Criminal Personality

Page 1 of 17




















Criminal Personality

Lori Kunch

Capella University

Master of Psychology

Fall 2009
















Criminal Personality

Page 2 of 17

Abstract

The science of predicting criminal recidivism has improved dramatically in the past few decades,
and there is far less emphasis on unstructured clinical judgment. There are now several actuarial
instruments in use that predict the risk of an offender committing a new offense of any kind and
a new violent and/or sexual offense in particular. Research has supported the use of these
instruments with people of various ages and races. Historically, risk assessment instruments have
been comprised of static information available on file (e.g., Statistical Information on
RecidivismRevised scale) or that require corroboration from collaterals or files, or information
that is received from the offender in an interview. Pen-and-paper self-report measures have been
considered unreliable because of their vulnerability to deception. This paper outlines the clinical
testing methodologies and the psychometrics used for predicting criminal recidivism.































Criminal Personality

Page 3 of 17


Introduction

The Minnesota Multiphasic Personality Inventory2 (MMPI-2) was developed originally by
Hathaway and McKinley (1943) to aid in providing diagnoses for medical and psychiatric
patients. However, shortly after its introduction, it became apparent that psychiatric patients
often elevated on more than one clinical scale, making it difficult to render clear and accurate
diagnoses from the test in isolation. Rather than abandoning its use, the approach to MMPI
interpretation was modified, focusing on an examination of the empirical correlates of the tests
clinical scales and code-types (configural patterns of elevations on the clinical scales). In
conjunction with data from other sources (e.g., relevant history and observation), the utilization
of empirical correlates allows the MMPI-2 to be used in psychological treatment formulation,
application, modification, and evaluation. Such an approach aids the clinician in understanding
the clients problems and weaknesses, as well as his or her resources and strengths (Forbey, &
Ben-Porath, 2002). The Personality Assessment Inventory (PAI) was designed to provide useful
assessment information in a wide variety of clinical contexts that translate smoothly to forensic
settings. The PAI provides information that can aid in offender classification, treatment planning,
and risk assessment. The PAI is a 344-item questionnaire that requires roughly 1 hour to
complete; the respondent is asked to check one of four response options indicating the extent to
which the item statement accurately describes him or her. The wording of these statements are
typically simple; the language is written at a fourth-grade level, and this reading level appears to
be among the lowest of comparable proceduresan important consideration in working in
forensic settings where educational levels may be lower than found in the general population
Criminal Personality

Page 4 of 17

(Morey, & Quigley, 2002). The MMPI and Personality Inventories can effectively predict
recidivism rates among juvenile and adult offenders.

Psychometric Qualities

Lack of trust in our social institutions and the concern of minorities have been reflected in
criticisms of the tests. Given the nature of the original sample that MMPI was validated on,
questions have been raised as to whether the instrument may be biased against certain ethnic and
racial groups. Reliability generalization (RG) is a meta-analytic technique that allows for the
systematic examination of variation in score reliability for different samples of test takers; this
procedure is based on the recognition that reliability is not a stable property of a test but is
sample dependent. Sample size was not significantly correlated with reliability estimates. RG
methods have the potential for deepening the level of understanding about the role of reliability
in the evaluation and use of personality tests. The measurement meta-analytic method called
"reliability generalization" to examine (a) the typical reliability, (b) the variability of score
reliability, and (c) the factors predicting variability of score reliability on the three MMPI
validity scales. Across studies the reliability coefficients on the three validity scales ranged from
.16 to .93, with the F scale having the most variability, as might be expected. Regression
analyses suggested that the age of participant best predicted the variation of reliability
coefficients (McCusker, 2007).
Test reliability

One of the most frequently administered psychometrics is the Minnesota Multiphasic Personality
Inventory-2 (MMPI-2). Occasionally, those participants taking the MMPI-2 will malinger or
Criminal Personality

Page 5 of 17

exaggerate their symptoms. Several malingering detection devices are available, and a significant
body of literature exists concerning their efficacy. However, little research is available
considering those factors that facilitate successfully evading detection as a malingerer. Some of
these studies have identified general intelligence and knowledge of the MMPI-2 as key variables
in the likelihood of escaping detection as a malingerer. The extant research considered the utility
of general intelligence and knowledge of the MMPI-2 as predictors in avoiding detection as a
malingerer. To detect malingering, the two traditional detection devices were employed: the F-
Scale and the F - K Index. Results indicate that intelligence and MMPI-2 knowledge contribute
significantly to the likelihood of successfully escaping detection as a malingerer (Pelfrey, 2004).


Reliability Indices of MMPI

The measurement meta-analytic method called

"reliability generalization" is used to examine (a)
the typical reliability,

(b) the variability of score reliability, and (c) the factors

predicting
variability of score reliability on the three MMPI

validity scales. Across studies the reliability co-
efficient

on the three validity scales ranged from .16 to .93, with the

F scale having the most
variability, as might be expected. Regression

analyses suggested that the age of participant best
predicted

the variation of reliability co-efficient (Gacono, 2002).

