This article was downloaded by: [Northeastern University]

On: 03 January 2015, At: 18:10

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK

Journal of Forensic Psychology

Practice
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/wfpp20

The MCMI-III in Child Custody

Evaluations
a b
Joseph T. McCann , James R. Flens , Vicky
c
Campagna , Pamela Collman , Thomas Lazzaro &
Ed Connor
a
United Health Services Hospitals , Bing-hamton,
NY, USA
b
Family Resource Center , Brandon, FL, USA
c
Youth Link of Fresno County
Published online: 20 Oct 2008.

To cite this article: Joseph T. McCann , James R. Flens , Vicky Campagna , Pamela
Collman , Thomas Lazzaro & Ed Connor (2001) The MCMI-III in Child Custody
Evaluations, Journal of Forensic Psychology Practice, 1:2, 27-44, DOI: 10.1300/
J158v01n02_02

To link to this article: http://dx.doi.org/10.1300/J158v01n02_02

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the
information (the Content) contained in the publications on our platform.
However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness,
or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the
Content should not be relied upon and should be independently verified with
primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages,
 and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the
Content.

This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan,
sub-licensing, systematic supply, or distribution in any form to anyone is
expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions
Downloaded by [Northeastern University] at 18:10 03 January 2015
 FULL-LENGTH ARTICLE

The MCMI-III
Downloaded by [Northeastern University] at 18:10 03 January 2015

in Child Custody Evaluations:

A Normative Study
Joseph T. McCann
James R. Flens
Vicky Campagna
Pamela Collman
Thomas Lazzaro
Ed Connor

ABSTRACT. This study provides normative data on the Millon Clinical

Multiaxial Inventory-III (MCMI-III) among 259 child custody examin-
ees who were obtained from private practice settings in four different
states. The mean MCMI-III profile in this sample was an elevation on

Joseph T. McCann is affiliated with the United Health Services Hospitals, Bing-
hamton, NY. James R. Flens is affiliated with the Family Resource Center, Brandon,
FL. Vicky Campagna is in Private Practice, Redwood City, CA. Pamela Collman is
affiliated with Youth Link of Fresno County. Thomas Lazzaro is in Private Practice,
Syracuse, NY. Ed Connor is in Private Practice, Covington, KY.
Address correspondence to: Joseph T. McCann, 151 Leroy Street, Binghamton,
NY 13905.
An earlier version of this paper was presented at the 5th Annual Millon Clinical
Inventories Conference on Personology and Psychopathology, Chicago, IL, October,
1998.
Journal of Forensic Psychology Practice, Vol. 1(2) 2001
E 2001 by The Haworth Press, Inc. All rights reserved. 27
 28 JOURNAL OF FORENSIC PSYCHOLOGY PRACTICE

Scale Y (Desirability), with subclinical elevations on Scales 4 (Histrion-

ic), 5 (Narcissistic), and 7 (Compulsive). Females scored significantly
higher than males on Scales 4, 5, and 7. Moreover, mean Base Rate (BR)
scores were very low for most scales. With the exception of Scales 4 and
7 for females, the frequency of clinically significant elevations (i.e., BR 
75) on MCMI-III scales was very low among child custody litigants.
Results from this study provide empirical support for the position that the
MCMI-III does not overpathologize child custody examinees. Some in-
terpretive guidelines are offered to reduce the likelihood of false positive
diagnoses on those few scales where such errors may arise. This study
Downloaded by [Northeastern University] at 18:10 03 January 2015

provides support for continued use of the MCMI-III in child custody

evaluations. [Article copies available for a fee from The Haworth Document
Delivery Service: 1-800-342-9678. E-mail address: <getinfo@haworthpressinc.
com> Website: <http://www.HaworthPress.com> E 2001 by The Haworth Press,
Inc. All rights reserved.]

KEYWORDS. MCMI-III, child custody, forensic assessment, norms

Over the past few decades, psychologists have become more in-
volved in the legal process by providing consultation to courts on a
variety of matters. One area that has been particularly adversarial is
the use of expert psychological testimony in cases involving disputes
over child custody and visitation. As a result of this growing area of
psychological expertise in court, there has been increased attention
given to the use of psychological tests in child custody proceedings.
The traditional approach to conducting child custody evaluations
has been a comprehensive one in which specific issues and questions
are posed to the examiner and multiple sets of data are integrated to
answer these questions (Halon, 1990). Typically, comprehensive psy-
chological evaluations are conducted on each parent, the individual
children, and potential step-parents and detailed observations are made
of parent-child interactions. Although there once was very little direc-
tion available for child custody examiners, recently a number of pro-
fessional standards have been developed to guide forensic mental
health practice and several texts on child custody evaluations have
appeared in the literature. For instance, the American Psychological
Association (1994) has drafted guidelines for conducting child custo-
dy evaluations, as have several states, including Georgia (Georgia
Psychological Association, 1990), Nebraska (Nebraska Psychological
Association, 1986), New Jersey (New Jersey Board of Psychological
Examiners, 1993), and Oklahoma (Oklahoma Psychological Associa-
 Full-Length Article 29

