Psychology
2011. Vol.2, No.6, 638-642
Copyright © 2011 SciRes. DOI:10.4236/psych.2011.26098
The Utility of MMPI-2 Scores with a Correctional
Population & Convicted Sex Offenders
Brianna Leigh Grover
Wright State University School of Professional Psychology, Dayton, OH, USA.
Email: grover.8@wright.edu
Received June 6th, 2011; revised August 11th, 2011; accepted September 2nd, 201 1.
The MMPI-2 is a widely used objective personality measure across all settings. It is especially useful in correc-
tional settings due to its objectivity with standardized administration and scoring This helps aid many factors in-
cluding volume, legal accountability, the nature of the clientele, and the need for security. Since assessments
provide very useful information about prisoners for correctional staff, these factors make the MMPI-2 a very
valuable test to use in prison. Studies have found no significant differences between African Americans and
Caucasians, however have found subtle differences between MMPI-2 scores of males and female offenders. One
specific area the MMPI-2 has been used for in prison is with sex offenders. Previous studies have aimed to use
the MMPI-2 to identify high and low risk sex offenders, as well as differentiate between general sex offenders
and internet sex offenders. Not only has it been used to identify certain sex offenders, it has been used in exam-
ining the predictors of completion of sex offender treatment programs.
Keywords: MMPI-2, Correctional Settings, Sex Offende rs
Introduction of MMPI-2
The MMPI-2 is the most widely used psychological test in
the United States and is often used in other countries around the
world (Graham, 2006). This test and the original MMPI, have
been the most widely researched personality assessment in-
struments. They have a research base of over 15,000 published
books and articles that explore the various applications of the
tests (Butcher, 2006). The original Minnesota Multiphasic Per-
sonality Inventory (MMPI) was introduced as a means of un-
derstanding the psychological problems of patients in psychiat-
ric and medical settings in 1940 (Butcher, 2006). Since then,
the test has been revised with the final version of the revised
MMPI (MMPI-2) including 567 true/false items. The MMPI-2
is similar in most ways to the original MMPI, as one goal in the
creation of the revised version of the test was to not lose the
thick research base that had already been completed (Graham,
2006).
In order to produce meaningful MMPI-2 results, the test
taker must be able to read at a sixth-grade reading level to
comprehend the content of the items and be able to respond
appropriately (Graham, 2006). The information gained from the
MMPI-2 can be very useful in providing a convenient overview
of personality style for the individual being tested (Butcher,
2006). The test yields validity scale scores which provide im-
portant information about test-taking attitudes of the individual
(Graham, 2006). While providing us useful information, if the
scores on the validity scales are too extreme, it could invalidate
our test information. If it is decided that the profile is valid, the
next step is to look at the clinical scales. The clinical scales
give us information on personality and psychopathology on the
dimensions each scale assesses for (Graham, 2006). When ele-
vations are present on the clinical scales, more information can
be gained from the content scales. The Harris-Lingoes sub-
scales provide us rationally based information of a grouping of
items where the content is similar within a standard clinical
scale (Graham, 2006). Since this information comes from face-
valid items, we are able to say the individual specifically en-
dorsed items pertaining to this content matter.
The RC, PSY-5, and supplementary scales can also provide
us with information about the personality style of an individual.
The major advantage of looking at the restructured clinical (RC)
scales is that demoralization was factored out from the rest of
the scales. With demoralization factored out, the core compo-
nent of each original clinical scale was determined and then the
creation of the RC scales was formed (Graham, 2006). The
Personality Psychopathology Five (PSY-5) scales tap a broad
area of personality traits that influence normal functioning and
clinical problems (Harkness & Mcnulty, 2006). Finally, the
supplementary scales are ones that provide additional, more
specific information about a certain area of personality func-
tioning that may be applicable to a certain individual (Graham,
2006). Some of these include the Addiction Acknowledgement
Scale (AAS), Addiction Potential Scale (APS), Overcontrolled-
Hostility (O-H) Scale, and many more.
