (k = 3) alone were small (see Table 7.1).

Results
The results of the meta-analysis, based on 13 studies and 1,208 patients, are summarized
in Table 7.1. The 13 studies yielded 14 mean ESs that addressed the fit or match be­
tween reactance and directiveness in treatment outcomes. One study included in the
2011 review (Karno & Longabuagh, 2004) generated two ESs in this analysis because
two separate samples were used within the study. The aggregate ES for the fit between
reactance and directiveness yielded a mean ES ( d ) across studies of 0.78 (standard
error = .1; p < .001; 95% confidence interval = .60-.97). That ES is considered large.
A smaller number of mean effect scores were available to test the independent role
of directiveness (k = 6) and resistance (k = 4). The ESs of directivenes and reactance,
as independent contributors to outcome, were (d) 0.40 and 0.54, respectively. These
ESs are considered moderate. The effective therapist may not only adjust his or her
Table 7.1. Summary of Studies in Meta-Analysis
Study Name N Design Measure Resistance

Calvert et al. (1988) 108 RCT D (FIR O -B )

Beutler, Engle, et al. (1991) 62 RCT D (M M PI)
Beutler, M ohr, et al. (1991) 63 RCT D (9 scales)
Beutler et al. (1993) 46 RCT D (M M PI)
Piper et al. (1999) 98 RCT D (Q O R )
Karno et al. (2002) 47 RCT D (M M PI)
Karno and Longabaugh (2004) 140 RCT D (anger)
138 RCT D (anger)
Karno and Longabaugh (2005a) 169 RCT D (obs)
Karno and Longabaugh (2005b) 139 RCT D (self-report)
Clarkin et al. (2007) 62 RCT I (BPD )

G regory et al. (2008) 30 RCT I (BPD )

W estra et al. (2009) 76 RCT D (C M O T S)

Aviram et al. (2016) 30 RCT D (V N IS and

obs)
Total N 1208
Measure Directiveness Number of ES MES M ES (Resist) MES (Fit) V o f Fit ESs
(Direct)
D (T O Q ) i 0.52 .019
I (C B T vs. F EP vS/Sd) 3 .34 0.88 .07
I (BEH vs. ND) 9 0.62 .065
D (TPRS) 1 .33 1.4 .11
I (Interp vs. insight) 4 .31 .43 0.64 .021
I (FS T vs. C B T ) 1 .46 .42 0.65 .044
D (TPRS) 3 1.16 .017
I (M I vs. C B T ) 2 0.43 .014
D (obs) 4 1.21 .014
I (TPR S) 6 1.12 .017
I (D B T v Pdyn vs. 3 0.14 .065
Support)
I (Pdyn vs. TAU) 4 0.52 .068
I (M I p rior to C B T vs. 3 .53 .92 0.9 .057
no M I)
I (M I) 5 .49 0.75 .014

( continued)
Table 7.1. Continued
Study Name N Design Measure Resistance Measure Directiveness Number of ES M ES M ES (Resist) M ES (Fit) V o f Fit ESs
(Direct)
ES Weighted by Sample Size .4 .54
ES, Random Effects Model (95% 0.78 (.6-.97)*
confidence interval)
Q (Random Effects Variance 52.48 (.086)*
Component)

Note. Design: RCT = randomized clinical trial.

Measures of resistance and directiveness: Measure of resistance and directiveness are either directly measured (D) or indirectly measured (I).
Specifically, D indicates the use of direct observational ratings of directiveness (obs) or a standardized trait measure (e.g., the MMPI [Minnesota Multiphasic Personality Inventory], QOR
[Quality of Object Relationships], FIRO-B [Fundamental Interpersonal Relations Orientation Scale, Behavior], VNIS [Vanderbilt Negative Indicators Scale], or Systematic Treatment
Selection-Clinician Rating Form) applied to each individual. I indicates that an indirect measure of resistance was used based upon a grouping variable such as patient diagnosis (e.g., BPD
[borderline personality disorder] or SAD [substance abuse disorder]) to indicate resistant groups.
Among measures of directiveness, D indicates the use of a direct rating of therapist acts in treatment (e.g., using an observational rating like the TPRS [Therapy Process Rating Form]) or
a simple observational rating (obs). I indicates the use of an indirect measure of directiveness, based on the general directiveness of the treatment model used. The following are identifiers
of the direct and indirect measures of the directive and nondirective treatments employed.
TOQ = Therapist Orientation Questionnaire—a measure of therapist directiveness; Pdyn = psychodynamic treatment—moderately directive; TAU = treatment as usual, nondirective;
BEH = Behavioral Tx, directive; ND = nondirective or reflective, nondirective; CBT = cognitive-behavioral therapy, directive; FEP = focused expressive therapy, low directive; Interp = in­
terpretive, highly directive; FST = family systems; NT = narrative therapy; MI = motivational interviewing, nondirective; DBT = dialectic behavior therapy, directive; Support = supportive
therapy, nondirective.
N ES/Study = number of effect sizes calculated for this study.
M ES (Direct) = the mean effect size attributable to the directiveness of the treatment—combining all treatments.
M ES (Resist) = the mean effect size attributable to the resistance variable—combining all varieties.
M ES (Fit) = the mean difference between effect sizes for “good” and “poor” fit.
V of Fit ESs = the variance of effect size for “good” and “poor” fit.
All ESs are expressed as d.
* p < .001.
 205 Reactance Level

therapeutic stance in response to the patient’s resistance but keep patient resistance low
and successfully use directive interventions independently of patient resistance levels.
The effects for the interaction of reactance and directiveness were not homogenous
(Q = 52.48, .08, p < .001). Thus between-study differences accounted for variability of
the effects for the fit between reactance and directiveness.
The weighted average ES for studies that used direct measures of resistance (k = 12)
were compared to those that used indirect measures of resistance (k = 2). The resulting
weighted average ES for the studies using direct measures was .88 and for those using
indirect measures was d = .26, which is notably smaller. The discrepancy between
these ESs indicate that indirect measures are less sensitive than direct measures, as was
found in the earlier meta-analysis (Beutler et al., 2011).
The addition of two studies to the 2011 sample did not substantially change
the findings obtained in the earlier meta-analysis. It is quite clear that when pa­
tient resistance is not met with confrontation and control, but with acceptance and
nondefensiveness, good things are more likely to happen in psychotherapy. Not quite
as clearly but suggested by the linearity and strength of the findings was the indication
that the reverse is also true. These results suggest that reactant patients have better
outcomes in nondirective treatments whereas directive interventions may be indicated
for patients with lower levels of resistance.

EVIDENCE FOR CAUSALITY

We can conclude with some caution that the relationship between therapist
directiveness and the patient’s reactance causally contributes to subsequent thera­
peutic improvement. This conclusion is supported by a line of controlled, prospective
research studies on this topic. Each of the 13 studies used well-established research
methods that consistently pointed to the patient x therapy fit as an important part of a
causal chain leading toward positive outcome.
The best evidence comes from a series of RCTs in which levels of therapist
directiveness were randomly assigned to patients who varied in resistance levels.
The previously described study by Beutler, Engle, et al. (1991) and four additional
randomized trials provide evidence for causality. The additional studies are Beutler,
Machado, et al. (1991), a Swiss replication of the foregoing study (Beutler, Mohr, et al.,
1991), Karno and Longabaugh (2005a) using the Project Match data, and a study
involving the treatment of alcoholic couples by Karno et al. (2002). These studies
differed in their randomization processes, clinical populations, and research measures
but nonetheless obtained consistent results.
O f course, it would be too simple if any one set of treatment, patient, and ther­
apist variables accounted for most of psychotherapeutic change. In the course of
studying the moderating effects of therapist directiveness on patient reactance, we
have also researched several other interactions that contribute to change. Results on
several of these are reported in this volume (e.g., coping style, functional impairment).
Understanding and enhancing the effects of psychotherapy may lie with identifying
many interacting contributors to change (Castonguay et al., in press).
 206 psy c h o th er a py rela tio n sh ips that w ork

LIMITATIONS OF THE RESEARCH

There are several limitations to consider when reviewing the results of the current
meta-analysis. We included only studies published in the English language and only
studies that utilized an RCT methodology. As long as one applies the findings within
an English-speaking environment, the findings are likely to generalize. We find no
immediate evidence to suggest that international studies published in non-English
journals and studies employing other research designs may detract from the current
results.
Additional limitations concern the types of measures used to quantify the patient
and therapy constructs in the analyzed studies. Table 7.1 reveals that, in the majority
of studies (k = 10) in this meta-analysis, reactance was measured indirectly, rather
than directly. The analysis demonstrated that these indirect measures are less sensi­
tive and probably less accurate than direct ones. Use of indirect or proxy measures
unnecessarily homogenizes the samples, since they are group measures of individual
differences. Given these limitations, we strongly recommend that researchers exploring
resistance/reactance and clinicians applying these principles use direct measures for
both therapist directiveness and patient resistance whenever possible.
Likewise, the use of brand-name psychotherapies as proxies for therapeutic
procedures contributes error and proves less sensitive. These indirect measures
may unwittingly lead practitioners to conclude that the results apply only to those
psychotherapies included in reviewed studies. For example, in the current meta­
analysis, one of the new research studies and three of the older ones employed m o­
tivational interviewing as a proxy for low-directive therapy. Thus our meta-analytic
results might reflect both the effects of low directiveness and some specific strength
or weakness associated with M I, at least in those studies. But this concern is present
when any indirect, proxy measure is used and when distinctly different patient groups
are unavailable.
In the future, we recommend that researchers ensure that a broad range of therapies
and at least two well-defined patient groups are represented in such analyses as
presented here. When possible, it would also prove valuable for research and practice
to move away from global brands and toward the use of clusters of like techniques
representing the principles that are linked to effective change.

DIVERSITY CONSIDERATIONS

Despite efforts in the research community to include culturally representative samples,

a large proportion of studies continue to focus on Western populations. The majority
of the studies included in the meta-analysis took place in a Western culture, mainly the
United States. Reactance may appear in a different form or at different levels among
non-Western cultures than they do in Western cultures. Further, individuals from one
of these non-Western societies may respond differently than that observed in this re­
view, especially as related to directive and nondirective approaches. For instance, re­
search shows that Asian Americans regard mental health professionals as authority
 207 Reactance Level

figures and welcome a more directive form of psychotherapy (e.g., Sue & Sue, 1999;
Wong et al., 2007). That observation may also prove to be the case for patients of other
cultural identities, such as gender, sexual orientation, socioeconomic status, and reli­
gious affiliation.
Some cross-cultural comparisons of reactance x directiveness have been explored
by research. While similarity to North American samples is the norm among South
American and European studies (e.g., Beutler, Mohr, et al., 1991; Corbella, Beutler,
et al., 2003), there are some indications of differences among Asian and non-Asian
populations (e.g., Beutler, 2009; Song et al., 2014). To date, the patterns related to reac­
tance seem to transcend geographic and ethnic boundaries, but some distinctiveness
appears also to be present among Asian groups.

TRAINING IMPLICATIONS
These meta-analytic results provide further support for psychotherapists to con­
sider patient uniqueness and to tailor treatment accordingly. Resistance is not
solely an impediment but a rich spring of information about a patient (Newman,
1994). Understanding how to assess, conceptualize, and treat reactance are critical
components of sound clinical practice.
We encourage clinicians in training to obtain an in-depth understanding of the
ways in which resistance manifests and how to adjust psychotherapy such that reac­
tance does not interfere with its progress. Resistance is more than a patient variable;
it is also a function of a clinician’s use of a directive and confrontive approach. Thus
trainees will probably benefit from receiving supervision in psychotherapies that vary
in levels of therapist control, structure, and directiveness.
Two recent studies examined the efficacy of an integrative psychotherapy super­
vision that emphasized tailoring therapy to patient reactance (Holt et al., 2015; Stein
et al., 2017). Supervisors departed from their usual supervision approach to guide
student therapists to utilize empirically established principles of change in electing a
treatment strategy. The use of particular techniques was based upon whether a pa­
tient presented as more or less reactant. Overall, patients whose therapists were trained
and supervised according to the Systematic Treatment Selection model demonstrated
greater therapeutic gains when compared to supervision as usual.
O f course, practitioners have long addressed patient “resistance,” and no psycho­
therapy course is complete without a discussion of this topic. Our meta-analysis adds
specificity to the educational plan by expanding the concept to reactance and thereby,
invoking therapist provocation as a stimulus for resistance. Mental health professionals
can be trained to resist the temptation of responding to patient dominance with more
dominance. A therapist’s optimal response to a patient’s resistance is to become less di­
rective and dogmatic and more supportive and reflective. Answering patient resistance
with less resistance of one’s own is likely to disarm the patient.
Taken further, we recommend that psychotherapy teachers and supervisors en­
courage therapists to see resistance as a contextual problem rather than as some weak­
ness or flaw within the patient. We hope that practitioners who read our work and
 208 p sy c h o th er a py rela tio n sh ips that w ork

come to view resistance as a response stimulated by threatened loss of interpersonal

control (reactance) will be better prepared to return control to the patient and reap
better outcomes by doing so.
Probably more difficult is training therapists to become more leading and direc­
tive with patients who present with low reactance. Conventionally, therapists within
some practice communities have been trained to avoid directiveness; they have fre­
quently been socialized to avoid being misled by patients who “appear” cooperative
when surely they are not. The research evidence demonstrates that most patients who
appear cooperative will benefit from structure and direct guidance. If the patient’s co­
operation subsequently is shown to be unreliable, the therapist can alter tactics and
become once again the interested source of support and encouragement rather than
the teacher and leader.
Finally, the results obtained here suggest that supervisors convey to the trainee that
it is the patients reactance level, not the therapist’s, that guides the optimal amount of
directiveness employed within sessions. In truth, the history of psychotherapy is replete
with therapeutic practices predicated on the clinician’s personality, preferences, and re­
actance levels. We suggest that a supervisor respectfully remind students and colleagues
alike that, when the research evidence strongly indicates a particular patient-treatment
match, it is their ethical and clinical responsibility to seriously consider that match and
to optimize patient outcomes (as opposed to addressing their own comfort levels).

THERAPEUTIC PRACTICES

♦ Assess routinely a patients reactance level (as a personality trait) and in-session
resistance behaviors (as an environment-specific state).
♦ Learn to recognize symptoms of state and trait resistance and come to differentiate
between them.
♦ Consider the possibility that the particular therapeutic approach itself may be
creating or magnifying patient resistance, beyond the ubiquity of the human aversion
to change. As evidenced by the meta-analytic review, intervention low-directiveness
is a counter to patient resistance.
♦ Maintain or re-establish a collaborative stance, which can prove an effective way to
approach resistance patients. This may involve an element of transparency—openly
naming the patient’s resistance and exploring how the therapist’s methods fuels such
resistance (Ellis, 2004).
♦ Respond thoughtfully and sensitively to resistance, including acknowledging the
patient’s concerns through reflecting, speaking candidly about the therapeutic
relationship, adjusting the therapeutic contract to include more patient control,
exploring underlying mechanisms that motivate reactance, and shifting from
resistance to change.
♦ Follow the research-supported match: more directive and structured therapy with
low-reactance patients. Become more of a guide and an oracle, and even a teacher,
but do so selectively. Doing so will generally bring better therapeutic results.
 209 Reactance Level

♦ Emphasize the patient’s self-control, employ a less directive stance, and consider
paradoxical interventions with highly reactant patients. A therapist may be less of a
technician who fixes things and more of a healer who understands and values things.
Doing so will also typically yield better therapeutic results.

REFERENCES
References m arked with an asterisk indicate studies included in the m eta-analysis.
Aviram, A., & Alice Westra, H. (2011). The impact o f m otivational interviewing on resistance
in cognitive behavioural therapy for generalized anxiety disorder. Psychotherapy Research,
21(6), 6 9 8 -7 0 8 . https://www.doi.org/10.1080/10503307.2011.610832
Aviram, A., W estra, H. A., Constantino, M. J., & Antony, M. M . (2016). Responsive m anage­
m ent o f early resistance in cognitive-behavioral therapy for generalized anxiety disorder.
Journal o f Consulting and Clinical Psychology, 8 4(9), 7 8 3 -7 9 4 . http://dx.doi.org/10.1037/
ccp0000100
Beutler, L. E. (2009). M aking science m atter in clinical practice: Redefining psychotherapy.
Clinical Psychology: Science and Practice, 16, 3 0 1 -3 1 7 .
*Beutler, L. E., Engle, D., Mohr, D., Daldrup, R. J., Bergan, J., M eredith, K., & Merry, W. (1991).
Predictors o f differential response to cognitive, experiential, and self-directed psychother­
apeutic procedures. Journal o f Consulting and Clinical Psychology, 59(2), 3 3 3 -3 4 0 . https://
www.doi.org/10.1037/0022-006X. 59.2.333
Beutler, L. E., Harwood, T. M ., M ichelson, A., Song, X ., & Holman, J. (2011). Resistance/re-
actance level. Journal o f Clinical Psychology, 67(2), 1 3 3 -1 4 2 . https://www.doi.org/10.1002/
jclp.20753
*Beutler, L. E., Machado, P.P., Engle, D., & Mohr, D. (1993). Differential patient x treat­
m ent m aintenance o f treatm ent effects am ong cognitive, experiential, and self-directed
psychotherapies. Journal o f Psychotherapy Integration, 3, 15-3 2 .
*Beutler, L. E., Mohr, D. C., Grawe, K., Engle, D., & M acDonald, R. (1991). Looking for dif­
ferential effects: Cross-cultural predictors o f differential psychotherapy efficacy. Journal o f
Psychotherapy Integration, 1 , 1 2 1-142.
Beutler, L. E., Moleiro, C., & Talebi, H. (2001). Resistance in psychotherapy: W hat conclusions
are supported by research. Journal o f Clinical Psychology, 58, 2 0 7 -2 1 7 .
Beutler, L. E., Rocco, F., M oleiro, C. M ., & Talebi, H. (2001). Resistance. Psychotherapy, 38(4),
4 3 1 -4 3 6 . https://www.doi.org/10.1037/0033-3204.38.4.431
Beutler, L. E., W illiam s, O. B., & Norcross, J. N. (2008). STS/Innerlife [Software program ].
Retrieved from www.innerlife.com. San Jose, CA.
Borenstein, M ., Hedges, H. V., Higgins, J. P. T., & Rothstein, H. (2009). Introduction to meta­
analysis. Chichester, England: Wiley.
Brehm , S. S., & Brehm , J. W. (1981). Psychological reactance: A theory o f freedom and control.
New York, NY: Academ ic Press.
Buboltz, W. J., Johnson, P , & Woller, K. P. (2003). Psychological reactance in college
students: Fam ily-of-origin predictors. Journal o f Counseling & Development, 8 1(3), 3 1 1 ­
317. https://www.doi.org/10.1002/;. 1556-6678.2003.tb 00258.x
Butcher, J. N. (1999). MMPI-2: A beginner’s guide to the MMPI-2. W ashington, DC: A m erican
Psychological Association.
 210 PSYCHO THERA PY RELATIONSHIPS THAT WORK

Butcher, J. N., Beutler, L. E., Harwood, T. M ., & Blau, K. (2011). The M M P I-2. In T. M.
Harwood, L. E. Beutler, & G. G roth-M arnat. (Eds), Integrative assessment o f adult person­
ality (3rd ed., pp. 1 5 2 -1 8 9 ). New York, NY: Guilford.
*Calvert, S. J., Beutler, L. E., & Crago, M. (1988). Psychotherapy outcom e as a function of
therapist-patient m atching on selected variables. Journal o f Social and Clinical Psychology,
6, 1 0 4 -1 1 7 .
Castonguay, L. G ., Constantino, M ., & Beutler, L. E. (Eds.). (in preparation). Principles o f psy­
chotherapy change that work: Applications, Vol. II. M anuscript in press.
Cham berlain, P., Patterson, G., Reid, J., Kava- nagh, K., & Forgatch, M . (1984). Observation
o f client resistance. Behavior Therapy, 15, 1 4 4 -1 5 5 . https://www.doi.org/10.1016/
S 0 0 0 5 -7 8 9 4 (8 4 )8 0016-7
*Clarkin, J. F., Levy, K. N., Lezenweger, M. F., & Kernberg, O. F. (2007). Evaluating three
treatm ents o f borderline personality disorder: A multiwave study. American Journal o f
Psychiatry, 164, 9 2 2 -9 2 8 .
Corbella, S., Beutler, L. E., Fernandez-Alvarez, H., Botella, L., M alik, M. L., Lane, G ., &
Wagstaff, N. (2003). M easuring coping style and resistance am ong Spanish and Argentine
samples: Development o f the System atic Treatm ent Selection Self-Report (ST S-SR ) in
Spanish. Journal o f Clinical Psychology, 5 9 , 9 2 1 -9 3 2 .
Dowd, E. T., M ilne, C. R., & W ise, S. L. (1991). The Therapeutic Reactance Scale: A measure of
psychological reactance. Journal o f Counseling & Development, 69(6), 5 4 1 -5 4 5 .
Ellis, A. (2004). M ethods to reduce and counter resistance in psychotherapy. In G. P. Koocher,
J. C. Norcross, & S. S Hill (Eds.), Psychologists’ desk reference (2nd ed., pp. 2 1 2 -2 1 5 ).
New York, NY: O xford University Press.
Firestone, R. W. (2015). The ultimate resistance. Journal o f Humanistic Psychology, 55(1), 77 -1 0 1 .
https://www.doi.org/10.1177/0022167814527166
Fisher, D. R., K arno, M . P., Sandow icz, M ., Albanese, A., & Beutler, L. E. (1995). Systematic
Treatment Selection Therapy Rating Scale manual. Santa Barbara: U niversity o f C alifornia.
*Gregory, R. J., Chlebowski, S., Kang, D., Remen, A. L., Soderberg, M. G., & Stepkovitch, J.
(2008). A controlled trial o f psychodynam ic psychotherapy for co-occu rring borderline
personality disorder and alcohol use disorder. Psychotherapy, 45(1), 2 8 -4 1 .
Hara, K. M ., W estra, H. A., Aviram, A., Button, M. L., Constantino, M . J., & Antony, M.
M. (2015). Therapist awareness o f client resistance in cognitive-behavioral therapy for
generalized anxiety disorder. Cognitive Behaviour Therapy, 44(2), 1 6 2-174.
Hathaway, S. R., & Mckinley, J. C. (1940). A Multiphasic Personality Schedule: C onstruction ofthe
Schedule. The Journal o f Psychology, 10(2), 2 4 9 -2 5 4 . d oi:10.1080/ 00223980.1940.9917000
Holt, H., Beutler, L. E., Kimpara, S., M acias, S., Haug, N. A., Shiloff, N., . . . Stein, M . (2015).
Evidence-based supervision: tracking outcom e and teaching principles o f change in clin­
ical supervision to bring science to integrative practice. Psychotherapy, 5 2 , 1 85-189.
*K arno, M . P., Beutler, L. E., & Harwood, M. (2002). Interactions between psychotherapy pro­
cess and patient attributes that predict alcohol treatm ent effectiveness: A prelim inary re­
port. Journal o f Alcohol Studies, 27, 7 7 9 -7 9 7 .
*K arno, P. M ., & Longabaugh, R. (2004). W hat do we know? Process analysis and the search
for a better understanding o f Project Match’s anger-by-treatm ent m atching effect. Journal
o f Studies in Alcohol, 65, 5 0 1 -5 1 2 .
*K arno, M . P., & Longabaugh, R. (2005a). An exam ination o f how therapist directiveness
interacts with patient anger and reactance to predict alcohol use. Journal o f Studies on
Alcohol, 66(6), 8 2 5 -8 3 2 . https://www.doi.org/10.15288/jsa.2005.66.825
 2 11 Reactance Level

*K arno, M . P., & Longabaugh, R. (2005b). Less directiveness by therapists improves drinking
outcom es o f reactant clients in alcoholism treatm ent. Journal o f Consulting and Clinical
Psychology, 7 3(2), 2 6 2 -2 6 7 . https://www.doi.org/10.1037/0022-006X.73.2.262
Kimpara, S., Regner, E., Satoshi, U., & Beutler, L. E. (Aug. 2015). How to monitor therapist
interventions and cross cultural differences between North America and Argentina. A poster
presented at the annual meeting o f the Am erican Psychological Association, Honolulu,
Hawaii.
Lipsey, M. W , & W ilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: SAGE.
M alik, M. L., Beutler, L. E., Gallagher-Thom pson, D., Thom pson, L., & Alim oham ed, S.
(2003). Are all cognitive therapies alike? A com parison o f cognitive and non-cognitive
therapy process and implications for the application o f em pirically supported treatm ents
(ESTs). Journal o f Consulting and Clinical Psychology, 71, 1 5 0-158.
Messer, S. B. (2002). A psychodynam ic perspective on resistance in psychotherapy: Vive la
résistance. Journal o f Clinical Psychology, 58(2), 1 5 7-163.
M itchell, C. W (2013). Reducing resistance in psychotherapy. In G. P Koocher, J. C. Norcross,
& B. A. G reene (Eds.), Psychologists’ desk reference (3rd ed., pp. 2 5 8 -2 6 2 ). New York,
NY: O xford University Press.
Newman, C. F. (1994). Understanding client resistance: M ethods for enhancing m otivation
to change. Cognitive and Behavioral Practice, 1(1), 4 7 -6 9 . https://www.doi.org/10.1016/
S 1077-72 2 9 (0 5 )80086-0
*Piper, W E., M cCallum , M ., Joyce, A. S., Azim, H. F., & Ogrodniczuk, J. S. (1999). Follow-up
findings for interpretive and supportive form s o f psychotherapy and patient personality
variables. Journal o f Consulting and Clinical Psychology, 67, 2 6 7 -2 7 3 .
Song, X ., Anderson, T., Beutler, L. E., Sun, G., Wu, G., & Kimpara, S. (2014). Cultural adapta­
tion in m easuring com m on client characteristics with an urban m ainland Chinese sample.
Psychotherapy Research, 51, 1-11.
Stein, M ., Beutler, L. E., Kimpara, S., Haug, N. A., Brunet, H., Som eah, K., . . .Macias, S. (2017).
The impact o f cross-interventions and principle-based supervision on trainee effectiveness.
M anuscript submitted for publication.
Strong, S. R., & M atross, R. P (1973). Change process in counseling and psychotherapy.
Journal o f Counseling Psychology, 20, 2 5 -3 7 .
Sue, D. W , & Sue, D. (1999). Counseling the culturally different: Theory and practice (3rd ed.).
New York, NY: Wiley.
*W estra, H. A., Arkowitz, H., & Dozois, D. J. A. (2009). Adding a m otivational interviewing
pretreatm ent to cognitive behavioral therapy for generalized anxiety disorder: A prelim i­
nary randomized control trial. Journal o f Anxiety Disorder, 2 3 (8 ), 110 6 -1 1 1 7 . https://www.
doi. org/10.1016/j.janxdis. 2009.07.014
W estra, H. A., Aviram, A., Kertes, A., Ahm ed, M ., & Connors, L. (2009). Manual fo r rating
interpersonal resistance. Unpublished m anuscript, York University, Toronto, ON.
W ilson, D. B. (2005). M eta-analysis m acros for SAS, SPSS, and Stata. Retrieved from http://
mason.gmu.edu/~dwilsonb/ma.html
W ong, E. C., Beutler, L. E., & Zane, N. W. (2007). Using m ediators and m oderators to test
assumptions underlying culturally sensitive therapies: An exploratory example. Cultural
Diversity and Ethnic Minority Psychology, 13(2), 1 6 9-177. https://www.doi.org/10.1037/
1099-9809.13.2.169
 8

R E L I G I O N AND S P I R I T U A L I T Y

Joshua N. Hook, Laura E. Captari, William Hoyt, Don E. Davis,

Stacey E. McElroy, and Everett L. Worthington Jr.

Religious or spiritual (R/S) beliefs and practices are woven into the fabric of many
people’s lives. Globally, about 68% of the world’s population views religion as an im ­
portant aspect of their day-to-day experience (Diener et al., 2011). In the United States,
89% believe in “God or a universal spirit" 75% describe religion as either “somewhat"
or “very” important to them, 80% pray regularly, and 50% belong to a local house of
worship (Pew Research, 2015).
An extensive body of research has documented a positive relation between R/S and
physical and mental health (Koenig et al., 2012). Specifically, R/S can foster increased
social belonging, connection, and support; adjustment to stressors through meaning­
making, coping, and resilience; and grounding of one’s identity through salient beliefs
and values (e.g., Paloutzian & Park, 2014; Pargament et al., 2013). However, for some
individuals, R/S can be a source of struggle and confusion or serve as a defense against
unresolved psychological conflicts (e.g., Cashwell et al., 2007; Exline & Rose, 2014).
Incorporating clients’ R/S identity within psychotherapy has the potential to impact
both change processes and treatment outcomes. According to Pargament (2011),
“When people walk into the therapist’s office, they don’t leave their spirituality behind
in the waiting room. They bring their spiritual beliefs, practices, experiences, values,
relationships, and struggles along with them” (p. 4).
Many patients who experience R/S as a salient part of their identity hope for their
therapist to integrate these beliefs and values within psychotherapy (Vieten et al.,
2013). While some individuals forthrightly state this, others do not. Such clients may
be reticent to disclose R/S-related aspects of their struggles in a setting they presume
to be limited to secular considerations, potentially hampering therapeutic outcomes.
A number of patients stand to benefit through secular psychotherapy but may expe­
rience additional gains if treatment were contextualized within their R/S values. For
others, R/S struggles play a significant role in their psychological and emotional dis­
tress, making it vital to address such topics in therapy.
In addition to patients’ unique needs, R/S adaptations in psychotherapy may be
complicated by several practitioner characteristics. Notably, psychotherapists as a
whole are less likely to identify as R/S compared with the general population. In one

212
 213 Religion and Spirituality

survey, 35% of psychologists—in contrast with 75% of the public— described their ap­
proach to life as significantly influenced by R/S (Delaney et al., 2007). Furthermore,
relatively few psychotherapists receive explicit training and supervision in how to eth­
ically and sensitively address patients’ R/S beliefs in assessment and treatment (Schafer
et al., 2011). Thus, although R/S is increasingly recognized as an important aspect of
multicultural competency, psychotherapists may be unsure how and in what way to
best facilitate integration of such concerns, which can result in “spiritually avoidant
care” (Saunders et al., 2010, p. 355).
In this chapter, we examine the role of R/S in psychotherapy. We begin by defining R/
S and discussing common measures of these constructs. We also offer clinical examples
illustrating treatment accommodations in psychotherapy, considering the patient’s R/
S beliefs and worldview, and review results of landmark studies and previous meta­
analyses. Next, we present data from the current meta-analysis examining the efficacy
of R/S-adapted psychotherapies and explore patient, study, and treatment character­
istics that may moderate therapeutic effects. We also consider research evidence for
the causal role of R/S adaptations in predicting psychotherapy outcome and note lim­
itations in the research to date. The chapter concludes with diversity considerations,
training implications, and therapeutic practices based on the research evidence.

DEFINITIONS
Historically, the terms religion and spirituality were closely linked, and at times, used
interchangeably (Sheldrake, 1992). Religion can be defined as adherence to common
beliefs, behaviors, and practices associated with a particular faith tradition and com­
munity, which provides guidance and oversight (Hill et al., 2000). In contrast, spiritu­
ality is a broader concept describing the subjective, embodied, emotional experience
of closeness and connection with what is viewed as sacred or transcendent. This often
constitutes either (a) a divine being or object or (b) a sense of ultimate reality or truth
and can be understood within the framework of implicit relational knowledge.
Spirituality has been further defined within four main categories, based on the sa-
cred/ transcendent object (Davis et al., 2015). Religious spirituality involves a felt sense
of closeness and connection with a higher power or worship tradition as described by a
specific religion (e.g., Christianity, Islam, Judaism, Buddhism). Many people’s spiritual
experiences occur within this context; however, for others, spirituality is independent
of any religious ties.
A growing number of Americans are religiously unaffiliated, self-identifying as
“nothing in particular” (16%), agnostic (4%), or atheist (3%; Pew Research, 2015).
A sizeable majority of the religiously unaffiliated endorse feeling a deep sense of “spir­
itual peace and well-being” (40%) or “wonder about the universe” (47%; Pew Research,
2016). These individuals may experience the sacred primarily through connection with
humanity, nature, or the cosmos. Humanistic spirituality involves a sense of closeness
and connection with other human beings, such as feeling compassion, lovingkindness,
care, or altruism. Nature spirituality involves a sense of closeness and connection with
the environment or aspects of nature, such being awestruck at a beautiful sunset or the
 2 14 PSY C H O TH E R A PY R ELA TIO N SH IPS THAT WORK

grandeur of a mountain landscape. Finally, cosmos spirituality involves a sense of close­

ness and connection with the universe, such as contemplating the vastness of outer
space, one’s sense of being within the cosmos, or awe over the miraculous complexity
of the human genome.
Adapting psychotherapy to a patient’s R/S framework may influence the treatment
in several ways: conceptualization, treatment goals, intervention, and interpersonal
process. First, understanding the patient’s R/S aids in conceptualizing causes of psy­
chological distress and identifying key risk and protective factors. This might include
exploring the role of R/S in the patient’s history, identity, and current functioning, as
well as any areas of difficulty (e.g., spiritual struggles, spiritual bypass).
Second, R/S-tailored psychotherapy provides a broader context within which to un­
derstand the patient’s reasons for attending psychotherapy. In addition to symptom
reduction and self-development, R/S clients may identify additional goals, like de­
veloping a closer relationship with Jesus Christ, faithfulness to Allah, following the
teachings of Buddha, or greater connection with the transcendent. Psychotherapeutic
outcomes such as increased spiritual well-being and positive religious coping may
prove important considerations when patients’ goals extend beyond the psychological
to the R/S aspects of their lives.
Third, the patient’s R/S can be integrated within traditional interventions (e.g., be­
havior activation, challenging negative thoughts, distress tolerance, mindfulness skills)
aimed at reaching treatment goals. R/S interventions might incorporate methods con­
sistent with a client’s R/S culture (e.g., prayer, meditation, religious imagery, sacred
scriptures, religious rituals or services) that may be positive coping resources.
Finally, although difficult to measure, R/S integration may occur implicitly in the
“being with” process of psychotherapy, as a patient experiences and internalizes the
consistent, attuned, and caring presence of the psychotherapist. Helping R/S patients
reflect on the ways in which the psychotherapy relationship mirrors or challenges their
perceived relationship with the sacred may positively impact how they relate to God
or their higher power.

MEASURES
For the purposes of the present meta-analysis, we examined psychotherapeutic change
in two patient dimensions: psychological outcomes and spiritual outcomes. Nearly
every treatment study in our meta-analysis included at least one psychological out­
come measure. For example, studies evaluating R/S-adapted psychotherapy for depres­
sion usually administered the Beck Depression Inventory-II (Beck et al., 1996) or the
Center for Epidemiological Studies Depression scale (Radloff, 1977), whereas those
examining R/S treatments for anxiety often used the Hamilton Anxiety Rating Scale
(Hamilton, 1959) or the Symptom Checklist-90-R (Derogatis, 1994).
Patients’ R/S was typically measured as a single demographic question (e.g., iden­
tification with a particular faith tradition or denomination), but some studies also
assessed R/S outcomes of accommodated psychotherapy. The Multidimensional
Measure of Religiousness/Spirituality (88 items; Fetzer Institute, 1999) and Brief
 215 Religion and Spirituality

Multidimensional Measure of Religiousness/Spirituality (38 items; Fetzer Institute,

1999) were the most comprehensive assessments used. These measures consider 12
domains of R/S, including daily spiritual experiences, meaning, values, beliefs, forgive­
ness, private religious practices, R/S coping, religious support, R/S history, commit­
ment, organizational religiousness, and religious preference. From this larger battery,
many researchers only administered the Daily Spiritual Experiences Scale (16 items;
Underwood & Teresi, 2002), which lists spiritual experiences some people have (e.g.,
feeling joy, peace, or gratitude; God’s presence, love, or comfort). Clients responded
based on how often they currently had each experience.
Another frequently used measure of R/S outcomes was the Functional Assessment
of Chronic Illness Therapy—Spiritual Well-Being Scale (12 items; Peterman et al.,
2002), originally developed to assess quality of life in survivors of cancer and other
chronic illnesses. Patients responded to items such as, “I find comfort in my faith/spir-
itual beliefs” and “My illness has strengthened my faith/spiritual beliefs.”
The Spiritual Well-Being Scale (20 items; Paloutzian & Ellison, 1991) was some­
times utilized and includes questions about one’s relationship with God (e.g., close­
ness, care) as well as existential well-being (e.g., sense of purpose, direction). Other
researchers measured R/S outcomes using the similarly-named Spiritual Well-Being
Questionnaire (20 items; Gomez & Fisher, 2003), which assesses personal (e.g., inner
peace, meaning), communal (e.g., love, respect for others), environmental (e.g.,
sense of wonder and connection with nature), and transcendental (e.g., faith, prayer)
well-being.
Some outcome studies measured participants’ R/S coping styles using the Religious
Coping Scale (99 items; Pargament et al., 2000) or the Brief Religious Coping Scale (14
items; Pargament et al., 2011). Patients responded to items assessing various aspects of
negative religious coping (e.g., spiritual discontent, reappraisals of God’s punishment,
pleading for rescue) and positive religious coping (e.g., spiritual surrender, seeking
support and connection, reappraisals of God’s protection and care).
Considering diversity in R/S practices, particularly across cultures, several
researchers used the Relational Well-Being scales (I: 30 items; McCubbin et al., 2010;
II: 14 items; McCubbin et al., 2013) to assess spiritual outcomes. Clients rated the fre­
quency of their engagement with activities such as connecting with nature, practicing
the traditions of ancestors, being active in a spiritual community, and having faith that
everything will work out. The Body Mind Spirit Well-Being Inventory (56 items; Ng
et al., 2005) was used in similar circumstances when a broader conceptualization of
R/S was preferred. Participants responded to items about their philosophy and core
values, related emotions (e.g., contentment, gratitude), and daily energy level and
concentration.
Some studies measured fidelity or adherence to the R/S treatment adaptation.
This generally included audio or video recordings of the sessions, a portion of which
were transcribed and evaluated by independent professionals not otherwise involved
in the study. In a few studies, the researchers used additional measures, such as the
Adherence Rating Scale (Waltz et al., 1993) or Therapist Competency Adherence Scale
(Weisman et al., 1998) to further quantify R/S adaptations.
 2 16 psy c h o th er a py r ela tio n sh ips that w ork

CLINICAL EXAMPLES

R/S adaptations of psychotherapy are as unique as each patient who walks through the
door. The following case examples illustrate several of the complex ways that R/S can
intersect with other cultural identities and influence the course of treatment. We focus
especially on the systematic ways that researchers have formally integrated R/S within
treatments.

Case 1: Religiously Integrated CBT for Depression

Religiously integrated cognitive behavior therapy (RCBT) is founded on the cogni­
tive model while contextualizing interventions within the patient’s religious frame­
work. The psychotherapist makes “explicit use of the client’s own religious tradition as
a major foundation to identify and replace unhelpful thoughts and behaviors to reduce
depressive symptoms” (Pearce et al., 2015, p. 58). Some of the major tools of RCBT
include (a) renewing the mind by replacing negative self-talk with excerpts from or
paraphrases of sacred scriptures; (b) meditating on sacred writings and engaging in
contemplative prayer; (c) considering religious beliefs and resources while challenging
thoughts; (d) cultivating forgiveness, hope, gratitude, and generosity through daily re­
ligious practices; (e) identifying and making use of R/S resources in line with one’s
faith tradition; and (f) altruistic involvement in one’s religious community. Ten-session
treatment manuals, and associated patient workbooks, have been developed to guide
RCBT with Christian, Muslim, Jewish, Buddhist, and Hindu patients.
Katina (age 42) presented to psychotherapy with depression. She self-identified
as an African American, cisgender woman, and her most salient identity was her
Christian faith. Katina was somewhat reticent about how traditional psychotherapy
could help her.
Picking up on this, her therapist explored Katina’s religious history and the role that
R/S played in her day-to-day life. They discussed ways she desired to incorporate this
into treatment. Katina reported growing up in a strict religious family, where she felt
she could never be good enough. She recounted episodes of physical and emotional
abuse by her father throughout childhood and the untimely death of her mother from
cancer when she was a teenager. Katina identified her Christian faith as a source of
coping and hope in the midst of this loss, but as psychotherapy progressed, she also
became more aware of her anger toward God because she believed that God took her
mother away or (at best) allowed bad things to happen. Katina identified her most
problematic core beliefs as (a) I am worthless and no one can ever love and accept
me as I am and (b) bad things keep happening to me and God does not stop them, so
I cannot fully trust him. As Katina and her psychotherapist modified these negative
core beliefs in the context of her faith, she found comfort in meditating on Scripture
passages about God’s presence and unconditional love. Katina identified that listening
to Christian music, journaling, taking reflective prayer walks, and attending a small
group at her church were all ways she could incorporate daily spiritual practices, rather
than being controlled by negative self-talk.
 217 Religion and Spirituality

Case 2: Spiritual Self-Schema Therapy for Addiction

Spiritual self-schema therapy integrates CBT with Buddhist psychological principles,
guided by a 12-session treatment manual (Avants & Margolin, 2004). The goal is to de­
crease impulsive behavior through modifying a client’s habitual self-schema, described
as a “highly automatized system of knowledge or beliefs about one’s intentions or
capacities, stored in long-term memory, that mediates the environment and interper­
sonal behavior” (Margolin et al., 2007, p. 982). When a habitual self-schema is acti­
vated, negative beliefs about the self can motivate self-destructive behaviors, such as
drug use. This psychotherapy attempts to facilitate a shift from an addict self-schema
to a spiritual self-schema by fostering mindfulness, self-compassion, and commitment
to do no harm to self or others. Spiritual self-affirmations, meditation, prayer, and
self-schema check-ins are used to foster spiritual awareness. Each session also focuses
on the development of a spiritual quality, including generosity, strong determination,
lovingkindness, and truth.
Tom (age 28) sought treatment at an inpatient drug rehabilitation facility at the
urging of his parents. He self-identified as a White, cisgender man, and spirituality was
a salient personal identity. He grew up in a family with a Jewish cultural heritage but
had explored other perspectives and worldviews. Tom recently lost his job after testing
positive for cocaine on a drug screening. He had begun experimenting with alcohol
and drugs in middle school and soon began dealing to other athletes on his sports
team. Tom had high hopes of playing football in college but received drug charges as a
high school senior that forced him to forfeit his scholarship. Though Tom tried to “get
clean” and start fresh, he was not able to maintain sobriety. He was shaken by losing
several friends to drug overdoses and expressed a desire to change but was resistant to
acknowledging the full impact of his addiction.
During psychotherapy, Tom was taught about the wandering nature of the mind,
what his psychotherapist called “monkey mind,” and how this contributed to his addict
self-schema. If Tom did not work to control his mind, he usually thought about using
drugs. Tom’s therapist introduced him to a meditation technique called anapanasati,
which involves sitting silently with eyes closed and practicing nonjudgmental aware­
ness of thoughts, feelings, and sensations while breathing naturally. Tom practiced this
coping strategy whenever he felt the overwhelming urge to use drugs. As treatment
progressed, Tom became more aware of the ways in which he turned anger and hatred
toward his parents inward, leading him to impulsive behavior that put him in harm’s
way. Tom began to notice moments throughout his day when his cravings subsided,
and he felt more inner peace and calm. Over time, he came to understand this as his
core spiritual self, which offered him wisdom, groundedness, and lovingkindness.

Case 3: Religious Cultural Psychotherapy for Anxiety

Religious cultural psychotherapy retains Beck’s cognitive model and accommodates
treatment for working with Muslim clients (Razali et al., 1998). This approach draws
on passages from the Holy Koran and Hadith (sayings and customs of the prophet) to
 2 18 psy c h o th er a py rela tio n sh ips that w ork

(a) examine the evidence for and modify automatic negative thoughts; (b) facilitate
the development of positive religious coping skills, including prayer, acceptance, and
adherence to Islamic customs; and (c) help clients understand their anxiety symptoms
within the context of their cultural and religious beliefs to reduce mental health stigma.
Clients are encouraged to cultivate feelings of closeness to Allah, read and reflect on
the truths of the Koran, and express their worries and fears in prayer.
Abdul (age 50) sought psychotherapy after being diagnosed with generalized anx­
iety disorder by his physician and resisting a referral for medication. He identified as a
Palestinian American, cisgender man, and his Muslim faith was his most salient iden­
tity. Within his religious tradition, taking medication for psychological difficulties was
discouraged, so he sought psychotherapy to learn to manage his symptoms. Although
a successful businessman, Abdul was constantly tense, consumed with racing thoughts
and fears that made it difficult to concentrate at work and home. Even during daily
prayer, he could not focus. This compounded his fears that Allah would punish him
for his lack of faithfulness.
In psychotherapy, Abdul acknowledged that he did not believe the world was a safe
place. He felt that he must constantly prepare himself and his family for the worst-case
scenario. He also worried about how rising political tensions and Islamophobia in the
United States might impact his safety and that of his wife and three children. The psy­
chotherapist validated Abdul’s fears while working with him to identify how his faith
could be a positive source of support. Abdul found that it eased his worries to meditate
on the beliefs that Allah was always in control and that he could trust Allah to take care
of him and his family. As Abdul’s shame over his symptoms subsided, he attended the
mosque more frequently and gained comfort from connection with others in his faith
community.

LANDMARK STUDIES

The first randomized clinical trial of R/S accommodated therapy occurred in the 1980s
(Propst, 1980). This seminal study involved patients who reported both (a) mild to mod­
erate depression and (b) moderately high levels of religiosity. Researchers randomized
patients (N = 44) into one of four groups: two group psychotherapy conditions (R/S in­
tegrated CBT, standard CBT) and two control conditions (nondirective R/S discussion
group plus self-monitoring, self-monitoring only). Groups met twice weekly for eight
one-hour sessions. R/S adaptations included utilizing religious imagery and coping
statements. The graduate student clinicians providing treatment did not identify as
R/S to control for expectancy effects. Both R/S-adapted CBT and the R/S discussion
group (with no CBT components) resulted in greater alleviation of depression than did
standard CBT or self-monitoring control. However, these between-group differences
were not maintained at six-week follow-up.
Years later, Propst and colleagues (1992) examined the efficacy of R/S CBT for clin­
ical depression in a community sample. The researchers randomized patients (N = 59)
who self-identified as Christian into one of four groups: two psychotherapy conditions
(R/S CBT or standard CBT) and two control conditions (pastoral counseling or
 219 Religion and Spirituality

waitlist). Patients received eighteen 50-minute sessions of individual psychotherapy,

guided by treatment manuals. Both R/S CBT and pastoral counseling resulted in
greater reductions in depression than the standard CBT and waitlist control groups.
These differences were maintained at 3- and 24-month follow-ups at a smaller magni­
tude than at termination.
Although these early studies were fairly sophisticated in design, they used mainly
Christian participants. Thus researchers began to examine adaptations of psycho­
therapy in Muslim populations (Azhar et al., 1994). In conjunction with medication,
Muslim patients (N = 62) with an anxiety disorder were randomly assigned to either
weekly R/S or supportive psychotherapy, receiving between 12 and 16 sessions. The
R/S accommodated treatment drew on passages from the Holy Koran and Hadith,
and incorporated prayer and discussions of religious issues. This study was replicated
among depressed patients (Azhar & Varma, 1995b); in both cases, R/S-accommodated
treatment resulted in greater initial symptom reduction compared with a matched sec­
ular treatment. At follow-up, this difference was nonsignificant.
More recently, researchers employed a mixed-method design, beginning with the
development of a spiritually integrated psychotherapy specifically for Muslim patients
(Ebrahimi et al., 2013). Subsequently, an intervention study was conducted to eval­
uate its efficacy in the treatment of dysthymia. Patients (N = 62) were randomized
to R/S CBT, standard CBT, a medication-only condition, or waitlist control group.
Psychotherapy consisted of eight 45-minute sessions held once a week. Patients re­
ceiving R/S treatment reported greater symptom reduction than those on medication
or in the waitlist control group. They also reported similar symptom reduction to those
in the standard CBT group. Notably, R/S-tailored psychotherapy was more effective
than standard CBT or medication in modifying dysfunctional attitudes that may un­
derpin dysthymia. In contrast with some previous studies, these additive effects of R/S
intervention were maintained at three-month follow-up.
The studies reviewed here developed and tested the efficacy of treatments for a
particular mental disorder (e.g., depression, anxiety) tailored within a particular reli­
gious tradition (e.g., Christian, Muslim), often using a fairly small and homogeneous
treatment sample. In contrast, Pearce and colleagues (2015) designed a transreligious
integrative treatment “to assist depressed individuals to develop depression-reducing
thoughts and behaviors informed by their own religious beliefs, practices, and re­
sources” (p. 56). This approach was developed for use with patients who ascribe to
any of the five major world religions (Christianity, Islam, Judaism, Buddhism, and
Hinduism).
A multisite randomized controlled study was conducted to evaluate the efficacy of
RCBT (Koenig et al., 2015, 2016), compared with standard CBT, among chronically de­
pressed patients. Individuals who self-identified R/S as at least “somewhat important”
were eligible for inclusion. Recruiting through multiple locations resulted in a larger
(N = 132) and more religiously diverse sample than previous studies. Participants re­
ceived ten 50-minute psychotherapy sessions, delivered remotely (telephone, Skype) by
eight licensed psychotherapists over the course of 12 weeks. Similar reductions in de­
pression were noted in both treatment conditions, suggesting that R/S psychotherapy
 220 PSY C H O TH E R A PY RELA T IO N SH IPS THAT WORK

is empirically on par with secular approaches. This study also evaluated patients’ spir­
itual change in treatment and found that RCBT was more effective than secular treat­
ment in promoting spiritual well-being and connection with the transcendent.
Researchers have also developed broader spiritual accommodations, accessible to
individuals who experience the sacred through connection with humanity, nature,
or the cosmos. In one early study, a 12-week mind-body-spirit group was created for
women coping with breast cancer; key intervention components included spiritual
rituals, imagery, affirmations, guided meditation, yoga, and self-reflection (Targ &
Levine, 2002). Participants (N = 181) were randomized to this psychospiritual group or
a standard support group, which combined CBT strategies with group sharing. While
both groups showed similar increases in quality of life and decreases in depression and
anxiety, the spiritual accommodative group reported a greater sense of spiritual inte­
gration, growth, and ability to embrace life’s fullness despite adversity.

