Coolidge Axis II Inventory-Revised (CATI+):

Manual
Frederick L. Coolidge, PhD

Chapter 1

Introduction to the CATI+

Purpose
The Coolidge Axis II InventoryRevised (CATI+)
has a threefold purpose: (1) to assess clinical
syndromes on Axis I of the American Psychiatric
Associations (APA) Diagnostic and Statistical
Manual of Mental Disorders (DSM), such as
Anxiety,
Depression,
Post-Traumatic
Stress,
Schizophrenia, and other psychopathological
syndromes; (2) to evaluate neuropsychological
functioning
(including
Language
Functions,
Memory and Concentration, and Neurosomatic
Symptoms related to brain dysfunction); and (3)
to assess personality disorders according to the
strict criteria for personality disorders on Axis II
of the DSM.
The CATI+ was originally published as the
Coolidge Axis II Inventory (CATI: Coolidge, 1984;
Coolidge & Merwin, 1992). The new name is a
better refl ection of its threefold purpose. The
current CATI+ contains 225 items and is
answered on a four-point Likert scale ranging
from Strongly False to Strongly True. It takes
from 30 to 45 minutes to complete and is
designed for ages 15 and above. It comes in two
available forms: a self report and a signifi cantother report.
The CATI+ consists of 42 scales which are
organized into eight distinct sections:
DSM Axis 1 Disorders. This section includes
disorders of Anxiety, Depression, Post-Traumatic
Stress, Psychotic Thinking, Schizophrenia, Social
Phobia, and Withdrawal.

DSM-Based Personality Disorders. This section

includes 14 scales, ten of which are from Axis II
of the DSM-IV (APA, 1994): Antisocial, Avoidant,
Borderline, Dependent, Histrionic, Narcissistic,
Obsessive-Compulsive, Paranoid, Schizoid, and
Schizotypal. Two are from Appendix B of the
DSM-IV: Passive-Aggressive and Depressive. The
fi nal two, Sadistic and Self-Defeating, are from
the Appendix of the DSM-III-R.
Neuropsychological Dysfunctions. This section
is comprised of three scales measuring problems
related to brain damage: Language, Memory and
Concentration, and Neurosomatic.
Executive Functions of the Frontal Lobes . This
section includes three scales assessing a set of
behaviors involving goals. The scales are
Decision Diffi culty, Planning Problems and Task
Completion Diffi culty.
Personality
Change
Due
to
a
Medical
Condition. This section is based on a diagnostic
category in the DSM-IV. The fi ve subtypes
contained within this section are Aggression,
Apathy, Disinhibition, Lability, and Paranoid.
Hostility Scales. The
section
are
Anger,
Impulsiveness.

three scales in
Dangerousness,

this
and

Normative Scales. This section includes fi ve

additional clinical scales: Apathy, Emotional
Lability, Indecisiveness, Maladjustment, and
Introversion-Extraversion.
Validity Scales. This section includes four
scales assessing whether or not respondents are
answering the items carefully and honestly. The
scales are Answer Choice Frequency, Random
Responding, Tendency to Look Good or Bad, and
Tendency to Deny Blatant Pathology.

Professional Qualifications
A Ph.D. level degree is required to score and
interpret the CATI+. People with a Masters
degree in Clinical Psychology and with specifi c

course work in psychopathology may in some

circumstances be allowed to use the CATI+. Any
professional who uses the CATI+ should possess
the required knowledge necessary to interpret it.
The overall validity of the CATI+ depends to
some
extent
on
the
knowledge
of
its
administrators. As with any psychological test,
the CATI+s scores and interpretations are based
on psychological theory and actuarial research.
Under no circumstances should the CATI+ be
used by unqualifi ed persons, nor should the
CATI+ report be used in any clinical situation or
policy or decision-making process in the absence
of solid corroborating data such as behavioral
observations,
biographical
and
historical
information, clinical interview, current social and
personal circumstances of the client, and other
psychological test results. Use of the CATI+ by
persons without an extensive clinical background
is clearly inappropriate.

The Philosophy of Diagnosis of the CATI+

There are two currently popular methods of
measuring personality disorders: categorical vs.
dimensional. The categorical approach assumes
that personality disorders are discrete noncontinuous entities. Under this approach, a
person is seen as either having or not having a
disorder. The DSM-IV is based on a categorical
taxonomic approach. One weakness of the
categorical approach is that it often leads to low
diagnostic reliability. The DSM-IV and previous
versions address these diffi culties by specifying
a list of criteria that a person must meet in order
to receive a specifi c diagnosis. The dimensional
approach assumes that personality disorders are
at the extreme of a continuous scale of normal
personality
functioning.
The
dimensional
approach has the potential to provide precise
measurement with high diagnostic reliability
(Frances, 1982).

The CATI+ is based primarily on a dimensional

approach. Norms have been established on
purportedly normally functioning people, and
cut-off scores have been established at one and
two standard deviations. The greater a persons
score, the more likely he or she would behave in
a manner consistent with a personality disorder.
It should be noted that a persons score on a
dimension may be infl uenced by environmental
factors such as social and/or occupational
functioning. In fact, the DSM-IV indicates that a
personality disorder diagnosis requires that the
person be suffi ciently disrupted in social and/or
occupational functioning. Therefore, a diagnosis
of a personality disorder should be given only
after careful examination of other relevant
clinical data, and according to DSM-IV, after an
assessment is made of the extent of social
and/or occupational dysfunction in the persons
life.

Why all the Recent Attention to Personality

Disorders?
With the change to a multiaxial classifi cation
system in the 1980 version of the American
Psychiatric
Associations
Diagnostic
and
Statistical Manual of Mental Disorders (DSM-III),
the category of personality disorders received
renewed attention. While clinical syndromes
were placed on Axis I, the personality disorders
were placed on a separate axis, Axis II. This
change subtly forced clinicians to examine their
patients for personality disorders when they had
not necessarily done so in the past. A second
major change was that unlike the previous
edition of the diagnostic manual, DSM-II (1968),
the personality disorders now had specifi c
criteria listed for each disorder. In DSM-II,
personality disorders were simply listed with one
to three sentence descriptions. In DSM-III the
patients had to meet a specifi ed number of the
criteria to meet criterion for a disorder. No

criterion for a disorder was more important or

essential than any other criterion, thus, they
were described as polythetic criteria.
In 1987 with the publication of DSM-III-R, two
new personality disorders were added in the
appendix
under
the
heading,
Proposed
Diagnostic Categories Needing Further Study
(page 367). They were the Sadistic and SelfDefeating
personality
disorders,
and
are
included in the CATI+ scales.

Test Administration
A patient may be assessed either by completing
the CATI+ directly (Form S; self-report version),
or through the ratings of another individual who
knows the respondent well (Form R; signifi cantother version). The latter has been created to
permit an assessment of the patients behavior
by a person familiar with the patient. In the
initial use of the signifi cant other version,
(Coolidge, Bracken, Taylor, Smith, & Peters,
1985),
caretakers
of
Alzheimers
patients
reported on personality change in their patients.
The reliability and the validity of the signifi cantother version have been established in a recent
study of 52 married couples (Coolidge, Burns, &
Mooney, 1993).
There are two administration and scoring
options
for
the
CATI+,
paper-and-pencil
administration
with
mail-in
scoring
and
computerized-assisted
administration
and
scoring. For those wishing to use the mail-in
scoring service, the respondent completes the
CATI+ paper-and-pencil scannable form. The
completed scannable form is then mailed to the
address on the back and a computerized report
is generated and returned.
Alternatively, both Form S and Form R of the
CATI+ may be administered and scored by
computer using SigmaSoft for Windows Software .
This software runs on any IBM-compatible
computer running Windows 3.1 or higher. At

least 4MB of memory and 5MB of disk-space are

required. The SigmaSoft software handles all
aspects of test administration, from presenting
items and recording responses to generating
reports. The simple instructions, large fonts, and
oversize buttons make it suitable for even the
novice
computer
user
to
complete
the
assessment on the computer. Complete details
on administration and scoring are presented in
the SigmaSoft CATI+ for Windows Software
Manual.
The following general guidelines should be
followed when administering the CATI+.
1.

The respondents should

quite, private location.

be

seated

in

2.

The appropriate materials should be given

to the respondent. For example, if another
individual is evaluating a patient, Form R
(the signifi cant other version) should be
administered.

3.

There should be an examiner available to

answer questions throughout the session.

4.

It is important to establish that the

respondent understands the instructions. If
there is uncertainty about any point, the
issue should be clarifi ed before beginning
testing.

5.

If the respondent has a question about a

particular item, the examiner should be
careful to answer in such as way as not to
infl uence the respondents answer.

6.

The answer sheets should be carefully

inspected to ensure that the respondents
name and all other demographic information
is completed, and that there are no
excessive omissions in responding to items.

7.

Completed answer sheets and reports, as

well as other CATI+ materials should be
treated as highly confi dential and stored in

such a way that unauthorized persons do

not have access to them.

Reports
Two types of report for each of Form S (selfreport) and Form R (signifi cant other) are
available from our CATI+ mail-in service. These
are the Brief Report and the Narrative Report,
and each is described in detail below:
CATI+ Brief Report. This report provides raw
scores, percentiles, and T-scores for each CATI+
scale. The T-scores are also presented in a bar
graph format. The profi le also indicates whether
the respondents scores fall within the normal
limits for each scale. Also provided is a table of
raw scores, as well as the following fi ve
administrative
indices:
Answer
Choice
Frequency,
Unscorable
Responses,
Random
Responding, Tendency to Look Good or Bad, and
Tendency to Deny Blatant Pathology.
CATI+ Narrative Report. This report contains
all of the information from the Brief Report as
well as fi ve additional sources of information. It
lists any critical items and/or drug and alcohol
items that a respondent may have endorsed.
Narrative information is provided for each scale
that was outside of the normal limits. For
example, if a respondents score was elevated
on
the
Withdrawal
scale,
the
following
information may be presented:
The
responses
suggest
that
the
respondent appears emotionally cold,
embarrasses easily, is emotionally fl at,
feels aloof and distant from others, avoids
social
gatherings,
avoids
social
interactions, does not mind being alone, is
uncomfortable in social situations and is
physically unaff ectionate.
In addition, potential therapy issues and
diagnostic possibilities are listed for further
exploration.

The CATI+ for Windows software produces four

types of report, each of which is described
below.
Data
Report. This
report
contains
the
respondents raw scores on each scale in a
format designed for use by other programs. This
report
is
particularly
useful
for
research
purposes.
Brief Report. This report provides the same
information available from our Mail-in Scoring
service.
CATI+ Score Report with Signifi cant Item and
Responses. This report provides the same
information as the Brief Report, as well as fi ve
additional sources of information. It lists any
critical items and/or drug and alcohol items that
a respondent may have endorsed. It presents the
scales that are outside the normal limits and
indicates which of the critical items are
endorsed on those scales. A list of responses to
the items on the four non-normative scales is
also provided. Possible therapy issues and
diagnostic possibilities are provided for further
exploration.
CATI+ Narrative Report. This report provides
the same information available from our Mail-In
Scoring Service, which is described in detail on
the previous page.

Chapter 2

Descriptions of the CATI+ Scales

The following chart lists 42 CATI+ scales, which
are organized into 8 sections. There are also fi ve
non-normative scales which are described on
pages 20 & 21.