tion, 1988). Comprehensive guides to conducting child custody evalu-

ations include books by Ackerman (1995) and Gould (1998).
When traditional psychological tests are applied in forensic contexts,
it is the duty of the expert to not only use methods that are reliable and
valid, but also ones that have relevance to the ultimate legal issue being
litigated (Heilbrun, 1992). With respect to child custody evaluations,
traditional psychological tests are relevant if they provide clinical data
that will help guide determinations as to what is in the best interests of
Downloaded by [Northeastern University] at 18:10 03 January 2015

the children, since this is the common legal test for determining child
custody (Melton, Petrila, Poythress, & Slobogin, 1997).
Prior research on the testing practices of psychologists who con-
duct child custody evaluations reveal that the most commonly used
tests with adults are the Minnesota Multiphasic Personality Inventory
(MMPI), Rorschach Inkblot Method, and Wechsler Adult Intelligence
Scale-Revised (WAIS-R). Keilin and Bloom (1986) found that the
MMPI was used by about 71% of child custody evaluators who were
surveyed and Ackerman and Ackerman (1997) found that the MMPI-2
was used by about 92% in a later study of child custody evaluators.
Despite widespread use of the MMPI-2, there was little information on
how child custody litigants actually performed on the MMPI-2 until
Bathurst, Gottfried, and Gottfried (1997) conducted their large study.
Over the last fifteen years, the Millon Clinical Multiaxial Invento-
ry (MCMI; Millon, 1983, 1987, 1994; Millon, Davis, & Millon,
1997) has become a more established psychological assessment in-
strument. Although Keilin and Bloom (1986) did not cite the MCMI
as a frequently used instrument in child custody evaluations, a more
current study by Ackerman and Ackerman (1997) found that about
34% of child custody evaluators utilized the MCMI with adults. This
finding reflects a general trend that has found the MCMI to be fre-
quently used in forensic psychological evaluations (Borum & Grisso,
1995; McCann & Dyer, 1996).
Use of the Millon Clinical Multiaxial Inventory-III (MCMI-III;
Millon, 1994; Millon, Davis, & Millon, 1997) in child custody pro-
ceedings has been controversial. Some have argued that the clinical
nature of the instrument renders it inappropriate for use in child custo-
dy evaluations, which have been conceptualized by some as nonclini-
cal types of examinations (Ackerman, 1995; Ackerman & Ackerman,
1997). Moreover, because norms for the MCMI-III were developed in
clinical settings, they have been viewed by some as inappropriate in
 30 JOURNAL OF FORENSIC PSYCHOLOGY PRACTICE

child custody cases because they supposedly lead to the MCMI-III

narrative report producing interpretations that overpathologize indi-
viduals being evaluated in child custody disputes (Otto & Butcher,
1995). A significant issue that appears to underlie these concerns is
whether the MCMI-III is diagnostically accurate in child custody set-
tings.
However, there are several reasons to view child custody evalua-
tions as clinical in nature, such as the highly conflicted and conten-
Downloaded by [Northeastern University] at 18:10 03 January 2015

tious nature of marital dissolution and the need to consider the pos-
sible presence of psychopathology and its potential impact on
parenting. In addition, the MCMI-III was designed for use with indi-
viduals who evidence problematic emotional and interpersonal symp-
toms or who are undergoing professional psychotherapy or a psycho-
diagnostic evaluation (Millon, Davis, & Millon, 1997, p. 6, emphasis
added). Child custody disputes certainly fit in the realm of individu-
als who . . . are undergoing . . . a psychodiagnostic evaluation and the
revised MCM-III manual has also endorsed use of the instrument in
child custody evaluations (Millon, Davis, & Millon, 1997).
Despite the fact that McCann and Dyer (1996) recommended clini-
cians use the MCMI-II instead of the MCMI-III in forensic settings
because of scant validity data in the original edition of the MCMI-III
manual (Millon, 1994), there have been recent advances in the literature
which outdate this recommendation and which now support validity of
the MCMI-III and its use in forensic evaluations (Craig, R. J., 1999;
Craig & Bivens, 1998; Craig, Bivens, & Olson, 1997; Davis & Hays,
1997; Davis, Wenger, & Guzman, 1997; Dyce, OConnor, Parkins, &
Janzen, 1997; Dyer, 1997; Dyer & McCann, 2000). Moreover, the
MCMI-III manual has been revised and the most recent edition includes
a more expanded and detailed validity study which supports use of the
MCMI-III in forensic assessments (Millon, Davis, & Millon, 1997).
Recently, Rogers, Salekin, and Sewell (1999) criticized forensic use
of the Millon inventories by asserting that neither the MCMI-II nor the
MCMI-III were validated against DSM-IV or legal criteria and that the
instruments supposedly have poor convergent and discriminant validi-
ty. However, Dyer and McCann (2000) cited several methodological
shortcomings of the Rogers et al. (1999) study, including arbitrary
reversal of predictor and criterion, an incomplete review of the litera-
ture that failed to consider the most current edition of the MCMI-III
manual, inaccurate statements about content validity of the MCMI-III,
 Full-Length Article 31