RESULTS OF PREVIOUS META-ANALYSES

Over the last two decades, there has been a marked increase in the number, variety, and
rigor of empirical studies examining R/S-accommodative psychotherapies, as the use
of such treatments has increased dramatically. We organize findings of previous meta­
analyses in Table 8.1, and summarize the development of this line of research next.
Early meta-analyses sought to broadly establish whether R/S accommodations
in psychotherapy were effective. A small meta-analysis of five randomized outcome
studies (McCullough, 1999) found that Christian-accommodated CBT was not sig­
nificantly more effective in reducing depression compared to secular CBT (d = .18).
A later meta-analysis identified 31 intervention studies with both Christian and
Muslim clients (Smith et al., 2007). The researchers analyzed alternate treatment,
waitlist control, and single group pre-post studies together, resulting in a medium ef­
fect (d = .56) of R/S adaptation on reducing psychological distress. Explicitly teaching
spiritual concepts was a significant moderator, and the researchers concluded that R/
S psychotherapy may be particularly effective when clients “learn to apply their own
religious-spiritual beliefs to their mental health or well-being concerns” (p. 653). In
light of this emerging evidence, scientists began to explore what sorts of clients would
most likely desire and benefit from R/S adaptation in psychotherapy.
The chapter on R/S (Worthington et al., 2011) from the second edition of
Psychotherapy Relationships that Work identified 46 randomized outcome studies of R/
S-adapted treatment and conducted the most comprehensive meta-analysis to that date,
testing the influence of R/S adaptation on both psychological and spiritual outcomes.
R/S accommodations were found to be more effective than secular treatments in
improving psychological health compared to alternate (d = .26) conditions, and
equally effective to secular psychotherapy when controlling for theoretical orientation
and treatment duration (d = .13). In situations where patients’ goals extended beyond
symptom reduction, R/S psychotherapies offered distinct benefits including greater
spiritual well-being and increased connection with the divine, compared to alternate
Table 8.1. Previous Meta-Analyses of R/S-Accommodated Intervention
Meta-analysis k N Study Designs Inclusion Criteria Comparison
Group(s)
McCullough 5 111 Randomized Christian Standard
(1999) accommodative CBT
CBT

Smith et al. Randomized, R/S Control,

(2007) quasi- accommodative alternate,
experimental, psychotherapy and single
and group
naturalistic pre-post
(analyzed
together)

Worthington 46 3,290 Randomized R/S Control,

et al. accommodative alternate,
( 2011 ) psychotherapy and
additive
(analyzed
separately)
Presenting Psychological Spiritual Biological Significant Key Findings
Problem Outcomes Outcomes Outcomes Moderators
Depression d = .18 NA NA None tested Christian-
accommodated CBT
is similarly effective to
standard CBT.
Various d = .56 (control, NA NA Explicitly R/ S-accommodated
mental alternate, and teaching psychotherapy is at
and single group spiritual least as effective as
physical pre-post concepts traditional treatment.
health combined) (d = .69); Symptom reduction
concerns meditation may be heightened
(d = 32) when clients "learn
to apply their own
religious-spiritual
beliefs to their mental
health” (p. 653).
Various d’s = .45 ds = .51 NA None significant R/ S-accommodated
mental (control), .26 (control), psychotherapy is
and (alternate), .41 equally effective with
physical and .13 (alternate), an identical secular
health (additive) and .33 psychotherapy in
concerns (additive) reducing psychological
distress and more
effective in facilitating
spiritual well-being
at posttreatment and
follow-up.

(continued)
T a b le 8 .1 . Continued

Meta-analysis k N Study Designs Inclusion Criteria Comparison Presenting

Group (s) Problem
Oh & Kim 21 1,411 Randomized Broadly defined R/ NR NR
(2012) S intervention

Anderson 16 1,197 Randomized R/S Control and Depression

et al. accommodative standard
(2015) CBT CBT for
Christian,
spiritual,
and
Muslim
CBT
(analyzed
separately)
Psychological Spiritual Biological Significant Key Findings
Outcomes Outcomes Outcomes Moderators
d s ranging from ds ranging ds ranging None tested R/S interventions
.65 to .76 from .65 from (including but
to .76 .39 to not limited to
.51 psychotherapy) are
effective in relieving
psychological distress
and promoting
spiritual well­
being, aiding pain
management, and
increasing physical
functionality.
R/S vs. control NA NA None tested CBT accommodated to
ds = 1.47 Christian, Muslim,
(Christian), and general spiritual
.48 (spiritual), worldviews is more
and .30 effective than standard
(Muslim) CBT in reducing
R/S vs. standard depression; however,
CBT ds = .59 small sample sizes
(Christian), and researchers' R/
.55 (spiritual), S allegiance may bias
and .31 these findings.
(Muslim)
Gonçalves 23 2,721 Randomized Broadly defined R/ Control, Various Anxiety d = .43; NA NA For anxiety: R/S interventions
et al. S intervention alternate mental depression formal (including
(2015) (analyzed and d = .09 psychotherapy psychotherapy and
together) physical (d = .35); meditation) are
of anxiety health meditation more effective in
and concerns (d = .43) reducing anxiety than
depression secular therapeutic,
educational, and
waitlist controls.

Note. Meta-analyses are listed in order of year of publication. R/S = religious or spiritual; k = number of studies meta-analyzed; N = sample size across all studies; d = Cohen’s d effect size;
NA = not applicable; NR = not reported; CBT = cognitive-behavioral therapy.
 224 PSY C H O TH E RA PY RELATIONSHIPS THAT WORK

(d = .41) treatments. This effect persisted significantly when controlling for theoretical
orientation and treatment duration (d = .33).
Since 2011, meta-analyses of R/S accommodations have become increasingly more
specialized. For example, 21 outcome studies of broadly defined R/S intervention
were meta-analyzed considering not only psychological and spiritual but also biolog­
ical outcomes, including pain level and functional status (Oh & Kim, 2012). These
researchers reported a moderate effect of R/S intervention on psychological and spir­
itual outcomes (ds ranging from .65 to .76), and a slightly smaller effect for biological
measures (ds ranging from .39 to .51). Another recent meta-analysis looked at depres­
sion and anxiety outcomes separately in 23 randomized studies of broadly defined R/S
intervention, including psychotherapy, pastoral services, meditation, and audiovisual
resources (Gonsalves et al., 2015). The researchers analyzed secular therapeutic, edu­
cational, and waitlist control studies together, resulting in a medium effect (d = .43) of
R/S intervention on anxiety. Intervention type moderated this effect (psychotherapy
d = .35; meditation d = .43). Considering depression, R/S intervention was similarly
effective to secular therapeutic, educational, and waitlist controls (d = .09). O f note,
both of these meta-analyses used loosely defined criteria for R/S intervention, which
may have inflated effect sizes. Further, analyzing all studies together (e.g., those with a
comparison treatment vs. control group) makes it difficult to draw clear conclusions.
A meta-analysis of R/S-tailored CBT in the treatment of depression and anx­
iety examined 16 randomized controlled trials (Anderson et al., 2015). Studies were
analyzed separately based on Christian, general spiritual, and Muslim accommodative
focus. In the treatment of depression, R/S CBT displayed significant benefits across all
groups when compared with standard CBT (ds = .59, .55, and .31, respectively). Due
to insufficient data, it was not possible to meta-analyze studies of R/S adaptations for
anxiety; however, R/S CBT for anxiety largely outperformed control conditions and in
one study outperformed standard CBT.
Research consistently demonstrates the importance of attending to patients’ R/S
within psychotherapy, but much remains to be clarified. The majority of meta-analyses
to date have only measured psychological outcomes, but R/S integration in treatment
may impact patients’ lives and experience of the sacred in significant ways over and
above reducing clinical distress, such as improving quality of life and spiritual well­
being. Furthermore, few studies have accounted for clients’ level of religious commit­
ment, which could impact the effectiveness of accommodations.
For this meta-analysis, we aimed to provide a comprehensive evaluation of the ef­
ficacy of R/S adaptations in psychological treatment by utilizing all available studies
from 1980 to 2017. In addition to updating the literature review, we built on prior
meta-analytic reviews of the literature in several key ways. First, we assessed outcomes
at posttest and follow-up for both psychological and spiritual outcomes. Second,
we evaluated comparisons with (a) no-treatment control conditions, (b) alternate
treatments, and (c) additive treatments that used identical theoretical orientation and
therapy duration to isolate the impact of R/S accommodation. Third, we analyzed both
experimental and quasi-experimental designs and evaluated the evidence for causality.
Fourth, we examined a number of patient, study, and treatment characteristics that
 225 Religion and Spirituality

may moderate established effects of R/S intervention. Finally, we aimed to explore the
role of client religiosity as a key factor in determining response to treatment.

META-ANALYTIC REVIEW

Inclusion Criteria
In the present meta-analysis, we included outcome studies of psychotherapy broadly
defined (Norcross, 1990), published in the English language, which explicitly inte­
grated R/S throughout the psychotherapeutic process, either through incorporation
of R/S content within a secular technique (e.g., Christian cognitive therapy) or the
addition of R/S practices (e.g., prayer, meditation, reading sacred texts) as an adjunc­
tive to sessions. Additionally, all studies that we considered for inclusion compared
an R/S-accommodated treatment with either (a) a no-treatment control condition or
(b) an alternate treatment. Although the vast majority of studies used an experimental
(randomized) research design, a small group of studies employed a quasi-experimental
(nonrandomized) design due to limitations arising from the setting.
We excluded studies of peer-led support groups (e.g., Alcoholics Anonymous,
Celebrate Recovery) as well as stand-alone self-help interventions (e.g., meditation,
mindfulness, intercessory prayer) that were not contextualized within psycholog­
ical treatment. Because we were interested primarily in psychological and spiritual
outcomes of psychotherapy, we did not consider studies that examined physical health
as the primary outcome measure. However, we did include outcome studies in which
psychological intervention was provided to individuals with a medical problem (e.g.,
cancer, hypertension) who sought psychotherapy for associated psychological (e.g.,
depression, anxiety) or spiritual (e.g., meaningless, feeling far from God) difficulties.

Literature Search
We identified studies for analysis using a comprehensive approach of both back­
ward and forward search. First, we identified relevant psychotherapy outcome studies
through database searches of PsycINFO, PsycArticles, Psychology and Behavioral
Sciences Collection, SocINDEX, and Dissertation Abstracts International, as of May
15, 2017, using the key terms [counseling OR therapy] AND [religio* OR spiritu*]
to define our search criteria. Second, we used previous meta-analyses and systematic
reviews of the literature to identify additional outcome studies that met the aforemen­
tioned criteria. Finally, we reviewed all issues to date of Spirituality in Clinical Practice,
because this American Psychological Association (APA) journal focuses specifically
on R/S oriented interventions.
Subsequently, we contacted the corresponding authors for each study identified
through the previous methods to inquire about additional investigations they had
conducted or were aware of, including unpublished file-drawer studies. Compared
with findings supporting the null hypothesis, findings that support differences in
treatments have been found to be several times more likely to be published (Coursol
 226 PSY C H O TH E RA PY RELATIONSHIPS THAT WORK

& Wagner, 1986). Furthermore, effect sizes tend to be significantly larger in published
compared with unpublished studies (Lipsey & Wilson, 2001). To minimize the risk of
overestimating population effects and account for publication bias, we included both
published and unpublished findings.

Calculation of Effect Sizes

For each study, we computed the standardized mean difference, commonly known as
Cohen’s d, which expresses the posttest difference between the R/S condition and the
comparison (or control) condition in standard deviation (SD) units. To account for
small sample bias, we report Hedge’s g, which substitutes pooled weighted SD units. For
interpretation purposes, g = .5 indicates that patients in the R/S condition scored, on
average, .5 standard deviations higher than those in the comparison condition. Mean
differences were reversed for negatively valenced outcome measures (e.g., depression,
anxiety) so that positive values of effect sizes indicate more favorable outcomes for the
R/S condition relative to the comparison condition.
Some published studies did not include sufficient data to classify the study and cal­
culate effect sizes. In these situations, we contacted the corresponding author to re­
quest the missing data. In situations where the data were not available, we excluded
these studies from the meta-analyses.
Overall, we identified 102 independent samples from 97 studies with data available
for inclusion. Among these, 45 samples (44 studies) compared R/S psychotherapy to a
control condition only, 43 samples (40 studies) compared R/S psychotherapy to a com­
parison condition only, and 14 samples (13 studies) compared R/S psychotherapy to
both control and comparison conditions. Among those with a comparison condition,
24 samples (23 studies) used an additive design, in which the R/S and comparison
groups employed the same treatment approach and duration (e.g., religious CBT vs.
standard CBT). Considered together, we examined a total of 116 comparisons in our
analyses.
For each study, we extracted the sample size and associated statistical information
(e.g., means, standard deviations) necessary to determine the effect direction and cal­
culate effect sizes (d, g) and standard error. Considering potential moderators, we
coded a number of study, treatment, and patient characteristics. Study characteris­
tics involved data source (e.g., published or unpublished), use of randomization, time
lapsed to follow-up data collection point, outcomes measured (e.g., religious, spir­
itual), and whether or not the study employed an additive design, as outlined earlier.
Treatment characteristics included therapeutic approach (e.g., cognitive-behavioral,
psychospiritual), intervention modality (e.g., group, individual, couple), accommoda­
tive focus (e.g., Christian, Muslim, general spirituality), number of sessions, use of a
treatment manual, and psychotherapy fidelity checks. Patient characteristics included
age, gender, race/ethnicity, presenting problem, use of psychotropic medication, and
religious affiliation. More specific nuances of R/S, such as religious commitment,
attendance at religious services, and use of spiritual practices (e.g., reading, prayer,
 227 Religion and Spirituality

meditation), were only measured in a handful of studies and, as such, were not in­
cluded in moderator analyses.

Data Analysis
We used Comprehensive Meta-Analysis Version 3.0 (Borenstein et al., 2012). Random
effects models were used because we had no reason to believe that the population effect
sizes were invariant. Consistent with random effects models, studies were weighted
by the sum of the inverse sampling variance plus tau-squared. Separate analyses were
conducted for studies utilizing comparison and control conditions, as well as for psy­
chological and spiritual outcomes.
Some studies reported more than one outcome measure, for example, depression,
anxiety, spiritual well-being, and hope. We calculated the effect size for one psycho­
logical outcome and one spiritual outcome that best assessed the goal of the specific
psychotherapy. For example, if a study purported to examine R/S cognitive behavioral
therapy for depression, a measure such as the Beck Depression Inventory-II (Beck et al.,
1996) was chosen to account for psychological outcomes and a measure such as the
Spiritual Well-Being Scale (Paloutzian & Ellison, 1991) was chosen to represent spir­
itual outcomes. In this example, measures of anxiety and hope were not considered. In
this decision-making process, measures known for their psychometric properties were
chosen over non-peer-reviewed, single-item, or more recently developed measures.
In some cases, where the psychotherapeutic context involved patients coping with a
medical problem (e.g., cancer, HIV, hypertension, infertility), we extracted the primary
psychological and (if available) spiritual outcome measured. We did not consider or
code studies that primarily assessed physical health outcomes.

The Studies and Patients

We analyzed data from 7,181 patients (3,495 from R/S interventions, 1,634 from al­
ternate interventions, and 2,052 from no-ireatment/control conditions), which was
gathered from 102 independent samples. Descriptive information for all studies is
summarized in Table 8.2. Most participants were diagnosed with a mental health dis­
order, including depression (k = 26), anxiety (k = 10), trauma/posttraumatic stress
disorder (k = 7), schizophrenia (k = 3), substance use (k = 4), or an eating disorder
(k = 2). Other participants received psychotherapy targeting psychological symptom
distress and/or spiritual well-being, having been diagnosed with cancer (k = 14), HIV
(k = 3), heart disease (k = 3), infertility (k = 2), or another medical condition. Still other
participants reported premarital/marital conflicts (k = 5), spiritual problems (k = 3),
difficulties with forgiveness (k = 4), aging (k = 2), stress/burnout (k = 2), or similar
challenges in life adjustment.
Across studies, a number of diverse R/S perspectives were integrated in psycho­
therapy, with the majority being Christianity (k = 28), Islam (k = 18), and general
spirituality (k = 51). A handful of studies explicitly integrated Buddhist (k = 3),
Table 8.2. Descriptive Information for Studies Included
Study Pub­ Design Random­ NRS N N Belief Problem
lished ized Alt Ctl

Afazel et al. Y c Y 45 NA 45 M uslim Kidney failure

(2013)
Akuchekian Y c Y 45 NA 45 M uslim OCD
et al.
(2011)
Andrews N A Y 13 12 NA Christian Infertility
(2013)
Ano et al. Y c Y 25 NA 25 Christian Spiritual
(2017) issues
A m iento Y c Y 25 25 NA G eneral Depression
et al.
(2012)
Azhar & Y c Y 15 15 NA M uslim Depression
Varma
(1995a)
Azhar & Y c Y 32 32 NA M uslim Depression
Varma
(1995b)
Azhar et al. Y c Y 31 31 NA M uslim A nxiety
(1994)
Baker (2012) N c N 7 10 NA G eneral Depression
B a k e r(2000) Y c Y 47 NA 47 G eneral Depression
A pproach Treatm ent P sychological Spiritual P sychological Spiritual
M ode(s) O utcom e O utcom e O utcom e O utcom e
g (vs. Alt) g (vs. Alt) g (vs. Ctl) g (vs. Ctl)

Psychospiritual Individual NA NA .82 NA

Cognitive- NR NA NA .78 NA
behavioral

Cognitive- Group .40/.53 .24/.08 NA NA

behavioral
Psychospiritual Group NA NA .42 .50

Behavior Individual .33/34 .06 NA NA

activation

Cognitive- Individual .73 NA NA NA

behavioral

Cognitive- Individual .27 NA NA NA

behavioral

Cognitive- Individual .27 NA NA NA

behavioral
Psychospiritual Group .47 .43 NA NA
Pastoral care Individual NA NA .00/.00 .52/.00
Barron N A Y 20 19 NA General
(2007)
Bay et al. Y C Y 85 NA 85 G eneral
(2008)
Beheshtipour Y C Y 65 NA 70 G eneral
et al.
(2016)
Borm ann Y C Y 71 NA 75 G eneral
et al.
(2013)
Borm ann Y C Y 66 NA 70 G eneral
et al.
(2012)
Borm ann Y C Y 36 35 NA G eneral
et al.
(2009)
Borm ann Y C Y 14 NA 15 G eneral
et al.
(2008)
Borm ann Y C Y 37 37 NA G eneral
et al.
(2006)
Bowland Y C Y 21 NA 22 G eneral
(2012)
Depression Cognitive- Group .71 .68 NA NA
behavioral
Heart disease Pastoral care Individual NA NA .21/.14 .13/.20

C ancer Psychospiritual Group NA NA 1.73/1.73 NA

PTSD M ind-body Group NA NA .26 .74

PTSD M ind-body Group NA NA .45 1.11

H IV M ind-body Group -.04/.28 .47/.48 NA NA

PTSD M ind-body Group NA NA .70 .65

H IV M ind-body Group .25 .31 NA NA

PTSD Psychospiritual Group NA NA .55 .88

( c o n t in u e d )
Table 8.2. Continued

Study Pub­ Design Random­ NRS N N Belief Problem

lished ized Alt Ctl

Breitbart Y c Y 67 58 NA G eneral C ancer

et al.
(2015)
Breitbart Y c Y 40 37 NA G eneral C ancer
et al.
(2012)
Chan et al. Y c Y 92 NA 93 G eneral Depression,
(2017) insom nia
Chan et al. Y c Y 141 NA 110 G eneral In-vitro
(2012) fertilization
Chan et al. Y c Y 27 16 17 G eneral C ancer
(2006)
Chida et al. Y c Y 66 NA 65 Buddhist Depression
(2016)
Cole (2005) Y c Y 9 NA 7 G eneral C ancer
Com bs et al. Y c Y 30 NA 32 Christian M arital
(2000)
Davis & Hill Y A Y 29 22 NA G eneral Dreams
(2005)
Dik et al. Y A Y 57 42 44 Christian Career
(2015)
A pproach Treatm ent P sychological Spiritual P sychological Spiritual
M ode(s) O utcom e O utcom e O utcom e O utcom e
g (vs. Alt) g (vs. Alt) g (vs. Ctl) g (vs. Ctl)

Existential Group -.1 5 / -.2 5 -.1 5 / -.1 9 NA NA

Existential Individual .45/—.01 .26/.08 NA NA

Body-m ind- Group NA NA .37/.44 3.08/2.26

spirit
Body-m ind- Group NA NA .37/.29 .32/.24
spirit
Body-m ind- Group .68/.52 NA -.0 6 / -.0 2 NA
spirit
Psychospiritual Group NA NA 3.84/4.21 NA

Psychospiritual Group NA NA -.49/ .47 NA

Cognitive- Group NA NA .88 NA
behavioral
Cognitive- Individual NA .38/-.05 NA NA
experiential
Person- Group -.2 4 .44 .30 .15
environment
Djuric et al. Y C Y 11 11 N G eneral W eight loss Psychospiritual Individual .19 1.00 NA NA
(2009)
Ebrahim i Y C Y 27 NA 24 M uslim Depression, Psychospiritual Group NA NA 1.77 NA
et al. suicidality
(2014)
Ebrahim i Y A Y 16 16 15 M uslim Depression Cognitive- Individual .31/.42 NA 3.07/2.75 NA
et al. behavioral
(2013)
Em ery N A Y 11 7 8 G eneral Aging R em iniscence Group .17/1.15 .59/1.13 .00 .91
(2002)
Fauver N C Y 17 NA 17 G eneral C ancer Psychospiritual Group NA NA .11 .68
(2011)/
Guthrey
(2015)
Ford (2016) N A Y 36 42 NA Christian Depression M ind-body Group .47 .42 NA NA
Ghodsbin Y C Y 38 NA 36 G eneral Heart disease Positive Group NA NA NA .18/.22
et al. Psychology
(2015)
G ibbel N A Y 24 19 22 G eneral Depression Cognitive Individual .55/.16 .16/.09 .61/.27 .30/-.17
(2010)
Hart & N C Y 28 26 NA G eneral Unforgive­ Psychospiritual Group .77/.59 NA NA NA
Shapiro ness
(2002)
Hawkins N C N 52 NA 52 Christian M arital M arriage Group NA NA .17 NA
(2016) enrichm ent

( c o n tin u e d )
Table 8.2. Continued

Study Pub­ Design Random ­ N RS N N Belief Problem Approach Treatment Psychological Spiritual Psychological Spiritual
lished ized Alt Ctl Mode(s) Outcome Outcome Outcome Outcome
g (vs. Alt) g (vs. Alt) g (vs. Ctl) g (vs. Ctl)
Hawkins Y A N 18 11 NA Christian Depression Cognitive- Individual .47 1.25 NA NA
et al. behavioral and
(1999) group
Ho et al. Y C Y 26 33 NA G eneral C ancer Body-m ind- Group .09 .76 NA NA
(2009) spirit
Hosseini Y C Y 33 NA 33 M uslim Pre-surgery Psychospiritual Group NA NA 5.68 NA
et al. anxiety
(2013)
Huguelet Y C Y 42 NA 42 G eneral Schizophrenia Spiritual Individual NA NA .13 NA
et al. assessment
(2011)
Iler (2001) Y C Y 25 NA 24 G eneral Pulm onary Pastoral care Individual NA NA .60 NA
disease
Jackson N C Y 14 NA 13 Christian Unforgive­ Prom ote Group NA NA .88 NA
(1999) ness empathy
Jafari et al. Y C Y 34 NA 31 M uslim C ancer Psychospiritual Group NA NA 1.60 1.92
(2013)
Johnson et al. Y A Y 13 16 NA Christian Depression Rational­ Individual -.52/.75 .86/1.22 NA NA
(1994) emotive
Johnson & Y A Y 5 5 NA Christian Depression
Ridley
(1992)
Kamsani N C Y 46 NA 48 M uslim Orphaned
(2014)
Koenig et al. Y A Y 47 48 NA Christian Depression
(2016,
2015)
Koszycki Y C Y 11 9 NA G eneral A nxiety
et al.
(2014)
Koszycki Y C Y 9 9 NA G eneral A nxiety
et al.
(2010)
Lampton Y C N 42 NA 23 Christian Unforgive­
et al. ness
(2005)
Lee et al. Y c Y 69 NA 79 G eneral C ancer
(2009)
Liu et al. Y c Y 12 NA 16 G eneral C ancer
(2008)
Margolin Y c N 14 11 NA Buddhist Substance
et al. use, H IV
(2007)
Rational- Individual .29 .69 NA NA
emotive

Psychospiritual Group NA NA .57 .64

Cognitive- Individual .14/-.06 .16/.40 NA NA

behavioral

Psychospiritual Individual 1.50/.92 1.25/1.39 NA NA

Psychospiritual Individual .10/-.07 NA NA NA

REACH Group NA NA .93 NA

forgiveness

Body-m ind- Group NA NA .22 1.22

spirit
Body-m ind- Group NA NA .64 NA
spirit
Spiritual Individual .25 .22 NA NA
self-schem a

( c o n t in u e d )
Table 8.2. Continued

Study Pub­ Design Random ­ N RS N N Belief Problem Approach Treatment Psychological Spiritual Psychological Spiritual
lished ized Alt Ctl Mode(s) Outcome Outcome Outcome Outcome
g (vs. Alt) g (vs. Alt) g (vs. Ctl) g (vs. Ctl)
Margolin Y C Y 30 30 NA Buddhist Substance Spiritual Individual .63 1.20 NA NA
et al. use, H IV self-schem a and
(2006) group
M cCain et al. Y C Y 68 65 57 G eneral Stress, H IV Spiritual Group .24/.20 NA —1.58/ —1.63 NA
(2008) growth
M cCauley Y C Y 51 49 NA G eneral Chronic Spiritual Individual .39 NA NA NA
et al. illness coping
(2011)
M iller et al. Y C Y 27 27 NA G eneral Substance use Spiritual Individual -.4 0 / -.2 8 -.31/ -.2 6 NA NA
(2008)1 guidance
M iller et al. Y C Y 31 34 NA G eneral Substance use Spiritual Individual .16/.01 -.0 9 / -.1 8 NA NA
(2008)2 guidance
M oeini et al. Y C Y 26 NA 26 M uslim Hypertension Psychospiritual Group NA NA NA 1.93/2.19
(2016)
Musarezaie Y C Y 32 NA 32 M uslim C ancer Spiritual Individual NA NA 1.13 .26
et al. support
(2015,
2014)
Nedderman Y C N 20 NA 19 Christian Hopelessness Psychospiritual Group NA NA .10 NA
et al.
(2010)
N ohr (2001) N A Y 35 23 14 G eneral Stress

O lson et al. Y C N 32 NA 29 Christian Spiritual

(2016) issues
O m an & Y C Y 63 NA 69 G eneral PTSD
Borm ann
(2015)
O m an et al. Y c Y 27 N 31 G eneral Stress/
(2006) A nxiety
Pecheur & Y A Y 7 7 7 Christian Depression
Edwards
(1984)
Piderm an Y c Y 54 NA 63 G eneral C ancer
et al.
(2014)
Propst (1980) Y A Y 7 10 11 Christian Depression
Propst et al. Y A Y 10 9 11 Christian Depression
(1992)1
Propst et al. Y A Y 9 10 11 Christian Depression
(1992)2
Rasar (2012) N C Y 11 NA 9 Christian Spiritual
issues
Razali et al. Y C Y 45 40 NA Muslim A nxiety
(2002), 1
Cognitive- Group .02/.09 .1 1/.03 .29 .04
behavioral
G od image Group NA NA NA .35
narrative
M ind-body Group NA NA .52 NA

M ind-body Group NA NA .62/.83 NA

Cognitive Individual .53A52 NA 1.93 NA

Q uality o f life Group NA NA NA -.02/-

Cognitive Group .00/.82 NA .90/.80 NA

Cognitive- Individual -.2 8 / -.8 0 NA .90 NA
behavioral
Cognitive- Individual 1.37/.23 NA 1.41 NA
behavioral
G od image Group NA NA NA .10
narrative
Cognitive Individual -.3 5 NA NA NA

( c o n t in u e d )
Table 8.2. Continued

Study Pub­ Design Random­ NRS N N Belief Problem

lished ized Alt Ctl

Razali et al. Y c Y 42 38 NA Muslim A nxiety

(2002), 2
Razali et al. Y c Y 54 49 NA Muslim A nxiety
(1998), 1
Razali et al. Y c Y 52 48 NA Muslim Depression
(1998), 2
Renatala Y c Y 56 64 NA G eneral Depression
et al.
(2015)
Richards Y c Y 43 35 NA G eneral Eating
et al. disorders
(2006)
Ripley et al. Y A Y 30 24 NA Christian M arital
(2014)
Roland N c N 65 63 NA G eneral Depression
(2014)
Rosm arin N c Y 36 42 47 Jewish A nxiety
et al.
(2010)
Rye et al. Y A Y 50 49 50 Christian Unforgive­
(2005) ness
A pproach Treatm ent P sychological Spiritual P sychological Spiritual
M ode(s) O utcom e O utcom e O utcom e O utcom e
g (vs. Alt) g (vs. Alt) g (vs. Ctl) g (vs. Ctl)

Cognitive Individual .12 NA NA NA

Cognitive Individual .31 NA NA NA

Cognitive Individual .32 NA NA NA

Body-m ind- Group M l . 93 NA NA NA

spirit

Psychospiritual Group .51 .61 NA NA

H ope-focused Couple -.2 1 / -.2 7 NA NA NA

Cognitive- Individual 3.49 NA NA NA

behavioral
Cognitive- Individual .221.48 .631.68 .441.99 .447.41
behavioral

REACH Group -.03/.02 NA . 211.21 NA

forgiveness
Rye & Y A Y 19 20 19 Christian Unforgive­ REACH Group .35/.64 .14/. 18 1.47/1.35 .40/.64
Pargament ness forgiveness
(2002)
Scott (2001) N A N 15 3 NA Christian C ancer Cognitive- Group .20 NA NA NA
behavioral
Shafiee et al. Y C Y 32 32 NA Muslim Depression Psychospiritual Group 1.24/1.09 NA NA NA
(2016)
Sreevani Y C Y 15 15 NA G eneral Depression Body-m ind- Group NA NA .92/1.36 1.85/2.
et al. spirit
(2013)
Stalsett et al. Y A N 50 50 NA G eneral Depression, Existential- Group .54/.87 NA NA NA
(2012) personality narrative-
disorders dynamic
Stratton et al. Y C N 22 NA 29 Christian Unforgive­ REACH Group NA NA .09/-.01 NA
(2008) ness forgiveness
Tadwalkar Y C N 13 6 NA G eneral Heart failure Chaplain care Individual -.1 0 .89 NA NA
et al.
(2014)
Targ & Y C Y 72 60 NA G eneral C ancer Body-m ind- Group .14 .01 NA NA
Levine spirit
(2002)
Toh & Tan Y C Y 22 NA 24 Christian Various Lay counseling Individual NA NA .70 .59
(1997)
Tonkin N A Y 9 9 NA Christian Eating Cognitive- Group -1.91 .01 NA NA
(2005) disorders behavioral
Trathen N C Y 23 NA 22 Christian Prem arital PREP Couple NA NA .04 NA
(1995)1
(continued )
Table 8.2. Continued
Study Pub­ Design Random ­ N RS N N Belief Problem Approach Treatment Psychological Spiritual Psychological Spiritual
lished ized Alt Ctl Mode(s) Outcome Outcome Outcome Outcome
g (vs. Alt) g (vs. Alt) g (vs. Ctl) g (vs. Ctl)
Trathen N C Y 23 NA 22 Christian Prem arital PREP Couple NA NA .10 NA
(1995)2
Wahass Y C Y 3 NA 3 M uslim Schizophrenia Cognitive- Individual NA NA 2.00 NA
& Kent behavioral
(1997)
Wang et al. Y C Y 38 NA 20 G eneral Trafficked Trauma- Individual NA NA .51 .50
(2016) focused
cognitive-
behavioral
W eisman de Y C Y 25 21 NA G eneral Schizophrenia Culturally- Family .62 NA NA NA
M am ani informed
et al.
(2014)
Wu & Koo Y C Y 53 NA 50 G eneral Dem entia R em iniscence Group NA NA .42 .45
(2016)
Yong et al. Y C Y 24 NA 27 G eneral Burnout Psychospiritual Group NA NA .28 1.16
(2011)
Zhang et al. Y C Y 46 48 NA Taoist A nxiety Cognitive Individual .85 NA NA NA
(2002)

Note. RS = religious or spiritual psychotherapy; Y = Yes; N = No; C = comparative design; A = additive design; NA = not applicable; R = religious; S = spiritual; NR = not reported; g (vs.
Alt) = effect size when compared with alternate treatment; g (vs. Ctl) = Hedge’s g effect size when compared with control group; g/g = Hedge’s g effect size at end of therapy vs. at follow-up;
PTSD = posttraumatic stress disorder.
 239 Religion and Spirituality

Jewish (k = 1), and Taoist (k = 1) perspectives. Treatment was provided in individual

(k = 38), group (k = 57), individual + group (k = 2), and couple/family (k = 4) formats.
Psychotherapists utilized a variety of approaches, and there was significant variation in
the level of detail provided. Some treatments were clearly delineated, including cogni­
tive, cognitive-behavioral, behavioral activation, rational emotive behavior therapy, ex­
istential, narrative, and body-mind-spirit approaches. Other articles spoke broadly of
psychospiritual, spiritually integrated, or supportive psychotherapy, with less attention
to the specifics of the intervention itself. We categorized therapeutic approaches in six
broad areas: cognitive and/or behavioral (k = 33), existential and/or narrative (k = 7),
general psychospiritual (k = 33), mind-body (k = 17), REACH forgiveness (k = 4), and
supportive and/or pastoral (k = 8). Follow-up time ranged from 1 to 6 months for R/
S treatment-control studies (M = 3.10; SD = 2.28), and from 1 to 12 months for R/S
treatment-alternate studies (M = 3.10; SD = 2.28).

Results
Omnibus meta-analytic results for psychological and spiritual outcomes are summarized
in Table 8.3 (quasi-experimental and randomized studies) and Table 8.4 (randomized
studies only). These tables report separate effect sizes by outcome type (psychological
or spiritual), timing of outcome assessment (posttreatment or follow-up), and compar­
ison type (waitlist controls, alternative-treatment comparisons, or additive-treatment
comparisons). For clarity, results discussed here focus on randomized studies.

R/S Treatments Versus No-Treatment Controls

First, we examined whether or not patients who received R/S integrated treatment
showed greater improvement compared with patients in a no-treatment control

Table 8 .3 . Results for Psychological and Spiritual O utcom es, Including B oth R andom ized
and Q uasi-E xperim ental Studies

Posttest Follow-up
Comparison N k g 95% CI i2 N K G 95% CI I2
Psychological Outcomes

Control 3,923 54 .70 .50-.91 89.31 1,573 18 .76 .2 8 -1 .2 4 94.84

Alternate 2,343 34 .32 .2 0 -.4 5 56.97 896 14 .33 .0 9 -.5 7 66.92
Additive 1,080 23 .32 - .0 5 -.6 8 87.58 565 15 .30 .0 5 -.5 5 5.05
Spiritual Outcomes

Control 2,434 30 .72 .4 7 -.9 7 87.99 1,112 11 .71 .20-1.21 93.56

Alternate 877 16 .43 .2 2 -.6 4 55.97 404 7 .21 - .1 4 -.5 6 66.98
Additive 630 14 .37 .2 2 -.5 3 0 268 8 .32 .0 2 -.6 2 3.70

Note. N = sample size summed across studies; k = number of effect sizes summarized; g = Hedge’s g, a
measure of effect size, which corrects for potential bias in Cohen’s d; 95% CI = confidence interval for
g; I2 = percentage of the observed variance that reflects real differences in effect sizes.
 240 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

Table 8 .4 . Results for Psychological and Spiritual O utcom es, Including Random ized
Studies O nly

Posttest Follow- up
Comparison N k g 95% CI I2 N k G 95% CI I2
Psychological Outcomes

Control 3,664 50 .74 .5 2 -.9 6 89.92 1,522 17 .81 .30-1.31 95.10

Alternate 2,283 31 .33 .2 0 -.4 7 6.49 896 14 .33 .0 9 -.5 7 66.92
Additive 805 19 .13 - .0 9 -.3 4 53.79 465 14 .22 - .0 1 -.4 4 3.33
Spiritual Outcomes

Control 2,373 29 .74 .4 8 -.9 9 88.34 1,112 11 .71 .20-1.21 93.56

Alternate 817 13 .43 .1 9 -.6 6 63.45 404 7 .21 - .1 4 -.5 6 66.98
Additive 601 13 .34 .1 8 -.5 0 0 268 8 .32 .0 2 -.6 2 3.70

Note. N = sample size summed across studies; k = number of effect sizes summarized; g = Hedge’s g,
a measure of effect size, which corrects for potential bias in Cohen’s d; CI = confidence interval for g;
I2 = percentage of the observed variance that reflects real differences in effect sizes.

condition. Within randomized studies, R/S-adapted psychotherapy outperformed no­

treatment control conditions on both psychological (d = .75, g = .74) and spiritual (d =
.75, g = .74) outcomes. These gains were similar at follow-up (d = .82 and .72; g = .81 and
.71, respectively). Treated participants were better off than control patients by about .7
standard deviations on average for both sets of outcomes—a large effect that is typical of
treatment-control effect sizes for many forms of psychotherapy (Wampold & Imel, 2015).

R/S Treatments Versus Alternate Treatments

Second, we examined whether patients who received R/S integrated treatment dis­
played greater improvement on psychological and spiritual outcomes compared with
patients receiving any alternate (e.g., secular) form of psychotherapy. Studies that used
an identical theoretical orientation and therapy duration to isolate the impact of R/
S accommodation were examined separately (see next section). Within randomized
studies, R/S-adapted psychotherapy outperformed alternate treatments on both psy­
chological (d = .33, g = .33) and spiritual (d = .43, g = .43) measures. Psychological but
not spiritual gains were maintained at follow-up to the level of statistical significance
(ds = .34 and .21; gs = .33 and .21, respectively). At termination, participants who re­
ceived R/S accommodative psychotherapy were better off than those who received an
alternate secular treatment by about .3 standard deviations on average for both sets of
outcomes, which is a small-medium effect. At follow-up, these participants continued
to report less psychological symptoms by about .3 standard deviations.

R/S Treatments Versus Alternate Treatments

That Used an Additive Design
Third, we examined studies in which the R/S and alternate (e.g., secular) psychotherapy
conditions used the exact same theoretical approach and treatment duration. The goal
 241 Religion and Spirituality

here was to isolate the additive effects of R/S-specific intervention elements. Within
randomized studies, there was not a significant effect of R/S integration on psycholog­
ical outcomes directly following treatment (d = .13, g = .13) or at follow-up (d = .22,
g = .22). This means that R/S-accommodated treatments were as effective, but not more
effective, than standard psychotherapy approaches. However, R/S-adapted psycho­
therapy did outperform standard psychotherapy on spiritual outcomes, both directly
following treatment (d = .34, g = .34) and at follow-up (d = .33, g = .33). Participants
who received R/S accommodative psychotherapy reported greater spiritual well-being
by about .3 standard deviations on average, which is a small-medium effect.
In summary, these omnibus meta-analytic results provide substantial empirical
support for incorporating clients’ R/S into psychological treatment. Consistent with
previous meta-analyses, R/S-adapted psychotherapy resulted in greater improvement
in patients’ psychological and spiritual functioning, compared with no-treatment
conditions and alternative non-R/S psychotherapies. With more stringent criteria, R/S
treatments were equivalent to secular treatments on psychological outcomes and were
superior to secular treatments on spiritual outcomes, both at posttest and follow-up.
O f note, results at follow-up should be interpreted more cautiously in light of fewer
studies including this data, resulting in a smaller analyzable sample.

Publication Bias
We conducted a series of analyses to determine whether our results were affected by
publication bias. Publication bias refers to the tendency for studies available for inclu­
sion in a meta-analysis to be systematically different from studies that were unavail­
able, such that conclusions may be inaccurate. We conducted fail-safe N analyses in two
ways: One procedure estimates how many missing studies with a mean effect of zero
would need to be added to the present meta-analyses to negate findings (Rosenthal,
1979), and the other procedure (Orwin, 1983) estimates how many missing studies
with a trivial standardized difference in means (which we defined as d < .10) would
need to be added to the present meta-analyses in order to bring the overall effect size
to the same trivial level.
These fail-safe N results (see Table 8.5) suggest minimal impact of publication bias,
with the exception of follow-up analyses, which are more modestly prone to bias be­
cause of small sample sizes. If all existent file-drawer studies were retrievable, effect
sizes for follow-up analyses of spiritual outcomes, in particular, could be weakened or
become nonsignificant (Borenstein et al., 2009). The R/S treatment-alternate and R/S
treatment-additive comparisons of spiritual outcomes at follow-up may have been im ­
pacted by publication bias and should be interpreted with caution.
Additionally, we used the trim and fill procedure (Duval & Tweedie, 2000) to ex­
plore the effects of publication bias. The trim and fill procedure estimates the number
of missing studies on each side of the mean due to publication bias and statistically
imputes these studies, recalculating the overall effect size. Adjustment for overesti­
mation of effects is calculated by imputing studies to the left of the mean, while ad­
justment for underestimation of effects is calculated by imputing studies to the right
 242 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

Table 8 .5 . Results for Fail-Safe N Analyses

Posttest Follow-up
Rosenthal’s K+ z, p for Orwins K+ Rosenthal’s K+ z, p for Orwins K+
Observed Observed
Comparison Studies Studies
Psychological Outcomes

Control 4,133 17.26, .00 224 486 1.36, .00 64

Alternate 460 7.47, .00 69 67 4.71, .00 32
Additive 122 4.91, .00 57 32 3.45, .00 30
Spiritual Outcomes

Control 1,851 15.52, .00 166 224 9.05, .00 46

Alternate 136 6.02, .00 47 0 1.95, .05 4
Additive 71 4.81, .00 39 7 2.64, .01 17

Note. Rosenthal’s K+ = number of missing studies with a mean effect of zero that would need to be
added to the analyses to bring the p value to < .05; z = the overall z-score for observed studies; p = the
p value for observed studies; Orwins K+ = the number of missing studies with a nonexistent or trivial g
(in this case <.10) that would need to be added to the analyses to bring the overall g under .1.

o f th e m e a n . Tables 8 .6 and 8 .7 re p o rt ad ju sted g values created th ro u g h th is p ro c e ­

dure. C o n s id e rin g p o sttest and follow -u p d ata, im p u tatio n s w ere m ad e to a cco u n t
fo r u n d erestim a ted effect sizes o f p sy ch o lo g ica l and sp iritu al o u tco m e effect sizes
w ith in tre a tm e n t-c o n tro l analyses. W ith in p o sttest analyses, a d ju stm en ts w ere m ad e
fo r ov erestim a ted effects o f sp iritu al o u tco m es in altern ate and ad ditive analyses. A t
follow -u p , im p u ta tio n s a cco u n te d fo r ov erestim ated effects in ad ditive analyses o f p sy ­
ch o lo g ica l o u tco m e s.

Table 8.6. Results for Trim and Fill Analyses Evaluating Posttest Data
Adjustmentfor Overestimation of Effects Adjustmentfor Underestimation of Effects
Comparison K+ g adj 95% CI K+ g adj 95% CI
Psychological Outcomes

Control 0 .70 .50-.91 11 .93 .70-1.16

Alternate 0 .32 .19-.45 0 .32 .19-45
Additive 0 .32 -.05-.68 9 .72 .39-1.06
Spiritual Outcomes

Control 0 .72 .47-.97 6 .88 .64-1.12

Alternate 3 .31 .07-.54 0 .43 .22-.64
Additive 2 .31 .13-.49 0 .37 .21-.53

Note. Adjustment for overestimation of effects is calculated by imputing studies to the left of the mean.
Adjustment for underestimation of effects is calculated by imputing studies to the right of the mean.
The K+ is the number of the studies imputed by the trim and fill procedures. The symbol g adj is the
weighted mean g of the distribution of g that contains both the observed and the imputed effects.
 243 Religion and Spirituality

Table 8 .7 . Results for Trim and Fill Analyses Evaluating Follow-Up Data
Adjustmentfor Overestimation of Effects Adjustmentfor Underestimation of Effects
Comparison K+ g adj 95% CI K+ g adj 95% CI
Psychological Outcomes
Control 0 .76 .28-1.24 5 1.19 .63-1.76
Alternate 0 .33 .09-.57 0 .33 .09-.57
Additive 4 .11 -.17-.39 0 .30 .05-.55
Spiritual Outcomes
Control 0 .71 .20-1.21 1 .79 .30-1.29
Alternate 0 .21 -.14-.56 0 .21 -.14-.56
Additive 0 .32 .01-.62 1 .40 .10-.69

Note. Adjustment for overestimation of effects is calculated by imputing studies to the left of the mean.
Adjustment for underestimation of effects is calculated by imputing studies to the right of the mean.
The K+ is the number of the studies imputed by the trim and fill procedures. The symbol d adj is the
weighted mean d of the distribution of d that contains both the observed and the imputed effects.

In summary, while it may be more difficult for studies ofR/S-adapted psychotherapies

with small magnitude or negative results to be published, after statistically estimating
publication bias, our overall conclusions remained mostly unchanged. However,
follow-up analyses, particularly in alternate and additive studies, should be interpreted
with more caution due to the small number of studies in these groups.

MODERATORS
We categorized moderator variables into three groups: (a) patient characteristics,
(b) study characteristics, and (c) treatment characteristics most likely to impact re­
sponse to R/S-tailored treatment. Potential patient and study moderators were tested
on the between-group effect sizes, whereas treatment moderators were tested on the
within-group effect sizes.
Regarding patient characteristics, we examined moderation by age, gender (per­
centage female), race/ethnicity (percentage racial/ethnic minority), religious affiliation
(percentage religious), presenting problem (psychological, health, or spiritual), and
psychotropic medication (prescribed as part of treatment). We tested each moderator
individually on the post R/S treatment-control and post R/S treatment-alternate ef­
fect sizes. In Table 8.8, we report intercepts (B0) and slopes (B t) for race/ethnicity, the
only moderator that emerged as significant. The intercept reflects the estimated effect
size when the value of the moderator is zero (e.g., if the sample consisted solely of
White participants), and the slope quantifies the relative difference in estimated effect
size associated with a 1% increase in racial/ethnic minorities in the sample. Having
more people identifying as racial/ethnic minorities resulted in larger effect sizes for R/
S treatment-control comparisons on psychological (B1 = .009) and spiritual outcomes
(B1 = .007). This pattern was also found in R/S treatment-alternate comparisons, con­
sidering psychological outcomes (B 1 = .006) but not spiritual outcomes.
 244 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Table 8.8. Significant Single-Moderator Analyses

Comparison k B^ B2 95% CI (BJ z(B J P
Treatment vs. Control on Psychological Outcomes Posttreatment

1
O

Ö
Race/Ethnicity (% m inority) 41 .22 .009 2.82 .005

0
Treatment vs. Control on Spiritual Outcomes Posttreatment
Race/Ethnicity (% m inority) 27 .33 .007 .00-.01 2.00 .046
Treatment vs. Alternate on Psychological Outcomes Posttreatment
Race/Ethnicity (% m inority) 43 .00 .006 .00-.01 2.30 .021

Comparison k g 95% CI Q df p(Q) I2

Treatment vs. Control on Psychological Outcomes Posttreatment

Published 11.94 1 .001

No 11a .29 .1 4 -.4 5 1.55 10 .394 5.20
Yes 43b .82 .5 7 -1 .0 7 475.83 42 .000 91.17
Treatment vs. Control on Spiritual Outcomes Posttreatment

Published 3.45 1 .063

No 8a .47 .2 8 -.6 6 5.00 7 .660 .00
Yes 22b .82 .5 0 -1 .1 3 232.29 21 .000 9.96

Comparison k g 95% CI Q df p(Q) I2

Psychological Outcomes Pre-Post R/S Psychotherapy

Accom m odative Focus 32.21 2 .000

Christian 24 .94a .7 1 -1 .1 6 108.43 23 .000 78.79
G eneral Spirituality 48 .78a .6 1 -.9 6 518.61 47 .000 9.94
M uslim 15 3.14b 2 .3 5 -3 .9 4 35.26 14 .000 96.00
Treatm ent Approach 52.09 5 .000
Cognitive Behavioral 32 1.97a 1 .5 7-2.38 493.43 31 .000 93.72
Existential/Narrative 7 .52b .2 0 -.8 4 39.60 6 .000 84.85
G eneral Psychospiritual 24 1.22c .8 6 -1 .5 7 336.21 23 .000 93.16
M ind Body 15 .52b .3 4 -.7 0 74.38 14 .000 81.18
REACH Forgiveness 4 .85b-‘ .4 7 -1 .2 3 1.03 3 .018 7.08
Support ive/Pastoral 7 .56b .2 3 -.8 9 32.63 6 .000 81.61
M odality 39.53 2 .000
Individual 34 1.73a 1.33-2.03 584.98 33 .000 94.36
Group 49 .90b .7 3 -1 .0 8 499.94 48 .000 9.40
Couple or Family 4 .40b .1 8 -.6 2 3.62 3 .306 17.09
M edication Included in Tx 18.72 1 .000
No 78 .91a .7 8 -1 .0 5 70.16 77 .000 89.00
Yes 11 3.53b .3 6 -2 .3 5 371.77 10 .000 97.31
Treatm ent M anual 23.81 1 .000
No 25 2.01a 1 .5 6-2.46 589.09 24 .000 95.93
Yes 60 .83b .6 9 -.9 7 443.13 59 .000 86.69
Treatm ent Fidelity Check 8.01 1 .005
No 67 1.27a 1 .0 7 -1 .4 7 995.99 66 .000 93.37
Yes 20 .84b .6 1 -1 .0 6 11.26 19 .000 82.77
 245 Religion and Spirituality

Table 8.8. Continued

Comparison k g 95% CI Q df p(Q) I2
Spiritual Outcomes Pre-Post R/S Psychotherapy

Treatm ent Approach 9.84 4 .043

Cognitive Behavioral 16 .54a,c .3 5 -.7 2 41.67 15 .000 64.01
Existential/Narrative 5 .40a .1 5 -.6 5 7.75 4 .101 48.36
G eneral Psychospiritual 10 74a_c .4 7 -1 .0 2 4.06 9 .000 77.54
M ind Body 10 .47a .2 9 -.6 6 37.75 9 .000 76.16
Supportive/Pastoral 7 .18b -.0 7 -.4 3 2.29 6 .002 7.43
Treatm ent Fidelity Check 6.04 1 .014
No 35 .42a .3 0 -.5 4 124.25 34 .000 72.64
Yes 13 .64b .5 1 -.7 8 19.48 12 .078 38.41

Note. Univariate analyses used a mixed model (studies random-levels of moderator variables fixed).
k = number of studies; Bo = intercept; B1 = slope; CI = confidence interval; z(Bi) = z statistic for the
slope; g = effect size; Q = homogeneity test; p(Q) = p value for homogeneity test. Q for the moderator
assesses homogeneity between groups; Qs for the levels assess homogeneity within groups. Different
subscripts indicate effect sizes that differ significantly from each other.

Regarding study characteristics, we tested moderation by publication status and

randomization on the post R/S treatment-control and post R/S treatment-alternate
effect sizes. Publication status moderated R/S treatment-control comparisons on psy­
chological outcomes (unpublished g = .29, published g = .82). Publication status did not
moderate R/S treatment-control comparisons for spiritual outcomes or R/S treatment-
alternate comparisons for spiritual or psychological outcomes. Furthermore, random­
ization of study did not moderate effect sizes.
We also explored a number of treatment characteristics that may impact response
to R/S treatment. We tested the following potential treatment moderators: (a) accom­
modative focus (Christian, Muslim, or general spiritual), (b) therapeutic approach
(cognitive and/or behavioral, existential and/or narrative, general psychospiritual,
mind-body, REACH forgiveness, supportive and/or pastoral care), (c) treatment
format (individual, group, couple/family), (d) medication included as part of the treat­
ment, (e) number of sessions, (f) use of a treatment manual, and (g) treatment fidelity
checks. All of these analyses focused on pre-post R/S treatment effect sizes to explore
the impact of these variables within R/S-tailored psychotherapy in particular.
Accommodation focus moderated effect sizes on psychological outcomes.
Accommodation to Islam had the largest effect size (g = 3.14) and varied significantly
from adaptations to Christian beliefs and (g = .94) and general spirituality (g = .78).
(The high effect size for Muslim accommodative studies may be due to the fact that
44% of these studies also included medication as part of the treatment protocol,
compared with 0% of Christian accommodative studies and 6% of general spirituality
accommodative studies. In the subsequent meta-regression controlling for medica­
tion, this difference was no longer significant.) Therapeutic approach also moderated
pre-post changes in both psychological distress and spiritual well-being. Considering
 246 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

psychological functioning, CBT had the largest effect size (g = 1.97) and varied signifi­
cantly from all other approaches. General psychospiritual approaches (g = 1.22) varied
significantly from CBT, supportive/pastoral (g = .56), existential/narrative (g = .52),
and mind-body (g = .52) psychotherapies. Considering spiritual outcomes, general
psychospiritual therapies had the largest effect size (g = .74) and varied significantly
from supportive/pastoral approaches (g = .18). CBT (g = .54) also varied significantly
from supportive/pastoral approaches.
Regarding modality, individual psychotherapy had the largest effect size (g = 1.73)
and differed significantly from group (g = .90) and couple/family (g = .40) therapies for
psychological outcomes. Treatments that did not use a manual showed larger pre-post
changes (g = 2.01) than did studies that used a manual (g = .83). On psychological
outcomes, studies that employed fidelity checks had smaller pre-post changes (g = .84)
compared with studies that did not (g = 1.27). However, the reverse was true for spir­
itual outcomes (fidelity checks g = .64; no fidelity checks g = .42).
In Table 8.9, we summarize analyses examining moderators simultaneously
using meta-regression (Viechtbauer, 2007). This analysis controlled for potential
confounding among moderator variables that emerged as statistically significant and
allowed us to examine unique effects. We dummy-coded polychotomous categorical
moderators, indicating our reference groups as Christian (accommodative focus),
cognitive-behavioral (treatment approach), and individual therapy (modality).
The significant moderators in the meta-regression were as follows: Unpublished
studies trended toward smaller effects in symptom reduction compared with those
published (B = -.5 1 ). Treatments including psychotropic medication predicted larger
effects than those not including medication on both psychological (B = -1 .5 5 ) and spir­
itual (B = -1 .6 3 ) outcomes. Although in individual moderator analyses, therapeutic
accommodations within Islam resulted in stronger psychological effect sizes, this was
attenuated to a nonsignificant level when controlling for concurrent use of psycho­
tropic medication. General spiritual accommodated psychotherapy was significantly
less effective than Christian-accommodated psychotherapy on spiritual outcomes
(B = -.2 6 ) but not on psychological outcomes. In terms of treatment approach, sup-
portive/pastoral therapy showed weaker effects than CBT on symptom reduction
(B = -1.11) and spiritual well-being (B = -.5 1 ). Mind-body psychotherapy approaches
predicted slightly smaller reductions in psychological distress compared with CBT
(B = - .79). Finally, group modalities trended toward weaker effects than individual
psychotherapy on psychological outcomes (B = -.57).