Use of the MMPI-2 in a Correctional Setting
One of the appealing things about the MMPI-2 is that it can
be used in a variety of settings including inpatient and outpa-
tient mental health settings, medical centers, employment
screenings, and correctional settings (Graham, 2006). There are
many differences between correctional and other settings.
Megargee (2006) discusses the four major differences as the
volume of cases, the degree of legal accountability required, the
nature of the offender population, and the consequent need to
maintain security.
Volume
As of December 31, 2009, state and federal correctional in-
stitutions housed 1,613,656 prisoners which was an increase of
3,897 prisoners from yearend 2008 (West, 2010). Offenders
who are detained or incarcerated in jails and prisons must be
screened upon entry. They are screened for a variety of possi-
B. L. GROVER 639
bilities including mental illness, developmental disabilities,
self-destructive tendencies, and their potential for dangerous or
violent behavior. Following this screening process, case ma-
nagers must determine which facilities and programs are best
suited for each individual’s specific needs (Megargee, 2006).
Research tells us that many offenders are emotionally disturbed
or mentally ill. The Bureau of Justice Statistics reported that
283,800 or 16% of all state and federal inmates could be con-
sidered mentally ill, and another 15% - 20% requires some sort
of mental health services at some point during their incarcera-
tion (Anno, 1991).
Legal Accountability
Prisoners must completely rely on the institution for their
health care which means the correctional health care system is
subject to external and even judicial review (Megargee, 2006).
Due to this, strict standards have been established for the as-
sessment and treatment of prisoners. These standards require
that prisoners have access to mental health services and place
emphasis on the need for assessment by capable and creden-
tialed mental health professionals. Assessments also should use
instruments that are reliable, valid, and suitable for correctional
settings (American Association for Correctional Psychology,
2000). Also, psychologists in a correctional setting may be
called on to testify in court regarding their evaluations so they
must be prepared to explain their findings to a judge or jury and
be able to defend their choice of tests as well as administration
and interpretation (Pope, Butcher, & Seelen, 2000).
The Nature of the Clientele
Obviously, the clientele in a correctional setting is going to
be different than in most other settings. The most significant
difference between corrections and other settings is probably
the antisocial nature of the individuals. In most settings, we can
assume that the individual being assessed will be cooperative
and will share the basic goal with the clinician of developing an
accurate picture of the person to help aid in treatment (Megar-
gee, 2006). This is not usually the case with prisoners, and they
may act deviant on purpose. They may do this because of the
authority nature a psychologist holds in prison and therefore,
the clinician may need to adapt administrations and interpreta-
tions of testing in this setting (Megargee, 2006).
Need for Security
The need to maintain security is one of the main goals of any
correctional institution. This affects correctional mental health
practice in a few ways. First, the physical safety of mental
health professi onal staff must be ma intaine d, which means test s
must be administered in a suitable setting by well-trained indi-
viduals that can maintain control of the testing situation. Also
the security of test data, reports, and assessment files must be
maintained so inmates should not be used to administer, score
or file test data. Finally, reports should not be left anywhere
that they could be seen or stolen by anyone including inmates
and cleaning crews (Megargee, 2006).
Reasons for Assessment in Correctional Settings
Most assessments in corrections focus on three issues that
include mental health screening, risk assessment, and needs
assessment (Megargee, 2006). The AACP (2000) standards for
mental health services in jails and prisons mandates that each
new intake must receive a mental health screening upon enter-
ing the facility and before being placed in a cell or holding area
with other inmates. This includes both a direct examination of
offenders as well as a review of their records. If the screening
shows any indication that an offender is in need of mental
health care, a more thorough examination is done. Next, risk
assessment is essential for correctional populations. It involves
evaluating the dangers the offender poses to the community at
large, the correctional staff, other inmates, and to themselves
(Megargee, 2006). Offenders must be assigned to the least-
restrictive facility that matches their needs for security and
control: open camps, low security, medium security, or maxi-
mum security facilities. Identifying the potential for violent
behavior is a very difficult and complex task for correctional
staff to take on, and one broad measure cannot be sure to cap-
ture the risk itself (Megargee, 2006). The final major assess-
ment purpose in correctional facilities is a needs assessment.