misleading criticisms about use of the MCMI-III for evaluating legally

relevant issues, and serious errors in the multitrait-multimethod proce-
dures employed by Rogers et al. (1999) that render their findings
meaningless. Therefore, while forensic application of the Millon in-
ventories in forensic settings is a topic of debate, there remains sub-
stantial support for continued use of the MCMI-III in many types of
forensic evaluations (Dyer, 1997; Dyer & McCann, 2000)
This paper reports findings from a normative study of how child
Downloaded by [Northeastern University] at 18:10 03 January 2015

custody litigants perform on the MCMI-III. We sought to answer three

general questions in this study: (1) How do individuals undergoing a
child custody evaluation typically score on the MCMI-III; (2) Does
the MCMI-III overpathologize child custody examinees; and (3) Are
there any modifications in interpretation of the MCMI-III that may be
required in child custody settings?

METHOD

Subjects

The subjects for this study were 259 individuals who completed the
MCMI-III as part of a child custody evaluation that was being adjudi-
cated. Subjects were obtained from six private practice settings in
which forensic mental health services, including child custody evalua-
tions, were conducted and in which the MCMI-III is routinely
administered. The sample was obtained from four different states,
namely Florida, New York, California, and Kentucky. There were 127
males and 130 females in the sample, with two subjects whose gender,
while available for scoring the MCMI-III protocol, was not available
when data were compiled for this study. Table 1 provides information
on the demographic characteristics for males, females, and the total
sample on such variables as age, education, marital status, number of
children at issue, and race. Overall, the data in Table 1 indicate that the
child custody examinees in this study tended to be white, in their late
thirties, college educated, and separated at the time of the evaluation.
There were no prominent differences noted between males and fe-
males on demographic characteristics. Nearly all of the MCMI-III
protocols were obtained in cases where both parents were being ex-
amined. In a very few cases, only one parent was examined. Thus, a
 32 JOURNAL OF FORENSIC PSYCHOLOGY PRACTICE

TABLE 1. Demographic Characteristics of Sample

Variable Males Females Total Sample

(n = 127) (n = 130) (n = 259)

Age
Mean 39.13 36.15 37.62
Standard Deviation 8.55 7.46 8.15
Education
Downloaded by [Northeastern University] at 18:10 03 January 2015

High School 45 53 94
College 45 41 86
Graduate School 17 13 30
No Data Available 24 23 49
Marital Status
Single 7 7 14
Married 5 9 14
Remarried 9 10 19
Divorced 19 18 37
Separated 34 34 68
No Data Available 53 52 107
Children at Issue
Mean 1.82 1.92 1.84
Standard Deviation 0.83 0.86 0.85
Race
White 79 80 159
African-American 1 0 1
Hispanic 8 3 11
Other 3 0 3
No Data Available 39 44 85

majority of cases in this study reflect a standard child custody evalua-

tion in which both parents are evaluated.
Procedure
The MCMI-III protocols were collected from each private practice
setting along with case relevant information, such as demographic
characteristics of each individual, any custody or visitation decision
that had been rendered by the court, and any reports of domestic
violence. The demographic characteristics in Table 1 are presented in
terms of raw data, rather than percentages, because of missing data for
several subjects. Computation of percentages based on either the total
 Full-Length Article 33

sample or the number of subjects for whom data were available might
be misleading. Data were missing for large portions of the sample
because certain demographic data were not available for some of the
private practice settings. All MCMI-III protocols had been computer
scored and were valid in that the Validity Index had a raw score of zero
and Scale X raw scores were within the acceptable range as defined in
the test manual (Millon, Davis, & Millon, 1997). The demographic
information and BR scores from the MCMI-III scales were entered
Downloaded by [Northeastern University] at 18:10 03 January 2015

into a computerized data base for analysis.

RESULTS
The mean MCMI-III profiles for males, females, and the combined
sample are provided in Table 2. For the entire sample, the mean
MCMI-III profile was characterized by a clinically significant eleva-
tion (i.e., BR  75) on Scale Y (Desirability); all other mean scale
elevations were below this clinically significant level. However, the
mean profile for the total sample masks a statistically significant dif-
ference that existed between males and females. For males, the mean
MCMI-III profile closely matched the mean profile for the total sam-
ple, with a clinically significant elevation on Scale Y, but there were
no other significant elevations. For females, the mean MCMI-III pro-
file consisted of clinically significant elevations on Scales Y and 4
(Histrionic), and an elevation on Scale 7 (Compulsive) that approached
clinical significance.
Given these observed differences between male and female pro-
files, statistical comparisons were made between MCMI-III scale
means for each of these two groups using t-tests for independent
groups. Because 27 scale comparisons were made, there was a risk of
making Type I errors in that at least one comparison could be expected
to be significant by chance alone. The required level of statistical
significance was therefore adjusted downward in order for a compari-
son to be considered significant by dividing the standard (p < .05)
level by the number of statistical tests conducted (i.e., 27). This re-
sulted in adoption of a more conservative level (p < .001) that was
required for statistical significance. Table 2 reports the t values for
each comparison between male and female MCMI-III scale means. On
scales 4, 5, and 7, females scored significantly higher than males. It is
important to note that although these differences were statistically
 34 JOURNAL OF FORENSIC PSYCHOLOGY PRACTICE