Relation between Patient R/S and Psychotherapy Outcome

It has been theorized that a client’s level of R/S affiliation, practice, and/or commitment
may impact the effectiveness of R/S psychotherapeutic accommodations. The majority
of outcome studies we identified assessed patient R/S solely as a demographic variable
(e.g., participants self-identified as Christian, Muslim, Jewish, or spiritual). Thus, in
most cases, patient R/S was viewed as a dichotomous rather than continuous variable
and used for informational purposes only. We tested religious affiliation (percentage
Table 8 .9 . M ultiple M oderator A nalyses o f Treatm ent C haracteristics P re -P o st R/S
Psychotherapy

Comparisons k B 95% CI z(B) Q df p

Psychological Outcomes
O verall Model: 64 12.03 14 .000
Intercept 2.89 1.85-3.93 5.43* .000
Race/Ethnicity .00 -.0 0 -.0 1 .69 .487
Published -.5 1 - 1 .1 0 -.0 9 -1 .6 8 + .093
Accom m odative Focus 1.01 2 .603
G eneral Spiritual .02 - .6 1 -.6 4 .06 .956
Muslim .40 - .5 5 -1 .3 6 .83 .407
Treatm ent Approach 14.77 5 .011
Existential/Narrative -.7 0 -1 .5 8 -.1 7 -1 .5 8 .114
G eneral Psychospiritual .11 - .5 7 -.8 0 .33 .742
M ind Body -.7 9 - 1 .5 8 -.0 0 -1.95+ .051
REACH/Forgiveness -.1 2 - 1 .1 1 -.8 8 -.2 3 .821
Supportive/Pastoral -1.11 - 2 .0 0 - -.2 1 - 2 .4 2 * .016
Modality 3.52 2 .172
Group - .5 7 - 1 .1 8 -.0 4 -1.83+ .067
Couple or Family -.6 4 - 1 .8 2 -.5 4 -1 .0 7 .286
M edication Included in T x -1 .5 5 - 2 .3 2 - - .7 7 - 3 .9 3 * .000
Treatm ent Manual .22 - .3 7 -.8 2 .73 .467
u>
bo
1

Treatm ent Fidelity Check .22 .69 .491

1
0

Spiritual Outcomes
O verall Model: 38 41.06 13 .000
Intercept .64 .3 6 -.9 2 4.50* .000
Race/Ethnicity -.00 - .0 1 -.0 0 -1 .0 8 .278
Published -.14 - .4 4 -.1 6 -.9 1 .365
Accom m odative Focus 8.33 2 .016
G eneral Spiritual -.26 - .5 2 -.0 0 -1 .9 6 * .050
Muslim .26 - .2 3 -.7 5 1.05 .296
Treatm ent Approach 1.39 5 .065
Existential/Narrative .07 - .3 1 -.4 5 .37 .713
G eneral Psychospiritual .30 -.1 2 -.7 1 1.40 .162
M ind-Body .21 - .2 1 -.6 4 .98 .330
oo
\
A
1

REACH/Forgiveness .05 .14 .893

o
0

Supportive/Pastoral -.51 - .9 1 — .11 -2 .5 0 * .012

M edication Included in T x -1.63 - 2 .5 6 - -.7 1 -3 .4 6 * .001
Modality -.30 -.7 0 -.1 1 -1 .4 4 .150
Treatm ent Manual .12 - .1 4 -.3 9 .90 .366
Treatm ent Fidelity Check -.18 -.4 6 -.1 1 -1 .2 4 .217

Note. R/S = religious or spiritual; k = number of studies; B = slope; CI = confidence interval; z(B) = z
statistic for the slope; Q = homogeneity test; p = two-sidedp value indicating statistical significance for
each level of the model. Reference groups are as follows: accommodative focus—Christian, treatment
approach—cognitive-behavioral, modality—individual, medication incorporated in treatment—yes,
treatment manual—yes, and treatment fidelity check—yes. Blank spaces indicate lack of qualifying
data to run analysis.
+p < .1.
* p < .05.
 248 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

of the sample religiously affiliated) as a potential moderator but did not find a signifi­
cant effect. Much is lost by analyzing treatment outcomes of all R/S individuals based
on identification alone, because there is tremendous variance in the extent to which
people are influenced by their R/S beliefs. For some, R/S may be little more than a cul­
tural affiliation, whereas as for others, R/S may be the driving force behind their core
values, life goals, and sense of being in the world. More specific measurements of the
strength of R/S are needed to better understand the relation between patient R/S and
psychotherapy outcome.
Regrettably, we were not able test moderation of the strength of religious or spir­
itual commitment because few studies measured or reported this information. Only
a few investigations used a measure of R/S beliefs or commitment (e.g., Religious
Orientations Scale [Allport & Ross, 1967]; Religious Commitment Inventory-10
[Worthington et al., 2003]) in the pretreatment screening process, identifying a m in­
imum cutoff score for inclusion to ensure that all patients in the study were at least
moderately R/S. This lack of specificity in research and measurement represents a sig­
nificant gap for future exploration.
Next we review four studies that incorporated and reported an R/S measure to com­
pare treatment effects across level of religiosity. First, Nohr (2001) found that clients in
both treatment conditions (R/S CBT, standard CBT) who placed higher value on spir­
ituality at intake were more likely to employ R/S coping strategies throughout and re­
ported greater gains in spiritual well-being at termination. Thus it appears that highly
R/S clients may turn to their faith as a source of support, whether or not it is explicitly
integrated in treatment.
A second study (Razali et al., 2002) found that more religious patients in the R/
S-adapted treatment condition showed significantly greater improvement in anxiety
symptoms than less religious patients, but this difference was nonsignificant at follow­
up. A third study (Ripley et al., 2014) tested religious commitment as a potential mod­
erator of treatment effectiveness in religion-accommodative strategic hope-focused
couples’ therapy. Clients who entered R/S treatment with lower religious commitment
(a) displayed less denial in a conflict discussion with their partner and (b) rated their
partner more positively when reviewing the tape of the conflict discussion.
Finally, a study of R/S psychotherapy for depression (Koenig et al., 2015,
2016) assessed organizational (e.g., attendance at religious gatherings) and
nonorganizational religiosity (e.g., prayer and scripture reading) using the Duke
Religion Index (Koenig et al., 1997) and Intrinsic Religious Motivation Scale (Hoge,
1972). A composite measure of level of religiosity was computed from these measures,
and outcome data was analyzed by “low religious” and “high religious” categories. An
interaction was found between patients’ level of religiosity and treatment group, and an
interesting pattern emerged: R/S treatment was more effective in reducing depression
among more highly religious individuals but was more effective in increasing daily
spiritual experiences among those who came into treatment with lower religiosity.
Although it is difficult to draw conclusions based on the few findings reviewed here,
we encourage researchers to examine the differential effects of R/S-accommodated
treatments based on the relative importance of R/S in patients’ lives.
 249 Religion and Spirituality

EVIDENCE FOR CAUSALITY

There is strong empirical evidence that tailoring psychotherapy to a patient’s R/S

beliefs and values can positively influence both psychological and spiritual outcomes.
As shown in Tables 8.3 and 8.4, there is evidence for both quasi-experimental studies
and randomized controlled trials. The latter designs afford firm casual conclusions;
that is, R/S accommodative therapies yield significant benefits in patients’ psycholog­
ical and spiritual outcomes. R/S adaptations can also offer unique additive benefits
that persist over time, including greater spiritual well-being and connectedness with
the sacred.
Our multiple moderator analyses (see Table 8.9) revealed some interesting findings
that warrant further exploration. Treatment approach and accommodative focus were
significant moderators within R/S integrative psychotherapy. CBT resulted in the
greatest improvement in psychological symptoms, while supportive/pastoral coun­
seling was the least effective. O f note, however, much of the R/S integration research
has been conducted on CBT, and some orientations (e.g., psychodynamic) were not
represented in the R/S psychotherapy outcome literature. In individual moderator
analyses, accommodations within Islam resulted in greater reductions in psycholog­
ical symptoms, compared with accommodations to Christianity or general spirituality,
but this difference became nonsignificant when we accounted for psychotropic medi­
cation included as part of treatment. As such, the stronger effect size initially apparent
for Muslim-tailored psychotherapy may be due to concurrent medication. Therapeutic
accommodations made to Christianity were significantly more effective than general
spiritual integration in promoting spiritual well-being and connection with the sacred.
This could be due to common beliefs and practices (e.g., prayer, reading Scripture,
attending church) associated with Christianity that can be incorporated into therapy
sessions and daily life more readily than individually defined spirituality.

LIMITATIONS OF THE RESEARCH

Despite the growing empirical support for R/S-adapted treatments, some shortcomings
must be acknowledged. We only analyzed studies available in English, thus outcome
research in other languages was not accounted for. Generalizability of findings has
been limited by the use of relatively homogeneous samples that fail to capture the
many aspects of diversity evident in patients presenting for treatment in the “real
world.” Furthermore, small sample sizes may result in low statistical power, making
it difficult to capture a true treatment effect. As such, future outcome research may
want to employ larger samples perhaps from multiple sites, and meta-analyses could
be conducted of extant studies in other languages.
Although the majority of studies used a randomized design, some did not. Because
religion is an emotionally charged topic for many people, random assignment to R/S
versus secular treatments can prove difficult, and highly religious people may be less
willing to receive a secular treatment. Despite this challenge, it is best practice to ran­
domize participants to avoid self-selection bias.
 250 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

Additionally, many studies did not use truly comparative secular and R/S treatments
(e.g., same theoretical orientation and duration). These types of studies are espe­
cially important because they best answer the pressing clinical question of whether
incorporating R/S beliefs within an existing psychotherapy improves outcomes for R/
S patients. Without isolating the R/S component, it is difficult to tease apart the rela­
tive impact of R/S from other features of psychotherapy. When designing an outcome
study, researchers can ensure their alternative treatment condition is identical in every
regard, with the exception of the R/S component.
Another limitation of this review is that researchers had varying perspectives
about what constitutes R/S integration. In this chapter, we have outlined four broad
ways in which treatment can be tailored (conceptualization, intervention, treat­
ment goals, and interpersonal process). We encourage researchers to be specific
about how and how much they integrate R/S and to make their treatment manuals
available for cross-validation studies. When making therapeutic adaptations to spir­
ituality, researchers can be specific in reporting how they define and incorporate
this construct; doing so would allow for more specific analyses, such as Eastern
versus Western spirituality or spirituality focused on humanity, nature, and/or the
cosmos. To date, the majority of studies have examined the efficacy of tailoring CBT.
More psychotherapy outcome studies are needed examining various therapeutic
approaches, as well as considering adaptations to Judaism, Hinduism, Buddhism,
and other major world religions.
Further, while nearly every study included a measure of psychological outcomes,
many did not assess spiritual outcomes. Particularly in light of the differences we have
noted between psychological and spiritual change trajectories, future research could
include measures of spiritual well-being. With a few exceptions, R/S has generally been
treated as a categorical rather than a continuous variable. The vast majority of studies
reported the R/S composition of their sample as a patient demographic but failed to
measure R/S variables such as religious orientation, religious commitment, and daily
spiritual practices. Future studies can gather more extensive data about the client’s R/
S and evaluate treatment effectiveness based on strength of religious commitment, as
well as daily spiritual experiences and practices, such as attendance at religious serv­
ices, prayer, and reading of sacred texts.
In most cases, only patient R/S has been assessed. Noting the bidirectional relation­
ship within the treatment dyad, further exploration of the relative importance of value
similarity between psychotherapist and client is important. Non-R/S psychotherapists
may feel unequipped to customize treatment, believing, for example, that Christian or
Muslim psychotherapy can be best provided by a Christian or Muslim psychotherapist,
which is not necessarily the case.
Despite employing a multipronged backward and forward search process, it is
possible that some empirical outcome studies were not retrieved, particularly those
published in other languages, included in books, or presented at conferences. Also, we
contacted the corresponding authors of all articles identified to inquire about addi­
tional unpublished studies, but there were likely some inaccessible file-drawer studies.
 251 Religion and Spirituality

DIVERSITY CONSIDERATIONS

Little attention has been accorded in the R/S research to the contributions and
interactions of diversity to psychotherapy outcomes. The majority of studies we
reviewed reported demographic information, such as participants’ age, gender, race/
ethnicity, and socioeconomic status, but rarely were these data integrated within
posttreatment analyses. We did not find differences in the moderator analyses for
gender, suggesting that both women and men profit equally from R/S-accommodated
psychotherapy.
The preliminary indications are that people identifying as racial/ethnic minorities
may experience slightly greater reductions in psychological symptoms than Whites,
but this requires replication. One hypothesis for future exploration is that R/S accom­
modation may prove most important when people have lower acculturation to a sec­
ular worldview and thus higher psychological stigma and mistrust of the mental health
system. A therapist’s explicit integration of such clients’ R/S worldview may build rap­
port and discourage early termination among racial/ethnic minority individuals who
identify as R/S.
Furthermore, little research has explored R/S diversity, including religious com­
mitment and daily spiritual practices. As the research base grows for incorporating
patients’ R/S beliefs and values in treatment, it is important for psychotherapists to
better understand what sorts of individuals are most likely to benefit. For example, it
may be that older individuals, women, those of lower socioeconomic status, or clients
of certain racial or ethnic backgrounds experience a greater or lesser benefit. Many
questions remain to be explored, including diversity aspects of not only the patient
but also the psychotherapist offering treatment. We encourage researchers to examine
these diversity variables as moderators when analyzing treatment outcomes.

TRAINING IMPLICATIONS
Psychotherapists’ personal beliefs and worldview can influence their clinical work
positively or negatively. In contrast with the general population, a majority of
psychologists do not identify as R/S (Delaney et al., 2007). This is reflected in many
training programs’ relative lack of attention to this element of diversity. In one study,
the vast majority of psychologists (76%) reported that R/S issues were not adequately
addressed in their professional development (Crook-Lyon et al., 2012). As of 2011,
only one in four psychology training programs offered a course in R/S, leaving trainees
to rely on informal sources for information (Schafer et al., 2011).
The APA Ethics Code (2017; Principle E) affirms religion as a key consideration
within diversity. “To ignore religion as a cultural issue may not only be unethical,
but also lead to malpractice” (Plante, 2014, p. 289). Failing to attend to R/S issues in
training may have a negative effect on patients’ experience in psychotherapy. Although
some student clinicians have been guilty of imposing their R/S values on clients,
others may not feel confident assessing or exploring R/S in psychotherapy and steer
clear of the topic altogether. This avoidance can prove counterproductive to alliance
 252 PS Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

formation, implicitly communicating that certain topics are “off limits” or unimpor­
tant. R/S-specific ethical competencies have been proposed (Vieten et al., 2013) and
can be used as a framework to guide the development of relevant attitudes, knowledge,
and skills among graduate students and supervisees.
Our meta-analytic findings suggest that attending to patients’ R/S values and beliefs
can significantly improve treatment outcomes. In light of this research, we recommend
that educators and supervisors explicitly discuss R/S considerations within assessment
and psychotherapy training, encouraging students to reflect on a client’s R/S as part
of conceptualization, intervention, treatment goals, and interpersonal process. This
can be done with thoughtful consideration and following the client’s lead. Some R/
S patients may explicitly request psychotherapy adaptations, while others may not be
as forthright because of reticence to self-disclose R/S-related aspects of their struggles
in a presumably secular setting. Discussion about the client’s potential desire to incor­
porate his or her R/S in treatment could be included as part of the informed consent
and initial consultation. As with any element of diversity, psychotherapists can adopt
an open, accepting stance.
Understandably, some clinicians may initially feel uncomfortable integrating a
patient’s R/S within treatment, particularly if they do not identify as R/S themselves.
Cultural humility provides a framework for both supervision and treatment that
facilitates an attitude of curiosity and a desire to understand the client’s R/S experi­
ence, rather than assuming the pressure of needing to be an expert (Hook et al., 2013,
2017). Several practical steps to increase competence in working with clients of diverse
religious backgrounds include (a) recognizing one’s own biases, (b) framing R/S as a
potentially important aspect of clients’ culture and identity, (c) seeking out resources
to learn, and (d) consulting with colleagues and clergy (Plante, 2014). Some evidence
suggests that non-R/S psychotherapists can be equally effective in tailoring treatment
to a client’s personal values and that overidentification with a client’s R/S values may
actually be counterproductive (Propst et al., 1992).
What does R/S accommodative psychotherapy look like in practice? Potential areas
to explore include: What does R/S mean to this client? What role does it play in their
day-to-day life? How do they perceive and interact with the divine? In what ways do
they perceive R/S to be a source of strength? W hat practices or activities might be
powerful coping resources? Conversely, how might R/S be a source of struggle or inner
turmoil? Are they experiencing difficulty reconciling previously-held R/S beliefs with
their lived experience? Have they been the target of R/S oppression or abuse? How
might this have impacted their relationship with the sacred?
Taking time to understand a patient’s R/S values and experiences communicates that
this aspect of their identity is welcome and an asset in the therapeutic process. R/S is an
important cultural lens that can be creatively applied in conceptualization, treatment
goals, intervention, and interpersonal process, remembering that R/S adaptations of
psychotherapy are as unique as each patient who walks through the door.
Training resources are available through APA’s Division 36: Psychology of Religion
and Spirituality, as well as the American Counseling Association’s Association for
Spiritual, Ethical, and Religious Values in Counseling. As well, we recommend several
 253 Religion and Spirituality

texts well-suited to therapists in training (e.g., Aten et al., 2011; Cashwell & Young,
2014; Johnson, 2013; Land, 2015; Pargament, 2011; Worthington et al., 2013).

THERAPEUTIC PRACTICES

To conclude, we offer several clinical practices based on the findings of this

meta-analysis.

♦ Treat religion and spirituality as a potentially important aspect of a client’s identity.

Express curiosity about each client’s lived experience. Explore R/S history, values, and
commitment as part of the intake process, and consider intersectionality with other
dimensions of diversity.
♦ Incorporate a patient’s R/S values and worldview in psychotherapy as requested and
when clinically indicated. Research shows that accommodative psychotherapies are
at least as effective as secular approaches in reducing psychological symptoms and
can be provided by therapists regardless of their personal R/S background.
♦ Consider R/S adaptations in psychotherapy for their unique benefits to clients’
spiritual lives, including greater spiritual well-being and increased connection with
the sacred. When a client’s treatment goals include not only symptom remission but
also spiritual development, integration of R/S within psychotherapy is a treatment of
choice.
♦ Tailor treatment especially when working with patients whose R/S is an influential
force in their day-to-day life. Preliminary findings suggest that accommodative
psychotherapies may result in the greatest symptom reduction among clients with a
high level of R/S commitment.
♦ Follow the client’s lead when incorporating R/S beliefs and practices into
psychotherapy. Avoid making assumptions based on religious identification, and
instead explore their unique desires, needs, and expectations.
♦ Practice respect and cultural humility when discussing patients’ religious worldviews
and practices. Be especially sensitive to one’s own potential biases about organized
religion and to each patients experience of the sacred.

REFERENCES
References m arked with an asterisk indicate studies included in the m eta-analysis.
*Afazel, M . R., Aghajani, M ., & M orassaie, F. (2013). The effects o f spiritual counseling on
the hope in hemodialysis patients: A clinical trial study. Journal o f Evidence-Based Care,
3 (9 ), 1 9 -2 8 .
*A kuchekian, S., Jamshidian, Z., Maracy, M . R., Almasi, A., & Jazi, A. H. D. (2011). Effectiveness
o f religious cognitive-behavioral therapy on religious oriented Obsessive Compulsive
Disorder and its co-m orbidity. Journal o f Isfahan Medical School, 2 8 (1 1 4 ), 1 -1 2 .
Allport, G. W., & Ross, J. M . (1967). Personal religious orientation and prejudice. Journal o f
Personality and Social Psychology, 5 , 4 3 2 -4 3 3 .
A m erican Psychological Association. (2017). American Psychological Association ethical prin­
ciples o f psychologists and code o f conduct. Retrieved from http://www.apa.org/ethics/code
 254 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Anderson, N., Heywood-Everett, S., Siddiqi, N., W right, J., M eredith, J., & M cM illan, D.
(2015). Faith-adapted psychological therapies for depression and anxiety: Systematic re­
view and m eta-analysis. Journal o f Affective Disorders, 176, 1 8 3-196.
*Andrews, S. L. (2013). Psychological effects of spiritually integrated therapyfor infertile women
(Unpublished doctoral dissertation). University o f Alabama, Birm ingham .
*A no, G. G., Pargament, K. I., Wong, S., & Pomerleau, J. (2017). From vice to virtue: Evaluating
a m anualized intervention for m oral spiritual struggles. Spirituality in Clinical Practice,
4 (2 ), 1 2 9 -1 4 4 .
*A rm ento, M. E., McNulty, J. K., & Hopko, D. R. (2012). Behavioral activation o f religious
behaviors (BA R B): Random ized trial with depressed college students. Psychology of
Religion and Spirituality, 4 (3), 2 0 6 -2 2 2 .
Spiritually oriented interventions
Aten, J. D., M cM inn, M. R., & W orthington, E. L. Jr. (2011).
for counseling and psychotherapy. W ashington, DC: A m erican Psychological Association.
Avants, S. K., & Margolin, A. (2004). Developm ent o f Spiritual Self-Schem a (3-S) Therapy
for the treatm ent o f addictive and H IV risk behavior: A convergence o f cognitive and
Buddhist psychology. Journal o f Psychotherapy Integration, 14(3), 2 5 3 -2 8 9 .
*Azhar, M . Z., & Varma, S. L. (1995a). Religious psychotherapy as m anagem ent o f bereave­
m ent. Acta Psychiatrica Scandinavica, 91, 2 3 3 -2 3 5 .
*Azhar, M. Z., & Varma, S. L. (1995b). Religious psychotherapy in depressive patients.
Psychotherapy and Psychosomatics, 63, 1 6 5-173.
*Azhar, M. Z., Varma, S. L., & Dharap, A. S. (1994). Religious psychotherapy in anxiety dis­
order patients. Acta Psychiatrica Scandinavica, 90, 1 -3 .
*Baker, D. C. (2000). The investigation o f pastoral care interventions as a treatm ent for de­
pression am ong continuing care retirem ent com m unity residents. Journal o f Religious
Gerontology, 12, 6 3 -8 5 .
*Baker, N. D. (2012). Effect o f group psychotherapy on individuals diagnosed with anxiety and
mood disorders in a family practice setting (Unpublished doctoral dissertation). Walden
University, M inneapolis, MN.
*Barron, L. W. (2007).Effect of religious coping skills training with group cognitive-behavioral
therapy for treatment of depression (Unpublished doctoral dissertation). N orthcentral
University, San Diego, CA.
*Bay, P S., Beckm an, D., Trippi, J., G underm an, R., & Terry, C. (2008). The effect o f pastoral
care services on anxiety, depression, hope, religious coping, and religious problem solving
styles: A randomized controlled study. Journal of Religious Health, 47, 5 7 -6 9 .
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Beck Depression Inventory manual. San
A ntonio, T X : Psychological Corporation.
*Beheshtipour, N., Nasirpour, P , Yektatalab, S., Karimi, M ., & Zare, N. (2016). The effect of
educational-spiritual intervention on the burnout o f the parents o f school age children
with cancer: A randomized controlled clinical trial. International Journal o f Community
Based Nursing and Midwifery, 4(1), 9 0 -9 7 .
Borenstein, M ., Hedges, L., Higgins, J., & Rothstein, H. R. (2012). Comprehensive meta­
analysis. Version 3.0. Englewood, NJ: Biostat.
Borenstein, M ., Hedges, L. V., Higgins, J. P T., & Rothstein, H. R. (2009). Introduction to meta­
analysis. W est Sussex, England: Wiley.
*B orm an n, J. E., Aschbacher, K., Wetherell, J. L., Roesch, S., & Redwine, L. (2009). Effects
o f faith/assurance on cortisol levels are enhanced by a spiritual m antram intervention in
adults with HIV: A randomized trial. Journal o f Psychosomatic Research, 66(2), 1 6 1-171.
 255 Religion and Spirituality

*B orm an n, J. E., Gifford, A. L., Shively, M ., Sm ith, T. L., Redwine, L., Kelly, A., . . . Belding, W.
(2006). Effects o f spiritual m antram repetition on H IV outcomes: A randomized controlled
trial. Journal o f Behavioral Medicine, 2 9(4), 3 5 9 -3 7 6 .
*B orm an n, J. E., Liu, L., Thorp, S. R., & Lang, A. J. (2012). Spiritual wellbeing mediates
PTSD change in veterans with m ilitary-related PTSD. International Journal o f Behavioral
Medicine, 19(4), 4 9 6 -5 0 2 .
*B orm an n, J. E., Thorp, S., Wetherell, J. L., & G olshan, S. (2008). A spiritually based group
intervention for com bat veterans with posttraum atic stress disorder: Feasibility study.
Journal o f Holistic Nursing, 2 6(2), 1 0 9-116.
*B orm an n, J. E., Thorp, S. R., Wetherell, J. L., G olshan, S., & Lang, A. J. (2013). M editation­
based m antram intervention for veterans with posttraum atic stress disorder: A randomized
trial. Psychological Trauma: Theory, Research, Practice, and Policy, 5(3), 2 5 9 -2 6 7 .
*Bow land, S., Edm ond, T., & Fallot, R. D. (2012). Evaluation o f a spiritually focused interven­
tio n with older traum a survivors. Social Work, 57(1), 73 -8 2 . https://www.doi.org/10.1093/
sw/swr001
*Breitbart, W , Poppito, S., Rosenfeld, B., Vickers, A. J., Li, Y., Abbey, J., . . . Cassileth, B. R.
(2012). Pilot randomized controlled trial o f individual m eaning-centered psychotherapy
for patients with advanced cancer. Journal o f Clinical Oncology, 30(12), 1 3 04-1309.
*Breitbart, W , Rosenfeld, B., Pessin, H., Applebaum, A., Kulikowski, J., & Lichtenthal, W. G.
(2015). M eaning-centered group psychotherapy: An effective intervention for improving
psychological well-being in patients with advanced cancer. Journal o f Clinical Oncology,
3 3(7), 7 4 9 -7 5 4 .
Cashwell, C. S., Bentley, P. B., & Yarborough, J. P. (2007). The only way out is through: The peril
o f spiritual bypass. Counseling and Values, 51(2), 1 3 9-148.
Cashwell, C. S., & Young, J. S. (2014). Integrating spirituality and religion into counseling: A
guide to competent practice. Alexandria, VA: A m erican Counseling Association.
*C han, C. H., Chan, C. L., Ng, E. H., Ho, P. C., Chan, T. H., Lee, G. L., & Hui, W. H. C.
(2012). Incorporating spirituality in psychosocial group intervention for women
undergoing in vitro fertilization: A prospective randomized controlled study. Psychology
and Psychotherapy: Theory, Research and Practice, 8 5(4), 3 5 6 -3 7 3 .
*C han, C. H., Ji, X . W , Chan, J. S., Lau, B. H., So, K. F., Li, A., . . . Chan, C. L. (2017). Effects
o f the integrative m ind-body intervention on depression, sleep disturbances and plasm a
IL-6. Psychotherapy and Psychosomatics, 86(1), 5 4 -5 6 .
*C han, C. L., Ho, R. T., Lee, P W , Cheng, J. Y., Leung, P. P, Foo, W , . . . Spiegel, D. (2006).
A randomized controlled trial o f psychosocial interventions using the psychophysio­
logical fram ework for Chinese breast cancer patients. Journal o f Psychosocial Oncology,
2 4 (1 ), 3 -2 6 .
*C hida, Y., Schrem pft, S., & Steptoe, A. (2016). A novel religious/spiritual group psycho­
therapy reduces depressive symptoms in a randomized clinical trial. Journal o f Religion
and Health, 5 5(5), 1 4 95-1506.
*C ole, B. S. (2005). Spiritually-focused psychotherapy for people diagnosed with cancer: A
pilot outcom e study. Mental Health , Religion, and Culture, 8 , 2 1 7 -2 2 6 .
*C om bs, C. W , Bufford, R. K., Cam pbell, C. D., & Halter, L. L. (2000). Effects o f cognitive-
behavioral m arriage enrichm ent: A controlled study. Marriage and Family: A Christian
Journal, 3, 9 9 -1 1 1 .
Coursol, A., & Wagner, E. E. (1986). Effect o f positive findings on subm ission and acceptance
rates: A note on m eta-analysis bias. Professional Psychology, 17(2), 1 3 6-137.
 2 56 PS Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Crook-Lyon, R. E., O ’Grady, K. A., Sm ith, T. B., Jensen, D. R., Golightly, T., & Potkar, K. A.
(2012). Addressing religious and spiritual diversity in graduate training and m ulticul­
tural education for professional psychologists. Psychology o f Religion and Spirituality, 4,
1 6 9 -1 8 1 .
Davis, D. E., Rice, K., Hook, J. N., Van Tongeren, D. R., DeBlaere, C., Choe, E., & W orthington,
E. L. Jr. (2015). Development o f the Sources o f Spirituality Scale. Journal o f Counseling
Psychology, 62(3), 5 0 3 -5 1 3 .
*Davis, T. L., & Hill, C. E. (2005). Spiritual and nonspiritual approaches to dream work: Effects
on clients’ well-being. Journal o f Counseling and Development, 8 3(4), 4 9 2 -5 0 3 .
Delaney, H. D., M iller, W R., & Bisono, A. M . (2007). Religiosity and spirituality among
psychologists: A survey o f clinician m em bers o f the A m erican Psychological Association.
Professional Psychology: Research and Practice, 38, 5 3 8 -5 4 6 .
Derogatis, L. R. (1994). SCL-90-R Symptom Checklist-90-R administration, scoring and
procedures manual. M inneapolis, M N: National Com puter Systems.
Diener, E., Tay, L., & Myers, D. G. (2011). The religion paradox: If religion makes people
happy, why are so m any dropping out? Journal o f Personality and Social Psychology, 101(6),
1 2 7 8 -1 2 9 0 .
*D ik, B. J., Scholljegerdes, K. A., Ahn, J., & Shim , Y. (2015). A randomized controlled trial of
a religiously tailored career intervention with Christian clients. Journal o f Psychology and
Christianity, 34(4), 3 4 0 -3 5 4 .
*D ju ric, Z., M irasolo, J., Kim brough, L., Brown, D. R., Heilbrun, L. K., Canar, L., . . . Sim on,
M. S. (2009). A pilot trial o f spirituality counseling for weight loss m aintenance in African
A m erican breast cancer survivors. Journal o f the National Medical Association, 101(6),
5 5 2 -5 6 4 .
Duval, S. J. & Tweedie, R. L. (2000). A non-param etric “trim and fill” m ethod o f accounting
for publication bias in m eta-analysis. Journal o f the American Statistical Association,
95, 8 9 -9 8 .
*E brahim i, A., Neshatdoost, H. T., Mousavi, S. G ., Asadollahi, G. A., & Nasiri, H. (2013).
Controlled randomized clinical trial o f spirituality integrated psychotherapy, cognitive-
behavioral therapy and m edication intervention on depressive symptoms and dysfunc­
tional attitudes in patients with dysthymic disorder. Advanced Biomedical Research, 2, 53.
https://www.doi.org/10.4103/2277-9175.114201
*E brahim i, H., Kazemi, A. H., Khoshknab, M. E , & Modabber, R. (2014). The effect o f spir­
itual and religious group psychotherapy on suicidal ideation in depressed patients: A
randomized clinical trial. Journal o f Caring Sciences, 3(2), 1 31-140.
*Em ery, E. (2002). Living history spiritually—or not? A comparison o f conventional and
spiritually-integrated reminiscence groups (Unpublished doctoral dissertation). Bowling
G reen State University, Bowling G reen, OH.
Exline, J. J., & Rose, E. D. (2014). Religious and spiritual struggles. In R. F. Paloutzian &
C. L. Park (Eds.), Handbook o f the psychology o f religion and spirituality (pp. 3 8 0 -3 9 8 ).
New York, NY: Guilford.
*Fauver, R. (2011). The healing wisdom within: A preliminary experimental trial o f psycho­
spiritual integrative therapy fo r people with cancer (Unpublished doctoral dissertation).
Institute o f Transpersonal Psychology, Palo Alto, CA.
Fetzer Institute. (1999). Multidimensional measure o f religiousness/spirituality fo r use in
health research. Kalamazoo, M I: Author. Retrieved from http://fetzer.org/resources/
m ultidim ensional-m easurem ent-religiousnessspirituality-use-health-research
 257 Religion and Spirituality

*Ford, K. M. (2016). The impact o f a Christian adaptation to mindfulness training on stress, reli­
gious coping, and God attachment: A randomized trial (Unpublished doctoral dissertation).
Liberty University, Lynchburg, VA.
*G hodsbin, F., Safaei, M ., Jahanbin, I., Ostovan, M. A., & Keshvarzi, S. (2015). The effect o f
positive thinking training on the level o f spiritual well-being am ong the patients with co r­
onary artery diseases referred to Im am Reza specialty and subspecialty clinic in Shiraz,
Iran: A randomized controlled clinical trial. ARYA Atherosclerosis, 11(6), 3 4 1 -3 4 8 .
*G ibbel, M. R. (2010). Evaluating a spiritually integrated intervention fo r depressed college
students (Unpublished doctoral dissertation). Bowling G reen State University, Bowling
Green, OH.
G om ez, R., & Fisher, J. W. (2003). Dom ains o f spiritual w ell-being and development and val­
idation o f the Spiritual W ell-Being Questionnaire. Personality and Individual Differences,
3 5(8), 1 9 7 5 -1 9 91.
G onsalves, J. P. B., Lucchetti, G ., M enezes, P. R., & Vallada, H. (2015). Religious and spiritual
interventions in m ental health care: A systematic review and m eta-analysis o f randomized
controlled clinical trials. Psychological Medicine, 45(14), 2 9 3 7 -2 9 4 9 .
*Guthrey, C. R. (2015). An empirical study o f the effect o f psycho-spiritual integrative therapy
on the spirituality o f cancer patients (Unpublished doctoral dissertation). Institute of
Transpersonal Psychology, Palo Alto, CA.
H am ilton, M. A. X . (1959). The assessm ent o f anxiety states by rating. British Journal o f
Medical Psychology, 32(1), 5 0 -5 5 .
*H art, K. E., & Shapiro, D. A. (2002, August). Secular and spiritual forgiveness interventions fo r
recovering alcoholics harboring grudges. Paper presented at the annual convention o f the
A m erican Psychological Association. Chicago, IL.
*Hawkins, J. L. II. (2016). Can participation in the A More Excellent Way: Marriage Enrichment
Program contribute to increased intimacy in marriage? (Unpublished doctoral dissertation).
New Orleans Baptist Theological Seminary, New Orleans, LA.
*Hawkins, R. S., Tan, S., & Turk, A. A. (1999). Secular versus Christian inpatient cognitive-
behavioral therapy programs: Im pact on depression and spiritual well-being. Journal o f
Psychology and Theology, 27, 3 0 9 -3 1 8 .
Hill, P C., Pargament, K. I., Hood, R. W. Jr., M cCullough, M . E., Swyeres, J. P , Larson, D. B., &
Zinnbauer, B. J. (2000). Conceptualizing religion and spirituality: Points o f commonality,
points o f departure. Journal fo r the Theory o f Social Behavior, 30, 5 1 -7 7 .
*H o, T. H., Lo, P H.Y., & Chan, C. L. W. (2009). The efficacy o f the body-m ind-spirit interven­
tio n and social support groups on Chinese breast cancer patients. In Y. M . Lee, C. L. W.
Chan, S. M. Ng, & P P Y. Leung (Eds.), Integrative body-mind-spirit social work: An em­
pirically based approach to assessment and treatment (pp. 2 1 7 -2 3 4 ). New York, NY: Oxford
University Press.
Hoge, D. R. (1972). A validated intrinsic religious m otivation scale. Journal fo r the Scientific
Study o f Religion, 11, 3 6 9 -3 7 6 .
Hook, J. N., Davis, D., Owen, J., & DeBlaere, C. (2017). Cultural humility: A guide to engaging
diverse identities in therapy. W ashington, DC: APA Books.
Hook, J. N., Davis, D. E., Owen, J., W orthington, E. L. Jr., & Utsey, S. O. (2013). Cultural hu­
m ility: M easuring openness to culturally diverse clients. Journal o f Counseling Psychology,
6 0(3), 3 5 3 -3 6 7 .
*H osseini, M ., Salehi, A., Fallahi Khoshknab, M ., Rokofian, A., & Davidson, P. M . (2013). The
effect o f a preoperative spiritual/religious intervention on anxiety in Shia Muslim patients
 258 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

undergoing coronary artery bypass graft surgery: A randomized controlled trial. Journal o f
Holistic Nursing, 31(3), 1 6 4-172.
*Huguelet, P., Mohr, S., Betrisey, C., Borras, L., G illieron, C., M arie, A. M ., . . . Brandt, P.
Y. (2011). A randomized trial o f spiritual assessment o f outpatients with schizo­
phrenia: Patients’ and clinicians’ experience. Psychiatric Services, 62(1), 7 9 -8 6 .
*Iler, W. L. (2001). The im pact o f daily visits from chaplains on patients with chronic obstruc­
tive pulm onary disease (C O PD ): A pilot study. Chaplaincy Today, 17, 5 -1 1 .
*Jackson, R. E. (1999). Reducing shame through forgiveness and empathy: A group therapy
approach to promoting prosocial behavior (Unpublished doctoral dissertation). Fuller
Theological Seminary, Pasadena, CA.
*Jafari, N., Farajzadegan, Z., Zam ani, A., Bahram i, F., Em am i, H., Loghm ani, A., & Jafari, N.
(2013). Spiritual therapy to improve the spiritual well-being o f Iranian women with breast
cancer: A randomized controlled trial. Evidence-Based Complementary and Alternative
Medicine, 2013, 1 -9 . https://www.doi.org/10.! 155/2013/353262
Johnson, R. (2013). Spirituality in counseling and psychotherapy: An integrative approach that
empowers clients. H oboken, NJ: Wiley.
*Johnson, W. B., DeVries, R., Ridley, C. R., Pettorini, D., & Peterson, D. R. (1994). The com ­
parative efficacy o f Christian and secular rational-em otive therapy with Christian clients.
Journal o f Psychology and Theology, 22, 1 3 0-140.
*Johnson, W. B., & Ridley, C. R. (1992). B rief Christian and non-C hristian rational-em otive
therapy with depressed Christian clients: An exploratory study. Counseling and Values, 36,
2 2 0 -2 2 9 .
*Kam sani, S. R. (2014). Religious identity, self-concept, and resilience among fem ale orphan
adolescents in Malaysia: An evidence-based group intervention (Unpublished doctoral dis­
sertation). Southern Illinois University, Carbondale.
Koenig, H. G ., King, D., & Carson, V. B. (2012). Handbook o f religion and health. New York,
NY: O xford University Press.
Koenig, H. G., Parkerson, G. R., & Meadow, K. G. (1997). Religion index for psychiatric re­
search. American Journal o f Psychiatry, 154(6), 8 8 5 -8 8 6 .
*K oenig, H. G., Pearce, M. J., Nelson, B., & Erkanli, A. (2016). Effects on daily spiritual
experiences o f religious versus conventional cognitive behavioral therapy for depression.
Journal o f Religion and Health, 55(5), 1 763-1777.
*K oenig, H. G ., Pearce, M. J., Nelson, B., Shaw, S. F., Robins, C. J., Daher, N. S., . . . Rosm arin,
D. H. (2015). Religious vs. conventional cognitive behavioral therapy for m ajor depression
in persons with chronic m edical illness: A pilot randomized trial. The Journal o f Nervous
and Mental Disease , 203 (4), 2 4 3 -2 5 1 .
*Koszycki, D., Bilodeau, C., Raab-M ayo, K., & Bradwejn, J. (2014). A m ultifaith spiritually
based intervention versus supportive therapy for generalized anxiety disorder: A pilot
randomized controlled trial. Journal o f Clinical Psychology, 70(6), 4 8 9 -5 0 9 .
*Koszycki, D., Raab, K., Aldosary, F., & Bradwejn, J. (2010). A m ultifaith spiritually based in ­
tervention for generalized anxiety disorder: A pilot randomized trial. Journal o f Clinical
Psychology, 66(4), 4 3 0 -4 4 1 .
*Lam pton, C., Oliver, G. J., W orthington, E. L. Jr., & Berry, J. W. (2005). Helping Christian
college students becom e m ore forgiving: An intervention study to prom ote forgiveness
as part o f a program to shape Christian character. Journal o f Psychology and Theology, 33,
2 7 8 -2 9 0 .
 259 Religion and Spirituality

Land, H. M . (2015). Spirituality, religion, and faith in psychotherapy: Evidence-based expressive

methods fo r mind, brain, and body. New York, NY: Oxford University Press.
*Lee, A. M ., Chan, C. L. W., Ho, A. H. Y., Wang, C. N., Tang, V. Y. H., Lau, S. S. M ., et al. (2009).
The efficacy o f the body-m ind-spirit intervention m odel on improving the quality o f life
and psychological well-being o f Chinese patients with colorectal cancer: A prelim inary
report. In Y. M. Lee, C. L. W. Chan, S. M . Ng, & P. P. Y. Leung. (Eds.), Integrative body-
mind-spirit social work: An empirically based approach to assessment and treatment (pp.
2 3 6 -2 4 3 ). New York, NY: Oxford University Press.
Lipsey, M. W., & W ilson, D. B. (2001). Practical meta-analysis (Vol. 49). Thousand Oaks,
CA: SAGE.
*Liu, C. J., Hsiung, P C., Chang, K. J., Liu, Y. F., Wang, K. C., Hsiao, F. H., . . . Chan, C. L.
(2008). A study on the efficacy o f b o d y -m in d -sp irit group therapy for patients with breast
cancer. Journal o f Clinical Nursing, 17(19), 2 5 3 9 -2 5 4 9 .
*M argolin, A., Beitel, M ., Schum an-Olivier, Z., & Avants, S. K. (2006). A controlled study o f
a spirituality-focused intervention for increasing m otivation for H IV prevention among
drug users. AIDS Education and Prevention, 18,311 -3 2 2 .
*M argolin, A., Schum an-Olivier, Z., Beitel, M ., Arnold, R. M ., Fulwiler, C. E., & Avants, S. K.
(2007). A prelim inary study o f spiritual self-schem a (3-S+ ) therapy for reducing impul-
sivity in HIV-positive drug users. Journal o f Clinical Psychology, 63, 9 7 9 -9 9 9 .
*M cC ain, N. L., Gray, D. P , Elswick, R. K. Jr., Robins, J. W., Tuck, I., Walter, J. M ., . . . Ketchum,
J. M . (2008). A randomized clinical trial o f alternative stress m anagem ent interventions in
persons with H IV infection. Journal o f Consulting and Clinical Psychology, 76, 4 3 1 -4 4 1 .
*M cCauley, J., Haaz, S., Tarpley, M ., Koenig, H., & Bartlett, S. (2011). A randomized controlled
trial to assess effectiveness o f a spiritually based intervention to help chronically ill adults.
International Journal o f Psychiatry in Medicine, 41(1), 9 1 -1 0 5 .
M cCubbin, H., Kehl, L., Strom , I., & M cCubbin, L. (2010). The physical and mental health o f
native Hawaiians: Monoethnic and multiethnic perspectives. Honolulu: C enter for Training,
Evaluation, and Research o f the Pacific University o f Hawaii at Manoa.
M cCubbin, L. D., M cCubbin, H. I., Zhang, W., Kehl, L., & Strom , I. (2013). Relational well­
being: An indigenous perspective and measure. Family Relations, 62(2), 3 5 4 -3 6 5 .
M cCullough, M. E. (1999). Research on religion-accom m odative counseling: Review and
m eta-analysis. Journal o f Counseling Psychology, 4 6 , 9 2 -9 8 .
*M iller, W. R., Forcehim es, A., O’Leary, M. J., & LaNoue, M. D. (2008). Spiritual direction in
addiction treatm ent: Two clinical trials. Journal o f Substance Abuse Treatment, 35, 4 3 4 -4 4 2 .
*M oeini, M ., Sharifi, S., & Kajbaf, M. B. (2016). Effect o f Islam -based religious program on
spiritual wellbeing in elderly with hypertension. Iranian Journal o f Nursing and Midwifery
Research, 2 1 (6 ), 5 6 6 -5 7 1 .
*M usarezaie, A., Ghasemipoor, M ., M om eni-G haleghasem i, T., Khodaee, M ., & Taleghani, F.
(2015). A study on the efficacy o f spirituality-based intervention on spiritual wellbeing o f
patients with leukem ia: A randomized clinical trial. Middle East Journal o f Cancer, 6(2),
9 7 -1 0 5 .
*M usarezaie, A., M oeini, M ., Taleghani, F., & M ehrabi, T. (2014). Does spiritual care program
affect levels o f depression in patients with leukem ia? A randomized clinical trial. Journal o f
Education and Health Promotion, 3(1), 9 6 -1 0 1 .
*N edderm an, A. B., Underwood, L. A., & Hardy, V. L. (2010). Spirituality group with female
prisoners: Im pacting hope. Journal o f Correctional Health Care, 16(2), 1 1 7-132.
 260 psy c h o th er a py rela tio n sh ips that w ork

Ng, S. M ., Yau, J. K., Chan, C. L., Chan, C. H., & Ho, D. Y. (2005). The m easurem ent o f body-
m ind-spirit well-being: Toward m ultidim ensionality and transcultural applicability. Social
Work in Health Care, 4 1(1), 3 3 -5 2 .
*Nohr, R. W. (2001). Outcome effects o f receiving a spiritually informed vs. a standard cognitive-
behavioral stress management workshop (Unpublished doctoral dissertation). M arquette
University, Milwaukee, W I.
Norcross, J. C. (1990). An eclectic definition o f psychotherapy. In J. K. Zeig & W. M. M union
(Eds.), What is psychotherapy? Contemporary perspectives (pp. 2 1 8 -2 2 0 ). San Francisco,
CA: Jossey-Bass.
O h, P. J., & Kim , Y. H. (2012). M eta-analysis o f spiritual intervention studies on biological, psy­
chological, and spiritual outcomes. Journal o f Korean Academy o f Nursing, 42(6), 8 3 3 -8 4 2 .
*O lson, T., Tisdale, T. C., Davis, E. B., Park, E. A., Nam, J., Moriarty, G. L., . . . Hays, L. W.
(2016). G od image narrative therapy: A m ixed-m ethods investigation o f a controlled
group-based spiritual intervention. Spirituality in Clinical Practice, 3(2), 7 7 -9 1 .
*O m an, D., & B orm ann, J. E. (2015). M antram repetition fosters self-efficacy in veterans for
m anaging PTSD : A randomized trial. Psychology o f Religion and Spirituality, 7(1), 3 4 -4 5 .
*O m an, D., Hedberg, J., & Thoresen, C. E. (2006). Passage m editation reduces perceived stress
in health professionals: A randomized, controlled trial. Journal o f Consulting and Clinical
Psychology, 74(4), 7 1 4 -7 1 9 .
O rw in, R. G. (1983). A fail-safe N for effect size in m eta-analysis. Journal o f Educational
Statistics, 8 (2 ), 1 5 7-159.
Paloutzian, R. F., & Ellison, C. W. (1991). Manual fo r the Spiritual Well-being Scale. Nyack,
NY: Life Advance.
Paloutzian, R. F., & Park, C. L. (Eds.). (2014). Handbook o f the psychology o f religion and spir­
ituality. New York, NY: Guilford.
Pargament, K. I. (2011). Spiritually integrated psychotherapy: Understanding and addressing the
sacred . New York, NY: Guilford.
Pargament, K. I., Exline, J. J., & Jones, J. W. (2013). APA handbook o f psychology, religion, and
spirituality: Vol 1: Context, theory, and research. W ashington, DC: A m erican Psychological
Association.
Pargament, K., Feuille, M ., & Burdzy, D. (2011). The B rief RCO PE: Current psychom etric
status o f a short measure o f religious coping. Religions, 2 (1 ), 5 1 -7 6 .
Pargament, K. I., Koenig, H. G., & Perez, L. M . (2000). The m any m ethods o f religious
coping: Development and initial validation o f the RCO PE. Journal o f Clinical Psychology,
5 6(4), 5 1 9 -5 4 3 .
Pearce, M . J., Koenig, H. G ., Robins, C. J., N elson, B., Shaw, S. F., C ohen , H. J., & King,
M . B. (2 0 1 5 ). Religiously integrated cognitive behavioral therapy: A new m ethod of
treatm ent for m ajo r depression in patients with ch ron ic m edical illness. Psychotherapy,
5 2 (1 ), 5 6 -6 6 .
*Pecheur, D. R., & Edwards, K. J. (1984). A com parison o f secular and religious versions of
cognitive therapy with depressed Christian college students. Journal o f Psychology and
Theology, 12, 4 5 -5 4 .
Peterman, A. H., Fitchett, G ., Brady, M. J., Hernandez, L., & Cella, D. (2002). M easuring
spiritual well-being in people with cancer: the functional assessment o f chronic ill­
ness therapy— Spiritual W ell-being Scale (FA CIT-Sp). Annals o f Behavioral Medicine,
24 (1), 4 9 -5 8 .
 26l Religion and Spirituality