This assessment aims to determine each offender’s manage-
ment and treatment needs. The goal of treatment is to help of-
fenders become responsible and productive members of society
and their families once they re-enter the community (Megargee,
2006). Needs assessments can help identify those who would
benefit from mental health interventions as well as identify
those who may benefit from educational programming and
vocational training (Megargee, 2006).
Why Use the MMPI-2 in Prison Then?
Given all the information about how a correctional setting
differs from other settings and the main reasons for assessment
in a correctional setting, why is the MMPI-2 a good option for
correctional populations? The MMPI-2 can be administered in a
group setting where a lot of offenders can do the test at the
same time, which takes care of the number of cases issue and
then the need for security can be handled within this one ad-
ministration of a handful of offenders. Also when looking at
legal accountability, since the MMPI-2 is a widely used object-
tive personality assessment measure, it is not difficult for cor-
rectional psychologists to justify its use with criminal offenders
and may actually be more difficult to defend omitting it from a
battery (Megargee, 2006). The MMPI-2 has a standard set of
items that are administered and scored in a standardized manner,
which leaves little subjectivity for the test (Butcher, 2006). Due
to the nature of offenders in correctional settings, the clinician
may need to adapt administrations and interpretations of testing
(Megargee, 2006). The MMPI-2 does a very good job at iden-
tifying those who are mentally ill or emotionally disturbed.
Since there is such a high number of people being brought into
the jails at a ll times of the day and night, t his volume of people
and limited resources available generally does not allow for the
use of the MMPI-2 in the initial screening process. However,
once a mental health screening shows indication of mental
health problems, the MMPI-2 can aid in the more thorough
mental health assessment (Megargee, 2006). When assessing
the potential for violent behavior, a very difficult and complex
task, personality tests such as the MMPI-2 are not very well
suited (Heilbrun & Heilbrun, 1995). So although Megargee’s
MMPI-2 offender classification system (Megargee et. al, 2001)
can separate more predatory offenders from those likely to be
victimized, the MMPI-2 is best used along with criminal his-
tory data and more narrowband instruments specifically fo-
cused on dangerous behavior when assessing risk. Finally, one
very useful part of the MMPI-2 that is especially relevant for
correctional settings are the validity scales designed to detect
deviant test-taking attitudes including defensiveness and ma-
lingering. These s ca les can help psychologists detect motivation
of an offender to present themselves in unrealistically positive
B. L. GROVER
640
and negative ways (Butcher, 2006).
Using the MMPI-2 with Ethnic Minorities in a
Correctional Setting
Several studies have addressed the possibility of MMPI-2
differences between African-Americans and Caucasians in cor-
rectional populations, but not much has yet been done on any
other minority group. One study looked at the association be-
tween race and MMPI-2 scores in African-American and Cau-
casian men who underwent court-ordered forensic psychologi-
cal evaluations and found that the two groups of men produced
highly comparable MMPI-2 scores. They found that there was
no significant difference on any of the validity or clinical scale
scores, but did find that the African-American men scored
higher than Caucasians on the Cynicism and Antisocial Prac-
tices content scales (Ben-Porath, Shondrick, & Stafford, 1995).
Other studies have produced similar results and although there
is only limited data comparing characteristics of the majority
and minority groups in correctional populations, Butcher (2006)
suggests that data from other settings would tell us that MMPI-
2 scores will be equally accurate for Caucasian and African
Americans in a correctional setting. It is important to note that
while there have been some studies done comparing African
American and Caucasian MMPI-2 scores in correctional set-
tings, there are also other minority groups in prisons including
Hispanics, American Indians, and Asian-Americans that should
be addressed.
Using the MMPI-2 with Different Genders in a
Correctional Setting
Megargee, Mercer, and Carbonell (1999) examined the dif-
ferences between male and female prisoners on the MMPI-2.
One of the three goals of the study was to compare MMPI-2
scores of male and female offenders. One difference they found
was on scale 5 where female offenders scored high and male
offenders scored low. This means that both groups, males and
females, scored in the masculine direction. Other than that dif-
ference, on the standard, content, and supplementary scales,
male and female profile patterns tended to be very close to one
another. However, one major difference between males and
females that was found is that the effect sizes for female of-
fenders were significantly higher than male offenders. This tells
us that female offender MMPI-2 scores deviated from the
norms more than male offender scores. Since women represent
less that 5% of the prison population, it makes sense that it is
more socially deviant for a woman to be incarcerated than a
man (Megargee, Mercer, & Carbonell, 1999).