Axis I
Anxiety

Schizophrenia

Depression

Social Phobia

Post-Traumatic Stress

Withdrawal

Psychotic Thinking

Axis II
Antisocial

Obsessive-Compulsive

Avoidant

Paranoid

Borderline

Passive-Aggressive

Dependent

Sadistic

Depressive

Schizoid

Histrionic

Schizotypal

Narcissistic

Self-Defeating

Neuropsychological Dysfunction
Overall Neuropsychological

Memory and Concentration

Language Functions

Neurosomatic Symptoms

Executive Functions of the Frontal Lobes

Overall Executive Functions

Planning Problems

Decision Difficulty

Task Completion Difficulty

Personality Change Due to a Medical Condition

Aggression

Emotional Lability

Apathy

Paranoid

Disinhibition

Hostility
Anger

Impulsiveness

Dangerousness

Other
Apathy

Maladjustment

Emotional Lability

Introversion-Extraversion

Indecisiveness

Validity
Answer Choice Frequency

Tendency to Look Good or

Bad
Random Responding
Pathology

Tendency to Deny Blatant

General Scale Descriptions

Axis I Scales
Anxiety
This scale is composed of 28 items and was
empirically derived from the 200 items in a
study by Hosman (1989). Although the
Anxiety scale of the CATI+ was not
specifi cally created to assess a specifi c Axis
I Anxiety Disorder, many of its items are
similar to the criteria of the Generalized
Anxiety Disorder. The items on the scale are
broadly based measures of anxiety including
an inability to relax, anxiety-provoking
social situations, lack of self-assurance,
insecurity in relationships, envy, future
worries, and troubled dreams.
Depression
This scale of 24 items was empirically derived in
a study by Lucero (1989). The items are
similar to the criteria presented in DSM for
Major Depressive Episode, although not all
of the criteria are represented by the scale.
The items deal with suicidal ideation, past
suicide
attempts,
depressive
thinking,
pessimism, disappointments with people,
and
troubled
dreams.
Because
the
psychopathological concepts of anxiety and
depression
have
some
overlapping
behavioral symptomatology, there are 11

overlapping items between the Anxiety and

Depression scales of the CATI+.
Post-Traumatic Stress Disorder
This scale consists of 14 items that partially
cover the criteria for Post-Traumatic Stress
Disorder (PTSD). Therefore, a diagnosis of
PTSD should not be based solely on the
PTSD scale of the CATI+. Consult the DSM
for a complete description of the PTSD.
Psychotic Thinking
This scale has 11 items, 10 of which are also on
the CATI+ Schizophrenia scale. The items
deal specifi cally with psychotic thought
processes such as ideas of reference,
paranoia, suspiciousness, bizarre somatic
complaints, and self-mutilating behavior.
Schizophrenia
This scale is composed of 45 items that cover
many
of
the
diagnostic
criteria
for
Schizophrenia in the DSM. The items include
the
general
categories
of
paranoia,
suspiciousness, hypersensitivity, ideas of
reference, strange and unusual thinking,
multiple somatic concerns, and eccentricity.
Social Phobia
This scale has 6 items dealing with fears of
interacting, speaking, or being embarrassed
in social situations. In addition, it covers
some of the general features section of DSM
for Social Phobia.
Withdrawal
This scale has 15 items (4 from the Social
Phobia scale) broadly covering not only
social aversion but also the withholding of
emotions and feelings, lack of desire for or
lack of enjoyment of close relationships,
preference for solitary activities, seeing
oneself as a loner, feelings of aloofness,
and possessing no close friends.

Axis II Scales
The items of the CATI+ for all 14 personality
disorders were created directly from the DSM
Axis II criteria. Additional items on the
personality disorder scales were also created
from the general and from the associated
features in the DSM. Generally, these additional
items refl ect or reinforce one of the specifi c
listed criteria. Item overlap among scales was
kept to a minimum, except in a few cases where
the criteria of the DSM themselves overlapped
(e.g., Criterion 5 of the Schizoid Personality
Disorder and Criterion 8 of the Schizotypal
Personality Disorder). Thus, a majority of the
personality disorder items of the CATI+ appear
on only one personality disorder scale. In order
to control for response bias (usually nay-saying),
some of the items of the CATI+ are scored in
reverse. The 10 personality disorders from Axis
II of DSM are described below.
Antisocial
There are 45 items on this scale which measures
a pervasive pattern of disregard for the
rights of others including such behaviors as
deceitfulness, impulsivity, aggressiveness,
recklessness, irresponsibility, a lack of
remorse, and evidence of a Conduct
Disorder.
Avoidant
There are 18 items on this scale which measures
a broad pattern of behaviors including
feelings of inferiority and inadequacy, social
inhibitions, hypersensitivity to criticism,
avoidance of, restriction and inhibition in
interpersonal
relationships,
feelings
of
inferiority, and an avoidance of novel
activities.
Borderline
There are 23 items on this scale which measures
a chronic and long-standing pattern of

unstable relationships, confused self image,

and aff ective instability. Aff ective instability
includes such behaviors as intense anger,
impulsivity, and recurrent suicidal gestures.
Dependent
There are 27 items on this scale which measures
the excessive and pervasive need to be
taken care of by somebody else. It includes
behaviors such as fears of abandonment,
excessive dependence on others, indecision,
and diffi culties working independently.
Histrionic
This scale has 29 items and measures extreme
emotionality,
the
excessive
need
for
attention,
sexual
seduction,
selfdramatization, and an over-reliance upon
physical appearances.
Narcissistic
This scale consists of 26 items and measures a
general pattern of excessive grandiosity,
self-entitlement, interpersonal exploitation,
lack of empathy, and envy.
Obsessive-Compulsive
Consisting of 33 items, this scale measures a
pervasive pattern of a preoccupation with
organization, perfectionism, devotion to
work, overconscientiousness, infl exibility,
rigidity, stubbornness, and interpersonal
control.
Paranoid
This scale, consisting of 20 items, measures a
long-standing
history
of
interpersonal
hypersensitivity including suspiciousness,
doubts of loyalty, fears of exploitation and
harm, and jealousy.
Schizoid
This scale, consisting of 10 items, measures a
chronic pattern of an absence of social

relationships, solitary activities, emotional

coldness, and detachment.
Schizotypal
There are 22 items on this scale which measures
a general pattern of odd beliefs and magical
thinking, unusual perceptual experiences,
suspiciousness,
inappropriate
or
odd
behavior, and social anxiety.
Two additional scales are
derived from
Appendix B of the DSM-IV (Criteria Sets for
Further Study) and are described below:
Depressive
This scale, consisting of 7 items, measures a
pervasive
pattern
of
depressive
symptomatology related to the patients
trait characteristics as well as their state
characteristics including persistent feelings
of dejection, low self concept, brooding,
negativism, pessimism, and feelings of guilt
and remorse.
Passive-Aggressive
This scale contains 23 items and measures a
chronic pattern of negative attitudes and
resistance
to
others
demands
with
behavioral symptomatology of sullenness,
argumentativeness, envy, resentment, and
feelings
of
being
misunderstood
and
unappreciated by others.
The CATI+ has also retained two personality
disorders from Appendix A of the DSM-III-R
(Diagnostic Categories for Further Study) that
may be useful in clinical evaluations and are
described below:
Sadistic Scale
There are 17 items on this scale which measures
a pervasive pattern of cruel and aggressive
behavior and other behaviors that demean
or humiliate others including physical
cruelty,
excessive
control
of
others,

fascination with violence, and amusement in

the suff ering of others (including animals).
Self-Defeating
Containing 21 items, this scale measures the
pervasive pattern of self-defeating behavior
including the chronic choice of situations
and people that lead to disappointment and
failure, rejection of chances to seek
pleasure, and excessive self-sacrifi ce that is
unsolicited by others.

Neuropsychological Dysfunction Scales

Eighteen items were created on a theoretical
and
clinical
basis
from
the
clinical
neuropsychology literature (e.g., Binder, 1986;
Gasparrini, Satz, Heilman, & Coolidge, 1978;
Lezak, 1983, 1987; McGlynn & Schacter, 1989;
Parker, 1990). The items were chosen for their
sensitivity and/or specifi city to a variety of types
of neuropsychological dysfunction. The CATI+
contains an 18-item Overall Neuropsychological
Dysfunction scale and three subscales: Language
Functions; Memory and Concentration; and
Neurosomatic Symptoms. The items for each of
these scales are as follows:
Language Functions
47. I have trouble understanding what I read.
79. When people talk to me it sounds like
they are mumbling.
106. I slur my words or I fi nd common words
diffi cult to pronounce.
129. People do not understand what I am
trying to say.
170. I have been told that my style of speech
is strange or vague.
Memory and Concentration
14. I think my memory has gotten worse in
the past few years.
121. I have trouble trying to remember the
names of common objects.

126. I tend to forget things I am supposed to

do.
132. I fi nd it diffi cult to memorize anything.
156. I often forget what I am about to say.
163. I have trouble concentrating.
216. I forget things I have just learned.
Neurosomatic Symptoms
147. When I try to go somewhere, I get lost
easily.
169. I have noticed a change in my sense of
taste or smell.
173. I have problems with my balance.
180. I think there is something wrong with my
mind.
199. I have headaches.
218. I have dizzy spells.

Executive
Scales

Function

of

the

Frontal

Lobes

A set of 16 items was selected from the CATI+

based on their face validity, consistent with the
theoretical understanding of the executive
functions of the frontal lobes (Lezak, 1982;
Luria, 1966; Welsh, Pennington, & Grossier,
1991). The executive functions have been
described as a set of behaviors that help to
achieve a future goal and may include planning,
organizing
ones
use
of
time,
searching,
strategizing, impulse control, inhibition, and
cognitive fl exibility. Damage to the frontal lobes
may impair these functions often aff ecting an
individuals social and occupational functioning,
but without concomitant evidence of a defi cit on
traditional
neuropsychological
assessment
measures (Lezak, 1993). A factor analysis of the
16 items was conducted on the CATI+ normative
sample (see Table 3-1). This resulted in the
development of an overall executive functions
scale (all 16 items) and the development of
three subscales. They are as follows:
Decision Diffi culty

16. I get advice or reassurance from others

before I make everyday decisions.
21. I have trouble making everyday
decisions.
32. Other people make most of my important
decisions.
36. I consider myself dependent on others.
49. I seldom let others make important
decisions in my life, like where to live, or
what job to take, etc. a
74. I enjoy making my own decisions without
help from others. a
107. I have no diffi culty starting projects on
my own. a
190. I avoid or postpone making decisions.
Planning Problems
28. I am not a procrastinator, that is, I dont
put off things that need to be done. a b
41. I like to make complete plans for my
vacation or leisure time. a
53. I am very concerned about details, lists,
or schedules before I begin a task.
189. I like to be really organized and have
everything in order before I get ready to
do something.
Task Completion Diffi culty
24. My perfectionism interferes with my
completing a task on time.
28. I am not a procrastinator, that is, I dont
put off things that need to be done. a b
38. I fail to accomplish tasks even when I
have the ability.
60. Often I cannot complete a task because I
set my standards too high.
190. I avoid or postpone making decisions. b
191. I have trouble fi nishing things on time
because I spend too much time getting
organized.
a
b

Reverse scored items.

Overlapping items.