Pew Research. (2015). Importance o f religion and religious beliefs. Retrieved from http://
w w w.pewforum .org/2015/11/03/chapter-1-im portance-of-religion-and-religious-beliefs/
#belief-in-god
Pew Research. (2016). Frequency of feeling spiritual peace and wellbeing.
Retrieved from http://www.pewforum.org/religious-landscape-study/
frequency-of-feeling-spiritual-peace-and-w ellbeing/
*Piderm an, K. M ., Johnson, M. E., Frost, M. H., A therton, P. J., Satele, D. V., Clark, M . M ., . . .
Rum mans, T. A. (2014). Spiritual quality o f life in advanced cancer patients receiving radi­
ation therapy. Psycho-Oncology, 2 3 (2 ), 2 1 6 -2 2 1 .
Plante, T. G. (2014). Four steps to improve religious/spiritual cultural com petence in profes­
sional psychology. Spirituality in Clinical Practice, 1(4), 2 8 8 -2 9 2 .
*Propst, L. R. (1980). The comparative efficacy o f religious and nonreligious imagery for the
treatm ent o f m ild depression in religious individuals. Cognitive Therapy and Research, 4,
1 6 7 -1 7 8 .
*Propst, L. R., O strom , R., Watkins, P , Dean, T., & M ashburn, D. (1992). Comparative efficacy
o f religious and nonreligious cognitive-behavioral therapy for the treatm ent o f clinical de­
pression in religious individuals. Journal o f Consulting and Clinical Psychology, 60, 9 4 -1 0 3 .
Radloff, L. S. (1977). The C E S-D scale: A self-report depression scale for research in the ge­
neral population. Applied Psychological Measurement, 1(3), 3 8 5 -4 0 1 .
*Rasar, J. D. (2012). The effect o f a manualized group treatment protocol on God image and at­
tachment to God (Unpublished doctoral dissertation). Liberty University.4, Lynchburg, VA.
*Razali, S. M ., A m inah, K., & Khan, U. A. (2002). Religious-cultural psychotherapy in the
m anagem ent o f anxiety patients. Transcultural Psychiatry, 39, 1 30-136.
*Razali, S. M ., Hasanah, C. I., A m inah, K., & Subram aniam , M. (1998). Religious-sociocultural
psychotherapy in patients with anxiety and depression. Australian and New Zealand
Journal o f Psychiatry, 32, 8 6 7 -8 7 2 .
*R entala, S., Fong, T. C., Nattala, P , Chan, C. L., & Konduru, R. (2015). Effectiveness o f bod y -
m in d -sp irit intervention on well-being, functional im pairm ent and quality o f life among
depressive patients: A randomized controlled trial. Journal o f Advanced Nursing, 71(9),
2 1 5 3 -2 1 6 3 .
*Richards, P S., Berrett, M . E., Hardman, R. K., & Eggett, D. L. (2006). Comparative effi­
cacy o f spirituality, cognitive, and em otional support groups for treating eating disorder
inpatients. Eating Disorders, 14, 4 0 1 -4 1 5 .
*Ripley, J. S., Leon, C., W orthington Jr, E. L., Berry, J. W., Davis, E. B., Sm ith, A., . . . Sierra,
T. (2014). Efficacy o f religion-accom m odative strategic hope-focused theory applied to
couples therapy. Couple and Family Psychology: Research and Practice, 3(2), 8 3 -9 8 .
*Roland, J. E. (2014). The relationship between the introduction o f spirituality into cognitive
behavioral therapy and depression recovery in African American women (Unpublished d oc­
toral dissertation). Capella University, M inneapolis, MN.
Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological
Bulletin, 8 6(3), 6 3 8 -6 4 1 .
*R osm arin, D. H., Pargament, K. I., Pirutinsky, S., & Mahoney, A. (2010). A randomized
controlled evaluation o f a spiritually integrated treatm ent for subclinical anxiety in the
Jewish community, delivered via the Internet. Journal o f Anxiety Disorders, 2 4(7), 7 9 9 -8 0 8 .
*Rye, M . S., & Pargament, K. I. (2002). Forgiveness and rom antic relationships in college: Can
it heal the wounded heart? Journal o f Clinical Psychology, 58, 4 1 9 -4 4 1 .
 262 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

*Rye, M . S., Pargament, K. I., Pan, W., Yingling, D. W., Shogren, K. A., & Ito, M . (2005). Can
group interventions facilitate forgiveness o f an ex-spouse? A randomized clinical trial.
Journal o f Consulting and Clinical Psychology, 73, 8 8 0 -8 9 2 .
Saunders, S. M ., Miller, M . L., & Bright, M . M . (2010). Spiritually conscious psychological
care. Professional Psychology: Research and Practice, 41(5), 3 5 5 -3 6 2 .
Schafer, R. M ., Handal, P. J., Brawer, P A., & Ubinger, M . (2011). Training and education
in religion/spirituality within APA-accredited clinical psychology programs: 8 years later.
Journal o f Religion and Health, 50, 2 3 2 -2 3 9 .
*Scott, S. (2001). Faith supportive group therapy and symptom reduction in Christian breast
cancer patients (Unpublished doctoral dissertation). Regent University, Virginia Beach, VA.
*Shafiee, Z., Zandiyeh, Z., M oeini, M ., & G holam i, A. (2016). The effect o f spiritual inter­
vention on postm enopausal depression in women referred to urban healthcare centers in
Isfahan: A double-blind clinical trial. Nursing and Midwifery Studies, 5(1), e32990.
Sheldrake, P (1992). Spirituality and history: Questions o f interpretation and method. New York,
NY: Crossroads.
Sm ith, T. B., Bartz, J., & Richards, P S. (2007). Outcom es o f religious and spiritual adaptations
to psychotherapy: A m eta-analytic review. Psychotherapy Research, 17, 6 4 3 -6 5 5 .
*Sreevani, R., Reddema, K., Chan, C. L. W., Leung, P. P. Y., Wong, V , & Chan, C. H. Y. C. (2013).
Effectiveness o f integrated body-m ind-spirit group intervention on the well-being o f Indian
patients with depression: A pilot study. The Journal o f Nursing Research, 21(3), 178-185.
*Stalsett, G., Gude, T., Ronnestad, M. H., & M onsen, J. T. (2012). Existential dynamic therapy
( “VITA”) for treatm ent-resistant depression with Cluster C disorder: M atched com parison
to treatm ent as usual. Psychotherapy Research, 2 2(5), 5 7 9 -5 9 1 .
*Stratton, S. P , Dean, J. B., N onnem an, A. J., Bode, R. A., & W orthington, E. L. Jr. (2008).
Forgiveness interventions as spiritual development strategies: Com paring forgiveness
workshop training, expressive writing about forgiveness, and retested controls. Journal o f
Psychology and Christianity, 2 7 , 3 4 7 -3 5 7 .
*Tadwalkar, R., Udeoji, D. U., Weiner, R. J., Avestruz, F. L., LaChance, D., Phan, A., . . . Schwarz,
E. R. (2014). The beneficial role o f spiritual counseling in heart failure patients. Journal o f
Religion and Health , 53(5), 1 5 75-1585.
*Targ, E. F., & Levine, E. G. (2002). The efficacy o f a m ind-body-spirit group for women with
breast cancer: A randomized clinical trial. General Hospital Psychiatry, 24, 2 3 8 -2 4 8 .
*Toh, Y., & Tan, S. (1997). The effectiveness o f church-based lay counselors: A controlled out­
com e study. Journal o f Psychology and Christianity, 16, 2 6 0 -2 6 7 .
*Tonkin, K. M . (2005). Obesity, bulimia, and binge-eating disorder: The use o f a cognitive behav­
ioral and spiritual intervention (Unpublished doctoral dissertation). Bowling G reen State
University, Bowling Green, OH.
*Trathen, D. W. (1995). A comparison o f the effectiveness o f two Christian premarital coun­
seling programs (skills and information-based) utilized by evangelical protestant churches
(Unpublished doctoral dissertation). University o f Denver, Denver, CO.
Underwood, L. G., & Teresi, J. A. (2002). The daily spiritual experience scale: Development,
theoretical description, reliability, exploratory factor analysis, and prelim inary construct
validity using health-related data. Annals o f Behavioral Medicine, 2 4 (1 ), 2 2 -3 3 .
Viechtbauer, W. (2007). A ccounting for heterogeneity via random-effects models and m od­
erator analyses in meta-analysis. Zeitschrift fa r Psychologies ournal o f Psychology, 215,
104 -1 2 1 .
 263 Religion and Spirituality

V ieten, C., Scam m ell, S., Pilato, R., A m m ondson, I., Pargament, K. I., & Lukoff, D. (2013).
Spiritual and religious com petencies for psychologists. Psychology o f Religion and
Spirituality, 5(3), 1 2 9-144.
*W ahass, S., & Kent, G. (1997). The m odification o f psychological interventions for persistent
auditory hallucinations to an Islam ic culture. Behavioural and Cognitive Psychotherapy,
2 5 (4 ), 3 5 1 -3 6 4 .
Waltz, J., Addis, M. E., Koerner, K., & Jacobson, N. S. (1993). Testing the integrity o f a psy­
chotherapy protocol: Assessm ent o f adherence and com petence. Journal o f Consulting and
Clinical Psychology, 61, 6 2 0 -6 3 0 .
W ampold, B. E., & Imel, Z. E. (2015). The great psychotherapy debate: The evidence fo r what
makes psychotherapy work. New York, NY: Routledge.
*W ang, D. C., Aten, J. D., Boan, D., Jean-Charles, W., Griff, K. P , Valcin, V. C., . . . Abouezzeddine,
T. (2016). Culturally adapted spiritually oriented traum a-focused cognitive-behavioral
therapy for child survivors o f restavek. Spirituality in Clinical Practice, 3(4), 2 2 4 -2 3 6 .
W eism an, A. G., Okazaki, S., Gregory, J., Goldstien, M . J., Tom pson, M . C., & Miklowitz, D. J.
(1998). Evaluating therapist com petency and adherence to behavioral family m anagem ent
with bipolar patients. Family Process, 37(1), 1 07-121.
*W eism an de M am ani, A., Weintraub, M . J., Gurak, K., & Maura, J. (2014). A randomized
clinical trial to test the efficacy o f a fam ily-focused, culturally inform ed therapy for schiz­
ophrenia. Journal o f Family Psychology, 2 8 (6 ), 8 0 0 -8 1 0 .
W orthington, E. L. Jr., Hook, J. N., Davis, D. E., & M cDaniel, M. A. (2011). Religion and spir­
ituality. In J. Norcross (Ed.), Psychotherapy relationships that work: Evidence-based respon­
siveness (2nd ed., pp. 4 0 2 -4 2 2 ). New York, NY: Oxford University Press.
W orthington E. L. Jr., Johnson, E. L., Hook, J. N., & Aten, J. D. (Eds.). (2013). Evidence-based
practices fo r Christian counseling and psychotherapy. Dow ner’s Grove, IL: IV P Academic.
W orthington, E. L. Jr., Wade, N. G., Hight, T. L., Ripley, J. S., M cCullough, M. E., Berry, J. W.,
. . . O ’Connor, L. (2003). The Religious C om m itm ent Inventory— 10: Development, re­
finem ent, and validation o f a brief scale for research and counseling. Journal o f Counseling
Psychology, 50, 8 4 -9 6 .
*W u, L. F., & Koo, M . (2016). Random ized controlled trial o f a six-week spiritual rem iniscence
intervention on hope, life satisfaction, and spiritual w ell-being in elderly with m ild and
m oderate dementia. International Journal o f Geriatric Psychiatry, 31(2), 1 2 0-127.
*Yong, J., Kim , J., Park, J., Seo, I., & Swinton, J. (2011). Effects o f a spirituality training program
on the spiritual and psychosocial well-being o f hospital m iddle m anager nurses in Korea.
The Journal o f Continuing Education in Nursing, 4 2(6), 2 8 0 -2 8 8 .
*Zhang, Y., Young, D., Lee, S., Li, L., Zhang, H., X iao, Z., . . . Chang, D. F. (2002). Chinese
Taoist cognitive psychotherapy in the treatm ent o f generalized anxiety disorder in contem ­
porary China. Transcultural Psychiatry, 39, 1 1 5-129.
 9

S E X U A L O R IE N T A T IO N

Bonnie Moradi and Stephanie L. Budge

The clinical need for lesbian, gay, bisexual, and queer (LGBQ+) affirma­
tive psychotherapies has been widely recognized (e.g., American Psychological
Association [APA], 2012; Fassinger, 2017; Harrison, 2000; King et al., 2007; Johnson,
2012; Pachankis & Goldfried, 2013). However, empirical research on the outcomes of
such therapies is nearly nonexistent (e.g., Bieschke et al., 2007; Harrison, 2000; King
et al., 2007). Among the barriers that impede research on the outcomes of LGBQ+
affirmative psychotherapies are complexities in defining whom LGBQ+ affirmative
psychotherapies are for and what key elements these therapies comprise.
Wrestling with these “who” and “what” questions is fundamental to forging best
practices for LGBQ+ affirmative psychotherapies. As such, we begin this chapter by
offering definitions to answer these who and what questions. We then describe prior
measures, clinical examples, and landmark studies shaping the evolution of LGBQ+
affirmative psychotherapies. Next, we summarize prior syntheses of studies on the
outcomes of LGBQ+ affirmative psychotherapies, describe our own search for data to
attempt a meta-analysis, and discuss the limitations and directions for research based
on our review. We end the chapter by delineating diversity considerations, training
implications, and clinical practices for advancing LGBQ+ affirmative psychotherapy
with clients of all sexual orientations.
Throughout this chapter, we avoid framing LGBQ+ affirmative psychotherapies as
a type of group-specific or “tailored” treatment applicable to some sexual orientations
and not others. We view such an approach as clinically and ethically problematic.
Instead, we conceptualize and advance LGBQ+ affirmative psychotherapies as a set
of practices that can be applied with all clients and, ultimately, in all psychotherapies.

DEFINITIONS

Sexual Orientation
We offer the following definitions of key sexual orientation constructs, noting that
these definitions reflect dominant cultural discourses in the United States and that

264
 265 Sexual Orientation

conceptualizations of sexual orientation and sexual minority status vary across cul­
tural communities (e.g., APA, 2012; Dillon et al., 2011; Moradi, 2016). In brief, sexual
orientation reflects the sex(es) and/or gender(s) to whom a person is attracted; it
includes multiple dimensions such as physical attraction, emotional attraction, and
sexual behaviors which may or may not align with one another at a given time or
across a person’s life. Sexual identity (or sexual orientation identity) captures people’s
identification or description of their sexual orientation to themselves and others (e.g.,
lesbian, gay, bisexual, queer, questioning, asexual). Sexual minority is an umbrella term
that captures a range of people whose sexual orientations and identities are stigmatized
and oppressed in current sociopolitical systems (i.e., LGBQ+).
Despite this multidimensionality, popular conceptualizations of sexual orientation
and identity are often grounded in a binary view of sex fused with gender (i.e., fe­
male = woman, male = man). Such a view has long been critiqued by feminist scholars
(e.g., Bem, 1993, West & Zimmerman, 1987) who distinguished sex, or the biological
and anatomical characteristics used to assign people at birth to sex categories (e.g.,
male, female, intersex) from gender, as the social meaning and collection of charac­
teristics prescribed to sex categories in a given society or culture. Nevertheless, based
on sex and gender binaries, sexual orientation is typically conceptualized such that
a person assigned female at birth is assumed to identify as a woman and present in
feminine ways. If she is attracted to other women, she is considered and compelled to
identify as lesbian, and if she is attracted to other men, she is considered and compelled
to identify as heterosexual or straight. Men are categorized in parallel fashion as gay
or heterosexual/straight. Problematically, within these binaries, bisexual, queer, and
other sexual orientations and identities are often rendered invisible or viewed as transi­
tory stops toward achieving an ultimate monosexual orientation and identity (i.e., gay/
lesbian or heterosexual/straight).
Estimates, based on willingness to self-identify on population-based surveys, sug­
gest that approximately 3.5% or 8 million adults in the United States identify as LGB,
8.2% or 19 million report having engaged in same-sex sexual behaviors, and 11% or
25.6 million report having some same-sex attraction (Gates, 2011).

LGBQ+ Affirmative Psychotherapies

Defining LGBQ+ affirmative psychotherapies first requires defining whom such
psychotherapies are for. A typical approach is to attempt to define LGBQ+ populations;
acknowledge their diversity across various sociodemographic characteristics (e.g.,
ability status, gender, class, ethnicity, race, sexual orientation); delineate the rich mul­
tidimensionality of sexual attraction, behaviors, orientations, and identities; and then
articulate what therapists are recommended to do with such clients. However, this ap­
proach still requires defining LGBQ+ communities based on complex definitions of sex,
gender, and sexual orientation that defy rigid and essentialist category boundaries. In
fact, these complexities in defining the boundaries of LGBQ+ populations has led some
to conclude that randomized controlled trials of LGBQ+ affirmative psychotherapies
 266 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

are not feasible (e.g., Johnson, 2012). One resolution to these complexities is to use self­
identification as LGBQ+, honoring clients’ self-definitions in shaping the boundaries
of who LGBQ+ people are and whom LGBQ+ affirmative psychotherapies are for. This
likely constitutes the most prevalent approach in clinical practice.
An alternative approach, which we favor, is to advance the principles of LGBQ+ af­
firmative psychotherapy with all clients (e.g., Matthews, 2007). This inclusive position
addresses important pragmatic realities. Specifically, practitioners may not be aware of
clients’ LGBQ+ identities. Even if therapists routinely assess LGBQ+ identities, clients’
may not want to or be ready to disclose these identities, especially in early phases of
psychotherapy. In fact, clients’ disclosure of LGBQ+ identities may be predicated on
therapists first creating the very conditions of LGBQ+ affirmativeness to facilitate such
disclosure (e.g., Dorland & Fischer, 2001). Moreover, many clients who do not iden­
tify as LGBQ+ may want (and warrant) LGBQ+ affirmative therapy (e.g., children of
LGBQ+ parents).
For these reasons, we endorse a conceptualization of LGBQ+ affirmative psycho­
therapy that acknowledges the unequal power inherent in the client-therapist dyad,
which may prove more pronounced in dyads involving LGBQ+ clients and hetero­
sexual therapists, and that places the responsibility of providing affirmative methods
on the clinician rather than on the client. Thus, we contend that the answer to the first
question “whom are LGBQ+ affirmative psychotherapies for?” is, simply, everyone.
This vision does not mean that LGBQ+ affirmative psychotherapy is “generic” psy­
chotherapy as currently practiced. Rather, LGBQ+ affirmativeness requires elevating
all psychotherapies to integrate elements that have been identified as LGBQ+ affirm­
ative. Drawing from classic and contemporary conceptualizations of psychotherapy
with LGBQ+ people (e.g., APA, 2012; Fassinger, 2017; Harrison, 2000; Johnson,
2012; King et al., 2007; Pachankis & Goldfried, 2013), we define LGBQ+ affirma­
tive psychotherapies as comprising four key themes: (a) counteracting anti-LGBQ+
therapist attitudes and enacting LGBQ+ affirmative attitudes, (b) acquiring accurate
knowledge about LGBQ+ people’s experiences and their heterogeneity, (c) calibrating
integration of accurate knowledge about LGBQ+ people’s experiences and their het­
erogeneity into therapeutic actions, and (d) engaging in and affirming challenges to
power inequalities. Across these themes, it is important not to confuse the absence of
inappropriate therapy (e.g., acting on anti-LGBQ+ bias or inadequate knowledge) with
the presence of affirmative therapy. We view these four defining themes as necessary
conditions for LGBQ+ affirmative psychotherapy.

MEASURES
Sexual Orientation

Psychotherapists and psychotherapy researchers can routinely assess sexual

orientation-related variables. Best practices involve assessing multiple dimensions,
such as sexual orientation identity, sexual attraction, and sexual behaviors, and in­
cluding open response options for people to self-describe beyond the predetermined
 267 Sexual Orientation

category options (e.g., SmArt, 2009). For example, in intake forms, self-identification
can be assessed with the question “How do you self-i dentify?” with options of bi­
sexual, gay, heterosexual/straight, lesbian, queer, and an open-response. It is helpful
to alphabetize response options to avoid unintentionally communicating a hierarchy
of identities. Sexual attraction can be assessed with the question “People vary in
their sexual attraction to other people. W hich best describes your attraction?” with
options to assess level of attraction to men, women, gender nonbinary people, and
an open-response option. Sexual behavior can be assessed with the question “Which
best describes your sexual partners?” with options to assess sexual behavior with men,
women, gender nonbinary people, no sexual behavior, and an open response option.
In therapy sessions or intake interviews, it is important to listen carefully to clients’
self-descriptions and the specific terms they use to refer to themselves (e.g., lesbian,
queer, bisexual) and their romantic partners (e.g., partner, spouse, wife, girlfriend) and
to mirror these terms. It can also be helpful to check in with clients in organic and open-
ended ways to facilitate clients’ personal descriptions. For example, broad questions
such as “What are some important aspects of who you are?” can be a starting point
for rich self-description of personal identities. Such questions can be followed with
more specific questions about sexual orientation identity and romantic relationships,
such as “What terms or identities do you prefer to describe your sexual orientation or
romantic attractions?” or “How do you prefer to refer to your partner?” Some clients
may resist such identity categories and labels altogether (e.g., “I don’t identify with any
sexual orientation label” or “my attractions are not based on gender”) and these are
also self-definitions to be respected and affirmed.
Best practices for assessing sexual orientation also require disaggregating sexual
orientation, sex, and gender. Assessing sex requires careful consideration to include
intersex individuals as well. One approach is to assess sex assigned at birth with the
options currently available on birth certificates (i.e., female, male) and using a separate
question that assesses whether individuals are also intersex (GenIUSS Group, 2014;
Intersex Society of North America, n.d.). Assessment of gender identity can include
categories for transwoman, transman, and nonbinary gender identities (e.g., gender-
queer) and an open response option for respondents to self-describe (for detailed
recommendations see GenIUSS Group, 2014). The aforementioned recommendations
for facilitating self-descriptions also apply to assessing sex and gender. Sexual orienta­
tion, sex, and gender variables can be assessed along with other demographics such as
age, ethnicity, race, and class.

LGBQ+ Affirmative Psychotherapies

A number of existing measures assess therapists’ or trainees’ self-reported perceptions of
their own competencies in working with LGBQ+ clients (see Table 9.1). Although these
are not measures of LGBQ+ affirmative psychotherapy ingredients per se, these meas­
ures are important advancements in research on LGBQ+ affirmative psychotherapies
and are the closest approximations of operationalizing those ingredients.
Table 9.1. Measures Intended to Assess LGBQ+ Affirmative Psychotherapy Competencies
Authors Measure Name Purpose and Description Psychometrics Sample Items
Bidell (2005) Sexual O rientation 29 items to measure ♦ Evaluated with 312 psychology ♦ Attitudes: “The lifestyle o f an LGB
Counselor counselors’ self­ undergraduate students, counseling client is unnatural or im m oral”
C om petency Scale reported attitudes, graduate students, counselor (reverse coded)
skills, and knowledge educators, and counseling supervisors ♦ Skills: “I have experience counseling gay
in working with LGB ♦ T hree-factor structure (attitudes, m ale clients”
clients skills, knowledge) derived from ♦ Knowledge: “Being born a heterosexual
principal axis factoring person in this society carries with it
♦ Cronbach’s alpha .90 overall, .88 certain advantages”
for attitudes, .91 for skills, .76 for
knowledge
♦ O ne-w eek test-retest reliability .84
overall, .85 for attitudes, .83 for skills,
.84 for knowledge
♦ LGB counselors scored higher than
heterosexual counselors
♦ Correlated as expected with
convergent validity indicators
Burkard et al. (2009) Lesbian, Gay, and 32 items to assess
Bisexual W orking counselor-trainees
Alliance Self- self-reported self-
Efficacy Scales efficacy in establishing
working alliance with
LGB clients

Crisp (2006) Gay Affirmative 30 items to assess

Practice Scale clinicians’ self-
reported beliefs and
behaviors in practice
with gay and lesbian
clients
Evaluated with two samples o f 303 ♦ Em otional bond: “I can express
and 229 counselor trainees em pathy for an LGB client”
T hree-factor structure (em otional ♦ Establishing tasks: “I can help LGB
bond, establishing tasks, setting goals) clients to establish social relationships
derived from principal axis factoring in the gay com m unity”
Cronbach’s alpha .98 overall, .97 for ♦ Cultural com petence: “An LGB
em otional bond, .96 for establishing client and I can m utually agree on an
tasks, .94 for setting goals in both im portant purpose for counseling”
samples
Three-week test-retest reliability
.83 overall, .90 for em otional bond,
.79 for establishing tasks, .63 for
setting goals
Correlated as expected with
convergent validity indicators
Evaluated with a sample o f 488 ♦ Beliefs: “Practitioners should educate
psychologists and social workers themselves about gay/lesbian lifestyles”
Tw o-factor structure evaluated with ♦ Behaviors: “I acknowledge to clients
confirm atory factor analysis, m odel fit the impact o f living in a hom ophobic
not reported society”
Cronbach’s alpha .95 overall, .93 for
beliefs, .94 for behaviors
Correlated as expected with
convergent validity indicators

( continued)
Table 9.1. Continued
A uthors M easu re N am e P u rp ose a n d D escription

Dillon & W orthington Lesbian, Gay, 32 items (original) and

(2003) and Dillon and Bisexual 15 items (short form )
et al. (2015) Affirmative to assess counselors’
Counseling self-reported self-
Self-Efficacy efficacy to perform
Inventory, original LGB affirmative
and short form counseling behaviors
Psychometrics Sample Items
♦ O riginal form evaluated with ♦ Application o f knowledge: “Assist LGB
three samples o f 336, 310, and 80 clients to develop effective strategies
counselor trainees and m ental health to deal with heterosexism and
practitioners, and a fourth test-retest hom ophobia”
reliability sample o f 36 graduate ♦ Relationship: “Establish a safe space for
students LGB couples to explore parenting”
♦ Short form evaluated with 575 ♦ Assessment: “Assess for posttraum atic
counselor trainees and m ental stress felt by LGB victim s o f hate crim es
health professionals and a test-retest based on their sexual orientations/
reliability sample o f 32 graduate identities”
students ♦ Advocacy skills: “Refer LGB clients to
♦ O riginal form with 5-factor affirmative legal and social supports”
structure (application o f knowledge, ♦ Self-awareness: “Identify my own
relationship, assessment, advocacy feelings about m y own sexual
skills, self-awareness) derived from orientation and how it may influence a
principal axis factoring client”
Logie et al. (2007) LG BT Assessment 26 items to measure
Scale social work graduate
students’ self-
reported “phobias,
attitudes, and cultural
com petence” in
working with LG BT
clients
♦ O riginal form Cronbach’s alphas
across samples were .86 to .96 for
overall items and ranged from .70 to
.96 across subscales
♦ Short form latent variable internal
consistency reliability estimates
ranged from .81 to .92 across
subscales
♦ O riginal form two-week test-retest
reliability was .51 overall
♦ Short form one-week test-retest
reliability was .80 overall
♦ Evaluated with a sample o f 173 ♦ Phobias: “I would feel com fortable
masters o f social work students working closely with a gay man”
♦ Cronbach’s alpha .93 for phobias, ♦ Attitudes: “Bisexuality is m erely a
.92 for attitudes, .30 for cultural different kind o f lifestyle that should
com petence not be condem ned”
Cultural com petence: “I am
knowledgeable about the issues and
challenges facing LG B T people and
feel com petent in m y ability to work
effectively with this population”
 272 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

These measures also have key limitations. Specifically, they tend to place greater
emphasis on assessing anti-LGBQ+ attitudes and feelings than on assessing LGBQ+
affirmative psychotherapy behaviors, some use problematic language (e.g., referring
to LGBQ+ identities as a lifestyle), some have psychometric limitations and gaps,
and all rely on therapists’ self-reports. None of these measures directly assesses
the fourth theme of LGBQ+ affirmative psychotherapy ingredients: engaging in
and affirming challenges to power inequalities. The Lesbian, Gay, and Bisexual
Affirmative Counseling Self-Efficacy Inventory (Dillon & Worthington, 2003) and its
short form (Dillon et al., 2015) are among the fullest in scope of coverage, assessing
the application of knowledge, therapy relationship, assessment, advocacy skills, and
self-awareness.
Though therapists’ self-reports are valuable, clients’ appraisal of the degree to which
LGBQ+ affirmative psychotherapy ingredients are present is particularly important to
assess, and observer ratings are needed as well. The Client Task Specific Change Scale-
Revised (Watson et al., 1998; Watson et al., 2010) is one example of directly assessing
clients’ perceptions. This measure assesses clients’ perceptions of change, such as “I
feel that I was able to successfully challenge my negative, automatic thoughts” and “I
became more compassionate toward myself, which I previously could not accept.” This
model could be adapted to assess clients’ perceptions of changes specifically targeted
by LGBQ+ affirmative psychotherapies, such as deeper understanding of how systems
of power operate in one’s life and how to engage in everyday resistance of oppres­
sive systems and dynamics. Such assessment could occur following sessions and upon
treatment termination.

CLINICAL EXAMPLES
Following are examples of how clinicians have implemented LGBQ+ affirmative
psychotherapies. These examples illustrate the transtheoretical, transdiagnostic, and
transpopulation scope of such psychotherapies.
Spengler et al. (2016) provide a useful example of counteracting anti-LGBQ+ thera­
pist attitudes and enacting LGBQ+ affirmative attitudes. In the following excerpt, they
illustrate how therapists can attend to power dynamics in the therapy relationship and
work to repair an instance of having offensively stereotyped an LGBQ+ client:

Therapist: I know we’ve been talking about your experiences “out there” with people
who do or say microaggressive things. I’m wondering, has there ever been anything
I’ve said or done in here that you would view as a microaggression?
Client: (silent for a moment, looks uncomfortable)
Therapist: Oh, maybe I’ll take that as a “yes”?
Client: (nods) But I do not want to say. I do not want to make you feel bad.
Therapist: Well, I certainly respect if you do not want to talk about it if you do not
feel comfortable. But, if you’ll allow me . . . I know that we’ve been talking about
how it’s difficult for you to speak up and say something “out there” when someone
does something that is hurtful to you. It may not be easy but I wonder if you could
 273 Sexual Orientation

practice speaking up in here, where it may feel safer. I understand I may have done
something I didn’t realize made you feel unsafe here, and I would like to have the
chance to explore that with you and try to make it right, if you are willing. It’s your
choice.
Client: It really was not a big deal. I think I’m just overreacting.
Therapist: Isn’t that what we’ve been talking about though? About how society makes
you feel like you’re overreacting to things that truly are offensive so that you will not
speak up? I want you to know that I do not think you’re overreacting.
(The client goes on to remind the therapist of a time that the therapist made an
off-hand remark that involved an offensive stereotype.)
Therapist: I appreciate you sharing that with me. I know that was difficult for you.
I care about our relationship and I was not aware of my comment’s impact on you.
Client: I just do not want things to be weird between us. I know you probably didn’t
even know that you hurt my feelings.
Therapist: You’re right, I didn’t know, but I am genuinely sorry I made a joke without
considering the true meaning of my words and how they might affect you. I appre­
ciate your perspective on the experience as it helps me learn and grow. I need to be
more aware of how my privilege affects my behaviors and my clients.
Client: You do not have to be sorry. I’m sorry I made you feel bad.
Therapist: You were the one who was hurt in this situation; you do not have to apolo­
gize to me. I do not know if you feel comfortable to do it now, or if you would rather
process the experience with another safe person, but I am interested to know how
that experience impacted you and how it impacted the way you perceived our rela­
tionship. (p. 365)

A second example illustrates the integration of accurate knowledge. Russell and

Hawkey (2017) offered an example of how to do this in a “stigma-informed” approach
to psychotherapy. Rather than focusing on a specific case, these authors delineated
how clinicians can integrate knowledge about anti-LGBQ+ stigma in their work across
clients. They emphasize the client’s appraisal of the stigma along with the use of posi­
tive coping strategies to mitigate the impact of stigma on the client.
First, this approach requires that the therapist acquire accurate knowledge about
the context of stigma that LGBQ+ people experience. In addition to the scholarly lit­
erature, public resources are available through national organizations such as the APA
Public Interest Directorate, APA Society for the Psychology of Sexual Orientation
and Gender Diversity, Lambda Legal, National LGBTQ Task Force, and the Williams
Institute. Next, grounded in such knowledge, the therapist builds on the individualized
understanding of the client to situate stigma and oppression in external contexts rather
than in the client’s self-blaming internalization. The therapist facilitates sociopolitical
analysis that connects heterosexism with other systems of stigmatization and oppres­
sion (e.g., racism, sexism). Finally, the therapist promotes positive coping and support
for the client; this could include the therapist and client engaging in immediate advo­
cacy and in broader social justice activism. Such strategies could include recognizing
and challenging signs of internalized stigma in oneself, drawing on social support
 274 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

systems and affirmative communities, and engaging in collective social and political
action.
Another clinical example of a stigma-informed psychotherapy is called Effective
Skills to Empower Effective Men (ESTEEM ), a 10-session individual treat­
ment designed as a transdiagnostic minority stress therapy for gay and bisexual
cisgender men. Pachankis (2014, 2015) developed this intervention by drawing from
propositions that minority stressors, including discrimination and oppression, vigi­
lance and anticipation of stigma, self-blaming internalization of stigma, and conceal­
ment of stigmatized identity, can tax LGBQ+ people’s mental health (Meyer, 2003).
In ESTEEM , psychotherapists work with clients to (a) normalize the adverse impact
of minority stressors, (b) facilitate emotion awareness and regulation, (c) reduce
avoidance of difficult and painful emotions, (d) promote assertive communication,
(e) restructure minority stress cognitions (e.g., anticipation of rejection), (f) validate
clients’ strengths, (g) build support relationships, and (h) affirm healthy and rewarding
expressions of sexuality. These treatment principles are posited to disrupt the pathways
between minority stressors and psychological symptomatology. Across these domains,
and especially with restructuring minority stress cognitions, it is important to remain
grounded in the reality of clients’ experiences of minority stressors and not to mini­
mize these experiences. Again, LGBQ+ affirmative psychotherapy involves calibrating
the integration of such strategies with an individualized understanding of the client’s
experiences.
As a final example, Fassinger (2017) offers a clinical illustration that moves away
from a preoccupation with affirming specific LGBQ+ identities to affirming the
client’s transgression of restrictive sexual orientation- and gender-related norms and
power inequalities. Fassinger described this as a transgression-affirmative nested-
narrative identity construction and enactment (NICE) therapy. Consider a client
with whom Fassinger worked for many years. The client was a 34-year-old, single,
professional African American woman who presented with job-related stress and
psychological symptomatology. Through the course of therapy, the client gradually
discussed her attraction to women, and the implications of this for various aspects
of her life such as her family, religious community, and career. There was not a single
moment of sexual orientation disclosure and invocation of LGBQ+ affirmative psy­
chotherapy behaviors. Rather, the entire process of therapy involved a feminist af­
firmative approach.
Fassinger (2017) explained, “I provided openness, collaboration, support, education,
and validation of whoever and wherever she was in her identity journey—which even­
tually led her to romantic relationships with women” (pp. 19-20). She also described
that the therapy could have been improved by a more deliberate transgression-
affirmative approach that involved “collaborative coconstruction of a life story needing
some deconstruction-examination and possible reconstruction/revision” (p. 44). Such
therapy aims to help clients build a coherent life narrative that includes their gender
and sexuality, rather than focusing on a sexual orientation identity label and tailored
therapy behaviors. O f paramount importance, this approach reclaims the transgres­
sion of systems of inequality (e.g., same-sex sexual attractions) as a strength and source
 275 Sexual Orientation

of power. This affirmation of clients’ transgressions of systems of inequality is a core

aim of the therapy and a form of social justice activism.

LANDMARK STUDIES

Although research on the outcomes of LGBQ+ affirmative psychotherapies is limited,

we highlight four landmark studies. The first two studies are significant landmarks that
profoundly shifted understanding of LGBQ+ people’s mental health and are founda­
tional to LGBQ+ affirmative psychotherapies. The two additional studies focus specif­
ically on psychotherapy with LGBQ+ people.
Evelyn Hooker’s (1957) groundbreaking research challenged the mental health
professions’ codification of LGBQ+ people as pathological. Observing that the pre­
vailing view of “homosexuality” was based on data from clinical or incarcerated gay
men, Hooker sampled nonclinical gay and heterosexual men matched on a variety of
sociodemographic characteristics. She administered standard projective tests to these
men and had clinical experts review the masked test results to determine the sexual
orientation of the participants. The study revealed that the experts could not accurately
identify the sexual orientation of participants and did not rate gay and heterosexual
men differently on psychological adjustment. This research, along with years of sub­
sequent activism, led iteratively to the removal of “homosexuality” and its variants as
mental disorders (Moradi, 2016).
Another landmark that shaped contemporary understanding of LGBQ+ mental
health and psychotherapy is Meyer’s (1995) study on minority stressors and mental
health. Drawing on theory and research with lesbian women (Brooks, 1981), Meyer
surveyed gay men about their psychological symptomatology and minority stressors,
including experiences of anti-LGBQ+ prejudice and discrimination, internalized
stigma, and anticipation and vigilance of stigma. Results of this study revealed that
each of these minority stressors accounted for unique variance in gay men’s psycho­
logical symptomatology. These findings and minority stress theory were extended to
LGBQ+ populations more broadly, forging new ground by identifying the stress of so­
cietal stigma and prejudice as a risk factor for psychological symptomatology among
LGBQ+ populations.
Building on these paradigm-s hifting studies, two more recent studies serve as
landmarks for understanding psychotherapy outcomes with LGBQ+ people and
outcomes of LGBQ+ affirmative psychotherapies. Each of these studies yielded valu­
able data and also illustrated current limitations to knowledge in these areas. The first
study aimed to compare psychotherapy outcomes between LGBQ+ and heterosexual
clients (Mondragon et al., 2015). The researchers compared pre- and posttherapy
outcomes for clients categorized as sexual minority with clients categorized as not
sexual minority who received mental health services at a university counseling center.
This study inferred sexual minority status in two ways. First, clients who indicated
on an intake form that they experienced distress related to sexual identity/orientation
were categorized as sexual minority. Second, therapy notes were coded for content
indicating a sexual minority client. Expectedly, there was some overlap between these
 276 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HA T WO RK

two groups. The pattern of results indicated that clients categorized as sexual minority
generally did not differ from the control groups on posttreatment distress, suggesting
comparable treatment outcomes across groups. Unsurprisingly, the sexual identity dis­
tress group had higher pretreatment distress than the control group not matched on
distress.
This study broke new ground by attempting to compare psychotherapy outcome
between LGBQ+ and heterosexual clients and suggesting that there are no differences.
However, these findings must be interpreted within the limitations of how the com­
parison groups were defined. Specifically, heterosexual people could experience sexual
identity distress, and sexual minority people may not experience sexual identity distress
or disclose sexual minority status in therapy. As such, the sexual identity distress group
in the study could have included some heterosexual clients reporting sexual identity
distress and excluded LGBQ+ clients not reporting sexual identity distress. Moreover,
LGBQ+ clients who did not discuss their sexual identity in therapy would have been
excluded from the therapy notes group and included in the control groups. Thus, while
this study forged new ground by suggesting no differences in psychotherapy outcomes
between LGBQ+ and heterosexual clients, the study’s operationalization of LGBQ+
versus heterosexual status warrants careful consideration and future refinement.
The final landmark study offers an example of an LGBQ+ affirmative psycho­
therapy that explicitly considered sociopolitical context (Pachankis et al., 2015). This
study evaluated outcomes of the aforementioned transdiagnostic ESTEEM with a
sample of young, self-identified gay and bisexual cisgender men. In this waitlist con­
trol study, participants in the immediate ESTEEM treatment group were compared
to participants in a waitlist control condition who received the ESTEEM treatment
subsequently. Results indicated that, relative to participants on the waitlist, those re­
ceiving immediate treatment improved on a range of symptomatology, including al­
cohol abuse, depressive symptoms, sexual compulsivity, condom use self-efficacy,
and anxiety. Many of these improvements were maintained at six-month follow-up.
A similar pattern of improvement was found in pooled analyses comparing pre- and
posttreatment across conditions. These results are promising in terms of the efficacy of
ESTEEM with cisgender gay and bisexual men. However, their generalizability to other
LGBQ+ populations remains to be examined. Moreover, ESTEEM has not yet been
evaluated against another form of psychotherapy.

RESULTS OF PREVIOUS META-ANALYSES

As the scope of the foregoing studies suggests, there is a paucity of controlled research
on the outcomes of LGBQ+ affirmative psychotherapies. Hence, there are no prior
meta-analyses on the outcomes of LGBQ+ affirmative therapies or on psychotherapy
outcomes with LGBQ+ people compared to outcomes with heterosexual people.
Indeed, a recent systematic review of randomized controlled trials (RCTs) of psycho­
logical interventions for anxiety and depression revealed that no studies reported ana-
lyzable data about sexual orientation, and only one study reported any information in
 277 Sexual Orientation

this regard (Heck et al., 2017). Moreover, key prior reviews of literature related to psy­
chotherapy with LGBQ+ populations concluded that no studies evaluated outcomes
of a particular psychotherapy approach with LGBQ+ people (e.g., Bieschke et al.,
2007; Harrison, 2000; King et al., 2007). Studies included in these reviews addressed a
range of themes such as LGBQ+ clients’ therapy utilization, views of helpful and un­
helpful therapy characteristics, and satisfaction with therapy. Finally, a recent system­
atic review of cultural adaptations of health and mental health services (Healey et al.,
2017) revealed only one study that focused on any subgroup of LGBQ+ populations.

META-ANALYTIC REVIEW

We conducted a new search for empirical studies of the outcomes of LGBQ+ affirmative
psychotherapies to determine the feasibility of a meta-analysis. We considered studies
that (a) compared the outcomes of LGBQ+ tailored or affirmative psychotherapies
with outcomes of another form of psychotherapy and/or (b) compared psychotherapy
outcomes for LGBQ+ people with outcomes for heterosexual people.
To identify relevant research, we established a number of inclusion criteria. In de­
fining psychotherapies, we focused on treatments carried out as psychotherapy or
counseling, based on psychological principles and addressing psychological symptoms
(as opposed to other forms of interventions such as psychoeducation, support groups,
or highly specific interventions focusing on HIV or sexual behaviors). We also focused
our search on studies reported in English after 1990, given substantial historical shifts
in conceptualizations and contexts for psychotherapy with LGBQ+ people.
We conducted preliminary keyword searches in the following databases: Academic
Search Premier, PsycINFO, PubMed, and ISI Web of Science. We completed searches
from each of our institutions separately and determined that the most comprehen­
sive search was via ProQuest’s PsycINFO. The final search combined two sets of terms
to capture (a) psychotherapy trials (e.g., counseling, psychotherapy, “random* clinic*
trial,” “therapy n5 effectiveness”) and (b) LGBQ+ populations (e.g., asexual, bisexual*,
gay, homosexual*, lesbian, sexual minority, queer. For LGBQ+ populations, we used
terms to capture LGBQ as well as transgender populations because studies often col­
lapse across these groups. In addition to this search, we distributed a call for unpub­
lished data to the following professional listserves: APA Divisions 12, 29, 17 (overall
and LGBT section), 44, 35, 49, and 51, as well as POWR-L, a feminist psychology list.

Results
As illustrated in Figure 9.1, the initial search on June 1, 2017, for psychotherapy trials
and LGBTQ+ studies resulted in k = 2,257. The call for data to professional listservs
and review of references from key articles yielded an additional three studies. After all
duplicates were deleted, the search resulted in k = 2,191. All abstracts were downloaded
and screened along the following categories: (a) may meet inclusion for meta-analysis,
(b) addresses psychotherapy and LGBTQ+ people without data, (c) includes data
 2 78 psy c h o th er a py r ela tio n sh ips that w ork

figu re 9 .1 P R I S M A flo w d ia g r a m .

related to psychotherapy with LGBTQ+ people, (d) addresses non-psychotherapy

interventions with LGBTQ+ people, and (e) discard.
A total of 22 abstracts, including all from category (a) and potentially relevant
abstracts from category (c) and (d) were retrieved for closer full-text evaluation. Full­
text review of these publications revealed that they did not meet the meta-analysis
inclusion criteria. That is, they did not compare the outcomes of LGBQ+ tailored
or affirmative psychotherapies with outcomes of another form of psychotherapy or
compare psychotherapy outcomes for LGBQ+ people with outcomes for heterosexual
people. Specifically, these publications comprised studies that did not use a control or
comparison group condition to evaluate the psychotherapeutic intervention (k = 4),
were correlational or nonempirical (k = 4), evaluated treatments for HIV risk reduc­
tion (k = 3), and included only participants with HIV+ status thereby precluding dis­
aggregation of sexual orientation from HIV+ status (k =3). Importantly, 16 of the 22
studies included only men whereas 6 included women and men; the studies gener­
ally did not specify inclusion or exclusion of transgender people, though one study
specified exclusion of transgender people.
 279 Sexual Orientation

Eight publications came closest to the inclusion criteria, though they did not meet
these criteria and were too diverse in focus and methodology for meta-analysis (see
Table 9.2). Two of these publications were those described in the Landmark Studies
section (Mondragon et al., 2015; Pachankis et al, 2015). The other six studies focused
on substance use. One of these studies compared cognitive-behavioral therapy, con­
tingency management, cognitive-behavioral therapy plus contingency management,
and a tailored cognitive-behavioral therapy that included specific content thought to
be relevant to methamphetamine-dependent gay and bisexual men (Shoptaw et al.,
2005). All treatments were associated with improved outcomes, with a few advantages
found for treatments that included contingency management; though, overall, there
were few significant differences between treatment conditions. Two follow-up studies
extended these data. The first follow-up compared the tailored cognitive-behavioral
therapy with a tailored social support therapy for substance abuse with gay and bi­
sexual men (Shoptaw et al., 2008); both treatments reduced substance use and sexual
risk behaviors, with a few advantages observed for the tailored cognitive-behavioral
therapy over the tailored social support therapy. The second follow-up used data from
the prior two studies to compare the tailored cognitive-behavioral therapy and the
tailored social support therapy with a tailored cognitive-behavioral therapy plus con­
tingency management in a new sample (Reback & Shoptaw, 2014). Again, there were
improvements in outcomes across treatments, with a few advantages in substance use
outcomes for the tailored cognitive-behavioral therapy and a few advantages in sexual
risk behavior outcomes for the tailored cognitive-behavioral therapy plus contingency
management.
The remaining studies focused on alcohol use. One study compared alcohol use
and relationship adjustment outcomes among gay and lesbian people with alcohol
use disorder who received individual plus couples therapy or individual therapy alone
(Fals-Stewart et al., 2009); this study found better outcomes for those who received in­
dividual plus couples therapy. Another study compared alcohol use outcomes among
men who have sex with men (MSM) and have alcohol use disorders who received
motivational interviewing (4 sessions), motivational interviewing plus coping skills
training (12 sessions), or declined treatment (Morgenstern et al., 2007); this study
found that posttreatment drinking was reduced across all conditions, and there was
no significant difference between treatment conditions. The final study compared
alcohol use outcomes among MSM with problem drinking who received a placebo,
naltrexone, behavioral self-control therapy, or naltrexone plus behavioral self-control
therapy (Morgenstern et al., 2012); this study found that behavioral self-control
therapy reduced problem drinking and there was no advantage to adding naltrexone.

LIMITATIONS OF THE RESEARCH

As the present review reveals, there is a dearth of research on the outcomes of LGBQ+
psychotherapies. Fundamental to advancing research in this area is addressing the lim­
itations discussed thus far, including the need to design and evaluate measures that
assess client, therapist, and observer appraisals of the presence of the key LGBQ+
Table 9.2. Characteristics of Eight Psychotherapy Studies
Authors Purpose Sample
Fals-Stewart et al. Com pared outcom es of Gay and lesbian people with
(2009) individual plus couples alcohol use disorder
therapy with outcom es of
individual therapy alone

M ondragon et al. Com pared outcom es of University counseling and career

(2015) psychotherapy for clients center clients
categorized as sexual
m inority with clients
not categorized as sexual
m inority

M orgenstern et al. Com pared outcomes HIV-negative M SM with alcohol

(2007) o f m otivational use disorders
interviewing (4
sessions), m otivational
interviewing plus coping
skills training (12
sessions), and declining
treatment
Primary Outcome Measures Primary Findings
♦ Percentage days o f heavy ♦ B etter drinking and
drinking from Tim eline relationship adjustment
Follow Back Interview (Sobell outcom es for those who
et al., 1996) received individual plus
♦ The Dyadic Adjustment Scale couples therapy than those
(Spanier, 1976) who received individual
therapy alone
♦ The O utcom e Q uestionnaire 45 ♦ Generally, no posttreatm ent
(Lam bert, et al., 2004) group differences on distress
♦ H igher pretreatm ent distress
for clients reporting distress
related to sexual orientation
than for control group not
m atched on pretreatm ent
distress
♦ Drinks per day from Tim eline ♦ Posttreatment drinking was
Follow Back Interview (Sobell reduced across all conditions
et al„ 1996) and there was no significant
♦ Short Inventory o f Problems difference between treatment
(M iller e ta l., 1995) conditions
Morgenstern et al. Compared outcom es o f Problem drinking M SM
(2012) naltrexone, behavioral seeking to reduce but not quit
self-control therapy, drinking
naltrexone plus
behavioral self-control
therapy, and placebo

Pachankis et al. (2015) Compared outcomes Gay and bisexual cisgender m en,
o f EST EE M , a 18-35 years old, English fluent,
trans diagnostic H IV-negative status, engaging
m inority stress adapted in H IV risk behaviors,
psychotherapy with experiencing symptoms of
outcom es o f waitlist depression or anxiety, not
control group receiving regular m ental
health services
Form 90 assessing lifetim e and ♦ Behavioral self-control
recent drug use severity (M iller therapy reduced problem
& Del Boca, 1994) drinking and there was
Short Inventory o f Problems no advantage to adding
(M iller e ta l., 1995) naltrexone
Frequency o f alcohol use from
Follow Back Interview (Sobell
et al„ 1980)
A lcohol Use Disorders Com pared to the waitlist
Identification (Saunders control condition,
et al„ 1993) treatm ent resulted in
C enter for Epidem iological improvements on a range of
Studies Depression Scale symptomatology, including
(Radloff, 1977) alcohol used problems,
Overall Depression Severity depressive symptoms, sexual
& Im pairm ent Scale (Bentley compulsivity, condom use
et al„ 2014) self-efficacy, and anxiety
Overall A nxiety Severity & M any improvements
Im pairm ent Scale (Norm an m aintained at six-m onth
et al., 2006) follow-up

( continued)
Table 9.2. Continued
A uthors P urpose Sam ple

Reback & Shoptaw Compared outcomes G ay and bisexual m en

(2014) o f tailored cognitive- seeking treatm ent for
behavioral therapy m etham phetam ine abuse
and tailored social
support therapy in the
samples from Shoptaw
et al. (2005, 2008) with
outcom es o f a tailored
cognitive-behavioral
therapy plus contingency
m anagem ent treatment
in a new sample
Primary Outcome Measures Primary Findings
♦ Sexual Compulsivity Scale ♦ Sim ilar improvement found
(Kalichm an & Rom pa, 2001) in pooled analyses comparing
♦ Safer Sex Self-Efficacy all participants pre- and
Q uestionnaire (Rendina, 2014) posttreatm ent
♦ 90-day Tim e Line Follow Back ♦ No treatm ent effects for
(Sobell & Sobell, 1992) m inority stress or general risk
♦ Measures o f m inority stress risk factors
factors and general risk factors
♦ Addiction Severity Index ♦ All treatm ent conditions were
(M cLellan et al., 1992) associated with improved
♦ U rine drug screening outcomes
♦ Behavioral Q uestionnaire- ♦ A few advantages in substance
Am phetam ine (C hesney et al., use outcomes for the tailored
1997) to assess sexual risk cognitive-behavioral therapy
behaviors ♦ A few advantages in sexual
♦ Retention risk behavior outcomes for the
tailored cognitive-behavioral
therapy plus contingency
managem ent
Shoptaw et al. (2005) Compared outcom es o f G ay and bisexual m en seeking
cognitive-behavioral outpatient treatm ent
therapy, contingency for m etham phetam ine
m anagem ent, cognitive- dependence
behavioral therapy
plus contingency
m anagem ent, and a
tailored cognitive-
behavioral therapy that
included group-specific
content for gay and
bisexual m en
Shoptaw et al. (2008) Compared outcomes G ay and bisexual m en seeking
o f tailored cognitive treatm ent for any stimulant
behavioral and tailored and/or alcohol abuse
social support substance
use treatm ents for gay
and bisexual m en

Note. M S M = m e n w h o h a v e s e x w ith m e n ; E S T E E M = E ffe c tiv e S k ills to E m p o w e r E ffe c tiv e M e n .