TABLE 2. Mean MCMI-III BR Scores of Child Custody Evaluation Examinees

Scale Males (n = 127) Females (n = 130) Total Sample (n = 259)

Mean S.D. Mean S.D. Mean S.D. t

X. Disclosure 30.45 15.50 32.73 16.64 31.64 16.10 1.14

Y. Desirability 75.83 13.50 75.64 11.73 75.56 13.06 0.12

Z. Debasement 20.39 19.28 25.70 23.78 23.10 21.87 1.97

Downloaded by [Northeastern University] at 18:10 03 January 2015

1. Schizoid 22.69 19.82 23.84 21.31 23.15 20.56 0.45

2A. Avoidant 19.07 19.65 23.26 21.18 21.05 20.53 1.64

2B. Depressive 20.89 22.81 17.82 17.18 19.25 20.16 1.22

3. Dependent 33.14 22.53 33.10 24.69 33.09 23.60 0.01

4. Histrionic 61.87 14.79 77.29 17.89 69.75 18.16 7.54***

5. Narcissistic 62.40 12.45 67.86 13.45 65.22 13.28 3.38***

6A. Antisocial 31.53 19.48 29.34 19.92 30.34 19.73 0.89

6B. Sadistic 27.48 22.79 29.30 24.17 28.44 23.56 0.62

7. Compulsive 63.61 12.91 72.97 18.88 68.37 15.70 5.00***

8A. Negativistic 16.78 16.19 18.64 18.66 17.73 17.48 0.85

8B. Self-Defeating 14.62 19.44 19.52 22.75 16.99 21.28 1.86

S. Schizotypal 16.02 20.10 16.52 20.30 16.34 20.22 0.20

C. Borderline 14.88 16.66 16.29 18.46 15.54 17.56 0.64

P. Paranoid 19.61 21.96 21.48 24.54 20.53 23.26 0.64

A. Anxiety 23.85 25.86 22.74 24.93 23.17 25.34 0.35

H. Somatoform 17.87 23.55 16.62 20.88 17.32 22.28 0.45

N. Bipolar: Manic 33.62 21.39 34.15 22.66 33.87 22.10 0.19

D. Dysthymic 15.49 19.92 12.36 15.70 13.86 17.92 1.40

B. Alcohol Dependence 25.50 20.25 24.05 25.54 24.68 23.04 0.50

T. Drug Dependence 29.98 21.03 29.36 22.74 29.54 21.89 0.23

R. Posttraumatic Stress 15.06 17.62 18.92 20.73 16.98 19.30 1.61

SS. Thought Disorder 15.02 19.60 13.90 16.85 14.39 18.21 0.49

CC. Major Depression 12.82 18.58 13.32 18.12 13.10 18.32 0.22

PP. Delusional Disorder 20.53 23.21 19.55 25.81 19.99 24.50 0.32

***p < .001

 Full-Length Article 35

significant, only Scale 4 (Histrionic) involved females scoring in the

clinically significant range (i.e., BR  75) and the mean for females
approached clinical significance on Scale 7 (BR = 72.97).
It is also interesting to note that with the exception of Scales Y, 4, 5,
and 7, the mean BR scores for all MCMI-III scales were extremely
low for both males and females. The typical scale had a mean BR
score that fell below BR = 35, which is well below the clinical median
level of BR = 60. It may be recalled that a BR score of 60 served as
Downloaded by [Northeastern University] at 18:10 03 January 2015

one of the anchor points for converting raw scores to BR scores during
development of the MCMI-III and this level represents the median raw
score from the MCMI-III normative sample (Millon, 1994). Also, a
BR score of 35 was observed to be the mean scale elevation for
nonclinical populations on the MCMI-II (Millon, 1987). Thus, child
custody examinees yielded mean MCMI-III scores on all scales, ex-
cept Y, 4, 5, and 7, that were at levels which are similar to those found
in nonclinical samples in the MCMI literature.
To examine the issue of whether or not the MCMI-III overpatholo-
gizes child custody examinees, we conducted a detailed examination
of the frequency with which subjects yielded clinically significant
elevations on each of the scales at varying BR score cutoffs. In Table
3, the frequency with which males, females, and the total sample
scored above or below a particular BR score is provided in the form of
percentages. In the first two columns of Table 3, the prevalence rate
data used to anchor MCMI-III BR scores at the 75 and 85 level are
listed for the presence and prominence, respectively, of personality
disorders and clinical syndromes. These values are noteworthy be-
cause the MCMI-III BR scores were calculated so that elevations at or
above a BR of 75 and 85 would occur at a frequency that is equal to
the prevalence rates at which a personality disorder or clinical syn-
drome was rated as present or prominent, respectively. If the MCMI-
III overpathologizes child custody examinees unfairly, then the fre-
quency of elevations at various BR scores should be higher than those
prevalence rates in the MCMI-III normative sample, assuming that
personality disorders and clinical syndromes do not occur at rates
greater than those observed in child custody evaluation settings as
opposed to a general clinical setting.
A survey of the results in Table 3 reveals that with the exception of
Scales Y, 4, and 7, the percentage of elevations above a BR of 75 and
85 are equal to or well below the prevalence rates cited in the MCMI-
 36 JOURNAL OF FORENSIC PSYCHOLOGY PRACTICE