Using the MMPI Profile to Identify Sex
Offenders
Davis and Archer (2010) conducted a review of the literature
to distinguish sex offenders from non-offender control groups
as well discriminate sex offenders from other types of offenders
based on objective personality measures. Most of the research
has been conducted on the MMPI, and their review updated the
literature and included the few studies that used the PAI and
MCMI-III, as well as the MMPI-2. Based on their review of 37
articles, they found that 33 of the articles utilized the MMPI to
assess psychopathological characteristics of sex offenders. In
all the studies reviewed, none of the validity or basic clinical
scales were found to have moderate or large effect sizes except
for the Pd scale. Graham (2006) notes that the Pd scale looks at
constructs involved with antisocial behaviors that include lying,
cheating, stealing, use of substances, and sexual acting out.
Davis and Archer (2010) pointed out that there are several rea-
sons why Pd scale elevations were found in the research they
reviewed with one major reason being that Pd elevations could
be reflective of general antisocial or criminal behaviors and not
specific characteristics of sex offenders. They also point out
that a higher Pd score could suggest the chronicity of criminal
behavior. In other words, the more times someone is arrested,
the higher that score would be elevated. Again, the Pd score
would be suggesting general criminal behavior and nothing
specific to sex offenders.
Since the Pd scale has five Harris & Lingoes subscales,
studies have been conducted looking at the subscales with sex
offenders, although very few. Results of the Panton (1978)
study found that offenders with no priors score lower on Au-
thority Problems subscale as well as the Sc basic scale. The
study also found a difference between child molesters, child
rapists, and adult rapists on the Pd subscales. Both of the rapists
groups scored similarly, endorsing items associated with Social
Alienation and Authority Problems subscales, while child mo-
lesters endorsed items associated with the Self-Alienation and
Familial Discord subscales. Davis & Archer (2010) noted a
major limitation when reviewing literature for their study was
that the available literature on sex offenders overwhelming
focuses on male Caucasian offenders, so this should be taken
into account when looking at the findings of their literature
review and other studies conducted.
Using the MMPI-2 to Identify High and Low
Risk Sex Offenders
Coxe and Holmes (2009) aimed to distinguish between high
and low risk sex offenders on 26 variables that included things
such as intelligence, age, criminal history, victim’s age, denial
patterns, measure sexual interest in children, admission of mea-
sure sexual interest, childhood history of sexual abuse, and
personality variables. They used a variety of measures includ-
ing the MMPI-2 for their study and found that only four of the
26 variables significantly predicted membership in the high-risk
group. These four variables were age, prior number of felonies,
the cognitive distortion score, and the MMPI-2 infrequency
scale score. When looking at the MMPI-2 results, Coxe and
Holmes (2009) found that the high-risk group scored signifi-
cantly higher than the low-risk group with an average score of
69 compared to 51. Graham (2006) notes that a high infer-
quency score could be an indication of persons who are very
psychologically disturbed. Coxe and Homes (2009) also found
differences between the clinical scales of high and low risk
offenders. For the low-risk group, they found no clinical scale
averages above the standard score of 65. However, for the high-
risk group, they found three scales elevated above the standard
score of 65 which were scale 4, 6, and 8. They concluded then
that the high-risk group is associated with a greater degree of
psychological maladjustment than the low-risk group.
Using the MMPI-2 to Identify General Sex
Offenders from Internet Sex Offende rs
Tomak, Weschler, Ghahramanlou-Holloway, Virden, and
B. L. GROVER 641
Nademin (2009) conducted a study aimed at comparing the
differences between general sex offenders and internet sex of-
fenders using the MMPI-2. In one section of their results, they
focused on demographic differences between the two groups of
offenders. They found that ISOs were more likely to be Cauca-
sian while GSOs were more likely to be Hispanic or African
American. They also noted that ISOs were more likely to be
married than GSOs who were more likely to have never been
married or widowed. In regards to education differences be-
tween the two groups, ISOs were more likely to have attempted
college while GSOs were more likely to have come close to
completing high school. It is important to note however, that
the GSOs were incarcerated while the ISOs were at a private
facility which may make a difference on demographic variables,
especially the ethnic group finding. Also, this study only in-
cluded 121 people, which should be kept in mind when exam-
ining the demographic differences between the two groups of
sex offenders.