♦ Addiction Severity Index ♦ All treatm ents were associated
(M cLellan et al., 1992) with improved outcomes
♦ U rine drug screening ♦ A few advantages observed
♦ Behavioral Q uestionnaire- for treatm ents that included
A m phetam ine (C hesney et al., contingency managem ent,
1997) to assess sexual risk though generally few
behaviors significant differences between
♦ Retention treatm ent conditions

♦ Addiction Severity Index ♦ Both treatm ent conditions

(M cLellan et al., 1992) were associated with
♦ U rine drug screening reductions in substance use
♦ Behavioral Questionnaire and sexual risk behaviors
(Chesney et al., 1997) to assess ♦ A few advantages observed
sexual risk behaviors for the tailored cognitive-
♦ Retention behavioral therapy over the
tailored social support therapy
 284 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

affirmative themes. Several additional limitations to be addressed involve broadening

the research boundaries in terms of the scope of applicability of LGBQ+ affirmative
psychotherapies, the range of psychotherapy outcomes considered, and the diversity
of LGBQ+ populations.
First, the scope of applicability of LGBQ+ affirmative psychotherapies has been
quite narrow, reflecting an implicit conceptualization that these are therapies tailored
specifically for LGBQ+ people and not others. Research is needed to investigate the
outcomes of LGBQ+ affirmative psychotherapies with all clients, across the spec­
trum of sexual orientations and identities. In such research, analyses would compare
the outcomes of psychotherapies with and without LGBQ+ affirmative ingredients;
though ethical practice limits experimental manipulation of some ingredients
with clients (e.g., anti-LGBQ+ attitudes). Similarly, research is needed to broaden
the range of presenting problems for which (a) psychotherapies are examined with
LGBQ+ populations and (b) LGBQ+ affirmative psychotherapies are examined
with all populations. W hile sexual risk behaviors and substance abuse are important
foci, there is a near exclusive focus on treatment evaluations with these presenting
concerns. Research is needed to investigate the full range of psychological symptoma­
tology and life challenges (e.g., depression, anxiety, body image, relationship distress,
vocational issues).
Second, the centrality of the context of oppression requires expanding the range of
psychotherapy outcomes considered in research. Traditional conceptualizations of fa­
vorable outcomes include decreases in symptomatology and increases in psychological
well-being. However, a functional response to anti-LGBQ+ discrimination and oppres­
sion may include increased anger, sadness, and discontent. In fact, these “symptoms”
may prove indicators of the important emotional work involved in recognizing op­
pressive systems, taking constructive action, and developing a social justice orientation
(e.g., Fischer & Good, 2004; Hercus, 1999; Moradi, 2012). Thus, such outcomes can
be measured and conceptualized as favorable psychotherapy outcomes. Measures of
symptom reduction can also be supplemented with measures of targeted outcomes,
such as minority stressors, including anticipation of stigma, internalized prejudice, and
concealment of sexual orientation (e.g., Pachankis et al., 2015). Additional outcomes
could include clients’ perceptions of key mechanisms for change, engagement in eve­
ryday activism and collective action, and social justice orientation.
Finally, sociodemographic and other forms of diversity among LGBQ+ populations
remain underexamined in psychotherapy research. Nearly all of the studies that
emerged for full-text examination in our review focused on cisgender men, mostly
recruited for HIV+ status and/or substance use problems. In these studies, the inclu­
sion or exclusion of transgender men was generally unacknowledged or transgender
men were explicitly excluded. Similarly, women (transgender inclusive) and people
with nonbinary or other gender identities were not included. Psychotherapy research
that includes LGBQ+ people of all genders and LGBQ+ populations beyond only
those with HIV+ status and substance abuse concerns is needed. Strategies delineated
for recruiting diverse samples across age, class, gender, ethnicity, race, and other
sociodemographics can be used (e.g., DeBlaere et al., 2010).
 285 Sexual Orientation

Beyond sociodemographic diversity, individual differences in levels of minority

stressors (e.g., experiences of discrimination, internalized stigma) could interact with
LGBQ+ affirmative psychotherapy ingredients to shape outcomes. Similarly, LGBQ+
affirmative psychotherapy ingredients may interact with clients’ expressed sexual ori­
entation such that outcomes differ for LGBQ+ identified clients compared to LGBQ+
nonidentified clients. Moderators such as sexual orientation identities and levels of m i­
nority stressors may be points for calibrating psychotherapy ingredients across clients.

DIVERSITY CONSIDERATIONS

LGBQ+ identities are diverse, culturally defined, and dynamic, as reflected in the
expanding inclusivity of sexual identities (e.g., L, G, B, Q, . . .). Moreover, LGBQ+
people as a group represent all ages, classes, genders, ethnicities, races, and other
sociodemographic characteristics. Acknowledging this diversity among LGBQ+
populations is critical.
To this end, it is helpful to distinguish strong intersectional analysis from superfi­
cial considerations of multiple/intersecting identities that involve blanket application
of group-l evel information or presumed cultural characteristics to individual clients
(Moradi & Grzanka, 2017). Strong feminist intersectional analysis (e.g., Collins, 1990/
2000; Crenshaw, 1989, 1991; Dill & Kohlman, 2012) requires understanding how mul­
tiple systems of oppression and privilege function simultaneously in clients’ lives and
developing interventions that attend to and challenge these systems of inequalities.
Psychotherapy practice and research informed by intersectional analysis will articu­
late, assess, and analyze the system-level constructs for which demographic variables
are implicit proxies (e.g., experiences of classism, heterosexism, racism, sexism).
Intersectional analysis may also challenge the epistemology and power inequalities
in how we evaluate psychotherapy outcomes (Moradi & Grzanka, 2017). This includes
valuing statistical significance of outcomes along with (rather than in lieu of) clients’
experiences and the benefits of psychotherapy to clients in real-world contexts. While
RCTs are considered the gold standard (Taylor & Asmundson, 2008), they present lim­
itations in evaluating LGBQ+ affirmative psychotherapies. For example, LGBQ+ af­
firmative ingredients are to be applied across psychotherapy methods, diagnoses, and
populations. This breadth can conflict with the pressures for a high degree of con­
trol in RCTs, for example in ensuring fidelity of interventions or in defining patient
populations in terms of demographics and/or diagnoses.
A complementary alternative is practice-based evidence (Barkham & Mellor-Clark,
2000), which focuses on high-quality data derived from clients and practitioners in
naturalistic settings with their contextual complexities. Indeed, data that capture rather
than control the complexity of LGBQ+ people and their presenting concerns can ad­
dress many of the limitations of prior research (e.g., inclusion criteria that focus nar­
rowly on cisgender men living with HIV). Practice-based evidence is also consistent
with intersectional analysis and the themes of LGBQ+ affirmative psychotherapies in
that it foregrounds context, clients, practitioners, and real-world complexity as integral
to evaluating psychotherapy, rather than as confounding factors to be controlled.
 286 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

TRAINING IMPLICATIONS

Training in LGBQ+ affirmative psychotherapies can first clarify that a prerequisite to

ethical practice is for clinicians to actively counteract their own biases that pathologize
and oppress LGBQ+ people (e.g., APA, 2012, Harrison, 2000; Johnson, 2012). This is
critical given the context of historical and contemporary sociopolitical stigmatization
of LGBQ+ people, both within and outside of psychotherapy (Moradi, 2016). Much
thoughtful deliberation has been offered about how trainees and practitioners can nav­
igate situations in which psychotherapists perceive a conflict between their personal
(often religious) values and working affirmatively with LGBQ+ clients (e.g., Fischer &
DeBord, 2007). We assert that any trainee or therapist who is not prepared to engage
in affirmative psychotherapy with client populations because of their gender, class, eth­
nicity, race, or sexual orientation and is not working to become prepared to do so is
not fit for the profession.
Second, training in LGBQ+ affirmative psychotherapies can emphasize acquiring
and translating accurate knowledge about LGBQ+ people’s experiences into ther­
apeutic action. Knowledge about LGBQ+ populations can be contextualized in
power dynamics around sexual orientation and its intersections with other systems
(e.g., class, gender, ethnicity, race) as well as in an individualized understanding of
the client’s experiences. Trainers can use critical reflection questions, such as those
delineated by Moradi (2017), to promote therapists’ self-analysis and intersectional
analysis in working with LGBQ+ (and other) populations. These questions can en­
courage therapists to challenge the implicit prototypes that they hold about LGBQ+
people and intersectionality; consider sexuality, culture, and intersectionality with all
clients; and engage in intersectional analysis to explicitly address systems and power
dynamics, rather than focus only on intrapersonal identities and demographics.
We caution against an “apply knowledge and stir” approach in which knowledge
about LGBQ+ populations is invoked uncritically with all (assumed) LGBQ+ clients.
Such an approach is objectifying and treats such clients as a stereotypic monolith.
Instead, the integration of any group-specific knowledge can be thoughtfully calibrated
to the specific experiences and needs of the client.
Finally, engaging in and affirming challenges to power inequalities is a necessary
component of LGBQ+ affirmative psychotherapies, and we view training in this area
as a growth edge competency for the field as a whole. Many articulations of LGBQ+
affirmative therapy end with aspirational statements about addressing sociopolitical
inequalities but stop short of calling for social justice activism (see Bieschke et al.,
2007; DeBord et al., 2017). Others include such activism as part of LGBQ+ affirma­
tive therapeutic interventions outside of the therapy room (e.g., Harrison, 2000). Yet
others caution that LGBQ+ affirmative therapies are inadequate because their goal is
to promote individual adjustment to an oppressive system; instead, such views suggest
that sociopolitical action is the only responsible form of LGBQ+ affirmative interven­
tion (e.g., Perkins, 1997). The former perspectives may align more comfortably with
traditional individualistic views of psychotherapy, and the latter perspective may elicit
reactions that social justice activism is beyond the scope of psychotherapy.
 287 Sexual Orientation

In our view, it is not realistic to train psychotherapists to engage in LGBQ+ affirm­

ative attitudes, knowledge, and actions in the therapy room but leave them to engage
in non-affirmative practices upon leaving the room. As a gay client stated about his
former therapist’s purported neutrality on his sexual orientation: “I see the false neu­
trality that existed. I don’t think you can believe in a pathological and a normative
view of homosexuality at the same time. I don’t think that being neutral exists as a
compromise between these two states” (Shidlo & Schroeder, 2001 as cited in Shidlo &
Gonsiorek, 2017).
Principles of LGBQ+ affirmative psychotherapies have to be practiced and fostered
by everyday actions and in all relationships. Drawing from Steinem’s (1983) examples
of anti-sexist “outrageous acts and everyday rebellions,” LGBQ+ affirmative actions
in and outside of psychotherapy might include challenging anti-LGBQ+ humor and
discourse, displaying LGBQ+ affirming images and information in the office, using
LGBQ+ inclusive language in assessment materials and conversations, asking one’s or­
ganizations and communities about their stance and actions toward LGBQ+ people,
refusing to contribute to or participate in organizations with anti-LGBQ+ practices, or
publicizing or joining local social justice efforts (e.g., bookstore, community center).
We believe that such everyday acts are essential to training in and practice of LGBQ+
affirmative psychotherapies.

THERAPEUTIC PRACTICES

We conclude this chapter by advancing therapeutic practices along the four key themes
of LGBQ+ affirmative psychotherapies. We recommend these practices based on the
available literature but reiterate the need for research that evaluates the outcomes of
these four themes and LGBQ+ affirmative treatments.

Counteracting Anti-LGBQ+ Therapist Attitudes and Enacting

LGBQ+ Affirmative Attitudes
♦ Counteract biases that may pathologize and oppress LGBQ+ identities and people.
Research with LGBQ+ clients delineates concrete examples of anti-LGBQ+ biases
to be avoided in therapists’ behaviors. Examples of such behaviors include assuming
that sexual orientation is the cause of presenting concerns, avoiding or minimizing
discussions of sexual orientation, overidentifying with LGBQ+ clients in ways that
are defensive or objectifying (e.g., “I have a gay friend”), operating on stereotypes
about LGBQ+ people or on heteronormative assumptions and biases, and enacting
pathologizing assumptions that LGBQ+ people need therapy or that LGBQ+
identities are dangerous or problematic (Shelton & Delgado-Romero, 2011).
♦ Use inclusive language and clients’preferred terms. In clinical contacts and measures,
until the client’s preferred terminology is assessed, use inclusive language such as
“your partner or spouse” rather than language that assumes partner/spouse gender.
Use of such inclusive language is associated with LGB people’s positive views of
therapists, greater willingness and comfort to disclose sexual orientation identity, and
 288 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HA T WO RK

greater likelihood to return to psychotherapy (Dorland & Fischer, 2001). Moreover,

contrary to speculation that such LGBQ+ inclusive language might confuse or
alienate heterosexual clients, inclusive language was found to be unrelated to
heterosexual peoples perceptions of therapist credibility and utilization intent and
was related positively to their willingness to disclose (Ross et al., 2013). In addition,
pay attention to and inquire about clients’ preferred language to describe themselves
and their relationships (e.g., “How do you prefer to describe your sexual orientation
or romantic attractions?” “How do you prefer to refer to your partner?”) and mirror
clients’ preferred language.
♦ Demonstrate an inclusive stance in professional materials and in the therapy
environment. There is evidence that LGBQ+ clients notice heteronormative bias in
the therapy environment, for example, in the absence of LGBQ+ representation in
displays and materials, and experience such bias as a form of subtle discrimination
in therapy (Shelton & Delgado-Romero, 2011). Examples of strategies for
demonstrating an inclusive stance include assessing sexual orientation using best
practices (see recommendations in the Measures section), and using brochures,
displays, and artwork that are inclusive and affirmative of LGBQ+ people.

Acquiring Accurate Knowledge about LGBQ+ People’s

Experiences and Their Heterogeneity
♦ Acquire accurate knowledge about the sociopolitical stigmatization and oppression o f
LGBQ+ people as well as their broader life experiences. Each of the clinical examples of
LGBQ+ affirmative psychotherapies and outcome studies in this review emphasizes
the need to integrate knowledge about anti-LGBQ+ stigma and minority stress. In
addition, acquire knowledge about a range of life issues such as identity formation
and management (e.g., identity concealment and disclosure strategies and their
consequences), family structures (e.g., families of choice, marriage, partnerships,
consensual nonmonogomy, parenting within and outside of legal status), workplace
experiences (discriminatory and antidiscrimination policies and laws, hostile or
affirming careers), LGBQ+ affirmative support communities, and the heterogeneity
of LGBQ+ people and their needs across dimensions of sociodemographic diversity
(e.g., APA, 2012, Balsam et al., 2006).
♦ Recognize LGBQ+ people’s strengths and resilience that may promote well-being and
mitigate minority stress and engage these strengths in psychotherapy. Research with
LGBQ+ people reveals a range of strengths, including actively pursuing authenticity
in self-definition, developing freedom from gender-specific roles, cultivating
cognitive flexibility, and being involved in social justice activism (e.g., Brewster et al.,
2013; Riggle et al., 2008). As one example, cognitive flexibility buffered bisexual
people’s mental health against experiences of antibisexual prejudice (Brewster et al.,
2013). Thus therapists can explore how LGBQ+ clients’ life experiences may foster
cognitive flexibility and how this strength can be integrated in interventions to
promote positive psychotherapy outcomes.
 289 Sexual Orientation

Calibrating Integration of Accurate Knowledge about

LGBQ+ People’s Experiences and Their Heterogeneity
into Therapeutic Actions
♦ Individualize understanding and treatment o f a given client, without overemphasizing
or underemphasizing the centrality o f LGBQ+ status. Individualization of knowledge
about LGBQ+ people is consistent with an “informed not-knowing stance”
(Anderson & Goolishian, 1992; Laird, 2000) whereby the therapist expresses genuine
curiosity to understand the client more deeply, rather than assume a preconceived
understanding. This ability to understand the client is coupled with, and in fact
predicated on, therapists taking responsibility to acquire accurate background
knowledge about LGBQ+ people, rather than expect clients to educate them.
♦ Consider individual differences in how stigma and minority stressors shape clients’ lives
and psychotherapy. Draw from the stigma-informed approach to therapy (Russell &
Hawkey, 2017) and the promising empirical evidence for ESTEEM (Pachankis, 2014,
2015) to assess and integrate individual differences in minority stress experiences.
In these approaches, practitioners first acquire knowledge about anti-LGBQ+
stigma and oppression, then work with clients to develop a rich individualized
understanding of how such stigma manifests in clients’ experiences.

Engaging In and Affirming Challenges to Power Inequalities

♦ Affirm clients’ challenging o f power inequalities and engage in such challenging outside
o f therapy. Within therapy, consider using the transgression-affirming NICE model
(Fassinger, 2017), which emphasizes fostering an affirmative stance toward the
client’s transgression of restrictive sexual orientation- and gender-related social
norms and power inequalities. Outside of therapy, practice and foster the themes
of LGBQ+ affirmation in everyday actions and relationships. Such actions in and
outside of psychotherapy can include the described examples of “outrageous acts and
everyday rebellions.”
♦ Engage in sociopolitical analysis as a component o f LGBQ+ affirmative psychotherapies,
distinctfrom cultural tailoring. Without careful sociopolitical analysis, cultural
tailoring efforts may blur into cultural stereotyping identified by LGBQ+ clients as
subtle biases in psychotherapy (Shelton & Delgado-Romero, 2011). For example,
culturally tailored substance use treatments may discuss LGBQ+ bars and clubs
as cultural triggers for substance abuse or draw parallels between disclosing ones
drug problems and disclosing one’s sexual identity. Instead, sociopolitical analysis in
psychotherapy will ideally situate LGBQ+ bars as outcomes of heterosexist systems
that necessitate such safer social spaces, destigmatize these spaces by acknowledging
that heterosexual-dominant bars also trigger substance use, and frame coming out
as an outcome of heteronormative assumptions that make it necessary for LGBQ+
people, but not heterosexual people, to constantly judge the context of anti-LGBQ+
threat.
 290 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

♦ Apply LGBTQ+ affirmative principles to all clients. All clients have a narrative—
articulated or not—about how their life is shaped by gender, sexuality, and other
sociopolitical systems; all clients’ life narratives and how they transgress systems
of inequality can be examined constructively and collaboratively in therapy; and
transgressions that challenge systems of inequality can be affirmed in all clients.
This approach can help all clients strive for more self and collective authenticity,
actualization, and equity.

REFERENCES
A m erican Psychological A ssociation. (2012). Guidelines for psychological practice with les­
bian, gay, and bisexual clients. American Psychologist, 67, 1 0 -4 2 . https://www.doi.org/
10.1037/a0024659
Anderson, H. & Goolishian, H. (1992). The client is the expert: A not-know ing approach to
therapy. In S. M cNam ee & K. J. Gergen (Eds.), Therapy as social construction (pp. 2 5 -3 9 ).
Newbury Park, CA: SAGE.
Balsam, K. F., M artell, C. R., & Safren, S. A. (2006). Affirmative cognitive-behavioral therapy
with lesbian, gay, and bisexual people. In P. A. Hays & G. Y. Iwamasa (Eds.), Culturally re­
sponsive cognitive-behavioral therapy: Assessment, practice, and supervision (pp. 2 2 3 -2 4 3 ).
W ashington, DC: A m erican Psychological A ssociation Press.
Barkham , M ., & M ellor-Clark, J. (2000). Rigour and relevance: Practice-based evidence in the
psychological therapies. In N. Rowland & S. Goss (Eds.), Evidence-based counselling and
psychological therapies (pp. 1 2 7 -1 4 4 ). London, England: Routledge.
Bem , S. L. (1993). The lenses o f gender: Transforming the debate on sexual inequality. New
Haven, CT: Yale University Press.
Bentley, K. H., Gallagher, M . W., Carl, J. R., & Barlow, D. H. (2014). Development and valida­
tion o f the Overall Depression Severity and Im pairm ent Scale. Psychological Assessment,
2 6 (3 ), 8 1 5 -8 3 0 . https://www.doi.org/10.1037/a0036216
Bidell, M . P. (2005). The Sexual Orientation C ounselor Com petency Scale: Assessing attitudes,
skills, and knowledge o f counselors working with lesbian, gay, and bisexual clients.
Counselor Education & Supervision, 44, 2 6 7 -2 7 9 . https://www.doi.org/10.1002/).1556-
6978.2005.tb01755.x
Bieschke, K. J., Paul, P. L., & Blasko, K. A. (2007). Review o f em pirical research focused on
the experience o f lesbian, gay, and bisexual clients in counseling and psychotherapy. In
K. Bieschke, R. Perez, & K. D eBord (Eds.), Handbook o f counseling and psychotherapy
with lesbian, gay, bisexual, and transgender clients (2nd ed., pp. 2 9 3 -3 1 6 ). W ashington,
DC: A m erican Psychological Association.
Bieschke, K. J., Perez, R. M ., & DeBord, K. A. (2007). Handbook o f counseling and psychotherapy
with lesbian, gay, bisexual, and transgender clients (2nd ed.). W ashington, DC: Am erican
Psychological Association.
Brewster, M. E., M oradi, B., DeBlaere, C., & Velez, B. L. (2013). Navigating the borderlands: The
roles o f m inority stressors, bicultural self-efficacy, and cognitive flexibility in the m ental
health o f bisexual individuals. Journal o f Counseling Psychology, 60, 5 4 3 -5 5 6 . https://www.
doi.org/10.1037/a0033224
Brooks, V. R. (1981). Minority stress and lesbian women. Lexington, M A: Lexington Books.
 291 Sexual Orientation

Burkard, A. W., Pruitt, N. T., Medler, B. R., & Stark-Booth, A. M. (2009). Validity and reliability
o f the lesbian, gay, bisexual working alliance self-efficacy scales. Training and Education in
Professional Psychology, 3 , 3 7 -4 6 . https://www.doi.org/10.1037/1931-3918.3.1.37
Chesney, M. A., Cham bers, D. B., & Kahn, J. O. (1997). Risk behavior for H IV infection among
participants in preventive H IV vaccine trials. Journal o f Acquired Immune Deficiency
Syndromes and Human Retrovirology, 16, 2 6 6 -2 7 1 .
Collins, P. H. (2000). Black feminist thought: Knowledge, consciousness, and the politics o f em­
powerment. New York, NY: Routledge. (O riginal work published 1990)
Crenshaw, K. W. (1989). Dem arginalizing the intersection o f race and sex: A Black fem inist
critique o f antidiscrim ination doctrine, fem inist theory and antiracist politics. University
o f Chicago Legal Forum, 140, 1 3 9-167.
Crenshaw, K. W. (1991). Mapping the margins: Intersectionality, identity politics, and vio­
lence against women o f color. Stanford Law Review, 46, 124 1 -1 2 9 9 . https://www.doi.org/
10.2307/1229039
Crisp, C. (2006). The G ay Affirmative Practice Scale (G A P): A new m easure for assessing cu l­
tural com petency with gay and lesbian clients. Social Work, 51, 1 1 5 -1 2 6 . https://www.doi.
org/10.1093/sw/51.2.115
DeBlaere, C., Brewster, M. E., Sarkees, A., & M oradi, B. (2010). Conducting research with LGB
people o f color: M ethodological challenges and strategies. The Counseling Psychologist, 38,
331 -3 6 2 . https://www.doi.org/10.1177/0011000009335257
D eBord, K. A., Fischer, A. R., Bieschke, K. J., & Perez, R. M. (2017). Handbook o f sexual ori­
entation and gender diversity in counseling and psychotherapy. W ashington, DC: A m erican
Psychological Association.
Dill, B. T., & Kohlm an, M . H. (2012). Intersectionality: A transform ative paradigm in fem ­
inist theory and social justice. In S. N. H esse-Biber (Ed.), The handbook o f feminist re­
search: Theory and praxis (2nd ed., pp. 1 5 4 -1 7 4 ). London, England: SAGE. https://www.
doi.org/10.4135/ 9781483384740.n 8
Dillon, F. R., Alessi, E. J., Craig, S., Eber-sole, R. C., Kumar, S. M ., & Spadola, C. (2015).
Development o f the lesbian, gay, and bisexual affirmative counseling self-efficacy
inventory— short form (L G B -C SI-SF ). Psychology o f Sexual Orientation and Gender
Diversity, 2 (1 ), 8 6 -9 5 . https://www.doi.org/10.1037/sgd0000087
Dillon, F. R., & W orthington R. L. (2003). The lesbian, gay, and bisexual affirmative coun­
seling self-efficacy inventory (LG B -C SI): Development, validation, and training
implications. Journal o f Counseling Psychology, 50, 2 3 5 -2 5 1 . https://www.doi.org/10.1037/
0 0 2 2 -0167.50.2.235
Dillon, F. R., W orthington, R. L., & M oradi, B. (2011). Sexual identity as a universal process. In
S. J. Schwartz, K. Luyckx, & V. L. Vignoles (Eds.), Handbook o f identity theory and research
(pp. 6 4 9 -6 7 0 ). New York, NY: Springer.
Dorland, J. M ., & Fischer, A. R. (2001). Gay, lesbian, and bisexual individuals’ perceptions: An
analogue study. The Counseling Psychologist, 29, 5 3 2 -5 4 7 . https://www.doi.org/10.1177/
0 0 1 1 0 0 0 001294004
Fals-Stewart, W., O’Farrell, T. J., & Lam, W. K. (2009). Behavioral couple therapy for gay and
lesbian couples with alcohol use disorders. Journal o f Substance Abuse Treatment, 37, 3 7 9 ­
387. https://www.doi.org/10.1016/j.jsat.2009.05.001
Fassinger, R. E. (2017). Considering constructions: A new m odel o f affirmative therapy. In
K. A. DeBord, A. R. Fischer, K. J. Bieschke, & R. M. Perez (Eds.), Handbook o f sexual
 292 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

orientation and gender diversity in counseling and psychotherapy (pp. 1 9 -5 0 ). W ashington,

DC: A m erican Psychological Association.
Fischer, A. R., & D eBord, K. A. (2007). Perceived conflicts between affirm ation o f religious
and sexual diversity. In K. J. Bieschke, R. M. Perez, & K. A. D eBord, (Eds.), Handbook o f
counseling and psychotherapy with gay, lesbian, bisexual, and transgender clients (2nd ed.,
pp. 3 1 7 -3 3 9 ). W ashington, DC: A m erican Psychological Association.
Fischer, A. R., & G ood, G. E. (2004). Women’s fem inist consciousness, anger, and psycholog­
ical distress. Journal o f Counseling Psychology, 51, 4 3 7 -4 4 6 . https://www.doi.org/10.1037/
0 0 2 2 -0167.51.4.437
Gates, G. J. (2011). How many people are lesbian, gay, bisexual, and transgender? Los Angeles,
CA: W illiam s Institute, University o f California, Los Angeles School o f Law.
Harrison, N. (2000). G ay affirmative therapy: A critical analysis o f the literature. British Journal
o f Guidance & Counselling, 28, 3 7 -5 3 . https://www.doi.org/10.1080/030698800109600
Healey, P, Stager, M . L., Woodm ass, K., Dettlaff, A. J., Vergara, A., Janke, R., & Wells, S. J. (2017).
Cultural adaptations to augment health and m ental health services: A systematic review.
BMC Health Services Research, 17, 8. https://www.doi.org/10.1186/s12913-016-1953-x
Heck, N., M irabito, L., LeMaire, K., Livingston, N., & Flentje, A. (2017). O m itted data in
randomized controlled trials for anxiety and depression: A systematic review o f the in ­
clusion o f sexual orientation and gender identity. Journal o f Consulting and Clinical
Psychology, 85, 7 2 -7 6 . https://www.doi.org/10.1037/ccp0000123
Hercus, C. (1999). Identity, em otion, and fem inist collective action. Gender and Society, 13,
3 4 -5 5 . https://www.doi.org/10.! 177/089124399013001003
Hooker, E. (1957). The adjustm ent o f the m ale overt hom osexual. Journal o f Projective
Techniques, 21, 1 8 -3 1 . https://www.doi.org/10.1080/08853126.1957.10380742
Intersex Society o f North A merica. (n.d.). About. Retrieved from http://www.isna.org
Johnson, S. D. (2012). Gay affirmative psychotherapy with lesbian, gay, and bisexual
individuals: Im plications for contem porary psychotherapy research. American Journal o f
Orthopsychiatry, 82, 5 1 6 -5 2 2 . https://www.doi.org/10.1111/j.1939-0025.2012.01180.x
Kalichm an, S. C., & Rompa, D. (2001). The Sexual Compulsivity Scale: Further development
and use with HIV-positive persons. Journal o f Personality Assessment, 76(3), 3 7 9 -3 9 5 .
https://www. doi.org/10.1207/S153 27752JPA 7603_02
King, M ., Semlyen, J., Killaspy, H., Nazareth, I., & O sborn, D. (2007). A systematic review o f
research on counseling and psychotherapy fo r lesbian, gay, bisexual & transgender people .
Leicestershire, England: British A ssociation for Counseling and Psychotherapy. Retrieved
from https://www.peter-ould.net/wp-content/uploads/king-systematic-review-1 .pdf
Laird, J. (2000). Theorizing culture: Narrative ideas and practice principles. Journal o f Feminist
Family Therapy, 11, 9 9 -1 1 4 . https://www.doi.org/10.1300/J086v11n04_08
Lam bert, M . J., Gregersen, A. T., & Burlingam e, G. M. (2004). The Outcom e Questionnaire-45.
In M. E. M aruish (Ed.), The use o f psychological testingfo r treatment planning and outcomes
assessment: Instruments fo r adults (pp. 1 9 1 -2 3 4 ). Mahwah, NJ: Erlbaum.
Logie, C., Bridge, T. J., & Bridge, P. D., (2007). Evaluating the phobias, attitudes, and cultural
com petence o f m aster o f social work students toward the LG BT populations. Journal o f
Homosexuality, 53, 2 0 1 -2 2 1 , https://www.doi.org/10.1080/00918360802103472
Matthews, C. R. (2007). Affirmative lesbian, gay, and bisexual counseling with all clients. In K.
J. Bieschke, R. M. Perez, & K. A. D eBord (Eds.), Handbook o f counseling and psychotherapy
with lesbian, gay, bisexual, and transgender clients, (2nd ed., pp. 2 0 1 -2 1 9 ). W ashington,
DC: A m erican Psychological Association.
 293 Sexual Orientation

M cLellan, A.T., Kushner, H., Metzger, D., Peters, R., Sm ith, I., Grissom , G ., . . . Argeriou,
M . (1992). The fifth edition o f the Addiction Severity Index. Journal o f Substance Abuse
Treatment, 9 (3 ), 1 99-213.
Meyer, I. H. (1995). M inority stress and m ental health in gay men. Journal o f Health and Social
Behavior, 36, 3 8 -5 6 .
Meyer, I. H. (2003). Prejudice, social stress, and m ental health in lesbian, gay, and bisexual
populations: Conceptual issues and research evidence. Psychological Bulletin, 129, 6 7 4 ­
697. https://www.doi.org/10.1037/0033-2909.129.5.674
M iller, W. R., & Del Boca, F. K. (1994). M easurem ent o f drinking behavior using the Form 90
fam ily o f instrum ents. Journal o f Studies on Alcohol, Supplement, 1 2 , 1 1 2-118.
M iller, W. R., Tonigan, J. S., & Longabaugh, R. (1995). The Drinker Inventory o f Consequences
(DrInC): An instrument fo r assessing adverse consequences o f alcohol abuse. Test manual.
P roject M A TCH M onograph Series, Vol. 4. Rockville, M D : National Institute on Alcohol
Abuse and Alcoholism .
M ondragon, S. A., Lambert, M. J., Nielsen, S. L., & Erikson, D. (2015). Com parative psy­
chotherapy outcom es o f sexual m inority clients and controls. International Journal o f
Education & Social Science, 2 (4), 1 4 -3 0 . Retrieved from http://www.ijessnet.com
M oradi, B. (2 0 1 2 ). F em in ist social ju stice orientation: An ind icator o f optim al
fun ctioning? The Counseling Psychologist, 40, 1 1 3 3 -1 1 4 8 . https://www.doi.org/10.1177/
0011000012439612
M oradi, B. (2016). Lesbian, gay, bisexual, and transgender issues. In H. S. Friedm an (Ed.),
Encyclopedia o f mental health (2nd ed., Vol. 3, pp. 1 9 -2 4 ). Oxford, England: Elsevier.
M oradi, B. (2017). (Re)focusing intersectionality in psychology: From social identities back
to systems o f oppression and privilege. In K. DeBord, R. M . Perez, A. R. Fischer, & K.
J. Bieschke (Eds.). The handbook o f sexual orientation and gender diversity in counseling
and psychotherapy (3rd ed., pp. 1 0 5 -1 2 7 ). W ashington, DC: A m erican Psychological
Association.
M oradi, B., & Grzanka, P. R. (2017). Using intersectionality responsibly: Toward critical epis­
temology, structural analysis, and social justice activism. Journal o f Counseling Psychology,
64, 5 0 0 -5 1 3 . https://www.doi.org/10.1037/cou0000203
M orgenstern, J., Irwin, T. W., W ainberg, M. L., Parsons, J. T., M uench, F., Bux, D. A., . . .
Schulz-H eik, J. (2007). A randomized controlled trial o f goal choice interventions for al­
cohol use disorders am ong m en who have sex with men. Journal o f Consulting and Clinical
Psychology, 75, 7 2 -8 4 . https://www.doi.org/10.1037/0022-006X.75.L72
M orgenstern, J., Kuerbis, A. N., Chen, A. C., Kahler, C. W., Bux, D. A., & Kranzler, H. R.
(2012). A randomized clinical trial o f naltrexone and behavioral therapy for problem
drinking m en who have sex with m en. Journal o f Consulting and Clinical Psychology, 80,
8 6 3 -8 7 5 . https://www.doi.org/10.1037/a0028615
N orm an, S. B., Cissell, S. H., M eans-Christensen, A. J., & Stein, M . B. (2006). Development
and validation o f an overall anxiety severity and im pairm ent scale (O A SIS). Depression
and Anxiety, 2 3(4), 2 4 5 -2 4 9 . https://www.doi.org/10.1002/da.20182
Pachankis, J. E. (2014). Uncovering clinical principles and techniques to address m inority
stress, m ental health, and related health risks am ong gay and bisexual m en. Clinical
Psychology: Science and Practice, 2 1 ,3 1 3 - 3 3 0 . https://www.doi.org/10.1111/cpsp.12078
Pachankis, J. E. (2015). A transdiagnostic m inority stress treatm ent approach for gay and
bisexual men’s syndemic health conditions: Archives o f Sexual Behavior, 44, 1 8 43-1860.
https://www.doi.org/10.1007/s10508-015-0480-x
 294 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

Pachankis, J. E., & Goldfried, M . R. (2013). C linical issues in working with lesbian, gay, and
bisexual clients. Psychology o f Sexual Orientation and Gender Diversity, 1, 4 5 -5 8 . https://
www.doi. org/10.1037/23 2 9 -0 3 8 2 .1.S.45
Pachankis, J., Hatzenbuehler, M ., Rendina, H., Safren, S., & Parsons, J. (2015). LGB-affirmative
cognitive-behavioral therapy for young adult gay and bisexual m en: A randomized
controlled trial o f a transdiagnostic m inority stress approach. Journal o f Consulting and
Clinical Psychology, 83, 8 7 5 -8 8 9 . https://www.doi.org/10.1037/ccp0000037
Perkins, R. (1997). Therapy for lesbians? The case against. Journal o f Lesbian Studies, 1, 257­
271. https://www.doi.org/10.1300/J155v01n02_08
Radloff, L. S. (1977). The C E S-D scale: A self-report depression scale for research in the ge­
neral population. Applied Psychological Measurement, 1(3), 3 8 5 -4 0 1 . https://www.doi.org/
10.1177/ 014662167700100306
Reback, C., & Shoptaw, S. (2014). Development o f an evidence-based, gay-specific cognitive
behavioral therapy intervention for m etham phetam ine-abusing gay and bisexual men.
Addictive Behaviors, 3 9 , 12 8 6 -1 2 9 1 . https://wwwdoi.org/10.1016Zj.addbeh.2011.11.029
Rendina, H. J. (2014). Hot and bothered: The role o f arousal and rejection sensitivity in dual pro­
cess sexual decision making fo r gay and bisexual men. Unpublished m anuscript, Graduate
C enter o f the City University o f New York, New York, NY.
Riggle, E. D. B., W hitm an, J. S., Olson, A., Rostosky, S. S., & Strong, S. (2008). The positive
aspects o f being a lesbian or gay man. Professional Psychology: Research and Practice, 39,
2 1 0 -2 1 7 . https://www.doi.org/10.1037/0735-7028.39.2.210
Ross, A. D., Waehler, C. A., & Gray, T. N. (2013). Heterosexual persons’ perceptions regarding
language use in counseling: Extending D orland and Fischer (2001). The Counseling
Psychologist, 41, 9 1 8 -9 3 0 .
Russell, G. M ., & Hawkey, C. G. (2017). Context, stigma, and therapeutic practice. In K. A.
DeBord, A. R. Fischer, K. J. Bieschke, & R. M. Perez (Eds.), Handbook o f sexual orien­
tation and gender diversity in counseling and psychotherapy (pp. 7 5 -1 0 4 ). W ashington,
DC: A m erican Psychological Association.
Saunders, J. B., Aasland, O. G ., Babor, T. F., de la Fuente, J. R., & G rant, M . (1993). Development
o f the alcohol use disorders identification test (A U D IT): W H O collaborative project on
early detection o f persons with harm ful alcohol consum ption-II. Addiction, 8 8(6), 791­
804. https://www.doi.org/10.1111 /j. 1360-0443.1993.tb02093.x
Shelton, K., & Delgado-Rom ero, E. A. (2011). Sexual orientation m icroaggressions: The expe­
rience o f lesbian, gay, bisexual, and queer clients in psychotherapy. Journal o f Counseling
Psychology, 58, 2 1 0 -2 2 1 . https://www.doi.org/10.1037/a0022251
Shidlo, A., & Gonsiorek, J. C. (2017). Psychotherapy with clients who have been through
sexual orientation change interventions or request to change their sexual orientation. In
K. A. DeBord, A. R. Fischer, K. J. Bieschke, & R. M . Perez (Eds.), Handbook o f sexual ori­
entation and gender diversity in counseling and psychotherapy (pp. 2 9 1 -3 1 2 ). W ashington,
DC: A m erican Psychological Association.
Shidlo, A., & Schroeder, M. (2001). [Data collected in the study “Changing sexual orienta­
tion: A consum er’s report” but not published in Shidlo & Schroeder (2002)]. Unpublished
raw data.
Shoptaw, S., Reback, C. J., Larkins, S., Wang, P, Rotheram -Fuller, E., Dang, J., & Yang, X.
(2008). O utcom es using two tailored behavioral treatm ents for substance abuse in urban
gay and bisexual m en. Journal o f Substance Abuse Treatment, 35, 2 8 5 -2 9 3 . https://www.
doi.org/10.10 1 6/j.jsat.2007.11.004
 295 Sexual Orientation

Shoptaw, S., Reback, C. J., Peck, J. A., Yang, X ., Rotheram -Fuller, E., Larkins, S., . . . Hucks-
O rtiz, C. (2005). Behavioral treatm ent approaches for m etham phetam ine dependence and
H IV-related sexual risk behaviors am ong urban gay and bisexual m en. Drug and Alcohol
Dependence, 78, 1 2 5 -1 3 4 . https://www.doi.org/10.1016/j.drugalcdep.2004.10.004
Sm A rt [Sexual m inority Assessm ent research team ]. (2009). Best practices fo r asking sexual
orientation on surveys. Los Angeles, CA: W illiam s Institute, U CLA School o f Law.
Sobell, L. C., Brown, J., Leo, G. I., & Sobell, M. B. (1996). The reliability o f the Alcohol
T im eline Followback when administered by telephone and by computer. Drug and Alcohol
Dependence, 42, 4 9 -5 4 .
Sobell, L. C., & Sobell, M. B. (1992). T im eline Follow -Back: A technique for assessing self­
reported alcohol consum ption. In R. Z. Litten & J. P. Allen (Eds.), Measuring alcohol con­
sumption: Psychosocial and biological methods (pp. 4 1 -7 2 ). Totowa, NJ: Humana.
Sobell, M . B., M aisto, S. A., Sobell, L. C., Cooper, A. M ., Cooper, T., & Saunders, B. (1980).
Developing a prototype for evaluating alcohol treatm ent effectiveness. In L. C. Sobell &
E. Ward (Eds.), Evaluating alcohol and drug abuse treatment effectiveness: Recent advances
(pp. 1 2 9 -1 5 0 ). New York, NY: Pergamon.
Spanier, G. B. (1976). M easuring dyadic adjustment: New scales for assessing the quality of
m arriage and sim ilar dyads. Journal o f Marriage and the Family, 3 8 , 15-2 8 .
Spengler, E. S., M iller, D. J., & Spengler, P. M. (2016). M icroaggressions: C linical errors with
sexual m inority clients. Psychotherapy, 5 3 , 3 6 0 -3 6 6 .
Steinem , G. (1983). Far from the opposite shore. In Outrageous acts and everyday rebellions
(pp. 3 4 1 -3 6 2 ). New York, NY: Holt.
Taylor, S., & Asm undson, G. J. (2008). Internal and external validity in clinical research. In
D. M cKay (Ed.), Handbook o f research methods in abnormal and clinical psychology (pp.
23-34). Los Angeles, CA: SAGE.
The G enIU SS Group. (2014). Best practices fo r asking questions to identify transgender and
other gender minority respondents on population-based surveys (J. L. Herm an, Ed.). Los
Angeles, CA: W illiam s Institute.
W atson, J. C., G reenberg, L. S., Rice, L. N., & G ordon, L. (1998). Client Task Specific Change
Measure-Revised. Unpublished manual, University o f Toronto, O ntario Institute for Studies
in Education, D epartm ent o f Adult Education and Counselling Psychology, Toronto, ON.
W atson, J. C., Schein, J., & M cM ullen, E. (2010). An exam ination o f clients’ in-session changes
and their relationship to the working alliance and outcome, Psychotherapy Research, 20,
2 2 4 -2 3 3 .
W est, C., & Zim m erm an, D. H. (1987). D oing gender. Gender and Society, 1 , 1 2 5 -1 5 1 . https://
w w w.doi.org/10.1177/0891243287001002002
 10

STAGES O F C H A N G E

Paul Krebs, John C. Norcross, Joseph M. Nicholson,

and James O. Prochaska

Individuals seeking psychotherapy do not arrive at a therapist’s doorstep with iden­

tical motivation, preparation, or capacity for behavior change. For most practitioners
and programs, patients are heterogeneous in terms of their desire and skill to change.
Virtually all psychotherapists readily acknowledge that a patient’s readiness for beha­
vior change profoundly influences the process and outcome of treatment.
In the transtheoretical model (TTM ), behavior change is conceptualized as a
process that unfolds over time and involves progression through a series of five
stages: precontemplation, contemplation, preparation, action, and maintenance. At
each stage of change, we propose that different change processes and relational stances
produce optimal progress. Adapting or tailoring psychotherapy to the individual pa­
tient thus requires matching the processes of change and the therapeutic relationship
to his or her stage of change. Further, as clients progress from one stage to the next, the
therapeutic relationship evolves accordingly.
In this chapter, we review the voluminous research evidence on the stages of change as it
applies to psychotherapy. We define the stages of change and related readiness constructs,
describe popular measures to assess them, and offer clinical examples and landmark
studies. Our meta-analysis is intended to address two specific aims: first, to assess the ability
of stages of change and related readiness measures to predict psychotherapy outcomes,
and second, to assess the outcomes from psychotherapy studies that matched treatment
to specific stages or readiness levels of change. We then analyze potential moderators of
the stages-outcome association, address limitations of the research evidence, and review
patient contributions and diversity considerations. The chapter concludes with training
implications and therapeutic practices for the stages of change in psychotherapy.

DEFINITIONS

Stages of Change
Following are brief descriptions of each of the five stages of change. Each stage
represents a period of time as well as a set of tasks needed for movement to the next

29 6
 297 Stages o f Change

stage. Although the time an individual spends in each stage will vary, the tasks to be
accomplished are assumed to be invariant.
Precontemplation is the stage at which there is no intention to change behavior
in the foreseeable future. Most patients in this stage are unaware or under-aware of
their problems. Families, friends, neighbors, or employees, however, are often well
aware that the precontemplators have problems. When precontemplators present for
psychotherapy, they often do so because of pressure from others. Usually they feel
coerced into changing by a spouse who threatens to leave, an employer who threatens
to dismiss them, parents who threaten to disown them, or courts that threaten to
punish them. Resistance to recognizing or modifying a problem is the hallmark of
precontemplation, which is frequently known to the public by the prejorative term
denial.
Contemplation is the stage in which patients are aware that a problem exists
and are seriously thinking about overcoming it but have not yet made a com­
m itm ent to take action. Contemplators struggle with their positive evaluations
of their dysfunctional behavior and the amount of effort, energy, and loss it will
cost to overcome it. People can remain stuck in the contemplation stage for long
periods. In one study we followed a group of 200 smokers in the contemplation
stage for two years; the modal response o f this group was to remain in the contem ­
plation stage for the entire tim e of the study without ever moving to significant
action (Prochaska & DiClem ente, 1983). Serious consideration of the problem
characterizes contemplation.
Preparation is a stage that combines intention and behavioral criteria. Individuals
in this stage are intending to take action in the next month and are frequently taking
small behavioral changes— “baby steps,” so to speak. While they have made some
reductions in their problem behaviors, patients in the preparation stage have not yet
reached the criterion for effective action, such as abstinence from alcohol abuse or
remission of depression. They are intending, however, to take such action in the im ­
mediate future.
Action is the stage in which individuals modify their behavior, experiences, and/or
environment to overcome their problems. Action involves the most overt behavioral
changes and requires considerable commitment of time and energy. Modifications of
the problem made in the action stage tend to be most visible and receive the greatest
external recognition, leading some to erroneously equate this single stage with the en­
tire change process. Individuals are in the action stage if they have successfully altered
the dysfunctional behavior for a period from one day to six months. Modification of the
target behavior to an acceptable criterion and overt efforts to change are the hallmarks
of action.
Maintenance is the stage in which people work to prevent relapse and consolidate
the gains attained during action. For some behaviors, such as addictions, maintenance
can last a lifetime; for other behaviors, maintenance can end at three to nine months
when patients remain free of the problem behavior and/or consistently engage in a
new incompatible behavior. Stabilizing behavior change and avoiding relapse are the
hallmarks of maintenance.
 29 8 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HA T WO RK

Processes of Change
The stages of change represent when people change; the processes of change entail how
people change. The processes of change represent an intermediate level of abstrac­
tion between metatheoretical assumptions and clinical techniques spawned by those
theories. While there are 500-plus ostensibly different psychotherapies (Pearsall, 2011),
we have identified only 8 to 10 different processes of change based on repeated prin­
cipal components analysis. Behavior change is conceptualized in terms of processes or
principles, not in terms of specific techniques.
Change processes are overt and covert activities that individuals engage in when they
attempt to modify problem behaviors. Each process is a broad category encompassing
multiple methods and relationship stances traditionally associated with dissonant the­
oretical orientations.
Table 10.1 presents the processes of change receiving the most research evi­
dence across 50 behavioral disorders along with their definitions and representative
interventions (the process of helping relationships has been deleted from the table).
A common and finite set of change processes has been repeatedly identified across
these diverse problems (Prochaska et al., 1985).

Table 10 . 1 . D efinitions and Representative Interventions o f the Processes o f Change

Process of Change Definition: Interventions
Consciousness Raising Increasing inform ation about self and problem : observations,
confrontations; interpretations, awareness exercises,
bibliotherapy
Self-Reevaluation Assessing how one feels and thinks about oneself with respect to
a problem : value clarification, imagery, corrective em otional
experience
Dram atic Relief/ Experiencing and expressing feelings about one’s
Em otional Arousal problem s and solutions: psychodram a, cathartic work, grieving
losses, role-playing, tw o-chair work
Self-Liberation Choosing and com m itm ent to act or belief in ability to
change: decision-m aking m ethods, m otivational interviewing,
com m itm ent-enhancing techniques
Counterconditioning Substituting alternative or incom patible behaviors for
problem : relaxation, desensitization, assertion, cognitive
restructuring, behavioral activation
Stimulus Control Avoiding or controlling stimuli that elicit problem
behaviors: restructuring one’s environm ent, avoiding high-risk
cues, fading techniques, altering relationships
Reinforcem ent Rewarding one’s self or reward by others for m aking
changes: contingency contracts, overt and covert
reinforcem ent, self-reward
 299 Stages o f Change

Stages x Processes
The TTM posits that different processes of change are differentially effective in cer­
tain stages of change. In general terms, change processes traditionally associated with
the experiential, cognitive, and psychoanalytic persuasions are most useful during the
earlier precontemplation and contemplation stages. Change processes traditionally as­
sociated with the existential and behavioral traditions, by contrast, prove most useful
during action and maintenance.
Consciousness raising helps clients progress from precontemplation to contem­
plation. In particular, patients need to increase their awareness of the advantages of
changing and the multiple benefits of psychotherapy. They also typically benefit from
enhanced awareness of themselves, their disorders, and their defenses.
Contemplation can be a safe haven for clients and therapists alike. Clients are in­
tending to make major changes, but not right now. First they need to continue to in­
crease consciousness. Reflecting, feeling, and re-evaluating how they have been and
how they might become can be hard work at times. But it can also prove meaningful
and even fun. And such sharing builds a therapeutic bond that can be hard to let go.
Who wants to give up such a close relationship? How can you fail as a therapist by
having such a good therapeutic relationship? The answer is by allowing your client to
stay stuck in contemplation.
The process of dramatic relief (emotional arousal) can include anticipatory grieving,
the sadness and loss of letting go of behaviors and relationships that no longer work.
Dramatic relief can also include facing the fear, guilt, or regret that would come from
not changing. If a patient clings tenaciously to safe and secure patterns that are also
self-defeating and self-destructive, how will he or she feel in the future?
As people progress from precontemplation to contemplation, they rely more on the
process of self-reevaluation. “How do I think and feel about myself as an angry or de­
pressed person? How will I think and feel about myself as a more active and mindful
person?” Reevaluation entails a courageous assessment of how one experiences and
prizes oneself with respect to the problem. As patients progress into the preparation
stage, they use more of self-liberation and its numerous methods. This is the belief that
they have the ability to change their behavior and the commitment to act on that belief.
During action, clients receiving adequate reinforcement for their efforts secure
better treatment outcomes. Clients may expect to be reinforced by others more than
others will reinforce them. Thus clients need to be prepared to depend more on self
rather than social reinforcements, including the psychotherapist.
Patients will learn and practice counterconditioning (reciprocal inhibition) as
they replace healthier behaviors for their problem behaviors. This process includes
the classic reciprocal inhibition methods: assertion to counter passivity, relaxation
to replace anxiety, cognitive substitutions instead of negative thinking, exposure to
counteravoidance, acceptance in place of hypercontrol.
As clients progress into the maintenance stage, they do not have to work as hard but
they have to apply change processes to prevent relapse. They particularly have to be
prepared for the situations that are most likely to induce relapse. Stimulus control in its
 300 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

multiple ways assist patients in avoiding or controlling triggers (emotions, thoughts,

places, people) that elicit problem behaviors. Across numerous studies, stimulus con­
trol is the change process least frequently used by psychotherapists and patients alike
(Prochaska & Norcross, 2018).
But for all patients, psychotherapy will probably terminate before the problem has
completely resolved. This is one reason clinicians and clients alike can feel anxious
about termination. They both know that under certain conditions the risk of relapse
is real. O f course, clients can return for further treatment if they lapse or relapse. They
can analyze what they did right, what mistakes they made, and what they need to do
differently to keep moving ahead.