Personality Change
Condition Scales

Due

to

Medical

The DSM-IV (APA, 1994) included a new section,

310.1 Personality Change Due to a Medical
Condition, whose purpose is to note personality
disturbances which are thought to be due to a
clear medical condition (e.g., brain tumors, head
trauma, cerebrovascular disease, epilepsy, etc.).
The DSM notes that the personality disturbances
should represent a change from the persons
previous
personality
patterns.
The
CATI+
includes fi ve subscales to specifi cally address
the fi ve subtypes defi ned in the DSM. They are
as follows:
Aggression
This scale has eight items measuring physical
fi ghts, anger, temper tantrums, tendency to
counterattack, and hurting oneself on
purpose.
Apathy
There are eight items on this scale measuring
indiff erence
to
criticism,
approval,
or
rejection; approval; lack of excitement in
life; absence of angry feelings; absence of
emotion; and emotional coldness.
Disinhibition
This

scale has seven items dealing with

recklessness, assertiveness, impulsiveness,
lack of planning, and boldness.

Emotional Lability
This scale has eight items measuring unstable
emotions, unstable relationships, sudden
depressive feelings, mood changes, and
lack of emotional control.
Paranoid
This scale has 10 items measuring expectations
of exploitation or harm, hypersensitivity,
suspiciousness, mistrustfulness, reluctance

to confi de in others, and doubts

faithfulness or loyalty in relationships.

of

Hostility Scales
Propensity towards hostile behavior is another
important
consideration
when
assessing
pathological
syndromes
and
personality
disorders. The three hostility scales of the CATI+
are described below:
Anger
This scale is comprised of 15 items concerned
with the issues of anger control, temper
tantrums, verbal hostility, and physical
fi ghts.
Dangerousness
This scale is comprised of 18 items concerned
with emotional lability, anger, cruelty,
fi ghting, paranoia, impulsiveness, a history
of juvenile delinquency, lack of empathy,
stealing, and a lack of remorse. Although
anger is a component of dangerousness,
only four items are shared between the
Anger and Dangerousness scales.
Impulsiveness
This scale is comprised of 7 items concerned
with
impulsivity
and
its
facets:
recklessness,
assertiveness,
boldness,
quitting a job without thinking about the
repercussions, and goal-less travel. The
scale has no overlapping items with the
Anger scale, and two overlapping items with
the Dangerousness scale.

Other Scales
The CATI+ assesses fi ve additional clinical scales
that do not fi t in the aforementioned categories.
Each is described below:
Apathy
This scale is comprised of 22 items measuring
immunity
to
criticism,
disapproval,

admiration, attention or praise; denial of

strong emotions; denial of anger; denial of
jealousy or envy; denial of sexual feelings
for
anyone;
and
self-perceived
low
emotionality. It is important to note that this
scale is distinguished from the Apathy scale
in the Personality Due to a Medical
Condition section in two ways. First, this
scale is more comprehensive, as it contains
22 dimensions from the Personality Change
Due to a Medical Condition scale and an
additional 14 items. Second, and perhaps
more importantly, the other Apathy scale is
examined only in individuals who have
experienced a clear medical condition.
Emotional Lability
This scale is comprised of 11 items dealing with
the general instability of emotions including
diffi culty in relaxing, unstable emotions,
intense but unstable relationships, tendency
to become depressed easily, feeling too
emotional, low frustration tolerance, and
quick mood changes.
Indecisiveness
This scale is comprised of 11 items dealing with
the inability to make major or minor
decisions in daily life. Theoretically, it may
be sensitive to brain dysfunction in the
frontal lobes of the cerebral cortex. This
view is based on Lezak (1982), who has
written of the executive functions of the
frontal lobes in tasks such as decisionmaking, organizing, planning, and effi cient
strategies for obtaining goals. The items on
the
Indecisiveness
scale
include
the
necessity for advice and reassurance,
trouble
with
decision
making,
procrastination, diffi culty starting projects,
having other people making important
decisions, failure
to accomplish
tasks
despite the ability, excessive worry about

doing something one hasnt done before,

and postponing decision making.
Maladjustment
This scale contains 71 items covering a wide
variety of pathological symptomatology
from the 14 personality disorder scales and
the Anxiety and Depression scales. There is
only one item overlap (question 180) with
the
Neuropsychological
Dysfunction
Neurosomatic subscale. Thus, in patient
cases of neuropsychological disturbances
the Maladjustment scale may be used as a
nonspecifi c
measure
of
general
psychopathology independent of (at least
scale-wise)
the
neuropsychological
disturbance itself.
Introversion-Extraversion
This scale is composed of 24 items assessing a
persons
inner
world
orientation
(introversion) or outer world orientation
(extraversion).
The
items
measure
enjoyment of solitary versus social activities
and gatherings, desire to please others,
public displays of feeling, and desire for
attention and praise from others. The scale
is bipolar with high scores indicating
introversion and low scores indicating
extraversion.

Validity Scales
There are four validity scales contained in the
CATI+, Answer Choice Frequency, Random
Responding, Tendency to Look Good or Bad, and
Tendency to Deny
Blatant Pathology. These
scales are designed to ensure that the
respondent is answering the items carefully and
honestly. For sample evaluations of the CATI+
validity scales (administrative indices), the
reader is directed to Chapter 4 of this manual
(page 33). Case Study 4 is particularly helpful
for the evaluation of the validity scales.

Validity Scale 1: Answer Choice Frequency

The fi rst validity scale examines the answer
choice frequency and the number of items
unanswered. Clinicians may wish to call into
question the reports for those patients who only
answer Strongly True and Strongly False, as well
as for those patients who do not use the
extremes of the answer scale. Clinicians may
also wish to examine the CATI+ answer sheet to
note the specifi c items that have been left
unanswered and to ask the patient the nature of
his/her
diffi culties
with
those
items.
Approximately 56% of the present normative
sample (N =937) left no items unanswered, and
approximately 99% had six or fewer unanswered
items.

Validity Scale 2: Random Responding

The second scale is a three-item validity scale
designed to assess whether the person has read
and understood the items or whether the person
is responding randomly. On these items it is
highly unlikely that anyone would answer
anything but Strongly False. The items are as
follows:
39. I played quarterback for the Denver
Broncos.
84. I was a member of the French Foreign
Legion.
196. I swam the English Channel.

Validity Scale 3: Tendency to Look Good or Bad

A third validity scale of 21 items was empirically
derived from 200 items through a series of
studies (Mitton, 1989) in an attempt to
discriminate
between
patients
who
are
intentionally trying to present themselves in a
good light and deny emotional diffi culties from
those who are intentionally trying to look bad or
pathological. However, to some extent all
dissimulation scales have a common problem. It
is sometimes diffi cult to diff erentiate between
those who are trying to look bad (malinger) from

those who genuinely are experiencing severe

psychological problems, and it is equally diffi cult
to
distinguish
between
those
who
are
psychologically healthy and those who are
denying emotional diffi culties. To this end,
caution should be exercised when interpreting
these results.

Validity Scale 4: Tendency to Deny Blatant

Pathology
The fourth validity scale has become one of the
most useful of the validity scales in order to
determine the validity of a patients profi le. The
TDBP consists of 97 items, about 43% of the
total number of items on the CATI+. The items
were chosen because they all assess obvious
pathology in the same direction. For example, if
a patient answers More True Than False or
Strongly True to an item on the TDBP, then the
patient is endorsing that pathology. If the
patient answers More False Than True or
Strongly
False,
then
he/she
is denying
pathology. The 97 items cover a broad range of
psychopathology, and there is no single content
area.
If
a
patient
endorsed
every
psychopathological item on the TDBP scale with
an answer of Strongly True, they would earn a
total raw score on the TDBP scale of 388 (i.e.,
four points times 97 items equals 388). Note
that a patient receives one point on the scale for
Strongly False, two points for More False Than
True, three points for More True Than False,
and four points for Strongly True. If a patient
denied every pathological item on the scale, he
or she would have a total raw score of 97 (i.e.,
one point times 97 items equals 97).
When the CATI+ was given to the purportedly
normal sample (N = 1,223; Coolidge & Merwin,
1992), their mean total TDBP scale score was
175 with a standard deviation of 30. This
indicates that the average normal person does
not answer Strongly False to every item on the
TDBP scale, but actually averages an answer

indicating More False Than True to nearly

every item. This also means that theoretically
approximately 68% of the population values
should fall within plus or minus one standard
deviation of the mean. In other words, most
normally functioning individuals would have
scores on the TDBP scale between 145 and 205.
Also, theoretically 95% of the population who
take the CATI+ should fall between plus or minus
two standard deviations from the mean, or 95%
of the sample should fall between scores of 115
and 235.
Interpretation of Scores on the TDBP
Extremely low scores (e.g., 115 or lower) may be
a strong indication that the patient is denying
much more than the average person. It is
possible, of course, that individuals obtaining
low scores could simply be happy and healthy.
However, there is no substantial evidence from
the literature that has validated the test-taking
attitudes of happy or self-actualized people.
Furthermore, the humanist Abraham Maslow in
his clinical opinion hypothesized that highly selfactualized people were not extremely happy.
Thus, the typical clinical experience on the
CATI+ with low scorers on the TDBP scale has
been
that
these
patients
are
resisting
psychological testing and refusing to cooperate
either consciously or unconsciously.
Examples of a low and a high scorer on the TDBP
are presented below.
Low Scorer A 36 year old male, arrested for
drugs, was sentenced by the legal system to
psychotherapy. His wife accompanied him to
therapy, and she told the therapist he was
reluctant to admit he had a problem with
drugs even though he had been arrested and
despite her warning to him that she thought
he was in denial of his problem. His TDBP total
score was 125 and this placed him in the 5 t h
percentile of all TDBP scores in the normative
sample. His CATI+ personality disorder scales

were all low (T scores between 30 and 45). His

CATI+ profi le was confi rmed during therapy as
an indication of denial. The therapist found
that, although he was willing to admit he had
been arrested for drugs and admitted that his
wife felt that his drug use was wrecking their
marriage, he still insisted he did not have a
problem with drugs. A short time later, the
patient was asked to take the test again
because it indicated some denial. He was
encouraged to be more open about himself.
The second testing revealed a TDBP score of
129. It appeared that the patient was still
denying and repressing, and his profi le is
probably best interpreted as a case of both
conscious
and
unconscious
defensive
processes operating.
High Scorer A 24-year- old female presented
herself for the fi rst time to her therapist
displaying numerous scars on each of her
forearms. The patient told the therapist in the
initial session she was a cutter borderline
schizophrenic
and
that
she
needed
a
therapist. Her CATI+ profi le showed elevations
on 9 of the 14 personality disorder scales and
on most of the other clinical scales (T score
greater than 60). Her TDBP total raw score was
289. The score placed her above the 99 t h
percentile for high scores on the TDBP scale
compared to the normative sample. Her
therapist, after a number of sessions, did not
feel that she was blatantly schizophrenic nor
did he fi nd her behavior in therapy excessively
disturbed. Her CATI+ profi le came to be
interpreted as refl ective of some nonspecifi c
personality disorder and as a cry for help.
Note that research has shown that high TDBP
scores are defi nitely associated with more
disturbed psychopathological syndromes, like
schizophrenia. However, clinicians should be
aware that high scores could also be an
indication of malingering, factitious disorders,