When comparing the validity and clinical scales of the two
groups, only four scores provided significant differences be-
tween the ISOs and GSOs. There were significant differences
found on the L and F validity scale scores in which the ISOs
scored significantly lower than the GSOs. When looking at the
clinical scale scores, the ISOs again scored significantly lower
than the GSOs on the Pd and Sc scales. The authors conclude
that although there are behavioral differences between GSOs
and ISOs, they have a similar goal of sexual gratification and
are heterogeneous in terms of personality characteristics. They
do not feel there is a personality profile common to the present
ISO sample or the GSO sample. However, they do note the
significantly lower scores of ISOs than GSOs on the L, F, Pd,
and Sc scales and suggest there are some differences that obvi-
ously exist between the two groups (Tomak, Weschler, Ghah-
ramanlou-Holloway, Virden, & Nademin, 2009).
Using the MMPI with Sex Offender Treatment
Programs
There has been little research done on factors that reduce re-
cidivism in sexual offenders. Some people may believe that sex
offenders cannot change; however, more often than not sex
offenders are released back into the community at some point in
their lifetime, and it can be beneficial to try to identify some
factors to help with treatment. Geer, Becker, Gray, and Krauss
(2001) aimed to identify predictors of treatment completion of
sex offenders serving prison sentences. They identified two
groups in their study as completers and non-completers. They
used a variety of assessment measures prior to voluntary par-
ticipation in a sex offender treatment program and then com-
pared the groups following the program. One measure they
used in the battery was the MMPI. Although they found sig-
nificant differences between the two groups on the L and PD
scales, the differences did not stand a statistical regression, and
they therefore did not place much emphasis on them. However,
with these results, comparisons are able to be made between
offenders who complete treatment and those who don’t, with
those who do not complete treatment having higher L and PD
scores on the MMPI.
Another study conducted by Miner, Marques, Day, Nelson
(1990) examined the Sex Offender Treatment and Evaluation
Project (SOTEP) that the California Legislature funded in 1984.
Although an older study, results provided valuable information
that can be used in future program development. The project
involved a pre-post assessment of change from intake to dis-
charge. In using the MMPI as one assessment measure, re-
searchers found significant changes in the Si, or social introver-
sion scale. This difference provides evidence that those who are
participating in the treatment program are learning to accept
responsibility and consequences of their behavior. The re-
searchers expected two results from this evidence. First, they
believed this change would allow the participants to have more
confidence in their own ability to positively change their lives
and second the acceptance of responsibility and consequences
of their behavior should decrease the likelihood of antisocial
behaviors.
Conclusions
The MMPI-2 is the most widely used objective personality
measure in all settings. It is especially useful in correctional
settings due to its objectivity with standardized administration
and scoring. In evaluating its use with sex offenders, it seems
that certain scales including the L, F, Pd, and Sc scales are ele-
vated with this population of offenders. However recent re-
search is differentiating between sex offender typologies and
finding that there are differences between groups of sex of-
fenders on their MMPI-2 scores. This information tells us that
although sex offenders may have the same goals but different
behaviors for meeting those goals, they also may have different
personality styles in general. When looking at the treatment of
sex offenders, there is not ne treatment program that will help
reduce recidivism across all offenders due to the different ty-
pologies and personality styles across these offenders. However,
some research shows differences in scores on the MMPI-2 in
regards to completion of a treatment program as well as which
scores change after successfully completing a treatment pro-
gram that may help aid in reducing recidivism for these offend-
ers.
References
American Association for Correctional Psychology, Standards Com-
mittee (2000). Standards for psychological services in jails, prisons,
correctional facilities, and agencies. Criminal Justice and Behavior,
27, 433-494.