MEASURES
Multiple assessment devices have been developed over the years to assess a person’s stage
of change or “readiness to change.” The measures vary in format (e.g., questionnaires,
algorithms, ladders, and interviews) as well as in specificity (e.g., generic measures for
multiple problems and disorder-specific measures).
The most frequent measure in psychotherapy research has been the University of
Rhode Island Change Assessment (URICA; McConnaughy et al., 1989). This 32-item
questionnaire yields separate scores on four continuous scales: Precontemplation,
Contemplation, Action, and Maintenance (precontemplators score high on both the
Contemplation and Action scales). Scores for each stage range from 8 to 40, with higher
scores indicating stronger endorsement of each subscale. Psychometric evaluation of
the URICA or the Stages of Change scale, as it is widely known, demonstrates a stable
four-factor structure (Pantalon et al., 2002) and subscale consistency (Cronbach’s
alphas .74-.88 [Petry, 2005]; 88-.89 [McConnaughy et al., 1983]).
Items used to identify precontemplation include “As far as I’m concerned, I don’t
have any problems that need changing” and “I guess I have faults but there’s nothing
that I really need to change.” Contemplators endorse such items as, “I have a problem
and I really think I should work on it” and “I’ve been thinking that I might want to
change something about myself.” Patients in the action stage endorse statements like, “I
am really working hard to change” and “Anyone can talk about changing; I am actually
doing something about it.” Representative maintenance items are, “I may need a boost
right now to help me maintain the changes I’ve already made” and “I’m here to prevent
myself from having a relapse of my problem.”
Other measures of change readiness include the Stages of Change and Treatment
Eagerness Scales (SOCRATES), developed for measuring readiness for change with
regard to problem drinking as an alternate measure to the URICA (Miller & Tonigan,
1996). This 19-item measure produces three continuous scales: Ambivalence,
Recognition, and Taking Steps, which represent continuously distributed motivational
processes. The SOCRATES has been found to be related to quit attempts for smoking
cessation (DiClemente et al., 1991), alcohol use (Isenhart, 1997; Zhang et al., 2004),
and drug use (Henderson et al., 2004).
 301 Stages of Change

In fewer research studies but more frequently in clinical practice, the stages are
assessed using a series of questions that result in a discrete categorization. The practi­
tioner asks if patients are seriously intending to change the problem in the near future,
typically within the next six months. If not, they are classified as precontemplators.
Clients who state that they are seriously considering changing the problem behavior in
the next six months are classified as contemplators. Those intending to take action in
the next month are in the preparation stage. Clients who state that they are currently
changing their problem are in the action stage.

CLINICAL EXAMPLES

The following exchange from a psychotherapy session demonstrates the relational

stance a transtheoretical therapist (Prochaska) would probably adopt with a patient in
the precontemplation stage. The client is a 32-year-old stockbroker in precontemplation
for chronic cocaine abuse. The stage of change dimension was briefly outlined and
then the client, “Donald,” was given feedback that his assessment indicated he was in
the precontemplation stage. Did he concur? “Yeh, probably.”

Therapist: We know that individuals in the precontemplation stage often feel coerced
into entering therapy rather than being there by choice. What pressures were there
on you to seek psychotherapy?
Client: Lots of people have been on my back. My girlfriend, my mother. My job may
be in jeopardy. They all think it’s caused by cocaine. But I’ve been using it for years
and it’s never been a problem.
Therapist: How do you react when people pressure you to quit cocaine when you’re
not ready?
Client: I get angry. I tell them to mind their own business.
Therapist: You get defensive.
Client: Sure, wouldn’t you? Nobody likes to be told what to do, to be treated like a kid.
Therapist: How would you react if I told you to quit cocaine?
Client: I would get angry. I would tell myself you’re just like all the others—think you
know better than me how to run my life.
Therapist: Would you want to drop out of therapy?
Client: Probably. I don’t react well to being controlled.
Therapist: I appreciate you sharing your reactions with me. Let me share my main
concern. I am concerned that you might drop out of therapy before I have a chance
to make a significant difference in your life.
I don’t want to coerce or control you. I do want to help you to be freer to do what is
best for your life. So will you let me know if I am pressuring you or parenting you?
Client: You’ll know.

Historically, confrontation was one of the recommended ways of relating to defen­

sive and resistant clients. By confronting patients’ defenses and resistance, therapists
 302 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

expected to break through the denial and other defenses. Research has consistently
shown, however, that a confrontational style of relating drives many patients away and
increases premature termination (Miller & Rollnick, 2012). Motivational interviewing,
by contrast, rolls with patient resistance and typically demonstrates large impacts
in a small number of sessions for precontemplators and contemplators (Lundahl
et al., 2010).
Later, in the same session, the psychotherapist adopts an affirming, Socratic style
and relies primarily on consciousness-raising methods that the research evidence
suggests will assist a patient to progress from precontemplation to contemplation. This
entails increasing awareness of the advantages of changing and the multiple benefits of
sticking with treatment.

Therapist: We know people are likely to complete therapy if they appreciate its many
benefits. Donald, how do you think people benefit from therapy?
Client: It makes the therapist better off.
Therapist: That’s good! And how about the client?
Client: I expect it helps them solve their problems.
Therapist: That’s true. And would that help them to feel better about themselves?
Client: Yeh, it should.
Therapist: And would that improve their moods?
Client: Sure.
Therapist: Would that improve their relationships?
Client: It should.
Therapist: And be more open and less defensive.
Client: I can see that.
Therapist: And do better in their job and make more money.
Client: I don’t know about that.
Therapist: It’s true. How about we make a deal. If your income goes up 10%, my fee
goes up 10%?
Client: That would be worth it.
Therapist: You might not believe this, but there’s only one other thing you could do for
an hour a week that would give you more benefits than therapy.
Client: What’s that?
Therapist: I’m not going to tell you because you might invest in that instead. (Client
laughs)

The psychotherapist’s stage-matched relational stance can be characterized as follows.

With patients in precontemplation, often the optimal role is like that of a nurturing
parent joining with a resistant and defensive youngster who is both drawn to and
repelled by the prospects of becoming more independent. With clients in contempla­
tion, the therapist stance is akin to a Socratic teacher who encourages clients to achieve
their own insights into their condition. With clients in the preparation stage, the stance
is more like that of an experienced coach who has been through many crucial matches
and can provide a fine game plan or can review the participant’s own plan. W ith clients
 303 Stages of Change

progressing into action and maintenance, the psychotherapist becomes more of a

consultant who is available to provide expert advice and support when action is not
progressing smoothly. As termination approaches in lengthier treatment, the thera­
pist is consulted less often as the patient experiences greater autonomy and enhanced
ability to live a life free from previously disabling problems.

LANDMARK STUDIES
The earliest studies on the stages of change involved a longitudinal examination of
self-change of tobacco smokers (Prochaska & DiClemente, 1982, 1983). The five stages
were identified within an integrative model of change, and that model was tested on
872 smokers. The processes of change were expected to receive differential emphases
during particular stages of change. Results indicated that self-changers: (a) used the
fewest processes of change during precontemplation, (b) relied on consciousness
raising during the contemplation stage, (c) emphasized self-reevaluation in both con­
templation and action stages, (d) employed self-liberation and reinforcement manage­
ment during the action stage, and (e) used counterconditioning and stimulus control
the most in both action and maintenance stages. These patterns of stage matching
among self-changers have been subsequently replicated in hundreds of studies across
disorders. These 1982 and 1983 articles launched the vocabulary and utility of the
stages of change, initially in addictive disorders and self-change and then eventually in
mental disorders and psychological intervention.
One of the earliest stage of change studies in psychotherapy concerned patient con­
tinuation and dropouts. Approximately one-quarter of patients prematurely discon­
tinue psychotherapy (Swift & Greenberg, 2012); however, the characteristics of these
dropouts have not been reliably known. In one pivotal study (Brogan et al., 1999), pre­
mature termination was predicted using traditional predictors—client characteristics
and problem characteristics, such as chronicity and intensity. These variables, however,
had zero ability to predict therapy dropouts. When the stages and processes of change
were used, 93% of the premature terminators—as opposed to therapy continuers and
early but appropriate terminators—were correctly identified. The stage of change pro­
file of the 40% who dropped out of therapy was that of precontemplators. The stage
profile of the 20% who terminated quickly but appropriately was that of people in the
action stage. The stage profile of the therapy continuers was that of contemplators.
In sum, the stage measure demonstrated its ability to identify and predict premature
dropouts.
In a study of dropout from drug treatment, the greater readiness for treatment as
measured by the SOCRATES was associated with reduced program dropout whereas
demographic variables (age, race, sex) were not predictive (Evans et al., 2009). In an­
other study of resident drug treatment retention, higher scores on the Contemplation,
Action, and Maintenance subscales of the URICA were associated with staying in the
program at least 30 days, whereas higher scores in the Precontemplation subscale was
associated with early termination (Choi et al., 2013). (These and other studies predicting
psychotherapy dropouts and outcomes are included in this chapter’s meta-analysis.)
 304 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

In another psychotherapy study involving male perpetrators of partner violence,

stage-tailored treatments were added to mandatory weekly group therapy. At the
six-month follow-up with the first 200 participants, the addition of stage matching
produced significant reduction in violence compared to the weekly group coun­
seling alone. In the stage-matched condition, only 3% of the female partners of the
perpetrators had been beaten in the past six months compared to 23% of the women
whose partners received only the group therapy (Levesque et al., 2012). As well, about
twice as many perpetrators had progressed to the action or maintenance stage during
group therapy. Matching treatments to patients’ stage of change has been found to
decrease dropouts, improve retention, and boost outcomes in multiple disorders, in­
cluding interpersonal violence.
A final landmark study reviewed here illustrates the T TM goal of increasing treat­
ment impacts with entire populations. This large study, adopting the World Health
Organization’s definition of health as more than the absence of physical or mental ill­
ness, aimed treatment at enhancing emotional well-being (Prochaska et al., 2012). The
intervention reached nearly 4,000 people in 39 states in the United States. The adult
sample averaged four chronic conditions and four risk behaviors, such as inadequate
exercise, unhealthy eating, depression, and poor stress management (Prochaska et al.,
2012). The participants also scored well below national averages on emotional and
physical well-being, and the majority were struggling rather than thriving. The success
rates for TTM -tailored telephone counseling, a TTM -tailored online program, and the
control condition were respectively 57%, 47%, and 37% for exercise; 75%, 65%, and
53% for stress management; and 31%, 26%, and 21% for healthy eating, which was
not treated. Those effectiveness patterns were clear and consistent for target behaviors.
In addition, the brief stage-matched interventions produced large impacts on overall
well-being. Comparing the TTM treatments to controls showed well-being increased
about twice as much for those using counselors and two-thirds as much for those using
the online program. It was striking that the percentage of patients thriving almost
doubled for the counseling condition (67% vs. 34% in control). This study exemplifies
how psychotherapy can raise the bar to help the population to feel better and live better.

RESULTS OF PREVIOUS META-ANALYSES

Empirical research on the stages of change has taken a number of tacks over the past
35 years (for a review, see Prochaska & Norcross, 2010), resulting in a vast literature.
In this section, we review the results of earlier meta-analyses on the integration of
the stages and processes, the ability of the stages of change to predict psychotherapy
outcomes, and the efficacy of tailoring treatments to the stages.

Integrating Stages x Processes

Years of research in behavioral medicine and psychotherapy converge in showing that
different processes of change are differentially effective in certain stages of change.
 305 Stages o f Change

A meta-analysis of 47 cross-sectional studies examined the relations of the stages and

the processes of change (Rosen, 2000). The studies involved smoking, substance abuse,
exercise, diet, and psychotherapy. The mean effect sizes (d) were approximately .70 for
variation in cognitive-affective processes by stage and .80 for variation in behavioral
processes by stage, both moderate to large effects. Effect sizes for stages by processes
did not vary significantly by the problem treated. For the five studies that examined
the change processes in psychotherapy, behavioral processes peaked in action while
cognitive-affective processes peaked in the contemplation or preparation stages.

Predicting Psychotherapy Outcomes

In the previous edition of this volume, we conducted a random-effects meta-analysis
of 39 psychotherapy studies involving 8,238 patients (Norcross et al., 2011). The stages
of change reliably predicted psychotherapy outcomes (d = .46). These results indicate
that patients beginning in the preparation and action stages (or with greater readi­
ness scores) do better than those beginning in precontemplation or contemplation (or
those with lower readiness to change scores). These studies in this meta-analysis, how­
ever, did not tailor interventions by assessing stage of change and providing different
combinations of intervention components by stage.

Tailoring Treatments to Stages

A large number of psychosocial treatments have been tailored to stage of change or
readiness for change. These have primarily been population-based studies delivered
via computer, mail, or phone, with a focus on health behaviors. Such interventions
have assessed and provided specific feedback by stage of change and other constructs,
such as self-efficacy. Results of these studies clearly show the effectiveness of tailoring
or matching psychological interventions to the patient’s stage of change.
We conducted a meta-analysis on 87 prospective, tailored interventions delivered
via computer or mail across smoking cessation, physical activity, healthy diet, and
mammography screening (Krebs et al., 2010). The mean effect size (d) of .18 (95% con­
fidence interval [CI] = .16-.20) represents a 39% increase (odds ratio = 1.39) over the
nontailored intervention or minimal care conditions and indicates a small to medium-
size effect for population-based interventions (Rossi, 2002). The subset of studies that
intervened on smoking cessation, for instance, resulted in an absolute increase of 6%
in quit rates, a rate comparable to that observed with four to eight individual in-person
counseling sessions (Fiore, 2008). Hence, use of stage tailoring proved more effective
than non-tailoring for health behaviors.
These stage-matching studies, however, did not include face-to-face psycho­
therapy nor did they address the disorders most commonly treated by mental health
professionals. Thus, in addition to updating our meta-analysis on the ability of the
stages of change to predict psychotherapy outcomes, we undertook a meta-analysis
specifically focusing on stage matching in psychotherapy.
 306 p sy c h o th er a py r ela tio n sh ips that w ork

META-ANALYTIC REVIEW: STAGES

PREDICTING OUTCOME
Here, we present the results of a meta-analysis, updated from the last edition of this
book, performed to gauge the effect of the stages of change on psychotherapy outcomes.

Search Strategy and Criteria

A medical librarian experienced in systematic reviews (Nicholson) conducted searches
in PubMed/MEDLINE, EMBASE (Ovid), PsycINFO (Ovid), CINAHL (Ebsco),
Cochrane CENTRAL (Ovid), and Web of Science. The search strategies included
MeSH and Emtree terms as well as keywords to reflect the three main concepts: stages
of change (e.g., readiness, motivation, as well as the measures of these constructs, such
as URICA, SOCRATES, Contemplation Ladder), psychotherapy (e.g., counseling,
therapy, intervention, psychosocial treatment), and individualized psychotherapy (e.g.,
stages of change, matching, tailored, treatment adaptation). To locate studies that may
have employed similar terms, we also conducted a forward search for articles that cited
identified studies, examined reference lists from published studies, and searched for
articles published by authors of studies deemed for inclusion. We also tracked citations
of selected references and hand-searched relevant sources to identify studies that had
not been identified by database searches.
A restriction of the search strategies concerned publication dates. We searched for
studies published from 2009 onward since the earlier meta-analysis searched from
2009 (Norcross et al., 2011). Otherwise, there were no language or other limits used in
the search. A total of 1,872 citations were retrieved from across the six databases. After
computerized de-duplication, the number of unique citations was 1,155.
Studies selected for the meta-analysis met the following criteria, which were con­
sistent with inclusion criteria for other meta-analytic reviews in this volume: (a) studies
reported results of behavioral/psychological face-to-face treatment, (b) treatment was
provided by mental health professionals, (c) patients had a diagnosable mental disorder
(Diagnostic and Statistical Manual o f Mental Disorders or International Classification
of Diseases criteria), (d) treatments consisted of at least three group or individual ses­
sions, (e) readiness to change measured prior to treatment was used to tailor or predict
treatment outcome, and (f) sufficient statistical information was available to calculate
an effect size. Studies were excluded if they (a) only used a computerized program to
provide feedback, (b) did not involve a mental health professional, (c) only involved
health behavior change counseling, and (d) did not include a measure of readiness to
change.
Abstracts of all 1,155 references were examined by two reviewers for possible in­
clusion according to the aforementioned inclusion criteria. O f these, 145 papers were
chosen for full text review as they potentially could meet criteria upon further review.
From the 145 studies, the reviewers excluded 108 largely for study design not meeting
criteria, insufficient information to code effect size, no measure of readiness to change,
and no translation available in English. Figure 10.1 presents a flowchart of study selec­
tion and the reasons for excluding studies.
 307 Stages of Change

figure 10.1 P R I S M A flo w c h a rt.

In the end, 37 new studies met the inclusion criteria and were analyzed in the pre­
sent review. Along with the 39 studies from the previous meta-analysis, that resulted
in a total of 76 studies for analysis.

Methodological Decisions
The primary database was created and the results were analyzed using the
Comprehensive Meta-Analysis software package (Biostat, 2006). Results reported as
correlations (r), mean differences (F or t), or tests of variance (X2) were transformed to
Cohen’s d (Lipsey & Wilson, 2001). Each obtained effect size estimate was weighted by
the inverse of the variance of the estimate, which gives greater weight to studies with
better estimates (for the most part, studies with larger sample sizes). If insufficient in­
formation was reported for effect size calculation, the study was excluded. Twenty-six
studies were randomized controlled trials (RCTs) while the remainder used a one-
group pre-post design. Regardless of study design, all effect sizes were calculated as the
relation between pretreatment stage of change and treatment outcome(s).
We employed a random effects model. This model assumes both study-level error
and variability among studies due to sampling of studies from a population of studies.
This model enables generalization to a population of studies.
Publication bias, the tendency for significant study results to be reported more often
than nonsignificant results, can upwardly bias effect size estimates in meta-analysis.
We assessed mean effects for degree of publication bias using two techniques: fail-safe
N and trim and fill. Fail-safe N calculates the number of unpublished studies with a
null effect size that would be needed to reduce the overall effect to nonsignificance.
 308 psy c h o th er a py rela tio n sh ips that w ork

Trim and fill (Duval & Tweedie, 2000) assesses the symmetry of a plot of effect size by
sample size (funnel plot) under the assumption that when publication bias exists, a dis­
proportionate number of studies will fall to the bottom right of the plot. This technique
then determines the number of asymmetrical outcomes, imputes their counterparts to
the left, and estimates a corrected mean effect size.
The 76 studies represented a variety of diagnoses and outcome measures with some
studies reporting more than one outcome (e.g., substance use and treatment dropout).
To ensure statistical independence of outcomes, when studies reported more than one
outcome, an overall mean effect size per study was included for calculating the overall
mean effect (using formulas by Borenstein et al., 2009).
To determine if moderator analysis was appropriate, variability between studies was
assessed via the Q test that employs weighted data and compares within- and between-
group heterogeneity using the Q statistic. A significant Q test indicates that there is
sufficient variability among the effect sizes of the studies to look for moderators that
could explain the variability.
Continuous moderators were examined using meta-regression. We conducted
moderator analyses for patient characteristics (adolescent vs. adult study populations;
>60% minority participants vs. not; percentage of female participants) and treatment
features (inpatient vs. outpatient setting; use of a treatment manual vs. not; number
of treatment sessions; theoretical orientation; RCT vs. nonrandomized design). We
also present effect sizes by psychotherapy outcomes: adherence to treatment, eating
disorder outcomes, substance use outcomes, and mood disorders/relational distress.

The Studies
Table 10.2 summarizes the attributes of the 76 studies, encompassing 21,424 psy­
chotherapy patients. We included data only from each study’s final assessment, most
of which were immediately upon treatment completion. Nine studies concerned
treatments for adolescents (ages 1 3 -18), while the others focused on adults (18+).
Sample sizes (N) ranged from 30 to 1,588, with an average of 278 participants at recruit­
ment and a 73% retention rate at follow-up. Most samples (k = 41) were comprised of
primarily White participants (>60%), 6 had primarily African American participants
(>60%), 10 studies recruited a racially mixed sample, and 16 did not report racial/
ethnic makeup. (Note that k denotes the number of studies, in contrast to N , which
refers to the number of participants in a study.) Patients on average were 45% female
(and ranged from 0% to 100%). Twenty-five studies conducted interventions in an in­
patient setting. The number of treatment sessions ranged from 1 (in some conditions)
to 30 with 13 being the mean. Thirty-one studies did not report the number of ses­
sions (most of these provided inpatient treatment). Thirteen studies reported using
a treatment manual, with cognitive-behavioral treatment the most common (k = 36),
followed by motivational enhancement (k = 9) and a combination of other orientations
(k = 14). The most frequent readiness measures were the URICA (k = 46) and the
SOCRATES (k = 10).
Table 10.2. Summary of Studies and Samples (k = 76) Included in the Meta-Analysis
Characteristic k %
C ountry
United States 46 60
Canada 9 12
Australia 3 4
Europe 16 21
Africa 2 3

Study Design
Single group p re-p o st 50 66
Random ized controlled trial 26 34

Patient Age
Adult (18+) 67 88
Adolescent (1 3 -1 7 ) 9 12

Patient Race/Ethnicity
W hite (>60% o f sample) 41 54
M ix (none greater than 60% o f sample) 10 13
A frican A m erican (>60% o f sample) 6 8
Data not reported 16 21

Treatm ent Setting

Outpatient 49 64
Inpatient 25 33
Not reported 2 3

Treatm ent M anual Used 13 17

N um ber o f Treatm ent Sessions

<10 17 22
1 0 -1 9 19 25
20+ 9 12
Data not reported 31 41

Treatm ent O rientation

Cognitive-behavioral 36 47
M otivational enhancem ent 9 12
12-step 5 7
O ther 12 16
Data not reported 14 18

Readiness Measure
University o f Rhode Island Change Assessment 46 61
Stages o f Change Readiness and Treatm ent Eagerness Scale 10 13
A norexia Stages o f Change Questionnaire 4 5
O ther 16 21
 310 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Effect Size
The 76 studies reported 137 separate data points, as a number of studies reported
associations between stage of change and multiple outcome variables (depression, anx­
iety, etc.). Results of the individual studies are summarized in Table 10.3.
The mean effect size was d = .41 with a 95% CI of .34 to .48 (range - .45 to 2.49),
Q(75) = 786.62, p < .001. Analysis of publication bias indicated a fail-safe N of 8,991.
By convention (Cohen, 1988), a d of .41 indicates a medium effect, demonstrating
that the stages of change is robustly associated with and predictive of outcomes in psy­
chotherapy. That is, the amount of progress clients make during treatment tends to be
a function of their pretreatment stage of change. For example, an intensive action- and
maintenance-oriented smoking cessation program for cardiac patients achieved suc­
cess for 22% of precontemplators and 43% of contemplators; 76% of those in action or
prepared for action at the start of the study were not smoking six months later (Ockene
et al., 1992).

Comparison to Previous Meta-Analysis

The results of the present meta-analysis parallel and extend those obtained eight
years earlier. That meta-analysis was based on fewer studies and produced an average
weighted d of .49 (95% CI = .35-.58). The present effect size of .41 (95% CI = .34-.48) is
in the same neighborhood as the previous meta-analysis but has a slightly lower mean
than found earlier.

Potential Moderators
The significant Q test for our meta-analysis indicated that there was sufficient varia­
bility among the effect sizes of the studies to examine moderators that might explain
this variability. We conducted moderator analyses for patient characteristics, treatment
features, and outcome measures.
With regard to patient characteristics, we found no statistically significant difference
between adolescent and adult populations (p = .96), nor by race/ethnicity (p = .90).
Effect size was not related to the relative number of male/female patients (p = .37).
With regard to treatment features, we found no differences in effect size between
inpatient (k = 25, d = .42) and outpatient treatment settings (k = 49, d = .38, p = .55),
between treatments that used a manual (p = .39) and those that did not, nor by number
of treatment sessions (p = 1.0). For studies reporting primary theoretical orientation,
the effect sizes were similar (p = .23): 12-step programs (k = 5, d = .42), cognitive-
behavioral treatment (k = 32, d = .42), motivational enhancement (k = 7, d = .18), and
combination (k = 14, d = .33). Randomized trials (k = 26, d = .33) did not differ from
pre-post designs (k = 50, d = .43, p = .19).
We analyzed the effect size by type of treatment outcome. For adherence to treat-
ment/premature dropout (k = 36) the mean effect size was d = 0.36 (95% CI = .26-.47).
Nine studies assessed the relation between baseline readiness to change and eating
Table 10.3. Effect Sizes by Study
Study Primary Diagnosis Readiness Measure

Abd Elbaky et al., 2014 Eating disorder A N SO CQ

Alexander & M orris, 2008 Dom estic abuse U RICA
Alexander et al., 2010 Dom estic abuse U RICA
Algars et al., 2015 Eating disorder M SCA RED
Allen, 1998 A lcohol abuse U RICA
Alosso, 2012 Obsessive-compulsive disorder U RICA
A m etller et al., 2005 Eating disorder A N SO CQ
Bachiller et al., 2015 Substance abuse U RICA
Bates, 2014 Eating disorder RM I + M SCA RED
Bauer et al., 2014 A lcohol dependence SO CRA TES
Berry, 2012 M ixed diagnoses U RICA
Blanchard et al., 2003 Substance abuse U RICA
Boswell et al., 2012 A nxiety disorders U RICA
Brodeur et al., 2008 Dom estic abuse U R IC A -D V
Callaghan et al., 2005 Substance abuse U RICA
Callaghan et al., 2008 (Budney Substance abuse U RICA
et al., 2000)
Callaghan et al., 2008 (Budney Substance abuse U RICA
et al., 2006)
Carpenter et al., 2002 Substance abuse U RICA
Castro et al., 2011 Eating disorder BN SO C Q
C hoi et al., 2013 Substance abuse U RICA
Chung & M aisto, 2009 Substance abuse Contem plation
Ladder
 N d SE 95% C l

L ow er Upper

63 .24 .26 - .2 7 .75

210 .45 .20 .07 .84
528 .12 .12 -.1 2 .36
32 .19 .37 -.5 4 .92
806 .28 .07 .14 .42
148 .18 .20 -.2 1 .56
70 .34 .12 .10 .58
46 .59 .33 -.0 5 1.24
46 1.09 .35 .41 1.77
805 .03 .01 .01 .06
163 .35 .16 .03 .66
252 .16 .13 -.1 0 .42
37 2.20 .58 1.06 3.34
302 .11 .12 -.1 2 .34
130 .74 .19 .37 1.11
60 .18 .40 -.6 1 .97

90 .93 .31 .33 1.52

174 .58 .17 .25 .92

40 .85 .41 .05 1.66
1317 .08 .03 .02 .14
142 .03 .23 -.4 3 .49

( c o n t in u e d )
Table 10.3. Continued
Study Primary Diagnosis

Clarke et al., 2012 Substance abuse

Connors et al., 2000a A lcohol abuse
Connors et al., 2000b A lcohol abuse
C ook et al., 2015 A lcohol abuse
Dale et al., 2011 A lcohol abuse
Dem m el et al., 2004 A lcohol abuse
Derisley et al., 2000 G eneral
Dove, 2016 Depression
Dozois et al., 2004 A nxiety
Eckhardt et al., 2008 Dom estic abuse
Em m erling et al., 2009 M ixed diagnoses
Evans et al., 2009 Substance abuse
Field et al., 2009 A lcohol abuse
Freyer et al., 2009 A lcohol abuse
Frias et al., 2016 Dysthymia
G eller et al., 2004 Eating disorder
Genders et al., 2010 Eating disorder
G om ez et al., 2012 Problem gambling
G ossop et al., 2007 Substance abuse
G ouse et al., 2016 Substance abuse
Haller et al., 2004 Substance abuse
Henderson et al., 2004 Substance abuse
Hewes & Janikowski, 1998 A lcohol abuse
Readiness Measure N d SE 95% Cl
Lower Upper
U RICA 138 .41 .19 .03 .78
U RICA 1187 .49 .08 .34 .65
U RICA 1187 .52 .10 .33 .70
U RICA 590 .62 .15 .33 .91
RCQ (T V ) 742 -.0 2 .01 -.0 3 .00
SO CRA TES 350 .58 .13 .33 .83
U RICA 60 1.30 .32 .68 1.92
SO C Q 439 .04 .15 -.2 5 .33
U RICA 81 .68 .24 .20 1.15
U R IC A -D V 199 .72 .17 .39 1.04
U RICA 93 .63 .22 .20 1.06
SO CRA TES 1588 .51 .04 .43 .58
U RICA 831 -.0 4 .11 -.2 5 .17
RCQ (T V ) 538 .37 .15 .06 .67
U RICA 61 .70 .30 .11 1.28
RM I 60 .75 .36 .04 1.45
M otivational ruler 30 1.04 .49 .07 2.01
U RICA 191 .82 .16 .51 1.13
SO CRA TES 1075 .90 .08 .74 1.06
SO CRA TES 986 .37 .18 .01 .72
U RICA 75 .87 .26 .36 1.38
U RICA 96 .68 .22 .25 1.10
SO CRA TES 58 2.49 .60 1.31 3.68
Hillen et al., 2015 Eating disorder A N SO CQ
Hunt et al., 2006 Posttraum atic stress disorder U RICA
and alcohol dependence
Ilagan et al., 2015 M ixed diagnoses Self-report scale
Isenhart 1997 A lcohol abuse SO CRA TES
Jakupcak et al., 2013 M ixed diagnoses U RICA
Kerns et al., 2005 Pain Pain SO C
Kinnam an et al., 2007 A lcohol abuse CSO C
Lewis et al., 2009 Depression Beilis SO C
Lewis et al., 2012 M ixed diagnoses SO C Q
Litt et al., 2013 M arijuana dependence RTC Q
M ahon et al., 2015 M ixed diagnoses U RICA
M ander et al., 2013 Eating disorder U RICA
M cKay et al., 2013 Substance abuse U RICA
M itchell, 2006 Substance abuse SO CRA TES
Myers et al., 2016 Substance abuse SO CRA TES
Pantalon et al., 2002 Substance abuse U RICA
Pantalon et al., 2003 Psychiatric inpatients U RICA
Petry et al., 2005 Gambling disorder U RICA
Ronan et al., 2010 V iolence SCQ
R ooney et al., 2005 Posttraum atic stress disorder U RICA
Scott & Wolfe, 2003 Dom estic abuse U RICA
Sherm an et al., 2016 Substance abuse SO CRA TES
Sm ith et al., 1995 G eneral therapy U RICA
Soberay et al., 2014 Gambling U RICA
Solem et al., 2016 Obsessive-compulsive disorder U RICA
 40 1.10 .45 .22 1.98
42 .68 .35 .00 1.36

331 .61 .23 .16 1.07

125 .69 .19 .32 1.07
104 .78 .21 .36 1.19
68 .00 .27 -.5 2 .52
120 .72 .20 .34 1.11
332 .30 .12 .08 .53
173 .32 .16 .01 .62
215 .27 .14 -.0 1 .54
124 .29 .35 -.3 9 .97
35 .18 .40 -.6 1 .97
268 .46 .14 .18 .74
357 .63 .11 .41 .85
335 .27 .15 -.0 2 .57
117 .46 .20 .06 .85
120 -.2 0 .09 -.3 8 -.0 1
234 .47 .15 .17 .78
262 .14 .16 - .1 7 .45
50 .63 .31 .03 1.23
194 .59 .21 .19 .99
175 -.0 5 .25 -.5 5 .44
74 1.84 .33 1.20 2.48
77 .26 .24 -.2 2 .74
121 .30 .19 - .0 7 .66

( c o n t in u e d )
Table 10.3. Continued
Study Primary Diagnosis Readiness Measure N d SE 95% CI
Lower Upper
Soler et al., 2008 Borderline personality disorder U RICA 60 .54 .61 - .6 7 1.74
Stotts et al., 2003 A lcohol and tobacco abuse U RICA 115 .49 .24 .03 .96
Tambling & Johnson, 2008 Relationship issue U RICA 469 -.1 3 .11 -.3 5 .08
Treasure et al., 1999 Eating disorder U RICA 125 .68 .29 .11 1.24
Wade et al., 2009 Anorexia A N SO CQ 47 .31 .31 -.2 9 .91
W illoughby et al., 1996 A lcohol abuse U RICA 152 -.1 5 .17 -.4 9 .18
Zem ore et al., 2014 Substance abuse U RIC A and TREA T 200 -.0 2 .20 -.4 2 .38
Overall Effect Size .41 .03 . 34 .48

Note. SE = standard error; CI = confidence interval; URICA = University of Rhode Island Change Assessment; SOCRATES = Stages of Change Readiness and Treatment Eagerness
Scale; RMI = Readiness and Motivation Interview.
 315 Stages o f Change

disorder outcomes. Studies employed the Eating Disorders Inventory, measures from
the European COST Action B6 Project, and count of relapse to assess outcomes. The
average effect size was d = .59 (95% CI = .34-.85). Twenty-two studies predicted sub­
stance use outcomes using baseline readiness to change. The most frequently used
outcome measures were the Addiction Severity Index, Severity of Dependence Scale,
Timeline Followback, and the Alcohol Use Questionnaire. The mean effect size was
d = .31 (95% CI = .2 0 - .42). Twenty studies assessed the relation between baseline
readiness to change and outcomes for mood disorder symptoms or relational distress,
which were deemed sufficiently similar to group together to increase reliability of the
estimate. Outcome measures included the State-Trait Anxiety Scale, Beck Depression
Inventory, Children’s Depression Rating Scale, and Outcome Questionnaire 45. The
mean effect size was d = .39 (95% CI = .27-.51).

MET-ANALYTIC REVIEW: STAGE-MATCHED TREATMENTS

Our second aim was to conduct a meta-analysis on psychotherapy studies that matched
treatment to specific stages or readiness levels of change. We were interested in learning
whether tailoring psychotherapy to the client’s stage of change produced the superior
results found in behavioral medicine and population-based studies reviewed earlier.
Unfortunately, we located no controlled group studies meeting our inclusion criteria
that matched psychotherapy to client stage or readiness. As a result, we could not per­
form a meta-analysis.
A number of studies did use in-person sessions and delivered treatment based
on stage or readiness to change but otherwise did not meet inclusion criteria in that
treatment either was a single session, provided by medical staff, or focused on health
behaviors such as smoking, physical activity, or diabetes management (Champion
et al., 2003; Chouinard & Robichaud-Ekstrand, 2007; Clark et al., 2004; Patten et al.,
2008; Van Sluijs et al., 2005; Wiggers et al., 2005). The one study that intervened on
mental and addictive disorders was not individually stage-tailored (James et al., 2004).
The failure to locate stage-matching studies in psychotherapy reflects, first, the
obvious dearth of such studies, and second, the pervasiveness of the medical model
in matching treatments to the patient’s disorder (Wampold & Imel, 2015). Third, the
paucity of such studies underscores the limited reach of conventional psychotherapy.
Psychotherapy has traditionally taken a passive and narrow approach to healthcare—
passively waiting for individuals suffering from mental health disorders in the contem­
plation or preparation stages to contact clinicians’ offices. In stark contrast, behavioral
medicine has adopted a proactive approach to recruiting and intervening with en­
tire populations in all stages of change. Not surprisingly, there are now hundreds of
published stage-matching studies in behavioral medicine.

EVIDENCE FOR CAUSALITY

In behavioral medicine and the addictions, the patient’s stage of change reliably predicts
outcomes, and matching treatments to the patient’s stage of change demonstrably
 3 16 psy c h o th er a py rela tio n sh ips that w ork

improves treatment outcomes. Dozens of RCTs and several meta-analyses provide ev­
idence of the causal link.
In psychotherapy, the stages of change are moderately associated with and reliably
predict patient outcomes, as evidenced in the current meta-analytic results. However,
there are insufficient RCTs to make any causal claim for the efficacy of stage matching
in psychotherapy at this time. O f course, the absence of evidence does not mean the
evidence of absence. Based on all the available research, stage-matching psychotherapy
likely produces similar benefits as in behavioral medicine and the addictions, but such
psychotherapy studies await completion.

LIMITATIONS OF THE RESEARCH

Although more than 4,000 research studies have been published on the stages of
change, none have directly and prospectively matched and mismatched psychotherapy
to the patient’s stage of change. Rather, the available research concerns the predictive
utility of the stages of change in terms of outcomes and dropouts, the differential use of
the processes of change at various stages of change, and the relative efficacy of assorted
forms of service delivery. Further, the majority of published research concern health
behaviors and addictive disorders, as contrasted to the wide range of mental disorders.
In the future, we anticipate controlled trials of such stage matching will be conducted
in psychotherapy proper. The merits and technologies of those RCTs are widely un­
derstood, as seen in controlled studies of treatment adaptations for patient cultures,
preferences, and reactance levels (see other chapters in this volume).
More broadly, we enthusiastically recommend that psychotherapy researchers join
the paradigm shift, in part initiated by the TTM , toward proactive outreach to entire
populations. Proactive outreach will markedly increase the percentage of high-risk and
suffering people receiving psychosocial treatment for behavioral disorders. Because
only a small minority of the population will be ready to take action, psychotherapists
will design treatments for the population at every stage: the 20% or less in the prepa­
ration stage, the 40% in the contemplation stage, and the 40% in the precontemplation
stage. By reaching out and customizing services to readiness to change, psychotherapists
can achieve a quantum increase in our ability to care for those suffering (Kazdin &
Rabbitt, 2013; Prochaska & Prochaska, 2016).

DIVERSITY CONSIDERATIONS

The stages of change have been found, in hundreds of studies, to apply to self-changers
and psychotherapy patients of diverse ages, cultures, disability statuses, ethnicities,
gender identities, races, religions, and sexual orientations. The moderator analyses
found that the stages of change evidenced similar outcome association and prediction
for patients of disparate ages, genders, and races/ethnicities. The stages are largely gen-
eralizable across cultures, disorders, and treatment settings as they represent, in our
view, the underlying structure of behavior change (Prochaska et al., 1992).
 3 17 Stages o f Change

Nonetheless, the majority of patient samples in our meta-analysis consisted pri­

marily of Whites, and over 90% of the studies were conducted in North America or
Europe (Table 10.2). Moreover, we were unable to include five potential studies in the
meta-analysis because they were published in non-English outlets. More studies on
the stages of change in psychotherapy from cultures and populations outside Western
developed countries are sorely needed.
One implicit cultural assumption of the TTM concerns the value of behavior change
itself. Change as progress is typically a Western and especially an American ideology.
Different cultures raise serious challenges to the belief that change represents progres­
sion, individually or culturally.
As with any transdiagnostic patient characteristic, practitioners cannot assume that
the stage of change defines the person’s experience. We respectfully discuss with the
client which factors, including cultures or intersections of cultures, prove fundamental
to tailoring psychotherapy. Automatically presuming that a client’s stage of change
should be the primary determinant of treatment selection is probably as hurtful as
ignoring it altogether.

TRAINING IMPLICATIONS
On the basis of the research evidence and our training experience, we offer the fol­
lowing recommendations for clinical training and supervision.

♦ Train students to assess the clients stage o f change. Probably the most obvious and
direct implication is to assess the stage of a client’s readiness for change and to tailor
treatment accordingly. In clinical practice, assessing stage of change typically entails
a straightforward question: “Would you say you are not ready to change in the next
six months (precontemplation), thinking about changing in the next six months
(contemplation), thinking about changing in the next month (preparation), or have
you already made some progress (action)?” Additionally, for specific diagnoses and
treatment settings, measures such as the URICA and or Anorexia Nervosa Stages of
Change Questionnaire (ANSOQ) can easily be administered.
♦ Help students expect variability in patients’ stages o f change. A useful guide is the
“4 0 -4 0 -2 0 rule” in the population at large (not in action-oriented treatment
programs): approximately 40% will be in precontemplation, 40% in contemplation,
and 20% in preparation or ready for action (Velicer et al., 1995).
♦ Train students integratively. Competing systems of psychotherapy have
promulgated apparently rival processes of change. However, ostensibly
contradictory processes become complementary when embedded in the stages
of change. Specifically, change processes traditionally associated with the
experiential, cognitive, and psychoanalytic persuasions prove most useful during
the precontemplation and contemplation stages. Change processes traditionally
associated with the existential and behavioral traditions, by contrast, are most
useful during the action and maintenance stages. Each psychotherapy system has
 3 18 psy c h o th er a py rela tio n sh ips that w ork

a place, but a differential place, in the therapeutic repertoire to assist clients to

traverse the stage of change.
♦ Teach students to predicate their therapeutic relationships more on the patients
characteristics (e.g, stage o f change, preferences) than on their theoretical
prescriptions. Do not ask what Freud, Rogers, or Beck theorized about the
therapeutic relationship. Instead, ask the consequential questions of what the
patient prefers, what matches his or her stage, and what the research indicates will
facilitate movement to maintenance and well-being.
♦ Provide integrative supervision that tailors supervision to the individual trainee as
he or she simultaneously adapts psychotherapy to individual clients. As students are
learning to match psychotherapy to their patient’s transdiagnostic features, such as
stage of change, culture, preferences, and reactance level, in parallel process their
supervisors are tailoring supervision to multiple student characteristics (Norcross
& Popple, 2017). “Example is always more efficacious than precept" as Samuel
Johnson observed.
♦ Train students to pursue and calculate societal impact, not only treatment
efficacy. Historically, psychotherapy outcome was evaluated by efficacy, the
percentage of patients who were successful at posttreatment or follow-up. The
TTM maintains that treatment success is more than efficacy alone. Impact is
defined as the participation rate x efficacy. If the best practice that produces 30%
efficacy generates 5% participation, its impact is 1.5%. If an alternative practice
that produces 20% efficacy generates 75% participation, its impact is 15%. The
apparently less effective practice (in terms of efficacy) actually has 10 times as
much impact on the population. From the beginning, psychotherapy students
should be mindful of the larger, bolder goal of impact.

THERAPEUTIC PRACTICES

Almost four decades of clinical research on the stages of change, including the meta­
analyses reviewed in this chapter, have identified a number of therapist behaviors that
will probably improve psychotherapy outcomes.

♦ Beware o f treating all patients as though they are in action. Professionals frequently
design excellent action-oriented treatments but then are disappointed when only a
small percentage of clients seek that therapy. The vast majority of patients are not
in the action stage, and thus professionals offering only action-oriented programs
are likely underserving or misserving the majority of their target population.
The therapeutic recommendation is to move from an action paradigm to a stage
paradigm.
♦ Set realistic goals by moving one stage at a time. A goal for many patients,
particularly in a time-limited managed care environment, is to set realistic goals,
such as helping patients progress from precontemplation to contemplation. Such
progress means that patients are changing if we view change as a process that
 319 Stages o f Change

unfolds over time, through a series of stages. Helping patients break out of the
chronic, stuck phase of precontemplation is a therapeutic success, since it almost
doubles the chances that patients will take effective action in the next six months. If
we help them progress two stages with brief therapy, we triple the chances they will
take effective action.
♦ Treatprecontemplators gingerly. We know that people in precontemplation
underestimate the pros of changing, overestimate the cons, feel defensive when
pressured, and are not particularly conscious of their defenses’ mistakes (Hall &
Rossi, 2008). Patients in preaction stages of change have lower expectations of
therapist acceptance, genuineness, and trustworthiness (Satterfield et al., 1995).
When psychotherapists try to impose action on these patients, they are likely to drive
them away, consequently blaming the clients for being resistant, unmotivated, or
noncompliant. Instead, match your relationships and change processes to the stage.
Motivational interviewing (Miller & Rollnick, 2012) has brilliantly incorporated
these lessons into its philosophical spirit and its treatment methods. A number of
studies included in this meta-analysis found that a few brief motivational sessions can
improve retention and ultimately outcome (Carroll et al., 2006; Sorsdahl et al., 2015).
♦ Tailor the processes to the stages. The research reliably demonstrates that patients
optimally progress from precontemplation and contemplation into preparation
by use of consciousness-raising, self-liberation, and dramatic relief/emotional
arousal. Patients progress best from preparation to action and maintenance by
use of counterconditioning, stimulus control, and reinforcement management.
To simplify: Use change processes traditionally associated with the insight or
awareness therapies for the early stages and change processes associated with the
action therapies for the later stages.
♦ Avoid mismatching stages and processes. A person’s stage of change provides
proscriptive as well as prescriptive information on treatments of choice. Action-
oriented therapies may prove quite effective with individuals who are in the
preparation or action stages. These same programs tend to be ineffective or
detrimental, however, with individuals in precontemplation or contemplation.
We have observed two frequent mismatches (Prochaska et al., 1995). First,
some therapists rely primarily on change processes most indicated for the
contemplation stage—consciousness raising, self-reevaluation—while they are
moving into the action stage. They try to modify behaviors by becoming more
aware, a common criticism of classical psychoanalysis: insight alone does not
necessarily bring about behavior change. Second, other therapists rely primarily on
change processes most indicated for the action stage—reinforcement management,
stimulus control, counterconditioning—without the requisite awareness, decision­
making, and readiness provided in the contemplation and preparation stages.
They try to modify behavior without awareness, a common criticism of radical
behaviorism: overt action without insight is likely to lead to temporary change.
♦ Prescribe stage-matched relationships o f choice as well as treatments o f choice.
Similar to using treatments of choice offering “therapeutic relationships of choice”
could enhance therapy outcomes (Norcross & Beutler, 1997). Once you know a
 320 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

patient’s stage of change, then you will know which relationship stances to apply
to help him or her progress to the next stage and eventually maintenance. These
relational matches, as reviewed earlier, entail a nurturing parent stance with a
precontemplator, a Socratic teacher role with contemplator, an experienced coach
with a patient in action, and then a consultant once into maintenance.
♦ Practice integratively. Psychotherapists moving with their patients through the
stages of change over the course of treatment will probably employ relational
stances and change processes traditionally emphasized by disparate systems of
psychotherapy. That is, they will practice de facto psychotherapy integration
(Norcross & Goldfried, 2005). Our research has consistently documented that
psychotherapists in their consultation rooms (and self-changers in their natural
environments) can be remarkably effective in synthesizing powerful change
processes across the stages (Connors et al., 2013).
♦ Anticipate recycling: Most psychotherapy patients will recycle several times through
the stages before achieving long-term maintenance. Accordingly, professionals and
programs expecting people to progress linearly through the stages of change are
likely to gather disappointing results. Be prepared to include relapse prevention
in treatment, anticipate the probability of recycling patients, and try to minimize
therapist guilt and patient shame over recycling (Prochaska et al., 2013).
♦ Integrate readiness to change into treatment resources. Readiness to change measures
can be built into self-help materials, health apps, online treatments, and similar
resources to enable tailoring of interventions in ways that improve outcomes. The
stages of change have been incorporated into several online assessments (e.g.,
ProChange [http://www.prochange.com/], InnerLife [http://www.innerlife.com/
]) and self-help books (e.g., Changing to Thrive by Prochaska and Prochaska, 2016;
Changeology by Norcross, 2015). But the opportunities for more are expanding
rapidly with the increased availability and popularity of health apps and online
treatments. These resources can complement and expand psychotherapy, as well as
reach underserved populations.
♦ Shift to an expanded view o f psychotherapy as proactive, population-based healthcare.
Psychotherapists need not discard effective means of assisting individuals suffering
from mental disorders. Instead, they can add to these invaluable services by
providing proactive recruitment and treatment of entire populations suffering from
chronic biobehavioral conditions. Such an expansion could produce unprecedented
impacts on the health and happiness of the populace.