TABLE 3. Percentage of MCMI-III Scale Elevations at Various BR Scores*

Scale BR  75 BR  85 BR $ 75 BR  75 BR  85 BR  90 BR  95 BR  100 BR  105

Prev.** Prev.**

X. Disclosure 28 12
Males 100 0 0 0 0 0 0
Females > 99 <1 <1 <1 <1 <1 0
Downloaded by [Northeastern University] at 18:10 03 January 2015

Total Sample > 99 <1 0 0 0 0 0

Y. Desirability 26 9
Males 47 53 18 11 4 3 0
Females 52 48 17 10 2 2 0
Total Sample 50 50 17 10 3 2 0
Z. Debasement 31 14
Males 100 0 0 0 0 0 0
Females 98 2 2 <1 0 0 0
Total Sample > 99 <1 <1 <1 0 0 0
1. Schizoid 19 5
Males 98 2 0 0 0 0 0
Females 98 2 <1 <1 <1 0 0
Total Sample 98 2 <1 <1 <1 0 0
2A. Avoidant 27 9
Males 97 3 0 0 0 0 0
Females 98 2 <1 <1 <1 0 0
Total Sample 97 3 <1 <1 <1 0 0
2B. Depressive 36 14
Males 97 3 0 0 0 0 0
Females 98 2 <1 0 0 0 0
Total Sample 97 3 <1 0 0 0 0
3. Dependent 48 14
Males 96 4 <1 <1 0 0 0
Females 91 9 5 4 3 <1 0
Total Sample 93 7 3 2 <1 <1 0
4. Histrionic 21 10
Males 81 19 8 6 0 0 0
Females 41 59 38 28 20 7 2
Total Sample 61 39 23 17 10 3 <1
 Full-Length Article 37

Scale BR  75 BR  85 BR $ 75 BR  75 BR  85 BR  90 BR  95 BR  100 BR  105

Prev.** Prev.**

5. Narcissistic 21 8
Males 84 16 4 2 <1 <1 0
Females 72 28 10 5 4 <1 0
Downloaded by [Northeastern University] at 18:10 03 January 2015

Total Sample 78 22 7 4 2 <1 0

6A. Antisocial 17 7
Males 98 2 <1 0 0 0 0
Females 98 2 <1 0 0 0 0
Total Sample 98 2 <1 0 0 0 0
6B. Sadistic 4 1
Males 97 3 0 0 0 0 0
Females 98 2 2 2 2 0 0
Total Sample 98 2 <1 <1 <1 0 0
7. Compulsive 21 4
Males 81 19 6 4 0 0 0
Females 47 53 21 8 3 <1 0
Total Sample 64 36 14 6 2 <1 0
8A. Negativistic 27 8
Males >99 <1 0 0 0 0 0
Females 98 2 0 0 0 0 0
Total Sample 98 2 0 0 0 0 0
8B. Self-Defeating 30 7
Males 98 2 <1 0 0 0 0
Females 94 5 2 0 0 0 0
Total Sample 97 3 2 <1 <1 0 0
S. Schizotypal 13 3
Males 100 0 0 0 0 0 0
Females 100 0 0 0 0 0 0
Total Sample 100 0 0 0 0 0 0
C. Borderline 27 10
Males 100 0 0 0 0 0 0
Females > 99 <1 <1 <1 <1 <1 0
Total Sample > 99 <1 <1 <1 <1 0 0
 38 JOURNAL OF FORENSIC PSYCHOLOGY PRACTICE

TABLE 3 (continued)

Scale BR  75 BR  85 BR $ 75 BR  75 BR  85 BR  90 BR  95 BR  100 BR  105

Prev.** Prev.**

P. Paranoid 13 4
Males > 99 <1 0 0 0 0 0
Females > 99 <1 0 0 0 0 0
Downloaded by [Northeastern University] at 18:10 03 January 2015