a cry for help, or someone genuinely disturbed

and crying for help.
Empirical Research on the TDBP Scale
Personality
Disorders. The
original
CATI+
clinical validation sample consists of 75 patients
who were in therapy, deemed by their therapist
to have at least one personality disorder, and
were without any major Axis I diagnosis, like
depression,
and
without
any
psychotic
disturbance,
like
bipolar
disorder
or
schizophrenia. For these 75 patients, their mean
TDBP was 190 and their standard deviation was
34. Compared to the normative sample, this
places the personality disordered sample at
about the 69 t h percentile. Thus, most of the
personality disordered patients are slightly
elevated compared to the normative sample;
however, there is substantial overlap between
the two distributions. Theoretically, 68% of the
population of personality disordered patients
would range between 156 to 224. The minimum
score in the 75 patient sample was 116, which
indicates extreme denial and/or repression. The
maximum score was 279. It is interesting to note
there were only two TDBP scores out of the 75
above 250 (a 270 and a 279).
All of the original 75 patients had voluntarily
sought therapy and none were court-ordered.
Their therapists (all were at the Ph.D. level) had
diagnosed them with the strong possibility of
having at least one personality disorder.
However, only 56% of the sample self-reported
at least one personality disorder scale above a T
score of 60 and only 43% self-reported at least
one T score above 70 on the personality disorder
scales. Surprisingly, 44% of the patients did not
endorse items indicating a personality disorder
since all of their T scores were below 60). It
could be argued in these cases that their
behavior is a classic hallmark of personality
disorders and that the patients are ego-syntonic

for this disorder and unaware of the eff ects of

their behavior upon themselves or others.
A separate analysis of these 33 patients (44%)
revealed that their mean TDBP score was 166
(SD = 22). The mean TDBP score for the group of
patients who reported at least one personality
disorder scale T score above 70 was 216 (SD =
26) and these two means were found to be
statistically signifi cantly diff erent from each
other, (t ( 6 3 ) = 8.36, p < .001).
In summary, if a patient is suspected of having
a personality disorder and he or she is
answering honestly (with some awareness of the
eff ect of their behavior upon themselves or
others) the patients TDBP score should be
above 200. It appears that a majority of
personality disordered patients will self-report a
personality disorder on the CATI+. On the other
hand, if a patient suspected of having a
personality disorder has a TDBP score below the
normative sample mean (175), then denial
and/or repression is very likely possible. If this
occurs, one eff ective strategy has been to ask
the patient to take the test again with
instructions to be more open or more honest. A
second strategy would be to use the signifi cantother form of the CATI+. It has been designed to
be given to a spouse or someone with a close
relationship to the patient. The signifi cant-other
form
has
been
psychometrically
validated
(Coolidge, Burns, & Mooney, 1995) and this form
can be scored with the current CATI+ scoring
system just like the self-report form.
TDBP and Schizophrenia. A study was conducted
with the CATI+ (Coolidge, Segal, Pointer, Knaus,
Yamazak,
&
Silberman,
in
press)
of
30
schizophrenic patients
in
a state
mental
hospital. Their mean TDBP score was 211 with a
standard deviation of 52. Their TDBP range of
scores was 124 to 324. Only two of the
schizophrenic patients TDBP scores (a 295 and a
324) exceeded the highest TDBP scores of the 75

patients in the personality disorder alone

sample. Theoretically, 68% of the population of
schizophrenic patients TDBP scores should fall
between 160 and 262. Again, when testing
schizophrenic patients, TDBP scores below 175
may indicate denial and/or repression. TDBP
scores below 175 occurred so frequently (27% or
8 out of the 30 patients) that a low TDBP score
in schizophrenia may actually refl ect the lack of
critical self-awareness that is symptomatic in
schizophrenia.
Are CATI+ scores for schizophrenics valid? In a
majority of these 30 cases (53% or 16 cases),
the schizophrenics TDBP scores were greater
than
one
standard
deviation
above
the
normative samples mean and the 16 cases
TDBP mean was 249. Thus, it appears that CATI+
scores of schizophrenic patients will more often
be valid than not.

A Summary of TDBP Scale Interpretation

The CATI+ is based on a dimensional approach
and has been designed to assess the likelihood
of the presence of a personality disorder. Under
this approach, a high score on a personality
disorder scale is associated with a high
probability of the presence of the personality
disorder in question. A low score suggests that a
person probably does not have the personality
disorder. As indicated in the DSM-IV, it is
important to remember that a person should not
be diagnosed as having a personality disorder if
the persons condition does not interfere with
keeping a job, going to school, or having
satisfactory interpersonal relationships. It is
quite possible (although not highly probable)
that a high score on a personality disorder scale
of the CATI+ may still not meet all of the criteria
for the offi cial DSM defi nition. It is also
important to remember that high and low scores
on the CATI+ may be aff ected by response bias
and attitudes that may not be directly related to
existing
psychopathology
(e.g.,
chronically

defensive people or extremely open and selfdisclosing people). The demand characteristics
of an assessment situation may also play a
strong role in the overall CATI+ profi le. For
example, environmental cues may have a strong
impact on the evaluation of parents in child
custody disputes, individuals eligibility for
parole, and psychological fi tness for military
duties. Despite all of these nonspecifi c factors,
the empirical clinical research is relatively clear:
overall, people answer items in psychological
testing based on the items content. The demand
characteristics of the testing situation or the
social desirability of an item play less signifi cant
roles.
In summary, CATI+ research has shown that
high scores on the TDBP scale (beginning at one
standard deviation above the normative mean,
or
about
205)
indicate
that
personality
disturbances are more likely and the patient is
not engaging in denial. The mean TDBP score for
patients with a personality disorder who are not
thought to be denying is about 216. Thus, scores
in this range may indicate that the patient is
attempting to be honest and truthful, but is also
indicative that a personality disturbance is
probably present. For patients with more blatant
Axis I disorders, like schizophrenia, it appears
that they are often able to validly report their
problems on the CATI+, and high TDBP scores
may
refl ect
their
openness
to
do
so.
Schizophrenic patients, who are not in denial,
have a mean TDBP score of 249, nearly 33 points
higher (about one standard deviation higher
than the patients with personality disorders).
Low scores on the TDBP scale (less than 145)
may indicate test taking denial, repression, or
lack of critical self-awareness. Patients with
personality disorders will probably have scores
one standard deviation above the normative
mean (205) while schizophrenic patients may be
expected to have scores above 250.

Non-Normative Scales
There are currently fi ve non-normative scales,
for which the items are not summed, nor are
they meant to be. However, the items on the
scales have been grouped on the basis of their
content for the clinicians convenience. A
patients answers to the items should be
examined individually. The scales are as follows:
Depersonalization
This scale contains 6 items assessing a sense of
self, helplessness when alone, feelings of
emptiness, boredom, and a troubled mind.
Drug and Alcohol
This scale contains 2 items assessing whether
the patient states that someone he or she
knows believes the patient himself/herself
may have a drug or alcohol problem and
whether the patient states a drug or alcohol
problem has gotten him or her into trouble.
Eccentricity
This scale has 4 items measuring odd, eccentric,
in appropriate appearance or behavior,
strange or vague speech, and talking out
loud to oneself.
Frustration Tolerance
This scale contains 3 items assessing frustration
level, irritability, and denial of gratifi cation.
Sexuality
This scale consists of 8 items assessing whether
the patient enjoys behaving or dressing in a
sexy or fl amboyant manner, engages in
sexual coercion, demonstrates enjoyment of
or interest in sex, displays monogamous
behaviors, and mistrusts a sexual partner.

Critical Items
Selected items were chosen that might require a
clinicians immediate attention. Items 211 and
214 are considered to be critical items, but do

not appear on any scale or subscale. The

remaining critical items also appear on other
scales. The critical items are as follows:
14. I think my memory has gotten worse in
the past few years.
17. Someone I know thinks I have an alcohol
or drug problem.
31. I feel like people are talking directly to
me or about me on the TV or radio.
97. I have never forced anyone to have sex
with me.
131. I have made more than one suicidal
threat or gesture in my life.
177. I have gotten into trouble because of a
drinking or drug problem.
178. I hear voices or see things that are not
really there.
180. I think there is something wrong with my
mind.
188. More than once I have hurt myself badly
on purpose, like cutting my wrists or
smashing my fi sts against a wall, etc.
198. Recently I have felt like killing myself.
211. I have a problem with food.
212. I have a memory that bothers me.
214. I have many physical problems.

Item Overlap
There is little item overlap on the 14 personality
disorders of the CATI+. The few exceptions are
in cases where the specifi c criteria in the DSM
overlap. There is no overlap between the
Depressive scales from Axis I and Axis II. There
is no overlap between the 18 items of the
Overall Neuropsychological Dysfunction scale
and the 16-item Overall Executive Functions
scales. In general, item overlap exists on the
CATI+ scales only when the scales themselves
overlap theoretically in the clinical literature or
in the DSM. For example, there are a few items
which
overlap
between
the
Anxiety
and

Depression
scales
because
depression
is
considered to be an ego dystonic mood state
frequently
accompanied
by
apprehension,
irritability, restlessness, agitation, and anxiety.
Psychometrically, it is preferable to have no
overlapping scales in a psychological inventory.
Item overlap can produce a statistical artifact of
increased similarity and reliability, or may
produce diffi culties with discriminant validity
between two or more diagnoses. However, if two
psychological disorders are thought to overlap
theoretically or to contain
highly similar
symptoms it is unnecessary to assess them with
diff erent items. In summary, item overlap exists
on the CATI+ when overlap is thought to exist
theoretically, there is overlap in the DSM
between two scales, or when the item overlap
poses no threat to the discriminant validity of
two scales (e.g., the Anxiety scale and the
Maladjustment scale).

Chapter 3

Standardization and Psychometric

Properties of the CATI+
The Normative Sample
The current CATI+ norms and scale reliabilities
are based on a sample of purportedly normal
subjects (see Table 3-1). The subjects were
recruited through college students and consisted
of their friends, relatives and acquaintances. The
college students were asked to recruit people
whom they deemed normal or psychologically
healthy. Specifi cally they were told not to test
anyone who had previously been in jail or prison,
anyone previously committed to a mental
hospital, or anyone mentally retarded. Table 3-2
presents the means, standard deviations and
scale reliabilities (Cronbachs coeffi cient alpha)
for the normative sample.
Table 3-1 The Normative Sample
SEX

Males
Females

ETHNICITY
Caucasian
Hispanic
Black
Asian

AGE

Mean Age
Age Range

EDUCATION

High School equivalent or greater

Some college

359
578

89%
7%
2%
1%
29.20
18-92

99%
70%

Bachelor degree or greater

8%

MARITAL STATUS

Single
Married or cohabiting
Divorced, separated, or widowed

TOTAL

57%
39%
5%

N = 937

Table 3-2
Means, Standard Deviations and Scale Reliabilities for the Normative
Sample (N = 937)
Scale

Mean

Standard Deviation

Reliability

Anxiety
Depression
Post-Traumatic Stress
Psychotic Thinking
Schizophrenia
Social Phobia
Withdrawal

54.90
41.90
30.20
18.20
85.00
12.70
31.60

11.60
9.80
7.00
4.20
14.30
3.40
6.30

0.89
0.89
0.81
0.73
0.89
0.74
0.80

Antisocial
Avoidant
Borderline
Dependent
Depressive
Histrionic
Narcissistic
Obsessive-Compulsive
Paranoid
Passive-Aggressive
Sadistic
Schizoid
Schizotypal
Self-Defeating

79.50
38.00
50.40
56.80
14.00
77.10
62.90
72.30
41.90
51.90
29.00
65.40
42.50
42.60