Anno, B. J. (1991). Prison health care: Guidelines for the management
of an adequate delivery system. Chicago: National Commission on
Correctional Health Care.
Ben-Porath, Y. S., Shondrick, D. D., & Stafford, K. P. (1995). MMPI-2
and race in a forensic diagnostic sample. Criminal Justice and Be-
havior, 22, 19-32.
Butcher, J. N. (2006). Pathways to MMPI-2 use: A practitioner’s guide
to test usage in diverse settings. MMPI-2: A practitioner’s guide (pp.
3-13). Washington, DC: American Psychological Association.
Coxe, R., & Holmes, W. (2009). A comparative study of two groups of
sex offenders identified as high and low risk on the static-99. Journal
of Child Sexual Abuse: Research, Treatment, & Program Innova-
tions for Victims, Survivors, & Offenders, 18, 137-153.
doi:10.1080/10538710902743925
Davis, K. M., & Archer, R. P. (2010). A critical review of objective
personality inventories with sex offenders. Journal of Clinical Psy-
chology, 66, 1254-1280. doi:10.1002/jclp.20722
Geer, T. M., Becker, J. V ., Gray, S . R., & Krauss, D. (2001). Predictors
of treatment completion in a correctional sex offender treatment
program. International Journal of Offender Therapy and Com-
parative Criminology, 45, 302-313.
Graham, J. R. (2006). MMPI-2: Assessing personality and psychopa-
thology (4th ed.). New York, NY: Oxford University Press.
Harkness, A. R., & McNulty, J. L. (2006). An overview of personality:
B. L. GROVER
642
Using the MMPI-2 personality psychopathology five (psy-5) scales.
MMPI-2: A practitioner’s guide (pp. 73-97). Washington, DC: Ame-
rican Psychological Asso c i ation.
Heilbrun, K., & Heilbrun, A. (1995). Risk assessment with MMPI-2 in
forensic evaluations. In Y.S. Ben-Porath, J. R. Graham, G. C. N. Hall,
R. D. Hirschman, & M. S. Zaragoza (Eds.), Forensic applications of
MMPI-2 (pp. 160-178). Thousand Oaks, CA: Sage.
Megargee, E. I. (2006). Use of the MMPI-2 in correctional settings.
MMPI-2: A practitioner's guide (pp. 327-360). Washington, DC:
American Psychological Association.
Megargee, E. I., Carbonell, J. L., Bohn, M. B., Jr., & Sliger (2001).
Classifying criminal offenders with the MMPI-2: The Megargee
system. Minneapolis: University of Minnesota Press.
Megargee, E. I., Mercer, S. J., & Carbonell, J. L. (1999). MMPI-2 with
male and female state and federal prison inmates. Psychological As-
sessment, 11(2), 177-185. doi:10.1037/1040-3590.11.2.177
Miner, M. H., Marques, J. K., Day, D. M., & Nelson, C. (1990). Impact
of relapse prevention in treating sex offenders: Preliminary findings.
Annals of Sex Research, 3, 165- 185. doi:10.1007/BF00850869
Panton, J. H. (1978). Personality differences appearing between rapists
of adults, rapists of children and non-violent sexual molesters of
female children. Research Communications in Psychology,
Psychiatry & Behavior, 3(4), 385-393.
Pope, K. S., Butcher, J. N., & Seelen, J. (2000). The MMPI, MMPI-2,
and MMPI-A in court: A practical guide for expert witness and
attorneys (2nd ed.). Washington, DC: American Psychological Asso-
ciation.
Tomak, S., Weschler, F. S., Ghahramanlou-Holloway, M., Virden, T.,
& Nademin, M. E. (2009). An empirical study of the personality
characteristics of internet sex offenders. Journal of Sexual Aggres-
sion, 15, 139-148. doi:10.1080/13552600902823063
West, H. C. (2010). Prisoners at yearend 2009- advanced counts.
Retrieved January 29, 2011 from US Bureau of Justice Statistics,
Washington, DC.
http://bjs.ojp.usdoj.gov/index.cfm?ty=pbdetail&iid=2272