Total Sample > 99 <1 0 0 0 0 0

A. Anxiety 62 --
Males 89 11 <1 0 0 0 0
Females 90 10 <1 0 0 0 0
Total Sample 90 10 <1 0 0 0 0
H. Somatoform 16 --
Males > 99 <1 0 0 0 0 0
Females 100 0 0 0 0 0 0
Total Sample > 99 <1 0 0 0 0 0
N. Bipolar: Manic 9 --
Males 100 0 0 0 0 0 0
Females > 99 <1 <1 <1 <1 <1 0
Total Sample > 99 <1 <1 <1 <1 <1 0
D. Dysthymic Disorder 36 --
Males 98 2 0 0 0 0 0
Females 100 0 0 0 0 0 0
Total Sample 99 1 0 0 0 0 0
B. Alcohol Dependence12 --
Males 98 2 0 0 0 0 0
Females 100 0 0 0 0 0 0
Total Sample > 99 <1 0 0 0 0 0
T. Drug Dependence 8 --
Males 97 3 0 0 0 0 0
Females 97 3 2 2 0 0 0
Total Sample 97 3 1 <1 0 0 0
R. Posttraumatic Stress 9 --
Males > 99 <1 0 0 0 0 0
Females 100 0 0 0 0 0 0
Total Sample > 99 <1 0 0 0 0 0
 Full-Length Article 39

Scale BR  75 BR  85 BR $ 75 BR  75 BR  85 BR  90 BR  95 BR  100 BR  105

Prev.** Prev.**

SS. Thought Disorder 9 --

Males 100 0 0 0 0 0 0
Females > 99 <1 0 0 0 0 0
Total Sample > 99 <1 0 0 0 0 0
CC. Major Depression 22 --
Downloaded by [Northeastern University] at 18:10 03 January 2015

Males 100 0 0 0 0 0 0
Females > 99 <1 <1 <1 <1 0 0
Total Sample > 99 <1 <1 <1 <1 0 0
PP. Delusional Disorder 3 --
Males 100 0 0 0 0 0 0
Females 97 3 2 2 <1 0 0
Total Sample 98 2 <1 <1 <1 <1 0

*Totals may not equal 100% due to rounding.

**Data are from MCMI-III manual (Millon, 1994).

III manual. With respect to Scale 4 (Histrionic), males yielded eleva-

tions that were similar to the presence (i.e., BR  75) and prominence
(i.e., BR  85) rate of histrionic personality disturbances in the
MCMI-III normative sample that also occurred at the BR  75 and
BR  85 level, respectively. However, females yielded elevations on
Scale 4 at rates that exceeded the normative sample BR cutoff rates. It
was not until females, as a group, achieved BR scores of 95 or above
that the frequency of Scale 4 elevations was about equal to the preva-
lence rate for presence of histrionic personality disorder traits. A BR
score of 100 or above for the female group yielded Scale 4 elevation
frequencies that were about equal to the rate for prominence of
histrionic personality disturbances. Similarly, Scale 7 (Compulsive)
elevations for males yielded frequencies at BR  75 and BR  85 that
were essentially equivalent to those found in the MCMI-III normative
sample. However, females produced a slightly greater frequency of
Scale 7 elevations, with BR  85 and BR  95 reflecting frequencies
that were equivalent to those found in the MCMI-III normative sample
for presence and prominence of compulsive personality disorder traits,
respectively.
 40 JOURNAL OF FORENSIC PSYCHOLOGY PRACTICE

DISCUSSION

The findings from this normative study reveal that the mean
MCMI-III profile for child custody examinees was a Scale Y eleva-
tion, which reflects a socially desirable response set. There were sub-
clinical elevations on Scales 4, 5, and 7 in the entire sample and
moderate elevations on Scales 4 and 7 among females. These results
are not surprising, given that there are high correlations between Scale
Downloaded by [Northeastern University] at 18:10 03 January 2015

Y and Scales 4 (r = 0.62), 5 (r = 0.73), and 7 (r = 0.62) as reported in

the MCMI-III manual (Millon, 1994; Millon, Davis, & Millon, 1997).
Research also reveals that Scales 4, 5, and 7 correlate in the positive
direction with measures of emotional health and in the negative direc-
tion with measures of psychological disturbance or maladjustment
(Craig, 1997; Craig & Weinberg, 1993). In studies that have used
control groups of presumably well-adjusted individuals (e.g., clergy,
Air Force pilots, college students), Scales 4, 5, and 7 tend to be
elevated. Therefore, moderate elevations on these scales in child cus-
tody litigants are not necessarily indicative of personality pathology.
Moreover, the finding of a socially desirable response set as the
mean MCMI-III profile in child custody cases was consistent with
results from other studies on response patterns of child custody exam-
inees on standardized self-report instruments (Bathurst, Gottfried, &
Gottfried, 1997). Furthermore, the results of this study revealed that
very low BR scores across most MCMI-III scales is common in child
custody settings. Elevations on severe personality (i.e., Scales S, C,
and P) and severe clinical syndrome (i.e., Scales SS, CC, and PP)
scales are very infrequent occurrences. As such, the findings reveal
that in general the MCMI-III does not overpathologize individuals
taking the test during the course of a child custody evaluation.
Among females, there are two scales that are exceptions to this
general conclusion and these scales warrant caution when they are
interpreted in child custody evaluations. On Scales 4 and 7, females
produced a higher frequency of elevations. We are unaware of any
research that indicates histrionic and compulsive personality disorders
are more prevalent in child custody settings than they are in general
clinical settings. Therefore, to reduce the likelihood of false positive
errors in which a person is viewed as having pathological levels of
histrionic or compulsive personality traits, an adjustment in BR scores
may be warranted when interpreting Scales 4 and 7 for females in child
 Full-Length Article 41