13.80
8.50
9.90
10.20
3.80
8.60
9.20
8.20
6.90
7.50
5.60
8.20
6.90
6.40

0.86
0.80
0.80
0.87
0.75
0.76
0.74
0.68
0.79
0.78
0.69
0.73
0.73
0.66

Overall Neuropsychological
Language Functions
Memory and Concentration
Neurosomatic Symptoms

31.80
8.30
13.40
10.30

8.00
2.50
3.90
2.90

0.83
0.74
0.74
0.50

Overall Executive Function

Decision Difficulty
Planning Problems
Task Completion Difficulty

32.60
14.70
9.20
11.80

6.40
4.40
2.50
3.20

0.76
0.75
0.62
0.70

Aggression
Apathy
Disinhibition
Emotional Lability
Paranoid

15.90
17.20
14.60
18.00
19.80

4.10
3.30
3.50
3.90
4.80

0.69
0.54
0.64
0.72
0.78

Axis I

Axis IIb

Neuropsychological Dysfunction a

Executive Functions of the Frontal Lobesc

Personality Change Due to a Medical Conditionc

Hostility Scalesa

Anger
Dangerousness
Impulsiveness

30.00
30.40
15.00

6.50
6.20
3.60

0.80
0.73
0.65

Apathy
Emotional Lability
Indecisiveness
Maladjustment
Introversion-Extraversion

51.20
27.90
21.80
128.10
71.30

7.70
5.50
4.80
22.50
9.40

0.78
0.79
0.77
0.92
0.84

Tendency to Deny Blatant Pathology

Tendency to Look Good or Bad
Answer Choice Frequency
Strongly False
More False than True
More True than False
Strongly True
Unanswered

175.50
60.50

29.00
7.90

0.84
0.84

61.20
58.40
49.50
29.80
1.30

22.90
20.20
29.80
13.90
2.00

Other Scalesa

Validity Scalesa

Note. aThe values in this section are based on the normative sample. bThe values in this section were obtained from a general
population sample (N=609). This sample was obtained in the same manner as the normative sample. cThe values in this section
were obtained from a general population sample (N=683). This sample was obtained in the same manner as the normative
sample.

Reliability
Internal Consistency
The scale reliabilities (Cronbach alpha) are
presented in Table 3-2. It can be seen that most
of the scales have adequate levels of internal
consistency.

Test-Retest Reliability
In a study by Merwin & Coolidge (1987), 39
college students (mean age =21) from an
introductory psychology course were recruited
and took the CATI+ during class time. One week
later, they were asked to take the test again.
They were instructed not to try to second
guess the experimenters but to take the test
under the same instructions that they took it
initially, (i.e., try to answer honestly). The
resulting
scale
reliabilities
were
.88
on
Depression, .89 on Anxiety, and .83 on
Neuropsychological Dysfunction.
The test-retest reliabilities for the personality
disorder scales of Axis II were obtained from the
same sample previously described in the Axis I

test-retest reliabilities. The mean test-retest

reliability of the Axis II scales was .90.

Validity
Axis IAnxiety Scale
Face Validity. From pilot studies with the
CATI+, 28 items were selected which had good
face validity with the concept of anxiety.
Convergent Validity. A sample of 92 college
students( 32 males, 60 females; Mean age =
22.2)
were
administered
the
CATI+,
the
Spielberger
State-Trait
Anxiety
Scale
(Spielberger, 1983), and the MMPI Anxiety scale.
The tests were administered during class time,
and participation was voluntary, although extra
credit was given. The CATI+ Anxiety scale
correlated .63 with the Spielberger State Anxiety
scale, .86 with the Spielberger Trait Anxiety
scale, and .83 with the MMPI Anxiety scale. All
three correlations were signifi cant at p < .01.
Additional evidence for the construct validity
of the CATI+ was provided in a study in which 30
subjects were divided into two groups on the
basis of their MMPI anxiety T scores. Nine
subjects had MMPI anxiety T scores greater than
or equal to 61. A control group (N =21) was
chosen of those subjects with a MMPI anxiety T
scores less than or equal to 38. An independent
t-test between the two groups revealed that the
mean depression score sum on the CATI+ for the
anxious group (M = 72.9) was signifi cantly
greater than the non-anxious groups mean (M =
42.0; t (2 8) =6.45, p <.001).
The subjects were also divided into two groups
on the basis of the Spielberger state anxiety
scores. The state anxiety group consisted of 8
subjects with Spielberger state anxiety T scores
greater than or equal to 60. The control group (N
= 14) had T scores less than or equal to 35. An
independent t-test revealed that the stateanxious groups anxiety sum mean on the CATI+

(M = 63.6) was signifi cantly higher than the non

state-anxious groups mean (M = 41.9; t ( 20 ) 5.46,
p <.001). The subjects were also divided into
two groups on the basis of their Spielberger Trait
Anxiety T scores using the same cutoff s as in the
state anxiety study. The trait anxious groups (N
=7) CATI+ anxiety sum mean (M = 75.6) was
signifi cantly higher than the non-anxious traitanxious groups (N =18) mean ( M = 43.3; t (2 3)
= 8.25, p <.001).

Axis IDepression Scale

Face Validity. A series of pilot studies with the
CATI+ lead to the selection of 24 items which
had good face validity with the concept of
depression.
Convergent Validity. To explore the convergent
validity of the Depression scale, a sample of 97
college students (46 males, 48 females, 3 nonspecifi ed; Mean age = 20.2) were administered
the CATI+, the Beck Depression Inventory (BDI;
Beck, Ward, Mendelson, Mock, & Erbaugh, 1961)
and the MMPI Depression scale. The tests were
administered during class time, and participation
was voluntary, although extra credit was given.
The CATI+ Depression scale score correlated .68
with the MMPI Depression scale T score and
correlated .69 with the BDI scale score. These
two correlations were signifi cant at p < .01.
As part of another study, the CATI+ Depression
scores of a sample of 24 personality disordered
patients in which 75% of the sample had an Axis
I disorder involving some type of depressive
symptomatology was compared to a control
group of 24 individuals. The mean of the clinical
group (53.0) was significantly greater than the mean
of the control group (37.0; t(46) = 5.29, p < .01.

Axis II Scales
Eleven licensed clinical psychologists (4 males
and 7 females; mean length of time in private
practice = 12.8 years) from the local community

were contacted and agreed to participate in the

validation of the CATI+ with the Millon Clinical
Multiaxial Inventory II (MCMI-II, Millon, 1987).
They were asked to select patients whom they
suspected of having personality disorders. They
were required to have known the patient for at
least fi ve sessions. They were told not to select
any patients with any psychotic syndromes
including
organic
mental
disorders,
schizophrenia, and bipolar disorders. Informed
consent was obtained from the clinicians and
their
patients,
and
the
patients
were
administered the CATI+ and MCMI-II by their
clinicians. Before administering the CATI+, the
clinicians fi lled out a personality disorder
checklist which allowed them to note which of
the 13 personality disorders they thought their
patients had. In addition, they fi lled out a
separate symptom checklist which contained a
randomized list of the 117 unique criteria for the
personality disorders from Axis II of DSM-III-R
and from the Appendix.
The initial analysis contained 24 patients
whose mean age was 38.0. There were 3 males
and 21 females. Eleven were married or living
with a signifi cant other, 8 were divorced or
separated, and 5 were single. All of the patients
had at least graduated from high school and 16
had some college experience (range = high
school to masters degree). For the 24 patients,
their suspected personality diagnoses, according
to the clinicians, were as follows: 2 Borderline, 4
Dependent, 4 Histrionic, 4 Passive-Aggressive, 2
Avoidant,
1
Obsessive-Compulsive,
1
Narcissistic, and 1 patient had no primary
diagnosis but had diagnoses of ObsessiveCompulsive and Self-Defeating traits. There were
fi ve patients with a suspected diagnosis of two
personality
disorders:
Borderline-Dependent,
Borderline-Narcissistic, Avoidant-Dependent, and
two Obsessive-Compulsive Passive-Aggressive
patients. The patients primary Axis I diagnoses

were as follows: 18 depressions, 4

disorders, and 2 adjustment reactions.

anxiety

For the CATI+, a criterion of one standard

deviation above the mean of the normative
sample was chosen to mark the presence of a
personality disorder. When this standard was
applied to the clinical sample and matched to
the
diagnoses
of
the
clinicians,
a
50%
concordance rate was obtained (12 of 24
patients). For the MCMI-II, at a base rate of 75 to
mark the presence of a personality disorder, a
63% concordance rate was obtained (15 of 24
patients). However, the higher concordance rate
of the MCMI-II was apparently attained at the
cost of a higher false positive rate. According to
the clinicians, the patients averaged 1.2 primary
diagnoses per person. The MCMI-II diagnosed an
average of 3.4 primary diagnoses (greater than
or equal to a base rate of 75) per patient while
the CATI+ averaged 2.8 diagnoses (greater than
or equal to one standard deviation above the
mean) per patient. For the 15 patients who were
diagnosed the same by the clinicians and the
MCMI-II, the average primary diagnoses per
patient was 4.7 on the MCMI-II.
The scale scores on the MCMI-II were
correlated with the corresponding scale scores
on the CATI+ (see Table 3-3). The median
convergent validity correlation was .58 for the
13 scales.

Table 3-3
Correlations Between CATI+ and MCMI-II Raw
Scores
Scales
Antisocial
Avoidant
Borderline
Dependent
Histrionic
Narcissistic
Obsessive Compulsive
Paranoid

Correlations
.57
.80
.87
.43
.72
.38
.10
.58

Passive Aggressive
Sadistic
Schizoid
Schizotypal
Self-Defeating

.86
.40
.22
.65
.67

Some of the low correlations between the

CATI+ and MCMI-II are not surprising in view of
the fact that Millon (1985) has previously argued
that the DSM-III-R criteria should not be
considered sacrosanct and that a valid test
instrument may be obtained without strict
adherence to the criteria (see also Widiger,
Williams, Spitzer, & Francis, 1985).
The base rate scores on the MCMI-II were also
correlated with the raw scores on the CATI+ (see
Table 3-4). The median convergent validity
correlation was .41 for the 13 scales.
For each of the 24 patients, a Spearmans
rank-order correlation coeffi cient was obtained
between the rankings of the 13 scales on the
CATI+ with the rankings of the 13 scales on the
MCMI-II.
The
median
convergent
validity
correlation was .42. They ranged from a high of .
82 to a low of -.07 for the 24 patients.
An additional analysis was made in order to
determine
whether
the
CATI+
71
item
Adjustment Index would be able to diff erentiate
between those who have personality disorders
and those who do not. It was found that the
mean score
(M = 140.7) on the adjustment index for the 24
clinical patients was signifi cantly greater than
the mean sum (M = 119.0) for an age, gender,
and marital status matched control group, ( t (4 6)
= 3.17, p <.003). The control group was chosen
from the large sample
(N = 609) which was
used to calculate personality disorder scale
reliabilities. Theoretically, a higher sum on this
scale indicates the endorsement of broad and
various aspects of psychopathology. It is
important to note, however, that the large
sample may have contained members who did
have personality disorders, and studies suggest

that the rate may be 6 to 10% of a sample of

convenience (Merikangas & Weissman, 1986).