REFERENCES
References m arked with an asterisk indicate studies included in the m eta-analysis.
*A bd Elbaky, G. B., Hay, P. J., le Grange, D., Lacey, H., Crosby, R. D., & Touyz, S. (2014). Pre­
treatm ent predictors o f attrition in a randomised controlled trial o f psychological therapy
for severe and enduring anorexia nervosa. BMC Psychiatry, 14(69). https://www.doi.org/
10.1186/ 1471-244X -14-69
 32 1 Stages o f Change

”"Alexander, P. C., & M orris, E. (2008). Stages o f change in batterers and their response to treat­
m ent. Violence and Victims, 2 3 (4 ), 4 7 6 -4 9 2 .
”A lexander, P. C., M orris, E., Tracy, A., & Frye, A. (2010). Stages o f change and the group
treatm ent o f batterers: A randomized clinical trial. Violence and Victims, 2 5(5), 5 7 1 -5 8 7 .
”Algars, M ., Ramberg, C., Moszny, J., Hagman, J., Rintala, H., & Santtila, P. (2015). Readiness
and m otivation for change am ong young women with broadly defined eating disorders.
Eating Disorders, 2 3 (3 ), 2 4 2 -2 5 2 . https://www.doi.org/10.1080/10640266.2014.1000100
”A losso, J. (2012). The relationship o f readiness to change and clinical outcom e during
residential treatm ent o f severe obsessive compulsive disorder. Dissertation Abstracts
International: Section B: The Sciences and Engineering, 7 3(4-B ), 2493.
”Ametller, L., Castro, J., Serrano, E., M artinez, E., & Toro, J. (2005). Readiness to recover in
adolescent anorexia nervosa: Prediction o f hospital admission. Journal o f Child Psychology
and Psychiatry and Allied Disciplines, 4 6(4), 3 9 4 -4 0 0 .
”Bachiller, D., Grau-Lopez, L., Barral, C., Daigre, C., A lberich, C., Rodriguez-Cintas, L., . . .
Roncero, C. (2015). M otivational interviewing group at inpatient detoxification, its influ­
ence in m aintaining abstinence and treatm ent retention after discharge. Adicciones, 27(2),
1 0 9 -1 1 8 .
”Bates, M . (2014). The development and evaluation o f a 5-week readiness for change precursor
to group cognitive-behavioral therapy for individuals with eating disorders. Dissertation
Abstracts International: Section B: The Sciences and Engineering, 7 5(6-B [E ]).
”Bauer, S., Strik, W , & Moggi, F. (2014). M otivation as a predictor o f drinking outcom es after
residential treatm ent program s for alcohol dependence. Journal o f Addiction Medicine,
8 (2 ), 1 3 7 -1 4 2 . https://www.doi.org/10.1097/ADM.0000000000000013
”Berry, C. C. (2012). Clinical factors related to and predictive ofpsychiatric inpatient length o f stay.
Dissertation Abstracts International: Section B: The Sciences and Engineering, 73(6-B ), 3943.
Biostat. (2006). Comprehensive meta-analysis. Englewood, NJ: Author.
”Blanchard, K. A., M orgenstern, J., Morgan, T. J., Labouvie, E., & Bux, D. A. (2003).
M otivational subtypes and continuous measures o f readiness for change: C oncurrent and
predictive validity. Psychology o f Addictive Behaviors, 17(1), 5 6 -6 5 .
Borenstein, M ., Hedges, L. V., Higgins, J. P T., & Rothstein, H. R. (2009). Introduction to Meta­
Analysis. Wiley. ISBN 9780470057247.
”Boswell, J. F., Sauer-Zavala, S. E., Gallagher, M. W , Delgado, N. K., & Barlow, D. H. (2012).
Readiness to change as a m oderator o f outcom e in transdiagnostic treatm ent. Psychotherapy
Research, 2 2 (5 ), 5 7 0 -5 7 8 .
”Brodeur, N., Rondeau, G., Brochu, S., Lindsay, J., & Phelps, J. (2008). Does the transtheoretical
m odel predict attrition in dom estic violence treatm ent programs? Violence and Victims,
23 (4), 4 9 3 -5 0 7 .
Brogan, M. M ., Prochaska, J. O., & Prochaska, J. M. (1999). Predicting term ination and co n ­
tinuation status in psychotherapy using the transtheoretical model. Psychotherapy, 36(2),
1 0 5 -1 1 3 .
Budney, A. J., Higgins, S. T., Radonovich, K. J., & Novy, P L. (2000). Adding voucher-based
incentives to coping skills and motivational enhancement improves outcomes during treatment
for marijuana dependence. Journal o f Consulting and Clinical Psychology, 68(6), 1051-1061.
Budney, A. J., M oore, B. A., Rocha, H. L., & Higgins, S. T. (2006). C linical trial o f abstinence-
based vouchers and cognitive-behavioral therapy for cannabis dependence. Journal o f
Consulting and Clinical Psychology, 74(2), 3 0 7 -3 1 6 .
 322 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

*Callaghan, R. C., Hathaway, A., Cunningham , J. A., Vettese, L. C., Wyatt, S., & Taylor, L.
(2005). Does stage-of-change predict dropout in a culturally diverse sample o f adolescents
admitted to inpatient substance-abuse treatm ent? A test o f the Transtheoretical Model.
Addictive Behaviors, 30(9), 1 834-1847.
*Callaghan, R. C., Taylor, L., M oore, B. A., Jungerm an, F. S., Vilela, F. A. D. B., & Budney, A. J.
(2008). Recovery and U R IC A stage-of-change scores in three m arijuana treatm ent studies.
Journal o f Substance Abuse Treatment, 35(4), 4 1 9 -4 2 6 .
*Carpenter, K. M ., Miele, G. M ., & Hasin, D. S. (2002). Does m otivation to change mediate
the effect o f D SM -IV substance use disorders on treatm ent utilization and substance use?
Addictive Behaviors, 2 7 (2 ), 2 0 7 -2 2 5 .
Carroll, K. M ., Ball, S. A., Nich, C., M artino, S., Frankforter, T. L., Farentinos, C., . . . Woody,
G. E. (2006). M otivational interviewing to improve treatm ent engagem ent and out­
com e in individuals seeking treatm ent for substance abuse: A multisite effectiveness
study. Drug and Alcohol Dependence, 81(3), 3 0 1 -3 1 2 . https://www.doi.org/10.1016/
j.drugalcdep.2005.08.002
*C astro-Fornieles, J., Bigorra, A., M artinez-M allen, E., Gonzalez, L., M oreno, E., Font, E., &
Toro, J. (2011). M otivation to change in adolescents with bulim ia nervosa m ediates clinical
change after treatm ent. European Eating Disorders Review, 19(1), 4 6 -5 4 . https://www.doi.
org/10.1002/erv.1045
Cham pion, V., M araj, M ., Hui, S., Perkins, A. J., Tierney, W., M enon, U., & Skinner, C. S.
(2003). Com parison o f tailored interventions to increase m am m ography screening in
nonadherent older women. Preventive Medicine, 36(2), 1 5 0-158.
*C hoi, S., Adams, S. M ., MacM aster, S. A., & Seiters, J. (2013). Predictors o f residential treat­
m ent retention am ong individuals with co-occurring substance abuse and m ental health
disorders. Journal o f Psychoactive Drugs, 4 5(2), 1 2 2 -1 3 1 . https://www.doi.org/10.1080/
0 2 7 9 1 0 72.2013.785817
Chouinard, M. C., & Robichaud-Ekstrand, S. (2007). Predictive value o f the transtheoretical
m odel to sm oking cessation in hospitalized patients with cardiovascular disease. European
Journal o f Cardiovascular Prevention and Rehabilitation, 14(1), 5 1 -5 8 .
*Chung, T., & M aisto, S. A. (2009). “W hat I got from treatm ent”: Predictors o f treatm ent co n ­
tent received and association o f treatm ent content with 6-m onth outcom es in adolescents.
Journal o f Substance Abuse Treatment, 37(2), 1 7 1-181.
Clark, M ., Hampson, S. E., Avery, L., & Sim pson, R. (2004). Effects o f a tailored lifestyle
self-m anagem ent intervention in patients with type 2 diabetes. British Journal o f Health
Psychology, 9 (3), 3 6 5 -3 7 9 .
*Clarke, N. (2012). The effects o f therapeutic alliance and client readiness to change on cog­
nitive behavior therapy treatm ent outcom es for a sample o f substance and non-substance
abusing psychiatric inpatient women. Dissertation Abstracts International: Section B: The
Sciences and Engineering, 7 3 (3-B ), 1843.
Cohen, J. (1988). Statistical Power Analysis fo r the Behavioral Sciences (2nd ed.). Hillsdale,
NJ: Lawrence Erlbaum Associates, Publishers.
*C onnors, G. J., D iClem ente, C. C., Derm en, K. H., Kadden, R., Carroll, K. M ., & Frone, M.
R. (2000). Predicting the therapeutic alliance in alcoholism treatm ent. Journal o f Studies
on Alcohol, 6 1(1), 1 3 9-149.
C onnors, G. J., D iClem ente, C. C., Valasquez, M. M ., & Donovan, D. M. (2013). Substance
abuse treatment and the stages o f change: Selecting and planning interventions (2nd ed.).
New York, NY: Guilford.
 323 Stages o f Change

*C ook, S., Heather, N., & M cCam bridge, J. (2015). Posttreatm ent m otivation and alcohol
treatm ent outcom e 9 m onths later: Findings from structural equation m odeling. Journal o f
Consulting and Clinical Psychology, 8 3(1), 2 3 2 -2 3 7 . https://www.doi.org/10.1037/a0037981
*D ale, V., Coulton, S., Godfrey, C., Copello, A., Hodgson, R., Heather, N., . . . Tober, G.
(2011). Exploring treatm ent attendance and its relationship to outcom e in a randomized
controlled trial o f treatm ent for alcohol problems: Secondary analysis o f the U K alcohol
treatm ent trial (U K A T T ). Alcohol and Alcoholism, 4 6(5), 5 9 2 -5 9 9 . https://www.doi.org/
10.1093/alcalc/agr079
*D em m el, R., Beck, B., Richter, D., & Reker, T. (2004). Readiness to change in a clinical
sample o f problem drinkers: Relation to alcohol use, self-efficacy, and treatm ent outcome.
European Addiction Research, 10(3), 1 3 3-138.
*Derisley, J., & Reynolds, S. (2000). The transtheoretical stages o f change as a predictor o f pre­
m ature term ination, attendance and alliance in psychotherapy. British Journal o f Clinical
Psychology, 39 (4), 3 7 1 -3 8 2 .
D iClem ente, C. C., Prochaska, J. O., Fairhurst, S. K., Velicer, W. F., Velasquez, M . M ., & Rossi,
J. S. (1991). The process o f sm oking cessation: An analysis o f precontem plation, contem ­
plation, and preparation stages o f change. Journal o f Consulting and Clinical Psychology,
5 9(2), 2 9 5 -3 0 4 .
*D ove, S. (2016). Adherence to treatm ent as a potential m ediator o f the relationship b e­
tween readiness to change and treatm ent outcom e in adolescent depression. Dissertation
Abstracts International: Section B: The Sciences and Engineering, 7 7(2-B [E ]).
*D ozois, D. J. A., Westra, H. A., Collins, K. A., Fung, T. S., & Garry, J. K. F. (2004). Stages
o f change in anxiety: Psychom etric properties o f the University o f Rhode Island Change
Assessm ent (U RIC A ) scale. Behaviour Research and Therapy, 42(6), 7 1 1 -7 2 9 .
Duval, S., & Tweedie, R. (2000). Trim and fill: A simple funnel-plot-based m ethod o f testing
and adjusting for publication bias in m eta-analysis. Biometrics., 56(2), 4 5 5 -4 6 3 .
*Eckhardt, C., H oltzw orth-M unroe, A., Norlander, B., Sibley, A., & Cahill, M . (2008).
Readiness to change, partner violence subtypes, and treatm ent outcom es am ong m en in
treatm ent for partner assault. Violence and Victims, 23 (4), 4 4 6 -4 7 5 .
*Em m erling, M . E., & W helton, W. J. (2009). Stages o f change and the working alliance in
psychotherapy. Psychotherapy Research, 19(6), 6 8 7 -6 9 8 . https://www.doi.org/10.1080/
10503300902933170
*Evans, E., Li, L., & Hser, Y.-I. (2009). Client and program factors associated with dropout
from court m andated drug treatm ent. Evaluation and Program Planning, 32 (3), 2 0 4 -2 1 2 .
https://www.doi.org/10.1016/j.evalprogplan.2008.12.003
*Field, C. A., Adinoff, B., Harris, T. R., Ball, S. A., & Carroll, K. M. (2009). Construct, concu r­
rent and predictive validity o f the U RICA : Data from two m ulti-site clinical trials. Drug
and Alcohol Dependence, 1 0 1 (1 -2 ), 1 15-123.
Fiore, M ., Jaen, C. R., Baker, T. B., et al. (2008). Treating tobacco use and dependence: 2008 up­
date. Rockville, M D : Agency for Healthcare Research and Quality.
*Freyer-A dam , J., Coder, B., O ttersbach, C., Tonigan, J. S., Rumpf, H. J., Joh n, U., & Hapke,
U. (2 0 0 9 ). The p erform ance o f two m otivation m easures and outcom e after alcohol de­
toxification. Alcohol and Alcoholism, 4 4 (1 ), 7 7 -8 3 . https://www.doi.org/10.1093/alcalc/
agn088
*Frias Ibanez, A., Gonzalez Vallespi, L., Palm a Sevillano, C., & Farriols Hernando, N. (2016).
M otivation for change as a predictor o f treatm ent response for dysthymia. Psicothema,
2 8 (2 ), 1 5 6 -1 6 0 . https://www.doi.org/10.7334/psicothema2015.211
 324 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

*Geller, J., Drab-H udson, D. L., W hisenhunt, B. L., & Srikameswaran, S. (2004). Readiness to
change dietary restriction predicts outcomes in the eating disorders. Eating Disorders: The
Journal o f Treatment & Prevention, 12(3), 2 0 9 -2 2 4 .
*G enders, R., & Tchanturia, K. (2010). Cognitive Rem ediation Therapy (C R T ) for anorexia in
group format: A pilot study. Eating and Weight Disorders, 15(4), e234 -e2 3 9 .
*G om ez-Pena, M ., Penelo, E., G ranero, R., Fernandez-Aranda, F., Alvarez-Moya, E.,
Santamaria, J. J., . . . Jim enez-M urcia, S. (2012). Correlates o f m otivation to change in
pathological gamblers com pleting cognitive-behavioral group therapy. Journal o f Clinical
Psychology, 68(7), 7 3 2 -7 4 4 . https://www.doi.org/10.1002/jclp.21867
*G ossop, M ., Stewart, D., & M arsden, J. (2007). Readiness for change and drug use outcomes
after treatm ent. Addiction, 102(2), 3 0 1 -3 0 8 .
*G ouse, H., M agidson, J. F., Burnham s, W , R em m ert, J. E., Myers, B., Joska, J. A., & Carrico,
A. W. (2016). Im plem entation o f cognitive-behavioral substance abuse treatm ent in Sub­
Saharan Africa: Treatm ent engagem ent and abstinence at treatm ent exit. PLoS ONE, 11(1),
e0147900.
Hall, K. L., & Rossi, J. S. (2008). M eta-analytic exam ination o f the strong and weak principles
across 48 health behaviors. Preventive Medicine, 46(3), 2 6 6 -2 7 4 .
*Haller, D. L., Miles, D. R., & Cropsey, K. L. (2004). Sm oking stage o f change is associated
with retention in a sm oke-free residential drug treatm ent program for women. Addictive
Behaviors, 2 9 (6 ), 1 2 65-1270.
*H enderson, M . J., Saules, K. K., & G alen, L. W. (2004). The predictive validity o f the University
o f Rhode Island Change Assessm ent questionnaire in a heroin-addicted polysubstance
abuse sample. Psychology o f Addictive Behaviors, 18(2), 1 0 6-112.
*Hewes, R. L., & Janikowski, T. P (1998). Readiness for change and treatm ent outcom e among
individuals with alcohol dependency. Rehabilitation Counseling Bulletin, 42(1).
*H illen, S., Dempfle, A., Seitz, J., H erpertz-D ahlm ann, B., & Buhren, K. (2015). M otivation to
change and perceptions o f the admission process with respect to outcom e in adolescent an­
orexia nervosa. BMC Psychiatry, 15, 140. https://www.doi.org/10.1186/ s12888-015-0516-8
*H unt, Y. M ., Kyle, T. L., Coffey, S. F., Stasiewicz, P R., & Schumacher, J. A. (2006). University
o f Rhode Island Change Assessm ent— Trauma: Prelim inary psychom etric properties in an
alcohol-dependent PTSD sample. Journal o f Traumatic Stress, 19(6), 9 1 5 -9 2 1 .
*Ilagan, G ., Vinson, M . L., Sharp, J. L., Ilagan, J., & O berm an, A. (2015). Exploring outcomes
and initial self-report o f client m otivation in a college counseling center. Journal o f American
College Health, 6 3(3), 1 8 7 -1 9 4 . https://www.doi.org/10.1080/07448481.2014.1003379
*Isenhart, C. E. (1997). Pretreatm ent readiness for change in m ale alcohol dependent
subjects: Predictors o f one-year follow-up status. Journal o f Studies on Alcohol, 58(4),
3 5 1 -3 5 7 .
*Jakupcak, M ., Hoerster, K. D., Blais, R. K., Malte, C. A., Hunt, S., & Seal, K. (2013). Readiness
for change predicts VA M ental Healthcare utilization am ong Iraq and Afghanistan war vet­
erans. Journal o f Traumatic Stress, 2 6 (1 ), 1 6 5 -1 6 8 . https://www.doi.org/10.1002/jts.21768
James, W , Preston, N. J., Koh, G., Spencer, C., Kisely, S. R., & Castle, D. J. (2004). A group in ­
tervention which assists patients with dual diagnosis reduce their drug use: A randomized
controlled trial. Psychological Medicine, 34 (6), 9 8 3 -9 9 0 .
Kazdin, A. E., & Rabbitt, S. M. (2013). Novel models for delivering m ental health services
and reducing the burdens o f m ental illness. Clinical Psychological Science, 1(2), 1 70-191.
https://www.doi.org/10.1177/2167702612463566
 325 Stages o f Change

*K erns, R. D., Wagner, J., Rosenberg, R., Haythornthwaite, J., & Caudill-Slosberg, M. (2005).
Identification o f subgroups o f persons with chronic pain based on profiles on the pain
stages o f change questionnaire. Pain, 116(3), 3 0 2 -3 1 0 .
*K innam an, J. E. S., Bellack, A. S., Brown, C. H., & Yang, Y. (2007). Assessm ent o f m otivation
to change substance use in dually diagnosed schizophrenia patients. Addictive Behaviors,
3 2(9), 1 7 9 8 -1 8 13.
Krebs, P., Prochaska, J. O., & Rossi, J. S. (2010). A m eta-analysis o f com puter-tailored
interventions for health behavior change. Preventive Medicine, 5 1 (3 -4 ), 2 1 4 -2 2 1 . https://
www.doi.org/10.1016/j.ypmed.2010.06.004
Levesque, D., Ciavatta, M ., Castle, P , Prochaska, J., & Prochaska, J. (2012). Evaluation o f a
stage-based, com puter-tailored adjunct to usual care for dom estic violence offenders.
Psychology of Violence, 2 , 3 6 8 -3 8 4 .
*Lewis, C. C., Sim ons, A. D., & Kim , H. K. (2012). The role o f early symptom trajectories
and pretreatm ent variables in predicting treatm ent response to cognitive behavioral
therapy. Journal o f Consulting and Clinical Psychology, 80(4), 5 2 5 -5 3 4 . https://www.doi.
org/10.1037/a0029131
Lewis, C. C., Sim ons, A. D., Silva, S. G., et al. (2009). The role o f readiness to change in response
to treatm ent o f adolescent depression. Journal o f Consulting and Clinical Psychology, 77(3),
4 2 2 -4 2 8 .
Lipsey, M ., & W ilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: SAGE.
*Litt, M. D., Kadden, R. M ., & Petry, N. M . (2013). Behavioral treatm ent for m arijuana de­
pendence: Random ized trial o f contingency m anagem ent and self-efficacy enhancem ent.
Addictive Behaviors, 38(3), 176 4 -1 7 7 5 . https://www.doi.org/10.1016/j.addbeh.2012.08.011
Lundahl, B., Kunz, C., Brownell, C., Tollefson, D., & Burke, B. (2010). A m eta-analysis o f m o ­
tivational interviewing: Twenty-five years o f em pirical studies. Research on Social Work
Practice, 20, 1 3 7-160.
*M ahon, M ., Laux, J. M ., M cGuire W ise, S., Ritchie, M . H., Piazza, N. J., & Tiamiyu, M . F.
(2015). B rief therapy at a university counseling center: W orking alliance, readiness to
change, and symptom severity. Journal o f College Counseling, 18(3), 2 3 3 -2 4 3 . https://www.
doi.org/10.1002/jocc.12017
*M ander, J., Teufel, M ., Keifenheim , K., Zipfel, S., & Giel, K. E. (2013). Stages o f change, treat­
m ent outcom e and therapeutic alliance in adult inpatients with chronic anorexia nervosa.
BMC Psychiatry, 13(111). https://www.doi.org/10.1186/1471-244X-13-111
McConnaughy, E. A., DiClem ente, C. C., Prochaska, J. O., & Velicer, W. F. (1989). Stages o f
change in psychotherapy: A follow-up report. Psychotherapy, 2 6 (4 ), 4 9 4 -5 0 3 .
McConnaughy, E. A., Prochaska, J. O., & Velicer, W. F. (1983). Stages o f change in psycho­
therapy: M easurem ent and sample profiles. Psychotherapy: Theory, Research & Practice,
20 (3), 3 6 8 -3 7 5 .
*M cKay, J. R., Van Horn, D., R ennert, L., Drapkin, M ., Ivey, M ., & Koppenhaver, J. (2013).
Factors in sustained recovery from cocaine dependence. Journal o f Substance Abuse
Treatment, 4 5 (2), 1 6 3 -1 7 2 . https://www.doi.org/10.1016/j.jsat.2013.02.007
Miller, W. R., & Rollnick, S. (2012). Motivational interviewing: Helping people change.
New York, NY: Guilford.
M iller, W. R., & Tonigan, J. S. (1996). Assessing drinkers' m otivation for change: The Stages of
Change Readiness and Treatm ent Eagerness Scale (SO C R A TES). Psychology o f Addictive
Behaviors, 10(2), 8 1 -8 9 .
 326 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

*M itchell, D., & Angelone, D. J. (2006). Assessing the validity o f the Stages o f Change Readiness
and Treatm ent Eagerness Scale with treatm ent-seeking m ilitary service members. Military
Medicine, 171(9), 9 0 0 -9 0 4 .
*M yers, B., van der Westhuizen, C., Naledi, T., Stein, D. J., & Sorsdahl, K. (2016). Readiness to
change is a predictor o f reduced substance use involvement: Findings from a randomized
controlled trial o f patients attending South A frican em ergency departments. BMC
Psychiatry, 16(35).
Norcross, J. C. (2015). Changeology: 5 steps to realizing your goals and resolutions. New York,
NY: Sim on & Schuster.
Norcross, J. C., & Beutler, L. E. (1997). Determ ining the therapeutic relationship o f choice in
brie f therapy. In J. N. Butcher (Ed.), Objective psychological assessment in managed health
care: A practitioner's guide. New York, NY: Oxford University Press.
Norcross, J. C., & Goldfried, M. R. (2005). Handbook o f psychotherapy integration (2nd ed.).
New York, NY: O xford University Press.
Norcross, J. C., Krebs, P. M ., & Prochaska, J. O. (2011). Stages o f change. In J. C. Norcross
(Ed.), Psychotherapy relationships that work (2nd ed., pp. 2 7 9 -3 0 0 ). New York, NY: Oxford
University Press.
Norcross, J. C., & Popple, L. M. (2017). Supervision essentials fo r integrative psychotherapy.
W ashington, DC: A m erican Psychological Association.
Ockene, J., Kristeller, J. L., Goldberg, R., Ockene, I., M erriam , P., Barrett, S., . . . Gianelly,
R. (1992). Sm oking cessation and severity o f disease: The C oronary A rtery Sm oking
Intervention Study. Health Psychology, 11(2), 1 1 9-126.
*Pantalon, M . V., Nich, C., Frankforter, T., & Carroll, K. M. (2002). The U R IC A as a measure
o f m otivation to change am ong treatm ent-seeking individuals with concurrent alcohol
and cocaine problems. Psychology o f Addictive Behaviors, 16(4), 2 9 9 -3 0 7 .
*Pantalon, M. V., & Swanson, A. J. (2003). Use o f the University o f Rhode Island change as­
sessment to m easure m otivational readiness to change in psychiatric and dually diagnosed
individuals. Psychology o f Addictive Behaviors, 17(2), 9 1 -9 7 .
Patten, C. A., Decker, P. A., Dornelas, E. A., Barbagallo, J., Rock, E., Offord, K. P , . . . Pingree,
S. (2008). Changes in readiness to quit and self-efficacy among adolescents receiving a brief
office intervention for sm oking cessation. Psychology, Health and Medicine, 13(3), 3 2 6 -3 3 6 .
Pearsall, P (2011). 500 therapies: Discovering a science fo r everyday living. New York,
NY: Norton.
*Petry, N. M. (2005). Stages o f change in treatm ent-seeking pathological gamblers. Journal o f
Consulting and Clinical Psychology, 73(2), 3 1 2 -3 2 2 .
Prochaska, J., Evers, K., Castle, P, Johnson, J., Prochaska, J., Rula, E., . . . Pope, J. (2012).
Enhancing multiple domains o f well-being by decreasing multiple health risk behaviors: A
randomized clinical trial. Population Health Management, 15(5), 2 7 6 -2 8 6 . https://www.
doi.org/10.1089/pop.2011.0060
Prochaska, J., & Norcross, J. (2018). Systems o f psychotherapy: A transtheoretical analysis (9th
ed.). New York, NY: Oxford University Press.
Prochaska, J., & Prochaska, J. (2016). Changing to thrive. C enter City, MN: Hazelden.
Prochaska, J. O., & D iClem ente, C. C. (1982). Transtheoretical therapy: Toward a m ore in ­
tegrative m odel o f change. Psychotherapy: Theory, Research & Practice, 19(3), 2 7 6 -2 8 8 .
https://www.doi.org/10.1037/h0088437
 327 Stages of Change

Prochaska, J. O., & D iClem ente, C. C. (1983). Stages and processes o f self-change of
sm oking: Toward an integrative m odel o f change. Journal o f Consulting and Clinical
Psychology, 51 (3), 3 9 0 -3 9 5 .
Prochaska, J. O., D iClem ente, C. C., & Norcross, J. C. (1992). In search o f how people change:
Applications to addictive behaviors. American Psychologist, 4 7(9), 1 1 02-1114.
Prochaska, J. O., D iClem ente, C. C., & Velicer, W. F. (1985). Predicting change in sm oking
status for self-changers. Addictive Behaviors, 10(4), 3 9 5 -4 0 6 .
Prochaska, J. O., & Norcross, J. C. (2010). Systems o f psychotherapy: A transtheoretical analysis
(7th ed.). Pacific Grove, CA: Brooks/Cole.
Prochaska, J. O., Norcross, J. C., & D iClem ente, C. C. (1995). Changing fo r good. New York,
NY: Avon.
Prochaska, J. O., Norcross, J. C., & D iClem ente, C. C. (2013). Applying the stages o f change. In
G. P. Koocher, J. C. Norcross, & B. A. Greene (Eds.), Psychologists' desk reference (3rd ed.,
pp. 1 7 6 -1 8 1 ). New York, NY: Oxford University Press.
*R onan, G., G erhart, J., Bannister, D., & Udell, C. (2010). Relevance o f a stage o f change anal­
ysis for violence reduction training. Journal o f Forensic Psychiatry & Psychology, 21(5),
7 6 1 -7 7 2 . https://www.doi.org/10.1080/14789949.2010.483285
*Rooney, K., Hunt, C., Humphreys, L., Harding, D., Mullen, M ., & Kearney, J. (2005). A test of
the assumptions o f the transtheoretical m odel in a post-traum atic stress disorder popula­
tion. Clinical Psychology and Psychotherapy, 12(2), 9 7 -1 1 1 .
Rosen, C. S. (2000). Is the sequencing o f change processes by stage consistent across health
problem s? A m eta-analysis. Health Psychology, 19(6), 5 9 3 -6 0 4 .
Rossi, J. S. (2002). Comparison o f the use o f significance testing and effect sizes in theory-based
health promotion research. Paper presented at the 43rd annual m eeting o f the Society for
Multivariate Experim ental Psychology.
Satterfield, W. A., Buelow, S. A., Lyddon, W. J., & Johnson, J. T. (1995). Client stages o f change
and expectations about counselling. Journal o f Counseling Psychology, 42 , 4 7 6 -4 7 8 .
Scott, K. L., & Wolfe, D. A. (2003). Readiness to change as a predictor o f outcom e in batterer
treatm ent. Journal o f Consulting and Clinical Psychology, 71(5), 8 7 9 -8 8 9 .
*Sherm an, B. J., Baker, N. L., & M cRae-Clark, A. L. (2016). G ender differences in can ­
nabis use disorder treatm ent: Change readiness and taking steps predict worse cannabis
outcomes for women. Addictive Behaviors, 60, 1 9 7 -2 0 2 . https://www.doi.org/10.1016/
j.add beh.2016.04.014
*Sm ith, K. J., Subich, L. M ., & Kalodner, C. (1995). The Transtheoretical Model's stages and
processes o f change and their relation to premature term ination. Journal o f Counseling
Psychology, 4 2 (1), 3 4 -3 9 .
*Soberay, A. D., Grimsley, P , Faragher, J. M ., Barbash, M ., & Berger, B. (2014). Stages of
change, clinical presentation, retention, and treatm ent outcom es in treatm ent-seeking
outpatient problem gambling clients. Psychology o f Addictive Behaviors, 2 8 (2 ), 4 1 4 -4 1 9 .
https: //ww w. doi.org/10.1037/a0035455
*Solem , S., Husby, A. S., Haland, A. T., Launes, G., Hansen, B., Vogel, P A., & Hagen, R. (2016).
The University o f Rhode Island Change Assessm ent as predictor o f treatm ent outcom e
and dropout in outpatients with obsessive-compulsive disorder treated with exposure and
response prevention. Psychother Psychosom, 8 5(2), 1 1 9 -1 2 0 . https://www.doi.org/10.1159/
000441361
 32 8 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

*Soler, J., Trujols, J., Pascual, J. C., Portella, M . J., Barrachina, J., Cam pins, J., . . . Pérez, V
(2008). Stages o f change in dialectical behaviour therapy for borderline personality dis­
order. British Journal o f Clinical Psychology, 47(4), 4 1 7 -4 2 6 .
Sorsdahl, K., Stein, D. J., Corrigall, J., Cuijpers, P , Smits, N., Naledi, T., & Myers, B. (2015).
The efficacy o f a blended m otivational interviewing and problem solving therapy interven­
tion to reduce substance use am ong patients presenting for em ergency services in South
Africa: A randomized controlled trial. Substance Abuse Treatment, Prevention, and Policy,
10, 46. https://www.doi.org/10.1186/s13011-015-0042-1
*Stotts, A. L., Schm itz, J. M ., & Grabowski, J. (2003). C oncurrent treatm ent for alcohol and
tobacco dependence: Are patients ready to quit both? Drug and Alcohol Dependence,
6 9(1), 1 -7 .
Swift, J. K., & Greenberg, R. P (2012). Prem ature discontinuation in adult psychotherapy: A
m eta-analysis. Journal o f Consulting and Clinical Psychology, 8 0(4), 5 4 7 -5 5 9 . https://www.
doi.org/10.1037/a0028226
*Tam bling, R. B., & Johnson, L. N. (2008). The relationship between stages o f change and out­
com e in couple therapy. American Journal o f Family Therapy, 36(3), 2 2 9 -2 4 1 .
^Treasure, J. L., Katzman, M ., Schm idt, U., Troop, N., Todd, G., & De Silva, P (1999).
Engagement and outcom e in the treatm ent o f bulim ia nervosa: First phase o f a sequential
design com paring m otivation enhancem ent therapy and cognitive behavioural therapy.
Behaviour Research and Therapy, 37(5), 4 0 5 -4 1 8 .
Van Sluijs, E. M . F., Van Poppel, M. N. M ., Twisk, J. W R., Brug, J., & Van M echelen, W (2005).
The positive effect on determ inants o f physical activity o f a tailored, general practice-based
physical activity intervention. Health Education Research, 2 0 (3 ), 3 4 5 -3 5 6 .
Velicer, W F., Fava, J. L., Prochaska, J. O., Abrams, D. B., Em m ons, K. M ., & Pierce, J. P (1995).
D istribution o f smokers by stage in three representative samples. Preventive Medicine,
2 4 (4 ), 4 0 1 -4 1 1 .
*W ade, T. D., Frayne, A., Edwards, S. A., Robertson, T., & G ilchrist, P (2009). M otivational
change in an inpatient anorexia nervosa population and implications for treatm ent.
Australian and New Zealand Journal o f Psychiatry, 43(3), 2 3 5 -2 4 3 .
W ampold, B., & Imel, Z. (2015). The great psychotherapy debate: Evidence fo r what makes psy­
chotherapy work (2nd ed.). Mahwah, NJ: Erlbaum.
Wiggers, L. C. W , O ort, F. J., D ijkstra, A., De Haes, J. C. J. M ., Legemate, D. A., & Sm ets, E.
M. A. (2005). Cognitive changes in cardiovascular patients following a tailored behavioral
sm oking cessation intervention. Preventive Medicine, 4 0(6), 8 1 2 -8 2 1 .
*W illoughby, F. W , & Edens, J. F. (1996). C onstruct validity and predictive utility o f the stages
o f change scale for alcoholics. Journal o f Substance Abuse, 8 (3), 2 7 5 -2 9 1 .
*Zem ore, S. E., & Ajzen, I. (2014). Predicting substance abuse treatm ent com pletion using a
new scale based on the theory o f planned behavior. Journal o f Substance Abuse Treatment,
4 6 (2 ), 1 7 4 -1 8 2 . https://www.doi.org/10.1016/j.jsat.2013.06.011
Zhang, A. Y., Harm on, J. A., Werkner, J., & M cC orm ick, R. A. (2004). Im pacts o f m otivation
for change on the severity o f alcohol use by patients with severe and persistent m ental ill­
ness. Journal o f Studies on Alcohol, 65(3), 3 9 2 -3 9 7 .
 11

p e r s o n a l iz in g p s y c h o t h e r a p y : r e s u l t s ,
c o n c l u s io n s , a n d p r a c t ic e s

John C. Norcross and Bruce E. Wampold

We shall not cease from exploration

And the end of all our exploring
Will be to arrive where we started
And know the place for the first time.
— T. S. Eliot ( “Little Gidding” in Four Quartets)

Having traversed dozens of meta-analyses in these two volumes, we feel like the tireless
traveler in Eliot’s poem: we have rediscovered psychotherapy relationships and respon­
siveness and know them, again, for the first time. This chapter allows us to integrate
the massive corpus of knowledge of the preceding chapters, to present the Task Force’s
formal conclusions and 28 recommendations, and to render some closing reflections
on the entire enterprise. Those recommendations, approved by the 10 members of the
Steering Committee, refer to the work in both this volume on treatment adaptations/re-
lational responsiveness and the first volume on therapy relationships. These statements
reaffirm and, in several instances, extend those of the earlier task forces (Norcross,
2002, 2011). We then conclude with some final thoughts on what works, and what does
not, in adapting psychotherapy to the individual patient.

CONCLUSIONS OF THE INTERDIVISIONAL TASK

FORCE ON EVIDENCE-BASED RELATIONSHIPS
AND RESPONSIVENESS
♦ The psychotherapy relationship makes substantial and consistent contributions to
patient outcome independent of the specific type of psychological treatment.
♦ The therapy relationship accounts for client improvement (or lack of improvement)
as much as, and probably more, than the particular treatment method.
♦ Practice and treatment guidelines should explicitly address therapist behaviors and
qualities that promote a facilitative therapy relationship.

329
 330 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

♦ Efforts to promulgate best practices and evidence-based treatments without

including the relationship and responsiveness are seriously incomplete and
potentially misleading.
♦ Adapting or tailoring the therapy relationship to specific patient characteristics (in
addition to diagnosis) enhances the effectiveness of psychological treatment.
♦ Adapting psychological treatment (or responsiveness) to transdiagnostic client
characteristics contributes to successful outcomes at least as much as, and probably
more than, adapting treatment to the client’s diagnosis.
♦ The therapy relationship acts in concert with treatment methods, patient
characteristics, and other practitioner qualities in determining effectiveness; a
comprehensive understanding of effective (and ineffective) psychotherapy will
consider all of these determinants and how they work together to produce benefit.
♦ The following list summarizes the Task Force conclusions regarding the
evidentiary strength of (a) elements of the therapy relationship primarily provided
by the psychotherapist and (b) methods of adapting psychotherapy to patient
transdiagnostic characteristics.

E le m e n ts o f th e R elatio n sh ip M e th o d s o f A d a p tin g

D e m o n s tra b ly Alliance in Individual Culture (race/ethnicity)

Psychotherapy
E ffe ctiv e Alliance in Child & Adol Religion/Spirituality
Psychotherapy
Alliances in Couple & Family Patient Preferences
Therapy Collaboration
Goal Consensus
Cohesion in Group Therapy
Empathy
Positive Regard and Affirmation
Collecting & Delivering Client
Feedback
P ro b a b ly E ffe ctiv e Congruence/Genuineness Reactance Level
Real Relationship Stages of Change
Emotional Expression Coping Style
Cultivating Positive Expectations
Promoting Treatment Credibility
Managing Countertransference
P ro m is in g b u t Repairing Alliance Ruptures Attachment Style
In su fficien t Immediacy
R e se a rch

I m p o r ta n t b u t N o t Sexual Orientation
Y e t In v e stig a te d Gender Identity
 331 Personalizing Psychotherapy

♦ The preceding conclusions do not constitute practice or treatment standards but

represent current scientific knowledge to be understood and applied in the context of
the clinical evidence available in each case.

RECOMMENDATIONS OF THE INTERDIVISIONAL

TASK FORCE ON EVIDENCE-BASED RELATIONSHIPS
AND RESPONSIVENESS

General Recommendations
1. We recommend that the results and conclusions of this third Task Force be widely
disseminated to enhance awareness and use of what “works” in the psychotherapy
relationship and treatment adaptations.
2. Readers are encouraged to interpret these findings in the context of the
acknowledged limitations of the Task Force’s work.
3. We recommend that future task forces be established periodically to review these
findings, include new elements of the relationship and responsiveness, incorporate
the results of non-English language publications (where practical), and update
these conclusions.

Practice Recommendations
4. Practitioners are encouraged to make the creation and cultivation of the therapy
relationship a primary aim of treatment. This is especially true for relationship
elements found to be demonstrably and probably effective.
5. Practitioners are encouraged to assess relational behaviors (e.g., alliance, empathy,
cohesion) vis-a-vis cut-off scores on popular clinical measures in ways that lead to
more positive outcomes.
6. Practitioners are encouraged to adapt or tailor psychotherapy to those specific
client transdiagnostic characteristics in ways found to be demonstrably and
probably effective.
7. Practitioners will experience increased treatment success by regularly assessing
and responsively attuning psychotherapy to clients’ cultural identities (broadly
defined).
8. Practitioners are encouraged to routinely monitor patients’ satisfaction with
the therapy relationship, comfort with responsiveness efforts, and response
to treatment. Such monitoring leads to increased opportunities to reestablish
collaboration, improve the relationship, modify technical strategies, and
investigate factors external to therapy that may be hindering its effects.
9. Practitioners are encouraged to concurrently use evidence-based relationships and
evidence-based treatments adapted to the whole patient as that is likely to generate
the best outcomes in psychotherapy.
 332 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

Training Recommendations
10. Mental health training and continuing education programs are encouraged to
provide competency-based training in the demonstrably and probably effective
elements of the therapy relationship.
11. Mental health training and continuing education programs are encouraged to
provide competency-based training in adapting psychotherapy to the individual
patient in ways that demonstrably and probably enhance treatment success.
12. Psychotherapy educators and supervisors are encouraged to train students in
assessing and honoring clients’ cultural heritages, values, and beliefs in ways that
enhance the therapeutic relationship and inform treatment adaptations.
13. Accreditation and certification bodies for mental health training programs are
encouraged to develop criteria for assessing the adequacy of training in evidence-
based therapy relationships and responsiveness.

Research Recommendations
14. Researchers are encouraged to conduct research on the effectiveness of therapist
relationship behaviors that do not presently have sufficient research evidence,
such as self-disclosure, humility, flexibility, and deliberate practice.
15. Researchers are encouraged to investigate further the effectiveness of adaptation
methods in psychotherapy, such as to clients’ sexual orientation, gender identity,
and attachment style, that do not presently have sufficient research evidence.
16. Researchers are encouraged to proactively conduct relationship and
responsiveness outcome studies with culturally diverse and historically
marginalized clients.
17. Researchers are encouraged to assess the relationship components using in­
session observations in addition to post-session measures. The former track the
client’s moment-to-moment experience of a session and the latter summarize the
patient’s total experience of psychotherapy.
18. Researchers are encouraged to progress beyond correlational designs that
associate the frequency and quality of relationship behaviors with client
outcomes to methodologies capable of examining the complex causal
associations among client qualities, clinician behaviors, and psychotherapy
outcomes.
19. Researchers are encouraged to examine systematically the associations among
the multitude of relationship elements and adaptation methods to establish a
more coherent and empirically based typology that will improve clinical training
and practice.
20. Researchers are encouraged to disentangle the patient contributions and the
therapist contributions to relationship elements and ultimately outcome.
21. Researchers are encouraged to examine the specific moderators between
relationship elements and treatment outcomes.
 3 33 Personalizing Psychotherapy

22. Researchers are encouraged to address the observational perspective (i.e.,

therapist, patient, or external rater) in future studies and reviews of “what
works” in the therapy relationship. Agreement among observational perspectives
provides a solid sense of established fact; divergence among perspectives holds
important implications for practice.
23. Researchers are encouraged to increase translational research and dissemination
on those relational behaviors and treatment adaptations that already have been
judged effective.
24. Researchers are encouraged to examine the effectiveness of educational,
training, and supervision methods used to teach relational skills and treatment
adaptations/responsiveness.

Policy Recommendations
25. APA Society for the Advancement of Psychotherapy, the APA Society for
Counseling Psychology, and all divisions are encouraged to educate its members
on the benefits of evidence-based therapy relationships and responsiveness.
26. Mental health organizations as a whole are encouraged to educate their members
about the improved outcomes associated with higher levels of therapist-offered
evidence-based therapy relationships, as they frequently now do about evidence-
based treatments.
27. We recommend that the APA and other mental health organizations advocate
for the research-substantiated benefits of a nurturing and responsive human
relationship in psychotherapy.
28. Finally, administrators of mental health services are encouraged to attend to and
invest in the relational features and transdiagnostic adaptations of their services.
Attempts to improve the quality of care should account for relationships and
responsiveness, not only the implementation of evidence-based treatments for
specific disorders.

WHAT WORKS FOR WHOM

As we did in the closing chapter to volume 1, we begin by explaining the process by

which the Task Force Steering Committee reached the preceding conclusions on the
evidentiary strength of the treatment adaptations. These characterizations tend to be
the most frequently cited findings. The conclusions represent the consensus of an ex­
pert panel composed of 10 judges who independently reviewed and rated the research
evidence. They indepdently evaluated the meta-analytic evidence for each adaptation/
responsiveness method on the following criteria: number of empirical studies, consist­
ency of empirical results, independence of supportive studies, magnitude of effect size
of the adaptation method, evidence for causal link between adaptation method and
treatment outcome, and ecological or external validity of research. The panel classified
 334 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

relationship elements as “demonstrably effective," “probably effective" “promising but

insufficient research to judge, “important but not yet investigated” or “not effective"
The experts’ ratings were then combined to render a consensus. In this way, we added
rigor and consensus to the process.
The 10 experts on the Steering Committee of the third Task Force, in alphabetical
order, were

Franz Caspar, PhD, University of Bern

Melanie M. Domenech Rodriguez, PhD, Utah State University
Clara E. Hill, PhD, University of Maryland
Michael J. Lambert, PhD, Brigham Young University
Suzanne H. Lease, PhD, University of Memphis
James W. Lichtenberg, PhD, University of Kansas
Rayna D. Markin, PhD, Villanova University
John C. Norcross, PhD, University of Scranton
Jesse Owen, PhD, University of Denver
Bruce E. Wampold, PhD, University of Wisconsin

Table 11.1 summarizes the meta-analytic findings on the effectiveness of the nine
adaptation/responsiveness methods. The meta-analyses employed the weighted d or g,
standardized mean differences between two treatments or conditions—in this case, the
difference between the conventional or unadapted therapy and the adapted or matched
therapy. In all of these analyses, the larger the value of d, the higher the effectiveness
of the specific adaptation or tailoring. As a reminder, a d of .30 in the behavioral sci­
ences is generally considered a small effect, .50 a medium effect, and .80 a large effect
(Cohen, 1988).
As seen in the Task Force conclusions and in Table 11.1, the expert consensus
deemed three of the methods as demonstrably effective, three as probably effective, one
as promising, and two as important but not yet sufficiently investigated. Practitioners
will find that fitting the therapy to clients’ racial/ethnic culture, religious/spiritual
identity, and therapy preferences will demonstrably improve treatment outcomes and
doing so to clients’ coping style, reactance level, and stages of change will probably do
so as well. Correlational research relating patient attachment security to psychotherapy
outcome is promising, but there are not yet any prospective matching studies. There
are indications from qualitative studies and a handful of uncontrolled quantitative
studies that attending to patients’ gender identity and sexual orientation may prove
efficacious, but the absence of controlled studies does not permit us to reach definitive
conclusions.
The meta-analytic effect sizes in Table 11.1 range from .13 to .78 (indicating a range
of small to large effects) and average about .50 (indicating a medium effect). Compare
those numbers to the 0.0 to .20 average effect sizes for the differential efficacy of one
bona fide psychotherapy over another for a particular mental disorder (Wampold &
Imel, 2015). That’s why the Task Force confidently concluded, “Adapting psychological
treatment (or responsiveness) to transdiagnostic client characteristics contributes to
 335 Personalizing Psychotherapy

Table 11 .1 . Su m m ary o f M eta-A nalytic Results on the Efficacy o f Treatm ent Adaptations/
R elational R esponsiveness to Patient T ransdiagnostic C haracteristics

Patient Characteristic # of Studies (k) # of Patients (N) Effect Size d Consensus on

or g Evidentiary Strength
Attachm ent Style 32 3,158 .35a Prom ising but
insufficient research
to judge
Coping Style 18 1,947 .60 Probably effective
Culture (race/ 99 13,813 .50 Dem onstrably
ethnicity) effective
G ender Identity NA NA NA Im portant but not
yet sufficiently
investigated
Therapy Preferences 51 16,269 .28 Dem onstrably
effective
R eactance Level 13 1,208 .78 Probably effective
Religion and 97 7,181 .1 3 -.4 3 Dem onstrably
Spirituality effective
Sexual O rientation NA NA NA Im portant but not
yet sufficiently
investigated
Stages o f Change 76 21,424 .41b Probably effective

N A = n o t a p p lic a b le ; t h e a u t h o r s d id n o t lo c a t e s tu d ie s to m e t a - a n a ly z e a n d in s te a d p e r f o r m e d a c o n ­
t e n t a n a ly s is o f s e le c t s tu d ie s.

a R e p r e s e n ts c o r r e la t i o n b e t w e e n p r e t r e a t m e n t s e c u r it y a t t a c h m e n t a n d p s y c h o th e r a p y o u t c o m e ; m o r e
s e c u r e a tta c h m e n t/ le s s i n s e c u r it y p r e d ic t e d b e t t e r t r e a t m e n t o u tc o m e s . b R e p r e s e n ts c o r r e l a t i o n b e ­
tw e e n p r e t r e a t m e n t s ta g e s o f c h a n g e a n d p s y c h o th e r a p y o u tc o m e ; p a tie n ts f u r t h e r a lo n g th e sta g es
e x p e r ie n c e b e t t e r t r e a t m e n t o u tc o m e s .

su cce ssfu l o u tco m es at least as m u ch as, and p ro b ab ly m o re th a n , ad ap tin g tre a tm e n t

to th e clie n t’s diagn osis.”
T h e m e ta -a n a ly tic fin d in gs o n ad ap tin g p sy ch o th erap y to p a tien t race/ethnicity,
p re fe re n ce s, and relig io n / sp iritu ality are p a rticu la rly im p ressive. F o r cu ltu ra l id e n ­
tity, th e rese a rch ers analyzed 99 stud ies in v o lv in g 1 3 ,8 1 3 p atien ts. T h e m e a n effect
size o f .50 in favor o f clien ts rec e iv in g cu ltu rally ad apted tre a tm e n ts d em o n strates
th a t “cu ltu ra l fit” w ork s. L ik ew ise, relig io u s/ sp iritu al-ad ap ted p sy ch o th erap y resu lted
in g rea ter im p ro v em en t in clien ts’ p sy ch o lo g ica l (g = .3 3 ) and sp iritu al (g = .43)
fu n c tio n in g co m p ared w ith n o n ad ap ted p sy ch o th erap ies. In m o re rig o ro u s additive
stu d ies, a cco m m o d a te d p sy ch o th erap ies w ere equ ally effective to stan d ard ap p roach es
in re d u cin g p sy ch o lo g ica l d istress (g = .1 3 ) b u t resu lted in g reater sp iritu al w ell-b ein g

(g = .3 4 ).
E ffe c t size n u m b e rs cap tu re and co n v ey lim ite d in fo rm a tio n . T h e sm all to m ed iu m
e ffect size (.2 8 ) o f a cco m m o d a tin g p sy ch o th erap y to p a tien t p re fere n ces proves a case
 336 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O RK

in point. In 28 studies, clients not receiving preferences were almost twice as likely to
drop out (odds ratio = 1.79) That constitutes an important impact clinically.
Unlike the elements of the psychotherapy relationship featured in volume 1, these
treatment adaptations enjoy evidence of direct causal impact. The adaptation methods
are smaller in number but more powerful in demonstrating causation. The meta­
analyses included largely randomized or quasi-randomized controlled trials, in con­
trast to the largely correlational research designs in the therapy relationship.
The number of evidence-based treatment adaptations/responsiveness methods is
smaller than the number of evidence-based treatment methods because the research
evidence for matching has been more elusive, probably for several reasons. First, pro­
spective matching studies on transdiagnostic patient characteristics are much rarer
and far more difficult to fund; we estimate that approximately 90% of federal research
grants for psychotherapy goes to comparing and disseminating manualized treatments
for specific mental disorders. Second, since the matching hypothesis is rarely the prin­
cipal objective of the randomized controlled trial (RCT), researchers resort to retro­
spective or post hoc analyses and indirect measures to investigate the potential (client)
aptitude by treatment interaction. That represents a weak research design and typically
underpowered statistical analyses. Third, experienced practitioners frequently adapt
to their patients as a part of ongoing treatment. Even the most manual-bound psycho­
therapist in a fixed-duration RCT will evidence responsiveness by respond differently
to, say, a patient in the precontemplation stage than one who is in the action stage or
a highly oppositional patient as opposed to a cooperative one (Chu & Kendall, 2009;
Hatcher, 2015). Aptitude by treatment interaction studies try to capture quicksilver in
the clinical setting (Dance & Neufeld, 1988).
The meta-analyses in this volume pertain to adapting psychotherapy, but adapting
other psychosocial treatments in these ways may well prove efficacious. Consider
the stages of change. Our expert panel opined that stage-matching face-to-face psy­
chotherapy was probably effective, but other meta-analyses of RCTs show that stage
matching in behavioral medicine and self-help interventions was demonstrably effec­
tive in (see Chapter 10 for details). Our conclusions pertain to psychotherapy, but we
encourage researchers and practitioners to test the generalizability and boundaries
of these adaptations to other treatments, such as self-help, Internet-mediated, and
psychopharmacological.
We present the treatment adaptations/responsiveness methods in this book as sepa­
rate, stand-alone practices, but every seasoned psychotherapist knows this is certainly
never the case in clinical work. The variance in outcomes for psychotherapy patients
is not easily partitionable into independent contributions of treatments, relationships,
therapists, and patients (Krause & Lutz, 2009). These adaptations never act in isolation
from the psychotherapy relationship, such as empathy, collaboration, or support. Nor
does it seem clinically possible to adapt psychotherapy in meaningful ways to the dis­
tinctive client and not routinely ascertain her feedback on the therapeutic process. All
treatment adaptations probably interconnect—if only in spirit and intent—and prove
symbiotic. In short, while the relationship elements and adaptation methods featured
in this two-volume book “work," they work together and interdependently. The design
 337 Personalizing Psychotherapy

and analysis of psychotherapy outcome studies need to be improved if we are to learn

who successfully treats whom and how (Baldwin & Imel, 2013; Krause & Lutz, 2009).
In our clinical presentations and workshops, we are frequently asked about the ad­
ditive benefits of simultaneously matching psychotherapy to several of these effective
adaptation methods. “What happens if you match to culture and stages of change to­
gether?” Alas, we do not know for certain; only a couple of studies by Larry Beutler
(2011) have investigated the effects of concurrent matching on two client qualities.
Those results indicated added benefit, but no definitive answers are available yet.
Nor has the discipline determined which particular adaptation/responsiveness
methods work best for any single patient. To some extent, it depends on the magnitude
or strength of the effect size of the adaptation. To some extent, it surely depends on the
salience that the client accords to that particular dimension or personal identity (e.g.,
race/ethnicity, gender, religion, sexual orientation). And, to some extent, it depends on
the clinical context and treatment goals. In all instances, success will largely depend on
therapist flexibility and monitoring the client’s experience of the intended responsive­
ness (Bohart & Wade, 2013; Levitt et al., 2016).
We expect that, in the future, psychotherapists will construct reliable assessments
o f their patient’s likely responses to these adaptations. These assessments will
follow the lead of personalized or precision medicine with treatment decisions
tailored to the individual patient based on their predicted response and cost-risk
considerations. Three exemplars of this approach are cultural adaptations, motive-
oriented therapeutic relationship (or plan compatibility; Caspar et al., 2005; Kramer
et al., 2014; Silberschatz, 2017), and the personalized advantage index (DeRubeis
et al., 2014; Hollon et al., 2014) for depression. Research and practice in cultural
adaptations has progressed to the point that, for both children and adults, clinicians
can identity which adaptation elements generally prove most efficacious (see
Chapter 3; Chu & Leino, 2017). In the second, therapists identify each patient’s par­
ticular conflicts, motives (plan), and problems and then try to effectively address
those—responsiveness (or plan compatibility). Multiple studies show that degree
of therapist responsiveness correlate substantially with and predict a variety of pa­
tient outcomes across psychotherapies. In the third exemplar, multivariable models
have determined response patterns for differentially predicting benefit for cogni­
tive therapy, interpersonal therapy, and medication for unipolar depression. Such
individualized assessments can assist psychotherapists in guiding patients to more
precise and effective adaptations.
As the evidence base on adaptations/responsiveness matures, we will know more
about their effectiveness for particular circumstances and conditions. In quant speak,
we will know more about their moderators and mediators. In adapting to culture, the
chapter authors discovered that studies employed multiple elements, such as language,
metaphors, and concepts. Their moderator analyses discerned that the single most ef­
fective adaptation was to use the client’s native or preferred language. Further, the more
cultural adaptations used in treatment, the larger the effect size. How well, then, does
cultural adaptation work in psychotherapy? It depends; it depends on the adaptation
and the circumstances.
 338 P S Y C H O T H E R A P Y RE L AT IO N S H IP S THAT WORK

Amid the torrent of meta-analyses, let us not lose the overarching goal of the
Task Force: to boost psychotherapy effectiveness. The meta-analyses establish that
responsiveness works. Take a mindful moment to consider the direct practice
implications: Adapting therapy to the entire person improves success and decreases
dropouts; the power of responsiveness exceeds that associated with T x Method A for
Disorder Z; this represents not clinical lore but established fact.
In the interminable debate on which psychotherapy works best, we are convinced
that the dispassionate, evidence-based answer is “It depends.” It depends in partic­
ular on the client, including diagnostic features but more importantly transdiagnostic
features. And it depends more on the relationship and responsiveness than a particular
therapy method.
Indeed, in our professional lifetimes, that is the sea change we have witnessed in
our beloved art and science of psychotherapy. The question is no longer “W hat is my
preferred theoretical orientation?” but rather “W hat relationship, adaptation, and ap­
proach will prove most effective with this particular client in this context?”