custody cases. The normative data from this study suggest that for
females only, Scale 4 scores could be reduced by about 15 BR points
and that Scale 7 scores could be reduced by about 10 BR points. This
process of developing MCMI-III profile adjustments is the same used
by Millon in his development of the four profile adjustments based on
disclosure level, response style, personality style, and inpatient status
(Millon, Davis, & Millon, 1997). Since the MCMI-III scoring programs
do not currently take such adjustments into account in child custody
Downloaded by [Northeastern University] at 18:10 03 January 2015

cases, the clinician should remain aware that modest elevations on

Scales 4 and 7 are common in these cases and are not necessarily
indicative of personality pathology. Unadjusted elevations in the BR =
75 to 85 range could instead reflect socially desirable personality char-
acteristics associated with Scales 4 (i.e., sociability, makes friends easi-
ly, etc.), 7 (i.e., rule oriented, hard-working, conscientious, etc.), and to
a lesser extent Scale 5 (i.e., confidence, self-assurance, etc.). For this
reason, the MCMI-III interpretive reports should be used with extreme
caution in child custody settings. In any case, the adjustments noted
above can be used flexibly and interpretation of MCMI-III results
should integrate data from the persons history, other test results, collat-
eral reports, and observations of parent-child interactions.
It is also worth noting that the comparisons of MCMI-III scale
elevation frequencies for child custody examinees against the clinical
prevalence rates in the MCMI-III manual may be viewed by some
individuals as inappropriate. That is, some may argue that child custo-
dy evaluations are not clinical in nature. This issue is beyond the scope
of this paper, but we believe there are several reasons to view child
custody examinations as clinical assessments because stress and con-
flict generated over disputes involving custody of children frequently
produce an intensification of inflexible and maladaptive coping strate-
gies that may constitute an exacerbation of subclinical personality
disturbances. Moreover, highly conflicted and contentious child custo-
dy disputes place a burden on the clinician conducting such evalua-
tions to consider the possibility that psychopathology or clinical
symptoms may be present and their impact on parenting must be
considered. However, for those who do not subscribe to the view of
child custody evaluations being clinical in nature, we have provided
sufficient normative data in Tables 2 and 3 to permit clinicians to
define their own BR cutoffs based on prevalence rates that are be-
lieved to be more representative of child custody settings. Another
 42 JOURNAL OF FORENSIC PSYCHOLOGY PRACTICE

comparable strategy would be to uniformly adopt a more conservative

BR cutoff (e.g., BR  85 or BR  90 without adjustments) for those
scales associated with a socially desirable response set, including
Scales Y, 4, and 7 and to a lesser extent Scale 5.
Results from this study provide empirical support for the position
that the MCMI-III does not overpathologize individuals who take the
test as part of a child custody evaluation. While Scales 4 and 7 tend to
over represent the severity of histrionic and compulsive personality
Downloaded by [Northeastern University] at 18:10 03 January 2015

traits in females, we suggest some adjustments that may reduce the

number of false positive diagnoses on these scales. Moreover, while
cross-validation of these findings is encouraged, there is support for
their generalizability in that subjects were selected from four different
states and were actually involved in a child custody dispute that was
being adjudicated.
The findings from this study also provide an empirical foundation
to support continued use of the MCMI-III in child custody evaluations
for those clinicians who feel that this instrument provides useful diag-
nostic information in such contexts. The MCMI-III can be useful to
screen for various personality disorders and clinical syndromes that
may have an impact on parenting or visitation. The presence or ab-
sence of psychopathology is not solely determinative of whether or
not a parent should ultimately be granted custody. Furthermore, the
clinician must integrate psychological test results with other data that
address these issues and we endorse the use of other assessment tech-
niques, in addition to the MCMI-III, that evaluate relevant issues in
child custody evaluations, such as those directly related to parenting
and the quality of the parent-child relationship. Toward that end, in the
future we plan to examine the relationship between MCMI-III results
and relevant external criteria, such as legal outcome and reports of
domestic violence.