Table 3-4
Correlations Between CATI+ and MCMI-II Base
Rate Scores
Scales
Antisocial
Avoidant
Borderline
Dependent
Histrionic
Narcissistic
Obsessive Compulsive
Paranoid
Passive Aggressive
Sadistic
Schizoid
Schizotypal
Self-Defeating

Correlations
.53
.69
.46
.38
.73
-.05
.56
.30
.74
-.12
.26
.41
.14

The CATI+ identifi ed 8 of the 24 cliniciandiagnosed patients as having no personality

disorders (i.e., none had any scales one standard
deviation
above
the
mean).
The
MCMI-II
identifi ed
three
patients
as
having
no
personality disorders. When this same criterion
was applied to the 24 matched control subjects,
12 subjects had one or more elevated scales,
and six subjects had two or more elevated
scales. This fi nding suggest at least two
overlapping possibilities for the CATI+: (a) the
criterion of one standard deviation above the
mean is too lenient resulting in an unacceptably
high false positive rate, or (b) that the presence
of personality disorders was greater than might
be expected in the control sample and in the
large sample.

Factor Analysis
A principal-components analysis with varimax
rotation
(Merwin
&
Coolidge,
1989)
was
performed upon the Axis II scales. This analysis
was based on the normative sample ( N = 937).
Three
components
were
extracted
with
eigenvalues above 1.00 (5.75, 2.37, and 1.60 for

the three components respectively). Component

1 accounted for 44.3% of the total variance,
while components 2 and 3 accounted for 18.3%
and 12.3% respectively. Table 3-5 presents the
factor loadings for the 13 scales on the three
components.
Component 1 has its highest loadings from the
Avoidant (.89)
Obsessive-Compulsive (.75),
Dependent (.72), Self-Defeating (.62), PassiveAggressive (.59), Paranoid (.56), and Schizotypal
(.55) scales. The primary theme of Component 1
appears to be avoidant behavior, general
maladjustment, and anxious features. Piersma
(1986), in a component analysis of the Millon
Clinical Multiaxial Inventory (MCMI; Millon 1977)
with 151 psychiatric inpatients, found a fi ve
component structure with the fi rst component
accounting for 50% of the total variance and the
avoidant scale had the highest scale loading
with factor 1. Millon (1987) also reported a
general maladjustment fi rst factor in an eight
factor structure on a general psychiatric sample
on the MCMI-II, and the fi rst factor accounted for
31% of the total variance. However, only four of
Millons eight factors dealt with the 13
personality disorders. The other four appeared
to deal with his nine DSM-III-R Axis I scales.
Similar to the present results, Millon also found
that the highest loadings for the Avoidant,
Passive Aggressive, and Self-Defeating scales
were on his fi rst factor. Also, the Avoidant scale
had the single highest loading of all possible
scales on Component 1 in both studies (.89 in
the present study, and .92 in Millons study).
This suggests that Component 1 on the CATI+
and the MCMI-II has prominent avoidant features
such as an active withdrawal from personal
relationships and a conscious and anxious
estrangement.

Table 3-5
Principal Components Analysis of the Axis II
Scales
Scale
Avoidant
Obsessive Compulsive
Dependent
Self-defeating
Passive Aggressive
Antisocial
Sadistic
Schizotypal
Paranoid
Schizoid
Histrionic
Narcissistic
Borderline

Component 1
Component 3

Component 2

0.89
0.75
0.72
0.62
0.59
-0.07
0.03
0.55
0.56
0.00
-0.18
0.28
0.30

0.04
0.02
0.01
0.53
0.43
0.90
0.85
0.63
0.60
-0.03
0.09
0.38
0.50

0.01
-.16
0.53
0.19
0.45
0.20
0.13
-.06
0.15
-0.94
0.91
0.64
0.62

Note: The depressive personality scale was not examined in the present
study.

Component 2 had its highest loadings from the

Antisocial (.90) and Sadistic scales (.85) with
contributions
from
the
Schizotypal
(.63),
Paranoid
(.60),
Self-Defeating
(.53),
and
Borderline (.50) scales. Thus, it appears to have
a strong antisocial/sadistic underpinning. The
other loadings may suggest a paranoid or
suspicious theme. Piersmas study had no clear
counterpart for this component, although his
third factor, accounting for 13% of the total
variance was labeled Impulsivity/Negativism.
Millon (1987) also found that the antisocial and
sadistic scales had their highest loadings on his
second factor and accounted for 29% of the total
variance. He concluded that his second factor
represented
acting-out
and
self-indulgent
behaviors.
In the present study, Component 3 had its
highest loadings with the Histrionic (.91) and
Schizoid (-.94) scales with minor loadings from
the
Narcissistic
(.64),
Borderline
(.62),

Dependent (.53) and Passive-Aggressive (.45)

scales. Thus, Component 3 is interpreted as a
pathological
continuum
of
introversionextraversion. Piersmas (1986) study had no
counterpart to the present Component 3. The
Schizoid and Histrionic scales in his study loaded
the highest (positive and negative, respectively)
on his Component 1; there was no clear parallel
component for the present Component 3 in
Millons (1987) study. Millons highest positive
loading for the Histrionic and highest negative
loading for the Schizoid scales occurred on his
second factor.

Summary of Axis II Results

The aforementioned results provide support for
the use of the CATI+ in the assessment of
personality disorders. The stability over time
and adequate level of internal consistency of
the personality disorder scales attests to
CATI+s test reliability. The initial validity
analyses also lend support to the CATI+s
convergent validity, and these results have been
confi rmed
on
52
additional
personality
disordered patients.
The issue of personality disorder scale validity
is predicated on many assumptions. It is
assumed
that
the
personality
disorders
themselves are valid entities, that the DSM-III-R
criteria are reliable and valid measures of these
entities (Melisop, Varghese, Joshua, & Hicks,
1982; Spitzer, Forman, & Nee, 1979), and that
people can reliably and validly assess their own
behavior (Mellor, 1987).
Given the tenuousness of the relationships
among these assumptions, the need for a
broadly based diagnostic approach cannot be
overemphasized. Thus, clinicians should never
base a diagnosis on a single diagnostic test.
Other psychological tests, clinical interviews,
and other corroborative information should be
used in the assignment of any DSM-III-R

diagnosis, along with careful adherence to the

DSM-III-R criteria.

Neuropsychological Dysfunction Scales

In a discriminant validity study (Coolidge &
Merwin, 1989; Coolidge, Merwin, & Philbrick,
1989), 17 closed head injured (CHI) patients
(mean age 36 years) volunteered to take the
CATI+. They were all enrolled in a cognitive
rehabilitation program for less than one year.
There were 13 females and 4 males, and their
mean full scale IQ was 91. All but two had high
school educations or greater. None of the
patients had any prior mental hospitalizations.
Thirteen of the patients had been in car
accidents, and 11 had no known period of
unconsciousness. Sixteen of the 17 patients had
been
administered
a
complete
neuropsychological test battery, and 13 of the
16 showed signifi cant impairment on the battery.
The other three patients had scores within the
borderline range of brain impairment. The one
patient who was not given the full battery
showed evidence of brain impairment on one of
the two neuropsychological tests given. The
median interval between injury and testing with
the CATI+ was 14 months.
The control group consisted of age and gender
matched controls obtained from a sample of
convenience who had no known prior head
injuries.
The mean sum (M = 38.9) on the Overall
Neuropsychological Dysfunction scale of the
CATI+ for the CHI group was signifi cantly higher
than the mean (M = 25.5) of the control group,
(t ( 32 ) = 4.75, p <.001). Trial and error revealed
that if a mean cutoff score of 30 was used to
identify neuropsychological dysfunction in these
two groups, 15 of the 17 (88%) CHI patients
would have been correctly identifi ed while 14 of
the 17 (82%) control group subjects would have
fallen below the cutoff .

A factor analysis with a Varimax rotation

(Merwin
&
Coolidge,
1989)
of
the
Neuropsychological Dysfunction scale on the
initial sample of 609 subjects revealed a three
factor structure, and a single factor accounted
for 32% of the total variance. The second and
third factors each accounted for 7% of the
variance. In the rotated factor matrix, Factor 1
appeared to measure concentration and new
learning,
Factor
2
appeared
to
involve
expressive and receptive speech, and Factor 3
appeared to measure global memory abilities.
In order to explore the discriminatory power of
the
individual
15
items
on
the
Neuropsychological
Dysfunction
scale,
a
multivariate analysis of variance (MANOVA) was
performed between the CHI and control groups.
The multivariate eff ect was signifi cant, ( p <.01).
It appeared that item I tend to forget things I
am
supposed
to
do,
had
the
highest
discriminatory power among the 15. Items about
concentration,
memorization,
reading,
and
balance had the next highest discriminatory
ability, respectively. Of the 15 items, 10 were
signifi cant
at
.019
or
less.
Of
the
5
nonsignifi cant items, three of them had been
designed to detect aphasia. This latter fi nding is
consistent with the literature that specifi c
aphasic syndromes are uncommon following CHI
in adults (Ewing-Cobbs, Levin, Eisenberg, &
Fletcher, 1987; Heilman, Safrran, & Geschwind,
1971). A subsequent item, I have headaches,
has been added to form the fi nal 16 items of the
present Neuropsychological Dysfunction scale of
the CATI+.

Executive Functions Scales

Factor Analysis
A principal components
Varimax
rotation
was
normative sample (N =
three-factor structure. The

factor analysis with

performed
on
the
937) This yielded a
fi rst factor, decision-

making diffi culties, accounted for 23% of the

total variance and had eight items with factor
loadings of .41 or greater. The second factor was
planning problems and accounted for 15% of the
total variance with four items showing factor
loadings of .45 or greater. The third factor, task
completion diffi culty, accounted for 9% of the
total variance and had six items with loadings of
.45 or greater. Only two of the 16 items loaded
on two factors.

A Study of Head Injury Patients

Coolidge and Griego (1995) performed a
multivariate analysis of variance (MANOVA) on
the 16-item Executive Functions scale and the
three subscales between 17 closed head injured
patients and a group-matched (age, gender, and
education) control group. Consistent with their
hypothesis,
the
multivariate
eff ect
was
signifi cant and the head injured patients scored
signifi cantly higher on the composite measure of
the Executive Functions Scale (indicating greater
dysfunction) and higher on the Decision-making
Diffi culty Subscale, but not on the other two
scales.

Chapter 4

Case Studies for the CATI+

The examination of specifi c instances of CATI+
results can be of assistance to understanding
the instruments properties and potential. The
following four case studies provide scenarios of
typical uses of the CATI+ for diagnostic and
treatment purposes.