WHAT DOES NOT WORK

Translational research proves both prescriptive and proscriptive; it tells us what works
and what does not. O f course, we could reverse the effective adaptations identified
in these meta-analyses. Thus what does not work is providing the identical therapy
to clients with externalizing and internalizing coping styles; ignoring race/ethnicity/
culture, gender identity, and sexual orientation altogether; disregarding patient
preferences; and responding to oppositional, highly reactant clients with strong direc­
tion and confrontation. In this section, we highlight a few broader therapist actions
that prove generally ineffective, perhaps even hurtful, in psychotherapy.

♦ Procrustean bed. We should all avoid the crimes of Procrustes, the mythological
Greek giant who would cut the long limbs of clients or stretch short limbs to fit his
one-size iron bed. The efficacy and applicability of psychotherapy will be enhanced
by tailoring it to the unique needs of the client, not by imposing a Procrustean bed
onto unwitting consumers of psychological services. Psychotherapists ought to be
adapting to clients, not the converse.
♦ Singularity. In the quest to adapt psychotherapy, some psychotherapists become
enamored with a single matching protocol and apply that match to virtually every
patient who crosses their path. They are convinced that a single adaptation, be it the
patient’s reactance, diagnosis, culture, or stage of change, is the exclusive means of
tailoring treatment to a successful outcome. However, the research appraised in this
book convincingly demonstrates that many adaptations succeed. We must also guard
against imposing the Procrustean bed when we adapt psychotherapy; one size, even
in adaptation or responsiveness, never works for all clients.
♦ Cultural arrogance. Psychotherapy is inescapably bound to the cultures in which
it is practiced by clinicians and experienced by clients. Arrogant impositions of
therapists’ cultural beliefs in terms of gender, race/ethnicity, sexual orientation, and
 339 Personalizing Psychotherapy

other intersecting dimensions of identity are culturally insensitive and demonstrably

less effective (Chapter 3). By contrast, therapists’ expressing cultural humility and
tracking clients’ satisfaction with cultural responsiveness markedly improve client
engagement, retention, and eventual treatment outcome.
♦ Flexibility without fidelity. The desire to be responsive with patients frequently
gives rise to a clinical dilemma (Norcross et al., 2017). Therapist flexibility to the
patient’s preferences, values, and cultures promises that psychotherapy “fits” the
patient but not necessarily that the resultant treatment has any research support.
Therapist fidelity to a research-supported treatment promises that psychotherapy
“works” but not necessarily with that particular client in that particular context.
Errors in either direction can portend clinical failure, but after an entire volume
dedicated to the benefits of treatment adaptation, we caution against ignoring the
research evidence on the effectiveness of psychological treatments. Focusing solely
on accommodating without addressing the client’s problems or distress will not
be optimally effective (Yulish et al., 2017). While the research supports adaptation
in many cases, the research also recommends fidelity to treatments as found
effective in controlled research. We need to balance flexibility with fidelity (Chu &
Leino, 2017).

CONCLUDING REFLECTIONS
Let us conclude, like T. S. Eliot, by “arriving where we started” and underscoring sev­
eral take-home lessons about leveraging individual differences among humans and
personalizing psychotherapy. In a technology-fueled and drug-filled world (Greenberg,
2016), there is a growing and pervasive tendency to standardize, industrialize, mech­
anize, and biologize what we do with our clients. Psychotherapists would do well to
heed the ancient wisdom in the Hippocratic Oath (modern version): “I will remember
that there is art to medicine as well as science, and that warmth, sympathy, and un­
derstanding may outweigh the surgeon’s knife or the chemist’s drug,” and “I will re­
member that I do not treat a fever chart, a cancerous growth, but a sick human being.”
Reaffirming the human element and attending to the patient’s totality in psychotherapy
stem from both a moral commitment and robust evidence.
The research evidence amounts to little if it is not enacted in practice and taught in
graduate programs. We implore our colleagues to progress beyond the well-intended
slogans of “different strokes for different folks,” “meet the clients where they are,” and
“a new therapy for each patient.” It is time to implement what we know what works
in in adaptations/responsiveness and simultaneously to avoid what does not. Build in
transdiagnostic responsiveness into graduate programs and provide training and up­
take in those adaptation methods that are demonstrably and probably effective.
When the discipline successfully does so, a bevy of benefits will assuredly accrue.
We reclaim the “psych” in psychotherapy. We transcend the limited and divisive “diag­
nosis only” approach to psychotherapy. We narrow the gap between research and prac­
tice. We embrace the clinical reality that patients respond differently. We rediscover the
individual differences that distinguish our field. We reorient from internecine conflict
 340 P S Y C H O T H E R A P Y R E L A T I O N S H I P S T HAT W O R K

to patient benefit. And, most consequentially, we become demonstrably more effective

with our psychotherapy clients.

REFERENCES
Baldwin, S. A., & Imel, Z. E. (2013). Therapist effects: Findings and m ethods. In M. J. Lam bert
(Ed.), Bergin and Garfield’s handbook o f psychotherapy and behavior change (6th ed.).
New York, NY: Wiley.
Beutler, L. E. (2011). Prescriptive m atching and systematic treatm ent selection. In J. C.
Norcross (Ed.), History o f psychotherapy: Continuity and change (2nd ed.). W ashington,
DC: A m erican Psychological Association.
Bohart, A. C., & Wade, A. G. (2013). The client in psychotherapy. In M . J. Lam bert (Ed.)
Bergin & Garfield’s handbook o f psychotherapy & behavior change (6th ed., pp 2 1 9 -2 5 7 ).
New York, NY: Wiley.
Caspar, F., Grossm ann, C., Unmussig, C., & Schram m , E. (2005). Com plem entary thera­
peutic relationship: Therapist behavior, interpersonal patterns, and therapeutic effects.
Psychotherapy Research, 1 5 ,9 1 -1 0 2 .
Chu, B. C., & Kendall, P. C. (2009). Therapist responsiveness to child engagement: Flexibility
within m anual-based C B T for anxious youth. Journal o f Clinical Psychology, 65, 7 3 6 -7 5 4 .
Chu, J., & Leino, A. (2017). Advancements in the m aturing science o f cultural adaptations
o f evidence-based interventions. Journal o f Consulting and Clinical Psychology, 85, 4 5 -4 7 .
Cohen, J. (1988). Statistical power analysis fo r the behavioral sciences (2nd ed.). Hillsdale,
NJ: Erlbaum.
Dance, K. A., & Neufeld, R. W. (1988). A ptitude-treatm ent interaction research in the clin­
ical setting: A review o f attempts to dispel the “patient uniform ity” myth. Psychological
Bulletin, 104, 1 9 2 -213.
DeRubeis, R. J., Cohen, Z. D., Forand, N. R., Fournier, J. C., Gelfand, L. A., & Lorenzo-Luaces,
L. (2014). The Personalized Advantage Index: Translating research on prediction into
individualized treatm ent recom m endations. A dem onstration. PloS ONE, 9 (1), e83875.
Greenberg, R. P. (2016). The rebirth o f psychosocial im portance in a drug-filled world.
American Psychologist, 71, 7 8 1 -7 9 1 .
Hatcher, R. L. (2015). Interpersonal com petencies: Responsiveness, technique, and training in
psychotherapy. American Psychologist, 70, 7 4 7 -7 5 2 .
Hollon, S. D., DeRubeis, R. J., Fawcett, J., Amsterdam , J. D., Shelton, R. C., Zajecka, J., . . . Gallop,
R. (2014). Effect o f cognitive therapy with antidepressant m edications vs antidepressants
alone on the rate o f recovery in m ajor depressive disorder: A randomized clinical trial.
JAMA Psychiatry, 7 1 , 1 1 57-1164.
Kramer, U., Kolly, S., Berthoud, L., Keller, S., Preisig, M ., Caspar, F., . . . Despland, J. N. (2014).
Effects o f m otive-oriented therapeutic relationship in a ten-session general psychiatric
treatm ent o f borderline personality disorder: A randomized controlled trial. Psychotherapy
and Psychosomatics, 83, 1 7 6-186.
Krause, M. S., & Lutz, W. (2009). Process transform s inputs to determ ine outputs: Therapists
are responsible for m anaging process. Clinical Psychology: Science and Practice, 16, 7 3 -8 2 .
Levitt, H. M ., Pomerville, A., & Surcase, F. I. (2016). A qualitative m eta-analysis exam ining
clients’ experience o f psychotherapy: A new agenda. Psychological Bulletin, 142, 8 0 1 -8 3 0 .
Norcross, J. C. (Ed.). (2002). Psychotherapy relationships that work. New York: Oxford
University Press.
 341 Personalizing Psychotherapy

N orcross, J. C. (Ed.). (2011). Psychotherapy relationships that work (2nd ed.). New York,
NY: O xford University Press.
N orcross, J. C., Hogan, T. P., Koocher, G. P., & Maggio, L. A. (2017). Clinician’s guide to
evidence-based practices: Behavioral health and addictions (2nd ed.). New York, NY: Oxford
University Press.
N orcross, J. C., & Wampold, B. E. (2011). W hat works for whom: Tailoring psychotherapy to
the person. Journal o f Clinical Psychology, 6 7 , 1 27-132.
Silberschatz, G. (2017). Improving the yield o f psychotherapy research. Psychotherapy
Research, 2 7 ,1 - 1 3 .
W ampold, B. E., & Imel, Z. (2015). The great psychotherapy debate (2nd ed.). Mahwah,
NJ: Erlbaum.
Yulish, N. E., Goldberg, S. B., Frost, N. D., Abbas, M ., O leen-Junk, N. A., Kring, M ., . . .
W ampold, B. E. (2017). The im portance o f problem -focused treatm ents: A meta-analysis
o f anxiety treatm ents. Psychotherapy, 54, 3 2 1 -3 3 8 .
 INDEX

Page num bers followed by f and t refer to figures and tables, respectively.

AA I (Adult Attachm ent Interview), 18, 20 anger, as resistance, 197

accom m odation focus, 2 4 5 -4 6 , 249 anxiety, 217, 219, 224, 248
accreditation bodies, 332 anxious-resistant (ambivalent)
action-oriented treatm ents, 318, 319 attachm ent style
action stage o f change, 297, 299, 300, 301 clinical examples, 22
activity preferences, 1 5 7 -5 8 , 180 defined, 1 7 -1 8
adaptations, treatm ent. See treatm ent m easuring, 1 9 -2 0
adaptations therapeutic practices for clients with, 47
Adapted Client Resistance Code, 192 See also preoccupied attachm ent style
Addiction Severity Index, 3 1 0 -1 5 APA. See A m erican Psychological
A dherence Rating Scale, 215 A ssociation
Adult Attachment Interview (AAI), 18-19, 20 APA Ethics Code, 2 5 1 -5 2
Adult A ttachm ent Prototype Rating, 21, 47 APA Public Interest Directorate, 2 7 3 -7 4
Adult Attachm ent Q -set, 19 APA Society for Counseling Psychology, 333
Adult A ttachm ent Scale, 21 APA Society for the A dvancem ent o f
affirm ing relational style, 302 Psychotherapy, 333
A frican A m erican clients, 9 5 -1 9 6 APA Society for the Psychology of
A frocentric literature, 92 Sexual O rientation and G ender
agency, individual, 138 Diversity, 2 7 3 -7 4
age o f client Applegarth, G ., 145 t
in attachm ent style study, 43, 45 aptitude by treatm ent interaction studies, 336
in cultural adaptations study, 1 0 8 -9 Argentina, 80
in preference accom m odation studies, 178 arousal, em otional, 2 9 8 1, 299
A insw orth, Mary, 1 5 -1 6 , 17, 18, 19 arrogance, cultural, 3 3 8 -3 9
alcohol use or abuse, 7 4 t, 197, 279 Asian Am ericans, 115, 2 0 6 -7
A lcohol Use Q uestionnaire, 3 1 0 -1 5 assertiveness, 138
allegiance effect, 1 A ssociation for Spiritual, Ethical, and
ambivalent attachm ent style. See anxious- Religious Values in Counseling, 2 5 2 -5 3
resistant attachm ent style assumptions, about gender identity,
A m erican Counseling Association, 2 5 2 -5 3 1 3 7 -3 8 ,1 5 1
A m erican Psychological A ssociation (APA), at-risk clients, in cultural adaptations
119, 157, 2 5 2 -5 3 , 333. See also entries study, 1 0 8 -9
beginning APA attachm ent continuity, 18
anapanasati, 217 landm ark studies, 2 4 -2 5
anchoring rating sheet, 192 as predictor o f change in outcome,
A nderson, N., 2 2 1 1, 224 33, 37, 42
androcentrism , 1 3 4 -3 5 and therapeutic practices, 48

343
 344 IN D E X

attachm ent style, 1 5 -4 8 Boswell, J. F., 1 5 9 -6 0

causality, evidence for, 44 Bowlby, John, 1 5 -1 8
clinical examples, 2 2 -2 4 Brehm , J. W , 188, 190, 191
defined, 17 Brehm , S. S., 188, 190, 191
diversity considerations with, 45 Brennan, K. A., 20
landm ark studies, 24 B rief M ultidim ensional Measure of
m easuring, 1 8 -2 0 Religiousness/Spirituality, 2 1 4 -1 5
m eta-analyses, 2 6 -4 2 B rief Religious Coping Scale, 215
m oderators, effects of, 43 Brogan, M . M ., 303
research lim itations, 45 Buddhism, 217
therapeutic practices, 47 Budge, S. L., 136
training implications with, 46 Burkard, A. W , 2 6 8 1
Attachm ent Style Q uestionnaire, 21
attachm ent theory, 1 5 -1 6 , 4 6 - 4 7 Callahan, J. L., 1 5 9 -6 0 , 168
autonomy, excessive, 21 Caspar, Franz, 334
Aviram, A., 2 0 0 -1 Cassel Hospital inpatient study, 2 4 -2 5
avoidant attachm ent style, 1 7 -1 8 , 1 9 -2 0 Castile, Philando, 9 0 -9 1 , 93
awareness, cultural, 88 categorical measures
Azhar, M . Z., 219 o f coping style, 59
o f resistance, 192 -9 3
Barber, J. P., 66, 70 Catharsis subscale (P E X ), 160
Bartholomew, K., 1 9 -2 0 , 21 CBT. See cognitive-behavioral therapy
Bartholom ew Attachm ent Interview, 20 C C C I-R (Cross-Cultural Counseling
Bartholom ew Attachm ent Rating Scale, 21 Inventory-Revised), 89
Beck Depression Inventory, 3 1 0 -1 5 C enter for Epidem iological Studies
Beck Depression Inventory-II, 214, 227 Depression scale, 214
behavioral drift, 11 certification bodies, 332
Behavioral reactance scale (T R S), 193 change
beliefs cultural value of, 317
client preferences vs., 159 readiness for, 320
religious or spiritual (See religious or resistance as aversion to, 188
spiritual beliefs) See also processes o f change
Berggraf, L., 76 “chase and dodge” sequences, 24
Bernal, G., 9 3 -9 4 , 102 Children’s Depression Rating Scale, 3 1 0 -1 5
Bess, J., 145t Choi, S., 303
Beutler, L. E., 65, 66, 196, 197, 205, 337 Christian-accom m odated CBT, 220, 249
biases Christian clients, 216, 2 1 8 -1 9 , 220, 2 4 5 -4 6
heteronorm ative, 288 cisgender (term ), 134
related to sexual orientation, 286, 287 cisnormativity, 134
self-selection, 249 client perceptions o f therapist com petence
Biddell, M. P , 2 6 8 1 in m ulticultural study, 1 1 1 -1 2
biological outcom es o f R/S-accommodated as predictor o f treatm ent outcome, 114
therapy, 224 research lim itations, 1 1 5 -1 6
Black clients, 95 therapist self-evaluations and, 119
Blumer, M . L., 145t client preferences, 157-81
Body M ind Spirit W ell-Being Inventory, 215 clinical examples, 163
Borenstein, M ., 73 and coping styles, 81
 345 IN D EX

defined, 157 consciousness raising, 298t, 299, 302

diversity considerations with, 179 C onstantine, M . G ., 95
effectiveness o f adaptations to, 335 consultant role o f therapist, 3 0 2 -3
landm ark studies, 1 6 5 -6 7 contem plation stage, 297, 299, 300, 301
m easuring, 1 5 9 -6 2 content adaptations, 88, 101
meta-analyses, 1 6 8 -7 5 context, adaptations to, 88, 101
m oderators and covariates, 1 7 5 -7 8 Cooper-N orcross Inventory o f Preferences
research lim itations, 178 (C -N IP ), 160, 161, 163
therapeutic practices, 180-81 coping skills, 58
training implications with, 180 coping strategies, 58
C lient Resistance Code (C R C ), 192 -9 3 coping style, 5 6 -8 1
C lient Task Specific Change causality, evidence for, 78
Scale-Revised, 272 clinical example, 62
clinical practice defined, 58
attachm ent style adaptations in, 47 diversity considerations with, 79
client preference adaptations in, 180-81 landm ark studies, 6 5 -6 6
coping style-related adaptations in, 81 m easuring, 5 9 -6 2
culture-related adaptations in, 118 meta-analyses, current, 6 8 -7 8
gender identity-related adaptations meta-analyses, previous, 67
in, 151 research lim itations, 79
reactance level-related adaptations in, 208 therapeutic practices, 81
religion/spirituality-related adaptations training implications with, 80
in, 253 Cornelius-W hite, J. H., 145t
sexual orientation-related adaptations correlational research designs, 332
in, 2 8 7 -8 9 cosm os spirituality, 2 1 3 -1 4
stage o f change-related adaptations in, 318 Costa, P T., 56
Task Force recom m endations for, 331 counterconditioning, 298t, 299
Cluster C personality disorders, 76 covariates, o f client preferences, 1 7 5 -7 8
C-NIP. See Cooper-N orcross Inventory o f Cram er’s V correlation coefficient, 28
Preferences C R C (Client Resistance C ode), 192 -9 3
cognitive-behavioral therapy (C B T ), 2 5 -2 6 Crisp, C., 268t
Christian-accom m odated, 220, 249 Cross-C ultural Counseling Inventory-
directiveness in, 1 8 9 -9 0 , 195, 200 Revised (C C C I-R ), 89
R/S-adapted, 2 1 8 -2 0 , 224 CRPs (cyclical relational patterns), 25
for traum a, 94 C T (cognitive therapy), 196
cognitive flexibility, 288 cultural adaptations, 86, 103/, 106t, 337
cognitive therapy (C T ), 196 in clinical practice, 1 1 8 -1 9
C ohens d. See d values defined, 8 7 -8 8
Cokley, K., 87 landm ark studies of, 93
collaboration, with client, 181, 208 with L G BQ + clients, 289
com petency-based training, 332 m easuring, 89
com plim entary style o f interaction, 6 m eta-analysis, 96, 100t
confounding variables, in multicultural cultural arrogance, 3 3 8 -3 9
com petency study, 1 1 5 -1 6 cultural com petence. See m ulticultural
confrontation com petence, therapists’
and reactance level, 188, 189, 198, 199 cultural experiences, attuning to, 117
and stage o f change, 3 0 1 -2 cultural fit, 118
 346 IN D E X

cultural hum ility D iClem ente, C. C., 303

about religious/spiritual beliefs, 252, 253 Differential Therapeutics in Psychiatry
improving/demonstrating, 117, (M ichels), 4 - 5
1 1 9 ,3 3 8 -3 9 Dillon, F. R., 268t
landm ark study of, 95 directiveness, therapist, 3
cultural identity, 317, 331, 335 causality, evidence for, 205
cultural knowledge, 88 clinical examples, 195
cultural skills, 88, 117, 118 defined, 190
culture, 8 6 -1 1 9 diversity considerations, 206
causality, evidence for, 113 landm ark studies of, 1 9 6 -9 7
clinical example, 90 m easuring, 191, 194
and coping styles, 7 9 -8 0 m eta-analyses, 1 9 9 -2 0 2 , 203t
cultural adaptations to treatm ents, 1 0 1 -8 reactance, 189
defined, 87 research lim itations, 206
diversity considerations with, 116 therapeutic practices, 208
landm ark studies, 9 3 -9 4 training implications with, 207
m easuring, 89 direct measures
m eta-analyses, 9 6 -1 1 3 o f coping styles, 60, 68
reactance level, 2 0 6 -7 o f directiveness, 191, 194, 195
research lim itations, 114 o f reactance, 192
therapeutic practices, 118 o f resistance, 190-91
therapist multicultural competence, 109-13 dismissing attachm ent style
training implications with, 117 clinical examples, 22, 24
cyclical relational patterns (CRPs), 25 m easuring, 1 8 -2 0
cyclothym ic disorders, 76 therapeutic practices, 47
disorder(s)
Daily Spiritual Experiences Scale, 2 1 4 -1 5 tailoring psychotherapy to
deactivating em otional strategies, 19 specific, 1 -2
decision aids, 160, 162 treatm ent m ethods based on, 9
de facto psychotherapy integration, 320 distress, 214, 276
Defensiveness subscale (P E X ), 160 “dodo bird verdict,” 66
delay-discounting m ethod, 1 5 9 -6 0 Dozier, M ., 24
denial, 297 dramatic relief, 2 9 8 1, 299
depression, clients with dropout, treatm ent
client preferences study of, 165 client preferences and, 3 3 5 -3 6
coping styles studies of, 6 5 -6 6 preference accom m odation and, 168,
cultural adaptations study of, 9 3 -9 4 173/, 173, 175
directiveness/reactance level study of, 196 pretreatm ent attachm ent as predictor
R/S-adapted therapy studies of, 216, of, 34, 37, 4 0 t, 42
2 1 8 -2 0 , 224, 248 stage o f change and, 303
Devine, D. A., 165 Drucker, Peter, 3
Diagnostic and Statistical Manual of Mental drug treatm ent, 217, 303
Disorders, 144 Duke Religion Index, 248
diagnostic measures d values, 7, 8t, 172, 202, 226, 310, 334
o f coping styles, 61 dynamic sizing, 8 8 -8 9 , 117
o f reactance, 191, 194 dysthymia, 219
 347 IN D EX

Eastern cultures, 7 9 -8 0 evidence-based responsiveness, 12

eating disorders, 3 1 0 -1 5 expectations, client preferences vs., 158
Eating Disorders Inventory, 3 1 0 -1 5 experienced coach role o f therapist, 3 0 2 -3
Ebrahim i, A., 219 Experiences in Close Relationships (ECR)
Ecological Validity Model, 8 7 -8 8 , 109, 114-15 scale, 20, 22, 47
E C R scale. See Experiences in Close experiential/existential therapy (FE P ), 196
Relationships scale Experiential/Insight-Oriented Activities
educational burdening, 140 (P C C I), 161
educational status, o f client, 4 3 -4 4 , 178 experiential learning, 117
Effective Skills to Em power Effective M en externalizing coping styles
(E ST E E M ), 274, 276, 289 defined, 5 6 -5 7
effect sizes, 7 landm ark studies on, 66
in attachm ent study, 28, 3 3 -4 2 measures of, 61
in client preferences study, 172 in m eta-analysis, 7 4 -7 5
in coping styles study, 73, 77 therapeutic practices for clients
in directiveness and reactance, 2 0 1 -2 with, 81
in R/S-accom modated therapy, 226 Eysenck, H. J., 56
in stage o f change, 310, 3 1 1 1
efficacy, 318 fail-safe N, 3 2 -3 3 , 241, 242Í, 3 0 7 -8
Egger’s regression test, 1 0 3 -4 false neutrality, 287
Elder, A. B., 145t Fals-Stewart, W., 280t
Eliot, T. S., 329, 339 familismo, 88
Elkin, I., 165 Fam ily Attachm ent Interview, 20
em otional arousal, 2 9 8 1, 299 Fassinger, R. E., 2 7 4 -7 5
em otional well-being, 304 fearful attachm ent style, 1 9 -2 0 , 22, 48
em pathic therapists, 3 feedback, on m ulticultural competency, 117
Engle, D., 65, 205 Fem inist Couple Therapy Scale, 135
enmeshed/preoccupied attachm ent Fem inist Family Therapist Behavior
pattern, 1 8 -1 9 Checklist, 1 3 5 -3 6
EST EE M . See Effective Skills to Em power Fem inist Psychology Institute, 150-51
Effective M en fem inist psychotherapy, 1 3 5 -3 6 , 138, 150
ethical standards, cultural com petence in, 89 Fem inist Therapy Behaviors-Revised
ethnicity, 86, 87 (F T B -R ), 1 3 5 -3 6
assessing client, 118 Fenway Institute, 151
client preferences for m atching, 1 5 9 -6 0 FEP (experiential/existential therapy), 196
in preference accom m odation studies, 179 Fernald, P S., 165
in R/S-accom modated therapy study, fidelity, therapist, 339
2 4 3 ,2 5 1 fit
See also culture cultural, 118
European C O S T P roject, 3 1 0 -1 5 reactance level, 191
Evans, E., 303 flexibility
evidence-based practice, 330 cognitive, 288
client preferences in, 157, 180 therapist, 339
in clinical practice, 331 Fonagy, P., 2 4 -2 5
psychotherapy m atching in, 5 -6 4 0 -4 0 -2 0 rule, 317
with transgender individuals, 141 Frank, Jerom e, 3
 348 IN D EX

freedom , reactance and, 1 8 8 -8 9 , 190 group psychodynam ic interpersonal therapy

Freud, Sigm und, 1, 56, 1 3 3 -3 4 (G P IP ), 2 5 -2 6
F T B -R (Fem inist Therapy group-specific knowledge, LG BQ +, 286, 288
Behaviors-Revised), 1 3 5 -3 6 group therapy, 119
Functional Assessm ent o f C hronic Illness group-wise indicators o f reactance, 192, 194
Therapy— Spiritual W ell-Being Guiando A Niños Activos, 94
Scale, 215 g values, 7, 8t, 334

GAF (Global Assessm ent of halo effect, 114, 1 1 5 -1 6 , 118

Functioning), 24 Ham ilton A nxiety Scale, 214
G A T (gender aware therapy), 136 Hawkey, C. G., 273
gatekeeping, by therapists, 141 Hazan, C., 1 9 -2 0 , 21
Gay Affirmative Practice Scale, 268t health, religious/spiritual belief and, 212
gender heteronorm ative assumptions, 1 3 7 -3 8
conceptualization of, 133 Hill, Clara E., 334
in cultural adaptation, 108 Hinton, D. E., 94
in preference accom m odation, 178, 179 H ippocratic Oath, 339
sex vs., 134, 2 6 4 -6 5 Hispanic Am ericans, 115
gender avoidance, 141 Hook, J. N., 95
gender aware therapy (G A T), 136 Hooker, Evelyn, 275
gendered language, 152 Hubers, M . J. H., 76
gender expression, 134, 135 hum anistic spirituality, 2 1 3 -1 4
gender-focused therapy, 144, 149 humility, cultural. See cultural hum ility
gender identity, 1 3 3 -5 2 Hunt, J., 145 1
assessing, 267 hyperactivating em otional strategies, 19
clinical examples, 136 hypochondriasis, 76
defined, 134
diversity considerations with, 150 identity(-ies)
landm ark studies, 139 cultural, 317, 331, 335
m easuring, 135 intersectional, 116
m eta-analyses, 1 4 2 -4 4 religious/spiritual beliefs as part of,
research lim itations, 149 212, 253
and therapeutic practices, 151 sexual/sexual orientation, 2 6 4 -6 5
training implications with, 150 See also gender identity
gender narrowing, 140 idiographic traditions, 2 -3
gender pathologizing, 141 inclusive language, 2 8 7 -8 8
gender repairing, 141 indirect measures
gender sim ilarities hypothesis, 141 o f coping styles, 60, 6 1 1, 68
G enlU SS Group, 135 o f directiveness, 191, 195, 206
Glass, C. R., 168 o f reactance, 194, 206
G lobal Assessm ent o f Functioning o f resistance, 190-91
(G A F), 24 individualizing, 2, 289
goals, treatm ent. See treatm ent goals “inform ed not-know ing” stance, 289
Gonçalves, J. P B., 2 2 1 1, 224 inhibition, reciprocal, 2 9 9 -3 0 0
G PIP (group psychodynam ic interpersonal insight-oriented psychotherapy
therapy), 2 5 -2 6 in coping style meta-analysis, 74
Greenfield, S. F., 1 3 9 -4 0 defined, 5 8 -5 9
 349 IN D EX

in landm ark coping style, 66, 67 Lambda Legal, 2 7 3 -7 4

therapeutic practices using, 81 Lam bert, M ichael J., 334
integrative supervision, 207, 318 language adaptations, 88, 101, 107,
integrative training, 3 1 7 -1 8 1 0 9 ,1 1 9
Interdivisional APA Task Force on Latinx A m ericans, 115
Evidence-Based Relationships and Lazarus, Arnold, 6
Responsiveness (third), 2 learning, experiential, 117
conclusions of, 329 Lease, Suzanne H., 334
frequently asked questions, 9 Lesbian, Gay, and Bisexual Affirmative
general recom m endations from , 331 Counseling Self-Efficacy Inventory,
lim itations o f work by, 7 2 6 8 t, 272
policy recom m endations from , 333 Lesbian, Gay, and Bisexual W orking Alliance
practice recom m endations from , 331 Self-Efficacy Scales, 268t
purpose and processes of, 3 Levesque, D., 304
research recom m endations from , 332 Levy, K. N., 25
training recom m endations from , 332 L G BQ + affirmative attitudes, 287
internalized transphobia, 1 3 4 -3 5 L G BQ + affirmative psychotherapies, 264
internalizing coping styles clinical examples of, 272
defined, 5 6 -5 7 defining, 265
landm ark studies on, 66 landm ark studies, 275
measures of, 61 measures o f therapist com petency in,
m eta-analysis, 7 4 -7 5 , 78 267, 268t
therapeutic practices for clients research lim itations, 279
with, 81 sociopolitical analysis in, 289
internal working models (IW M s), 17 training in, 286
International Classification o f L G BQ + populations
Diseases, 144 defining, 2 6 5 -6 6
interpersonal processes, in R/S-adapted in gender identity study, 142, 143
psychotherapy, 214 L G B T Assessm ent Scale, 268t
Interpersonal Process Recall Lichtenberg, James W., 334
interviews, 167 Lindheim , O., 169
interpersonal therapies, 48, 9 3 -9 4 Lipsey, M. W., 73
intersectional approach, 152, 285 local adaptations, 89
intersectional identities, 116 Logie, C., 2 6 8 t
intersex individuals, 135, 267 Longabaugh, R., 197, 205
intim ate partner violence, 304 Lundquist, C., 140, 145 t
Intrinsic Religious M otivation Scale, 248
Inward O rientation subscale (P E X ), 160 M achado, P. P., 205
IW M s (internal working m odels), 17 M ain, Mary, 17, 1 8 -2 0
m aintenance stage, 297, 2 9 9 -3 0 0
Jean B aker M iller Training Institute, 150-51 m ajor depression, 65
managed care, 1 1 -1 2
Kagan, Jerom e, 57 m anualized treatm ents, 246, 336
Kanam ori, Y., 145t marginalized clients, 332
Karno, M. P., 197, 205 M arkin, Rayna D., 334
Kim , Y. H., 2 2 1 t Mathy, R. M ., 145 t
Koenig, H. G., 2 1 9 -2 0 M atross, R. P , 189
 350 IN D EX

M CAS (M ulticultural Counseling Awareness m ood transcendence, 10

Scale), 8 9 -9 0 M oradi, B., 286
M cCabe, K., 94 M orgenstern, J., 2801
M cCullough, M. E., 2 2 1 1 m otivational interviewing (M I)
M C I (M ulticultural Counseling as directiveness m easure, 195, 206
Inventory), 8 9 -9 0 with precontem plators, 319
M cRae, R. R., 56 in reactance level m eta-analysis, 2 0 0 -1
m edical m odel, 315 M otivational Interviewing Treatment
m ental health policy, 333 Integrity Scale, 2 00-1
M eta-Analysis Reporting Standards, 4 m otive-oriented therapeutic
m eta-com petence, in preference relationship, 337
accom m odation, 180 Muenz, L. R., 66, 70
m etaphor adaptations, 88, 102, 107 M ulticultural Awareness Knowledge Skills
M exican A m erican families, 94 Scale, 89
Meyer, I. H., 275 m ulticultural com petence, therapists’,
M I. See m otivational interviewing 86, 111/
M ichels, Robert, 4 - 5 defined, 8 8 -8 9
microaggressions, 95 landm ark studies of, 94
M innesota M ultiphasic Personality m easuring, 89
Inventory (M M PI-1 and M M P I-2), m eta-analysis, 97, 109 -1 3
60, 193 training to improve, 117
m inority clients See also client perceptions o f therapist
preference accom m odation, 179 com petence
in R/S-accommodated therapy, 243, 251 M ulticultural Counseling Awareness Scale
m inority stressors, 149, 275, 285, 288, 289 (M C A S), 8 9 -9 0
M izock, L., 140, 145t M ulticultural Counseling Inventory
M M PI-1 and M M P 1-2 (M innesota (M C I), 8 9 -9 0
M ultiphasic Personality Inventory), M ultidim ensional M easure o f Religiousness/
60, 193 Spirituality, 2 1 4 -1 5
m odel o f others, attachm ent style multilingual therapists, 118
and, 1 9 -2 0 m ultivariate m eta-analysis, 32
m odel o f self, attachm ent style and, 1 9 -2 0 M uslim -accom m odated therapy, 249
m oderator(s) Muslim clients, 217, 219, 220, 2 4 5 -4 6
o f attachm ent style, 43
o f client preferences, 1 6 8 -6 9 , 172, 1 7 5 -7 8 National Institute o f M ental Health, 165 -6 6
in cultural adaptations, 99, 1 0 4 -8 , National Institute on A lcohol Abuse and
104t, 337 A lcoholism , 1 9 7 -9 8
gender as, 136 National L G BTQ Task Force, 2 7 3 -7 4
in m ulticultural competency, 97, 99, nature spirituality, 2 1 3 -1 4
111, 112t N EO inventories, 60, 61
o f religious/spiritual identity, 2 4 3 -4 6 , neuroticism , 56, 57
2441, 2471 Nohr, R. W , 248
o f stage o f change, 310 nom othetic traditions, 2 -3
Mohr, D. C., 205 nonbinary gender identities, 134
Moleiro, C., 66 Norcross, John, 198, 334
M ondragon, S. A., 2 7 5 -7 6 , 2801 nurturing parent role o f therapist, 3 0 2 -3
m ood disorders, 3 1 0 -1 5 Nuttall, J., 145 1
 351 IN D EX

observational perspective, 333 power systems in psychotherapy, 15 0 -1 5 2

obsessions, client with, 76 practice-based evidence, 285
odds ratio, 172 precontem plation stage, 297
O h, P. J., 2 2 1 1 indicators of, 300, 301
one size fits all therapy, 4 - 5 , 338 processes o f change for clients in, 299
oppositional behavior, 1 8 8 -9 0 , 193 therapeutic practices with clients in, 319
oppression, 1 3 5 -3 6 , 152, 284 predilections, client preferences and, 1 6 6 -6 7
Osler, W illiam , 2 Preference for College Counseling Inventory
outcom e measures (P C C I), 160
in cultural adaptations, 106, 1 0 8 -9 preferred term s, sexual-orientation
in preference accom m odation, 176, 178 related, 2 8 7 -8 8
in R/S-adapted therapy, 214 prem ature term ination. See dropout,
in stage o f change, 3 1 0 -1 5 treatm ent
O utcom e Q uestionnaire 45, 3 1 0 -1 5 preoccupied attachm ent style, 22, 23, 47.
Outward O rientation subscale (P E X ), 160 See also anxious-resistant (ambivalent)
overestimated effect sizes, 2 4 1 -4 2 attachm ent style
Owen, Jesse, 334 preparation stage, 297, 301
PRI (Patient Resistance Inventory), 193
PAAQ (Perceptions o f Adult Attachm ent proactive outreach, 316, 320
Q uestionnaire), 21 processes o f change
Pachankis, J. E., 274, 280t defined, 298, 298t
panic disorder, 44 in m eta-analysis, 304
Parent-Child Interaction Therapy (P C IT ), 94 stages o f change and, 299
Pargament, K., 212 therapeutic practice recom m endations
Patient Attachm ent Coding System, 20 on, 319
Patient Resistance Inventory (P R I), 193 Prochaska, J. O., 303, 304
Paul, G ordon, 1 Procrustean bed, 338
PC C I (Preference for College Counseling prognostic beliefs, 159
Inventory), 160 P roject M A TCH , 197
P C IT (Parent-C hild Interaction Therapy), 94 Propst, L. R., 2 1 8 -1 9
Pearce, M. J., 219 psychoanalytic psychotherapy, 1, 1 8 9 -9 0
Perceptions o f Adult A ttachment psychological outcomes
Q uestionnaire (PAAQ), 21 measures of, 214
personalismo, 88 in m eta-analysis o f R/S-accommodated
personality disorders, 48, 76 therapy, 227, 239t, 240t, 2 4 5 -4 6 , 249
personality measures, 6 0 -6 1 , 193 o f R/S-accom modated therapy, 2 2 0 -2 4
personalized advantage index, 337 psychological research, 332
personalizing, 2 Psychology o f Religion and Spirituality
P EX (Psychotherapy Preferences and (APA Division 36), 2 5 2 -5 3
Experiences Q uestionnaire), 160 psychospiritual support group, 220
Plante, T. G ., 2 5 1 -5 2 psychotherapy
pluralistic therapy, 167 Bowlby’s key tasks for, 16
Pope-Davis, D. B., 9 4 -9 5 client preferences in (See client
post-traum atic stress disorder (P T SD ), 94 preferences)
poverty, clients living in, 116 gender as topic in, 151
power, gender and, 133 role o f gender in, 1 3 3 -3 4 , 141
power inequalities, challenging, 286, 289 psychotherapy m atching, 4, 336
 352 IN D E X

Psychotherapy Preferences and Experiences resistance vs., 1 8 8 -8 9

Q uestionnaire (P E X ), 160 therapeutic practices, 208
psychotropic m edications, 246 o f therapist, 208
PTSD (post-traum atic stress disorder), 94 training implications with, 207
publication bias readiness to change, 320
and cultural adaptation, 103, 113-14 Reback, C., 280t
and stages o f change, 3 0 7 -8 R eciprocal Attachm ent Questionnaire
and studies o f R/S-adapted (RA Q ), 2 1 -2 2
therapy, 241 reciprocal inhibition, 2 9 9 -3 0 0
publication status, as potential moderator, recycling, 320
245, 246 refugees, 94
publication year, 43, 1 0 6 -7 , 176 reinforcem ent, 2 9 8 t, 299
Puerto R ican clients, 9 3 -9 4 relapse, 300
relational behaviors, 3 1 9 -2 0 , 331
Q -statistic, 172 relational style, flexibility of, 10
Q test, 308 Relational W ell-Being scales, 215
Relationship Q uestionnaire (R Q ), 20, 21
race, 86, 87 Relationship Style Questionnaire, 21
assessing client, 118 reliability coefficients, 112
client preferences on m atching, 1 5 9 -6 0 relief, dramatic, 298t, 299
in preference accom m odation, 178, 179 religion and spirituality, 2 1 2 -5 3
in R/S-accommodated therapy, 243, 251 causality, evidence for, 249
See also culture clinical examples, 2 1 6 -1 7
Rachlin, K., 140, 145t and culture, 116
Racial M icroaggression in Counseling defined, 213
Scale, 95 diversity considerations with, 251
random effects m odel, 227, 307 effectiveness o f adaptations to, 335
randomized clinical trials (RCTs), 1 ,3 3 6 landm ark studies, 218
causality evidence from , 79, 205 m easuring, 214
indirect measures o f reactance level m eta-analyses, current, 2 2 5 -4 1 , 228t
in, 194 m eta-analyses, previous, 220, 22 1 1
therapy focus in, 62 m oderators, 2 4 3 -4 6
RAQ (Reciprocal Attachm ent research lim itations, 249
Q uestionnaire), 2 1 -2 2 therapeutic practices, 253
Razali, S. M ., 248 training implications with, 251
R B C T (religiously integrated cognitive religiosity, 248
behavior therapy), 216 Religious Coping Scale, 215
RCTs. See randomized clinical trials religious cultural psychotherapy, 217
reactance level, 1 8 8 -2 0 9 religiously integrated cognitive behavior
causality, evidence for, 205 therapy (R B C T ), 216
clinical examples, 195 religious or spiritual beliefs
defined, 1 8 8 -9 0 cultural hum ility about, 252, 253
diversity considerations with, 206 as part o f identity, 212, 253
landm ark studies, 1 9 6 -9 7 o f therapists, 2 1 2 -1 3
m easuring, 1 9 0 -9 4 religious or spiritual com m itm ent, strength
m eta-analyses, 1 9 8 -2 0 2 , 203t of, 248
research lim itations, 206 religious spirituality, 213
 353 IN D EX

research design self-liberation, 2 9 8 t, 299

in cultural adaptations study, 1 0 5 -6 self-reevaluation, 298t, 299
in preference accom m odation studies, self-report measures
175, 178 o f cultural competency, 1 1 1 -1 2 , 114,
research-supported content, in m ulticultural 115, 117
training, 1 1 7 -1 8 o f resistance, 193
residential treatm ent programs, 44 self-selection bias, 249
resilience, 288 setting, treatm ent, 4 3 -4 4
resistance Severity o f D ependence Scale, 3 1 0 -1 5
causality, evidence for, 205 sex
defined, 188, 189 assessing, 135, 267
landm ark studies, 1 9 6 -9 7 gender vs., 134, 2 6 4 -6 5
m easuring, 1 9 0 -9 2 sexism, 1 3 4 -3 5
meta-analyses, 1 9 9 -2 0 2 , 203t sexual identity (sexual orientation
reactance vs., 1 8 8 -8 9 identity), 2 6 4 -6 5
resistant clients, 6 sexual identity distress, 276
therapeutic practices with, 208 sexual m inority, 2 6 4 -6 5 , 2 7 5 -7 6
training therapists to work with, 2 0 7 -8 sexual orientation, 2 6 4 -9 0
responsiveness, 2, 3, 338 clinical examples, 272
Richtberg, S., 76 defined, 2 6 4 -6 5
Ripley, J. S., 248 diversity considerations with, 285
Rodriguez, M elanie M . D om enech, 334 landm ark studies, 275
rom antic relationships, 19 m easuring, 2 6 6 -6 7
Rosselló, J., 9 3 -9 4 meta-analyses, 2 7 6 -2 7 7
RQ (Relationship Q uestionnaire), 20 research lim itations, 279
R/S-adapted CBT, 2 1 8 -2 0 , 224 therapeutic practices, 2 8 7 -8 9
Russell, G. M ., 273 training implications with, 286
Sexual O rientation Counselor Com petency
safe haven, 15 Scale, 2 6 8 t
Salierno, E. F., 1 3 6 -3 7 Shapiro, David, 56
satisfaction, client, 95 shared decision-m aking, 167
scientific m indedness, 8 8 -8 9 , 117 Shaver, P. R., 1 9 -2 0 , 21
secure attachm ent style, 16 Shoptaw, S., 280t
clinical examples, 22, 23 shyness, 75t
defined, 1 7 -1 8 Sm ith, T. B., 97, 221t
m easuring, 19 sm oking cessation, 303, 305
therapeutic practices for clients with, 47 snake phobia, 165
secure/autonomous attachm ent social context, for gender, 133
pattern, 1 8 -1 9 socialization, 1 3 5 -3 6
secure bond, 17 social justice, 152
security versus anxiety dim ension (Adult societal impact, 318
Attachm ent Q -set), 19 Society for the Psychology of
self-control, 209 W om en, 150-51
self-directed treatm ent (S/SD), 196 sociopolitical analysis, 289
self-identification SO C RA TES (Stages o f Change and
o f sex and gender, 267 Treatm ent Eagerness Scale), 300, 303
o f sexual orientation, 2 6 6 -6 7 Socratic teacher role, 3 0 2 -3
 354 IN D EX

Spengler, E. S., 2 7 2 -7 3 substance abuse, 66

Spengler, P. M ., 151 substance use, 279, 3 1 0 -1 5
spirituality. See Religion and spirituality substance use disorders, 1 3 9 -4 0
spiritually integrated psychotherapy, 219 supervision, 9 - 1 0 , 332
spiritual outcomes integrative, 207, 318
m easures of, 214 preference accom m odation in, 180
in meta-analysis of R/S-accommodated See also training
therapy, 227, 239t, 240t, 24 5 -4 6 , 249, 250 Support subscale (P E X ), 160
o f R/S-accommodated therapy, 2 2 0 -2 4 Swift, J. K., 1 5 9 -6 0 , 1 6 8 -6 9
spiritual self-schem a therapy, 217 Switzerland, 80
Spiritual W ell-Being Q uestionnaire, 215 Symptom C hecklist-90-R , 214
Spiritual W ell-Being Scale, 215, 227 sym ptom -oriented psychotherapy
S/SD (self-directed treatm ent), 196 in coping style meta-analysis, 74, 78
Stabb, S., 145 1 defined, 5 8 -5 9
stage o f change, 3, 2 9 6 -3 2 0 in landm ark coping style studies, 66, 67
causality, evidence for, 315 therapeutic practices using, 81
clinical examples, 301 Systematic Treatm ent Selection m odel, 207
defined, 2 9 6 -9 9
diversity considerations with, 316 Tao, K. W., 97
landm ark studies, 303 Tasca, G. A., 25
m easuring, 300 Task-Oriented Activities, 161
m eta-analyses, 3 0 4 -1 5 tem peram ent, 18
psychosocial treatment adaptations for, 336 term ination o f therapy, 300
research lim itations, 316 T FP (transference-focused psychotherapy), 25
therapeutic practices, 318 TG N C (transgender and gender-
training implications with, 317 nonconform ing) individuals, 140-41
Stages o f Change and Treatm ent Eagerness therapeutic approach
Scale (SO C R A T ES), 300, 303 as measure o f directiveness, 191, 194, 195
Stages o f Change scale, 300 reactance level and, 208
state-like reactance, 190, 208 in R/S-accommodated therapy, 24 5 -4 6 , 249
state measures o f reactance, 192 Therapeutic Reactance Scale (T R S), 193
State-Trait A nxiety Scale, 3 1 0 -1 5 therapeutic relationship
Steering C om m ittee, third Task Force, attachm ent in, 16
329, 3 3 3 -3 4 in clinical practice, 331
Steinem , G., 287 conclusions about, 329, 330
stigm a-inform ed approach to patient characteristics and, 318
psychotherapy, 2 7 3 -7 4 in psychotherapy focused on gender
stigmatization, for L G BQ + populations, identity, 150
288, 289 reactance/resistance, 189
stimulus control, 298t, 2 9 9 -3 0 0 research on, 332
Strange Situation experim ent, 1 7 -1 9 therapist(s)
stressors, minority, 149, 275, 285, 288, 289 as attachm ent figure, 16
Strong, S. R., 189 attachm ent style of, 46
Strupp, Hans, 6 client preferences about, 158, 180
ST S Clinician Rating Form , 61 ineffective actions by, 338
STS/Innerlife, 6 1 ,1 9 3 - 9 4 m ulticultural com petence of, 109 -1 3
ST S training, 80 religious/spiritual beliefs of, 2 1 2 -1 3 , 250
 355 IN D EX

Therapist Com petency Adherence Scale, 215 training on psychotherapy with, 150, 151
therapist preferences, 158 transgression-affirm ative nested-N IC E
Therapist Process Rating Scale, 62 therapy, 2 7 4 -7 5 , 289
therapy environm ent, heteronorm ative bias translational research, 333
in, 288 transphobia, 1 3 4 -3 5
Therapy Personalisation Form , 161 transreligious integrative treatm ent, 219
Therapy Process Rating Scale, 195 transtheoretical m odel (T T M ), 317
therapy process, in multicultural processes o f change in, 299
com petence, 97 stages o f change in, 296, 299, 304
Therapy with W om en Scale, 135 traum a, C B T adapted for, 94
Thum, L. S., 44 treatm ent adaptations, 2
tim e-lim ited dynamic psychotherapy, 64 conclusions about, 330
T im eline Followback, 3 1 0 -1 5 cultural, 1 0 1 -8
training, 9 - 1 0 effectiveness of, 333
about transdiagnostic responsiveness, 339 interconnections between, 3 3 6 -3 7
in attachm ent style, 46 and purpose o f Task Force, 3
in client preferences, 180 singularity for, 338
in coping style, 80 to stage o f change, 305
in culture, 117 treatm ent duration, 176
in gender identity, 150 treatm ent focus, 62, 71 t, 73
integrative, 3 1 7 -1 8 treatm ent form at (modality)
in reactance level, 207 in attachm ent style study, 43
in religion and spirituality, 251 in client preferences study, 1 5 7 -5 8
in sexual orientation, 286 in R/S-accom modated therapy study, 246
in stage o f change, 317 in stage o f change study, 310
Task Force recom m endations, 332 treatm ent goals
See also supervision cultural adaptations for, 88, 102, 107, 109
trait-like reactance, 192, 208 in R/S-adapted psychotherapy, 214
trait-like resistance, 190, 1 9 3 -9 4 , 196 stage o f change and, 3 1 8 -1 9
transdiagnostic adaptations, 3, 5, 330, 331. treatm ent m ethod adaptations, 88, 102, 107
See also specific adaptations Treatm ent o f Depression Collaborative
transdiagnostic patient characteristics, 3 Research Program , 66, 165
conclusions on, 330 treatm ent option, in preference
m atching psychotherapy to, 2, 9 accom m odation, 1 7 6 -7 7
overlap am ong, 7 - 8 treatm ent outcomes
See also specific characteristics attachm ent style and, 2 9 t, 3 3 -3 4 , 3 5 t
transference-focused psychotherapy (TFP ), 25 m ulticultural com petency o f therapist
transgender (term ), 134 and, 97
transgender-affirm ative m ethods, 152 preference accom m odation and, 168, 173,
transgender and gender-nonconform ing 174/, 175
(TG N C ) individuals, 140-41 pretreatm ent attachm ent as predictor
transgender individuals, 133, 136 of, 33, 35, 40
case example, 1 3 8 -3 9 religious/spiritual beliefs and, 2 4 6 -4 8
gender-focused therapy with, 1 4 4 -4 9 with R/S-accom modated therapy, 252
gender identity study with, 1 4 3 -4 4 stage o f change as predictor of, 305
landm ark studies, 1 4 0 -4 1 , 145 1 See also psychological outcomes
lim itations o f research with, 149, 150 Treatm ent Preference Interview, 160, 161
 356 IN D EX

treatm ent preferences, 158 W ampold, Bruce E., 334

treatm ent readiness (T R T ) subscale, of well-being, em otional, 304
M M PI, 197 Westra, H. A., 200
treatm ent responsiveness, 2 W hite privilege, 92
trim and fill technique, 1 0 3 -4 , 2 4 1 -4 2 , 2421, W illiam s Institute, 273 -7 4
2431, 3 0 7 -8 willingness to adapt, 11
Trimble, J. E., 97 W ilson, D. B., 73
T R S (Therapeutic Reactance Scale), 193 W ilson, G. T., 73
T R T (treatm ent readiness) subscale, of Women’s Recovery Group
M M PI, 197 (W R G ), 1 3 9 -4 0
T T M . See transtheoretical m odel W orld Health Organization, 304
W orld Professional A ssociation for
underestimated effect sizes, 2 4 1 -4 2 Transgender Health, 151
University o f Rhode Island Change W orthington, E. L., Jr., 2 2 0 -2 4 , 22 1 t
Assessm ent (U R IC A ), 300, 303 W orthington, R. L., 268t
unresolved attachm ent pattern, 1 8 -1 9 W R G (W om en’s Recovery Group), 1 3 9 -4 0

Vanderbilt psychotherapy research studies, 6 year o f publication, as potential

Varma, S. L., 219 moderator, 43, 1 0 6 -7 , 176
Verbal reactance scale (T R S), 193 Yüksel, §., 145t

Walters, N. T., 151 Zalaznik, D., 44