REFERENCES
Ackerman, M. J. (1995). Clinicians guide to guide custody evaluations. New York:
John Wiley & Sons.
Ackerman, M. J. & Ackerman, M. C. (1997). Custody evaluation practices: A survey
of experienced professionals (revisited). Professional Psychology: Research and
Practice, 28, 137-145.
American Psychological Association. (1994). Guidelines for child custody evalua-
tion in divorce proceedings. American Psychologist, 49, 677-680.
 Full-Length Article 43

Bathurst, K., Gottfried, A. W., & Gottfried, A. E. (1997). Normative data for the
MMPI-2 in child custody litigation. Psychological Assessment, 9, 205-211.
Borum, R. & Grisso, T. (1995). Psychological test use in criminal forensic evalua-
tions. Professional Psychology: Research and Practice, 26, 465-473.
Craig, R. J. (1997). A selected review of the MCMI empirical literature. In T. Millon
(Ed.), The Millon inventories: Clinical and personality assessment (pp. 303-326).
New York: Guilford.
Craig, R. J. (1999). Testimony on the Millon clinical multiaxial inventory: Review,
commentary, and guidelines. Journal of Personality Assessment, 73, 290-304.
Downloaded by [Northeastern University] at 18:10 03 January 2015

Craig, R. J. & Bivens, A. (1998). Factor structure of the MCMI-III. Journal of

Personality Assessment, 70, 190-196.
Craig, R. J., Bivens, A., & Olson, R. (1997). MCMI-III-derived typological analysis
of cocaine and heroin addicts. Journal of Personality Assessment, 69, 583-595.
Craig, R. J. & Weinberg, D. (1993). MCMI: Review of the literature. In R. J. Craig
(Ed.), The Millon clinical multiaxial inventory: A clinical research information
synthesis (pp. 23-70). Hillsdale, NJ: Lawrence Erlbaum Associates.
Davis, R. D., Wenger, A., & Guzman, A. (1997). Validation of the MCMI-III. In T.
Millon (Ed.), The Millon inventories: Clinical and personality assessment (pp. 327-
359). New York: Guilford.
Davis, S. E. & Hays, L. W. (1997). An examination of the clinical validity of the
MCMI-III depressive personality scale. Journal of Clinical Psychology, 53,
15-23.
Dyce, J. A., OConnor, B. P., Parkins, S. Y., & Janzen, H. L. (1997). Correlational
structure of the MCMI-III personality disorder scales and comparisons with other
data sets. Journal of Personality Assessment, 69, 568-582.
Dyer, F. J. (1997). Application of the Millon inventories in forensic psychology. In
T. Millon (Ed.), The Millon inventories: Clinical and personality assessment
(pp. 124-139). New York: Guilford.
Dyer, F. J. & McCann, J. T. (2000). The Millon clinical inventories, research critical
of their forensic application, and Daubert criteria. Law and Human Behavior, 24,
487-497.
Georgia Psychological Association. (1990). Recommendations for psychologists
involvement in child custody cases. Atlanta, GA: Author.
Gould, J. W. (1998). Conducting scientifically crafted child custody evaluations.
Thousand Oaks, CA: Sage.
Halon, R. L. (1990). The comprehensive child custody evaluation. American Journal
of Forensic Psychology, 8, 19-46.
Heilbrun, K. (1992). The role of psychological testing in forensic assessment. Law
and Human Behavior, 16, 257-272.
Keilin, W. G. & Bloom, L. J. (1986). Child custody evaluation practices: A survey of
experienced professionals. Professional Psychology: Research and Practice, 17,
338-346.
McCann, J. T. & Dyer, F. J. (1996). Forensic assessment with the Millon inventories.
New York: Guilford.
Melton, G. B., Petrila, J., Poythress, M. G., & Slobogin, C. (1997). Psychological
 44 JOURNAL OF FORENSIC PSYCHOLOGY PRACTICE

evaluations for the courts: A handbook for mental health professionals and law-
yers (2nd ed.). New York: Guilford.
Millon, T. (1983). Millon clinical multiaxial inventory manual (3rd ed.). Minneapo-
lis, MN: National Computer Systems.
Millon, T. (1987). Manual for the Millon clinical multiaxial inventory-II (MCMI-II).
Minneapolis, MN: National Computer Systems.
Millon, T. (1994). Millon clinical multiaxial inventory-III (MCMI-III). Minneapolis,
MN: National Computer Systems.
Millon, T., Davis, R., & Millon, C. (1997). MCMI-III manual (2nd ed.). Minneapolis,
MN: National Computer Systems.
Downloaded by [Northeastern University] at 18:10 03 January 2015

Nebraska Psychological Association. (1986). Guidelines for child custody evalua-

tion. Lincoln, NE: Author.
New Jersey State Board of Psychological Examiners. (1993). Specialty guidelines for
psychologists in custody/visitation evaluations. Newark, NJ: Author.
Oklahoma Psychological Association. (1988). Ethical guidelines for child custody
evaluation. Oklahoma City, OK: Author.
Otto, R. K. & Butcher, J. N. (1995). Computer-assisted psychological assessment in
child custody evaluations. Family Law Quarterly, 29, 79-96.
Rogers, R., Salekin, R. T., & Sewell, K. W. (1999). Validation of the Millon clinical
multiaxial inventory for Axis II disorders: Does it meet the Daubert standard?
Law and Human Behavior, 23, 425-443.

RECEIVED: 12/01/99
REVISED: 03/28/00
ACCEPTED: 04/25/00