Case 1: A Patient with Multiple Sclerosis

This CATI+ narrative report is of a 46-year- old
female diagnosed with multiple sclerosis, a
disease
that
interferes
with
nerve
cell
transmission in the brain. The primary symptoms
of multiple sclerosis are somatic, such as
impairment in ambulation and vision. Later
symptoms may include cognitive impairment,
either primarily as a function of brain cell
damage
or,
secondarily,
as
a
result
of
psychological trauma in reaction to a chronic,
debilitating disease.
The fi rst step in the evaluation of the CATI+ is
to
examine
the
administrative
indices.
Examining this patients CATI+ administrative
indices (see Figure 4-1) shows that her number
of
Strongly
False
responses
is
elevated
compared to the normative sample by over one
standard deviation. However, her More True Than
False responses are slightly elevated while her
Strongly True answers are highly elevated. This
appears to indicate that the patient is answering
openly and honestly. There are no unscorable
responses and, since on the random responding
scale her count is 3, there is no evidence for
randomness in her answers. On the Tendency to
Look Good or Bad Scale, the patient has

obtained a T score of 35, indicating one of two

possibilities: either she is trying to look bad or
she may genuinely feel bad. In the present case,
since she has been diagnosed with multiple
sclerosis, the latter interpretation seems more
probable. On the TDBP scale, the patient has a T
score of 57 and, although within normal limits,
this indicates that she has no tendency to deny
blatant pathology and, indeed, it appears that
the patient is answering truthfully.
At this point, since the administrative indices
of validity indicate a valid and reliable profi le,
the patients Axis I, Axis II, Neuropsychological
and other clinical scales may be examined (see
Figure 4-2). On Axis I, the patient is elevated on
the Post-Traumatic Stress disorder scale. The
latter fi nding is not surprising, since she is
coping with a potentially life-threatening chronic
disease. A review of the Critical Items section of
the CATI+ reveals that the patient has answered
Strongly True to item 212, indicating that she
has a memory that bothers her. This may
indicate that the elevation on the post-traumatic
stress disorder scale may not be due to the
diagnosis of multiple sclerosis, but the patient
may
have
experienced
some
other
lifethreatening experience such as rape or sexual
trauma. At this point it will be important for the
clinician to query the patient in regard to this
memory. Her second highest elevation on Axis I
is the depression scale and her T score of 55
places her in the 69 t h percentile for normal
individuals. Here it is important for the clinician
to investigate the critical item 198 in order to
determine if the patient is suicidal. In this case,
the patient has answered Strongly False,
although it still may be useful for the clinician to
query the patients attitudes about suicide.
On Axis II, there are only two signifi cant
elevations, one on the Histrionic scale and the
other on the Passive-Aggressive scale. Her T
score of 65 gives her an overall percentile of 93

compared to the normative sample. Since the

patient has answered that she enjoys social
gatherings and likes to be perceived as a
sexually
attractive
person
and
places
importance
on
physical
attractiveness,
a
potential
psychological
diffi culty
may
be
developing.
In
chronic
cases
of
multiple
sclerosis, patients can become incontinent and
non-ambulatory. Since our present patient places
such an emphasis on attractiveness, she may
have diffi culties coping with the realities of the
disease in the future. Again, it would be
important for the clinician to query the patient
in regard to this latter fi nding.
Not surprisingly, the major elevations in this
CATI+ profi le are on the neuropsychological
scales, all of which are in the 99 t h percentile.
Consistent with her disease, she is experiencing
not only somatic problems but memory and
language problems as well. Indeed, even the
scales that measure executive functions of the
frontal lobes are also highly elevated.
In terms of her personality change as a
function of her medical condition, there are two
important elevations on Disinhibition and her
Emotional Lability. In the former case, it would
be important to query the patient about the
areas in her life where she is experiencing this
disinhibition. The clinician may be helpful in
preventing
potentially
injurious
impulsive
behavior. On the Hostility scales, she is within
normal limits on anger and dangerousness.
However, her Impulsiveness scale is highly
elevated which reinforces the prior fi nding on
the Disinhibition scale.

Case 2: A Psychotherapy Patient

This patient is an 18-year- old female selfreferred for psychotherapy. As part of her initial
evaluation, she was administered the CATI+.
Examination
of
her
administrative
indices
reveals that she is within the normal limits (see

Figure 4-3). In terms of her answer choice

frequency, it can be seen that there is no
excessive tendency to choose the Strongly False
or More False Than True answers and, indeed,
she is somewhat elevated on her Strongly True
answers. There are no unscorable responses and
her Random Responding Scale has a count of 3,
which
indicates
no
evidence
of
random
responding. On the Tendency to Look Good or
Bad Scale she has a T score of 35, which (given
her self-referred for therapy circumstance)
indicates that she may be genuinely feeling bad
and not just trying to look bad. The TDBP scale,
although within normal limits, also indicates no
tendency whatsoever to deny blatant pathology;
thus, denial does not appear to be a problem in
this patients profi le.
There are many important considerations when
examining this patients CATI+ profi le (see
Figure 4-4). On Axis I, her scores are elevated on
virtually all scales. The highest of these is the
Depression scale. As with any elevation on the
CATI+, the clinician should individually examine
item 198: Recently I have felt like killing
myself. The CATI+ protocol will also produce
this item in the Critical Items section if it is
endorsed in a positive direction. In the present
case the patient has endorsed this suicidal item,
so it would be important for the clinician to
query the patient about her suicidal ideation as
quickly as possible to determine the likelihood of
her acting on her impulses. Her Post-Traumatic
Stress disorder scale is also elevated and on the
critical items she has endorsed item 212 that
reads, I have a memory that bothers me.
Again, it would be important for the clinician to
query her response.
On Axis II, consistent with the clinicians
preliminary diagnosis, the patient is elevated on
the Borderline scale. Interestingly, she has even
a higher elevation on the Avoidant scale, which
may
contribute
to
the
present
level
of

unhappiness that is perceived in this CATI+

protocol. Indeed, she has seven personality
disorder scales elevated beyond one standard
deviation above the mean. This is not an unusual
situation, since CATI+ research has shown that
75%
of
patients
with
clinician-diagnosed
personality
disorders
have
two
or
more
concurrent personality disorders.
On her neuropsychological scales, she has only
one subscale elevation regarding language. The
CATI+ interpretation reveals that she has trouble
speaking and that her speech is perceived by
others as strange or vague. The clinician should,
of course, query the patients report. On the
overall scale of neuropsychological dysfunction,
She has a slightly elevated T score (55) and the
report
indicates
symptoms
of
mild
brain
dysfunction. In this case, the elevation on the
overall neuropsychological dysfunction scale is
being caused apparently solely by her problems
on the language dysfunction subscale.
In regard to her executive functions, her
overall scale is not elevated. However, she is
reporting
decision-making
diffi culties
and
planning problems. CATI+ research has also
shown
a
relationship
between
executive
functions defi cits and the borderline personality
disorder. With regard to the hostility scales,
there is the expected elevation on the anger
scale (96 t h percentile), but she is within normal
limits on dangerousness and impulsiveness. The
latter fi ndings may be perceived as a positive
prognostic sign, given her borderline tendencies,
since her anger does not appear to make her
dangerous nor impulsive at this time, at least
according to her self-report.

Case 3: A State Mental Hospital Patient

This patient is a 40-year- old male confi ned by
court order to the forensic unit, maximum
security, of a state mental hospital. Examination
of his CATI+ administrative indices indicates an

overwhelming tendency to choose the More True

Than False and Strongly True answers (see Figure
4-5). Immediately, this should suggest to the
clinician that the patient has no tendency to
deny pathology and, indeed, the problem
appears to be the opposite: the patient appears
to choose the pathological position with regard
to almost any item. A TDBP score in this case of
327 may then indicate an extremely disturbed
patient or a disturbed patient who is crying out
for help. There are two unscorable items in his
profi le.
This patients administrative indices profi le
presents an unusual dilemma (see Figure 4-5).
The random responding count is 10, which
indicates that the profi le may be invalid. CATI+
research has shown that better than 85% of
inpatient schizophrenics can validly report their
symptomatology. In any CATI+ profi le, a Random
Responding Scale raw score greater than three
may be indicative of random or capricious
responding. In the present case, the count of 10
on the Random Responding Scale may indicate
random responding, but may be more indicative
of the active psychotic thought processes of a
fl oridly schizophrenic patient. This CATI+ profi le
may be valid in spite of a 10 on the Random
Responding Scale. The reasoning is as follows: If
a patient randomly responded to all of the items
of the CATI+, the TDBP score would be
approximately 243 (if all answer choices were
evenly divided across the 97 items of the TDBP
scale). In the present case, the TDBP score is
327 which might indicate that the patient is
indeed intentionally favoring the pathological
pole (strongly true) for most of the items
consistent, perhaps, with the patients very
disturbed self-image.
On the other validity scales, the Tendency to
Look Good or Bad Scale score is very low,
indicating that the patient may not only feel bad
but also is trying to make himself look even

worse. The latter circumstance (known as

psychotic malingering) is possible when a
patient in a state mental hospital prefers to
avoid criminal charges and avoid serving time in
a state prison. The TDBP scale, which is even a
more global and broad measure of the tendency
to look good or bad, indicates one of the highest
raw scores (327) and T scores ever obtained by
a schizophrenic patient on the CATI+. The latter
high scores indicate no tendency whatsoever to
deny pathology and, indeed, an extremely strong
tendency to endorse psychopathology as noted
earlier.
Figure 4-6 presents this patients CATI+ profi le.
Consistent
with
his
over-endorsement
of
pathology, there are signifi cant elevations on all
Axis I scales and on most of the Axis II scales.
Interestingly, the highest two elevations on Axis
I are for the psychotic thinking scale and the
schizophrenia scale. On Axis II, the single
highest elevation is on the sadistic scale. Given
the patients elevation on all three hostility
scales
(anger,
dangerousness,
and
impulsiveness), and his high elevation on the
sadistic personality disorder scale, the patient
appears to be in an appropriate setting ( a
maximum security ward in a state mental
hospital).
On the Overall Neuropsychological Scale, this
patient scored exceedingly high as well as very
high on all three subscales of language, memory
and concentration, and neurosomatic problems.
These results are not surprising in light of the
fact that a substantial portion of schizophrenic
patients report neuropsychological diffi culties. In
some of these cases, it may be a function of
their medications. Many antipsychotic drugs are
well known for their interference with attention
and memory functions. Other schizophrenics,
particularly those who may have been homeless
or on the streets for a long period of time, may
have indeed suff ered head injuries, been

exposed to hostile toxins (e.g., bad drugs), or

suff ered
from
malnutrition.
All
of
these
conditions may result in neurological damage
and subsequent neuropsychological dysfunction.
On the Executive Functions Scale, it can be
seen that this patient is not particularly
elevated, except on his decision diffi culty
subscale. In fact, he has scored rather low,
considering all his other scales, on the poor
planning subscale of executive functions. This
may actually be a pathological red fl ag for the
clinician because the patient scored so high on
the sadistic and antisocial scale and his hostility
scales are so elevated that this patient may take
time to plan his dysfunctional and dangerous
actions.
It is important that the clinician remember to
examine critical items, particularly item 198:
Recently I have felt like killing myself. In the
present case, an examination of the critical
items reveals that the patient has made more
than one suicidal threat or gesture in his life,
has gotten into trouble because of drinking or
drug problems, and is currently suicidal. It would
be very important for the clinician to assess by
querying the patient the likelihood that he would
commit suicide, given his present circumstances.

Case 4: A Patient with an Invalid Response

Profile
This CATI+ report is for a male patient who
demonstrated invalid administrative indices.
Careful examination of these indices reveals
several concerns (see Figure 4-7). The fi rst of
these scales, Answer Choice Frequency, shows
that, compared to the normative sample, he has
chosen the answer Strongly False at twice the
rate of the normative sample. This is a fi rst
indication that he has a tendency to deny
problems, since a majority of the items of the
CATI+ are written such that Strongly False
(answer 1) indicates denial of psychopathology

and Strongly True (answer 4) indicates an

endorsement. The next administrative index is
unscorable responses or, in other words,
unanswered responses. Here, he appears to have
answered all questions and there are no
unscorable responses. The next validity index is
the Random Responding Scale where it is
extremely unlikely that any normal person would
answer anything but Strongly False (or 1) to the
three items. Our research suggests that only in
the cases of random responding and in some
blatantly psychotic disorders will a patient
answer anything but Strongly False. In his case,
there is no evidence for random responding. His
count equals 3, which indicates Strongly False on
all three items. Anything other than a count of 3
would indicate random responding, psychotic
responding or, more rarely, the patient is
responding playfully. The next index is a
Tendency to Look Good or Bad and he has a T
score of 49, which indicates no excessive
tendency to either fake good or to fake bad. The
fi nal index, and one of the most useful, is the
Tendency to Deny Blatant Pathology Scale
(TDBP). Since the latter index has more than fi ve
times the total number of items than the
Tendency to Look Bad or Good Scale, the TDBP
scale is perhaps more useful and reliable in
determining the patients tendency to deny. Here
we fi nd that his T score is 41, essentially one
complete
standard
deviation
below
the
normative mean. While it places him within
normal limits on the scale, it is perhaps the
clearest indication other than answer choice
frequency that he may be less than honest about
his problems. This is particularly true if he is a
clinical patient, either inpatient or outpatient,
seeking psychological help. In cases of denial
such as this, the clinician may wish to take into
account the testing referral circumstances. For
example, it might not be unusual for this patient
to have such a low TDBP score if he has been
court-ordered for psychological services. In the

latter case, it is suggested that the clinician

readminister the CATI+ with the instructions that
the CATI+ is a sensitive test to faking.
Furthermore, the clinician, in the face of a test
protocol where there is evidence of denial on the
administrative indices, may confront the patient
with the fact of their denial and ask the patient
to take the test again and to be more
cooperative.
In summary, this patients test protocol is
relatively useless. There appear to be only two
possibilities: fi rst, he is completely normal,
although compared to the normal sample he has
denied far more than the average person or,
second, if he is in dire psychological straits he
has clearly adopted a resistant test-taking
attitude and he needs to be encouraged to be
more truthful in his self report.
Importantly, there is another option. The CATI+
is
one
of
the
few
psychological
and
neuropsychological inventories that contains a
signifi cant other form. If in the clinicians
judgment it is determined that he cannot
produce a valid profi le, then the clinician may
wish to look for signifi cant others in this
patients life who may be able to provide a
better picture of his psychological profi le.
Figure 4-1: CATI+ Administrative Indices for
Case 1

Figure 4-2: CATI+ Profi le for Case 1

Figure 4-3: CATI+ Administrative Indices for Case 2

Figure 4-4: CATI+ Profi le for Case 2

Figure 4-5: CATI+ Administrative Indices for Case 3

Figure 4-6: CATI+ Profi le for Case 3

Figure 4-7: CATI+ Administrative Indices for Case 4

References
American
Psychiatric
Association.
(1968).
Diagnostic and Statistical Manual of
Mental Disorders (2nd ed). Washington,
DC: Author.
American
Psychiatric
Association.
(1980).
Diagnostic and Statistical Manual of
Mental Disorders (3rd ed.). Washington,
DC: Author.
American
Psychiatric
Association.
(1987).
Diagnostic and statistical manual of
mental
disorders
(3rd
ed.
rev.) .
Washington, DC: Author.
American
Psychiatric
Association.
(1994).
Diagnostic and statistical manual of
mental disorders (4th ed.); DSM-IV .
Washington, DC: Author.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J.,
& Erbaugh, J. (1961). An inventory for
measuring
depression.
Archives
of
General Psychiatry, 4, 561-571.
Binder, L. M. (1986). Persisting symptoms after
mild
head
injury:
A
review
of
postconcussive syndrome. Journal of
Clinical
and
Experimental
Neuropsychology, 8, 323-346.
Coolidge, F. L. (1984). Coolidge Axis II Inventory.
U.S. Copyright TXU-026, Washington,
D.C.
Coolidge, F. L., Bracken, D. D., Taylor, L. R.,
Smith,
K.,
&
Peters,
B.
(1985).
Personality disorders in Alzheimers
disease: A hierarchical cluster analysis
approach.
Journal
of
Clinical
and
Experimental Neuropsychology, 7, 616.

Coolidge, F. L., Burns, E. M., & Mooney, J. A.

(1995). Reliability of observer ratings in
the
assessment
of
personality
disorders: A preliminary study. Journal
of Clinical Psychology, 51, 22-28.
Coolidge, F. L., Burns, E. M., & Mull, C. E. (1992).
Personality disorders in the elderly.
Clinical Gerontologist, 12, 41-55.
Coolidge, F. L., & Griego, J. A. (1995). Executive
functions
of
the
frontal
lobes:
Psychometric properties of a self-rating
scale. Psychological Reports, 77, 24-26.
Coolidge, F. L., & Merwin, M. M. (April, 1989).
Self-assessment of the cognitive and
emotional sequelae of brain damage.
Paper presented at the meeting of the
Western
Psychological
Association,
Reno, NV.
Coolidge, F. L. & Merwin, M. M. (1992).
Reliability and validity of the Coolidge
Axis II Inventory: A new inventory for
the
assessment
of
personality
disorders.
Journal
of
Personality
Assessment, 59, 223-238.
Coolidge, F. L., Merwin, M. M., Nathan, J. H., &
Schmidt,
M.
M.
(1996,
June).
Assessment
of
neurobehavioral
symptoms after traumatic brain injury.
Indian Journal of Psychological Issues,
V. 4, 1-9.
Coolidge, F. L., Merwin, M. M., & Philbrick, P. B.
(Nov., 1989). Validation of a self-report,
brain dysfunction scale of the Coolidge
Axis II Inventory. Paper presented at the
meeting of the National Academy of
Neuropsychologists, Washington, DC.
(title only)
Coolidge, F. L., Segal, C. L., Pointer, J. C., Knaus,
E. A., Yamazak, T. G., & Silberman, C. S.
(in press). Personality disorders in
elderly inpatients with chronic mental

illness.
Journal
Geropsychology

of

Clinical

Costanzo, R. M., & Becker, D. P. (1986). Smell

and taste disorders in head injury and
neurosurgery
patients.
In
H.
L.
Meiselman & R. S. Rivlin (Eds.), Clinical
Measurement
of
Taste
and
Smell
(pp.565-578). New York: Macmillan.
Cronbach, L. J. (1951). Coeffi cient alpha and the
internal
structures
of
tests.
Psychometrika, 16, 297-334.
Ewing-Cobbs, L., Levin, H. S., Eisenberg H. M., &
Fletcher,
J.
M.
(1987).
Language
functions following closed-head injury
in children and adolescents. Journal of
Clinical
and
Experimental
Neuropsychology, 9, 575-592.
Frances, A. (1982). Dimensional and categorical
systems of personality diagnosis: A
comparison. Comprehensive Psychiatry,
23, 516-527.
Gasparrini, W. G., Satz, P., Heilman, K., &
Coolidge, F. L. (1978). Hemispheric
asymmetries of aff ective processing as
determined
by
the
Minnesota
Multiphasic
Personality
Inventory.
Journal of Neurology, Neurosurgery, and
Psychiatry, 41, 470-473.
Heilman, K. M., Saff ran, E. & Geschwind, N.
(1971).
Closed
head
trauma
and
aphasia.
Journal
of
Neurology,
Neurosurgery and Psychiatry, 34 , 265269.
Hosman, P. A. (1989). Convergent validity of the
Spielberger Anxiety Scale, the MMPI
anxiety scale, and the Coolidge Axis II
Inventory. Unpublished honors thesis,
University
of
Colorado,
Colorado
Springs, CO.
Lezak, M.D. (1993). Newer contributions to the
neuropsychological
assessment
of

executive functions. Journal of

Trauma Rehabilitation, 8, 24-31.

Head

Lezak, M.D. (1982). The problem of assessing

executive
functions.
International
Journal of Psychology, 17, 281-297.
Lezak,

M.D.
(1983).
Neuropsychological
Assessment.
New
York:
Oxford
University Press.

Lezak,

M.D. (1987). Relationships between

personality
disorders,
social
disturbances, and physical disability
following traumatic brain injury. Journal
of Head Trauma Rehabilitation, 2, 5769.

Lucero,

D. L. (1989). The construction and

validation of a depression scale for the
CATI.
Unpublished
honors
thesis,
University
of
Colorado,
Colorado
Springs, CO.

Luria, A. R. (1966). Higher cortical functions in

man. New York: Plenum.
McGlynn, S. M., & Schacter, D. L. (1989).
Unawareness
of
defi cits
in
neuropsychological syndromes. Journal
of
Clinical
and
Experimental
Neuropsychology, 11, 143-205.
Mellor,

S. (1987). Self-other agreement in

personality ratings by committed and
uncommitted work partners. Perceptual
and Motor Skills, 63, 1295-1299.

Melisop, G., Varghese, F., Joshua, S., & Hicks, A.

(1982). The reliability of Axis II of DSMIII. American Journal of Psychiatry, 139,
1360-1361.
Merikangas, K. R., & Weissman, M. M. (1986).
Epidemiology
of
DSM-III
Axis
II
Personality Disorders. In A. J. Frances &
R. B. Hales (Eds.), American Psychiatry
Association: Annual Review: Vol. 5.
Psychiatry Update Series. Washington,
DC: American Psychiatric Press.

Merwin, M. M., & Coolidge, F. L. (March, 1987).

Coolidge Axis Two Inventory scale
reliabilities. Paper presented to the
meeting of the Colorado Psychological
Association, Denver, CO.
Merwin, M. M., & Coolidge, F. L. (Nov., 1989).
Factor
analysis
of
a
self-report
inventory for the assessment of brain
dysfunction. Paper presented to the
meeting of the Colorado Psychological
Association, Denver, CO.
Millon,

T. (1977). Millon Clinical

Inventory.
Minneapolis:
Computer Systems

Multiaxial
National

Millon, T. (1985). The MCMI provides a good

assessment of DSM-III disorders: The
MCMI-II will prove even better. Journal
of Personality Assessment, 49, 379-391.
Millon,

T. (1987). Millon Clinical Multiaxial

Inventory-II.
Manual.
Minneapolis:
National Computer Systems.

Mitton, N. M. (1989).
the
CATI:
Unpublished
of Colorado,

A scale of dissimulation on
Reliability
and
validity.
masters thesis, University
Colorado Springs, CO.

Parker, R. S. (1990). Traumatic Brain Injury and

Neuropsychological Impairment. New
York: Springer-Verlag.
Piersma, H. L. (1986). The factor structure of the
Millon
clinical
multiaxial
inventory
(MCMI)
for
psychiatric
inpatients.
Journal of Personality Assessment, 50,
587-594.
Spielberger, C. D. (1983). Manual for the StateTrait Anxiety Inventory. Palo Alto, CA:
Consulting Psychologists Press, Inc.
Spitzer, R. L., Forman, J. B. W., & Nee, J. (1979).
DSM-III fi eld trials: I. Initial interrater
diagnostic reliability. American Journal
of Psychiatry, 136, 815-817.

Welsh, M. C., Pennington, B. F., & Grossier, D.

(1991). A normative developmental
study of executive functions: A window
on prefrontal function in children.
Developmental
Neuropsychology,
7,
131-149.
Widiger, T. A., Williams, J. B. W., Spitzer, R. L., &
Francis, A. (1985). The MCMI as a
measure
of
DSM-III.
Journal
of
Personality Assessment, 49, 366-378.