Evaluations about Test Reliability of MMPI

According to Foley, et al, (2002), when the MMPI is used in the manner for which it was
designed and validated, it psychometrics properties are likely to be adequate for the clinical
research purposes. There are two issues related to the validity of MMPI-2 scores: incremental
Criminal Personality

Page 6 of 17

validity is supported if scaled scores provide information about a person's behavior, personality
features, where psychopathology features that is not provided by other measures. As for the
validity of cutoff scores, it is important to keep in mind that the optimal cutoff scores will vary
depending on the nature of the population of the patients sampled. MMPI-2 cutoffs were derived
using the distribution of scores from the normative sample. Therefore, these cutoffs may or may
not be appropriate in certain clinical context. A potential problem itself reported inventories,
including the MMPI-2, is their susceptibility to distortion through various test-taking attitudes or
responds sets. For example, some respondents may wish to place themselves at a favorable light;
other may fake bad to increase the likelier to receive aid. Obviously, if the condition is not
aware of these responses out in a given patient, the test interpretation can be a gross error. To
help detect malingering-faking bad- other response sets or tests-taking attitudes, and carelessness
or misunderstanding, the MMPI-2 continues to incorporate the traditional for validity scales that
were included in the original MMPI.

1. Scale. This is the number of items left unanswered.
2. F scale. The 60 items were seldom answered in the scored direction by the standardization
group. A high F score may suggest deviant behavior, or other hypothesis about extra tests
characteristics or behaviors.
3. L (lie) scale. This includes 15 items who's endorsed places the respondent of very positive
light.
4. K (Defensiveness) Scale. PS 30 items suggest defensiveness and admitting certain problems.
These items reportedly detect faking good.
5. Fb (Back-page Infrequency) Scale. These 40 items are scoring in the end of the test are
infrequently endorsed.
6. VRIN (Variable Response Inconsistency) Scale. This consisted of 67 pairs of items with
either similar opposite content. High BRIN scores suggest random responding.
7. TRIN (True Response Inconsistency) scale. This consists of 23 item pairs that are opposite in
content. High TRIN White scores suggest a tendency to give true responses indiscriminately; low
TRIN scores suggest a tendency to give false responses indiscriminately.
These seven MMPI-2 validity scales provide a means for understanding the test respondence
motivation and test-taking attitudes.
Criminal Personality

Page 7 of 17



Validation Process of MMPI

Among the most successful malingering detection devices are psychometrics, particularly those
with validity scales. Chief among these psychometrics are the Hare Psychopathy Checklist
(PCL) and the MMPI-2. Whereas the Hare PCL relies on an interview process for information
acquisition, the MMPI-2 is a self-administered, purely objective device. In addition to its 10
clinical scales, 3 validity scales accompany the MMPI-2: the L, F, and K Scales. Although useful
in identifying test takers who are attempting to mask personality flaws, the L Scale is not useful
in detecting sophisticated malingering attempts or in identifying those who are given instructions
to malinger.The F Scale, the K Scale, and the F K figure are the principal tools employed to
detect malingering attempts (Pelfrey, 2004).
The MMPI-2 has a number of validity scales, designed to measure an individuals approach to
test taking. The four original MMPI validity scales, Cannot Say (CNS), Infrequency (F), Lie (L),
and Defensiveness (K), were all maintained for use with the MMPI-2. In addition, several new
validity scales, Variable Response Inconsistency (VRIN), True Response Inconsistency (TRIN),
Infrequency-Back (Fb), Infrequency-Psychopathology (F[p]), and Superlative (S), have also been
developed for use in determining a test-takers approach to the MMPI-2. The scales are reviewed
here in the order they appear in the recently updated MMPI-2 materials.
Evaluations about Test validity of MMPI and Personality Tests
Although the F Scale is the single most useful detection device it is not perfect. Black
participants tend to elevate the F Scale, and suggested that those experiencing moderate stress or
young participant populations are likely to evidence elevations. Thus, measures other than the F
Criminal Personality

Page 8 of 17

Scale are necessary to delineate genuine reports and malingering attempts. Assessment and
testing play critical roles in the evaluation of criminal populations. Although psychometrics are
only a portion of a strong assessment process (which should, among other steps, include a review
of the participants medical, criminal, and psychological history; an interview; and a medical or
toxilogical screening), psychological tests represent one of the most valuable tools in identifying
those individuals who attempt to misrepresent their symptoms. Participants misrepresent their
symptoms in one of two waysfaking good and faking bad. Dissimulating, or faking good, may
be done to cover existing symptoms and avoid treatment, portray a better-than-reality image of
oneself (e.g., to gain parole or release), or avoid a commitment hearing. The converse is
malingering, or faking bad. Participants might malinger, or exaggerate their symptoms, for a
variety of reasons. They may be seeking a private cell or better treatment in a prison, they may
be trying to obtain benefits (e.g., a veteran faking post-traumatic stress disorder), or they may be
trying to avoid responsibility for a crime. Several indexes of faking strategies: These are (a) rare
symptoms, (b) symptom severity, (c) obvious versus subtle symptoms, and (d) symptom
selectivity. Participants usually fake along one of these indexes or occasionally adopt an
erroneous stereotype and adopt that perspective (Walters, 2006).

The Appropriateness of Test Norms

Standardization sample of the Minnesota Multiphasic Personality Inventory2 (MMPI-2) is
widely considered to be the most frequently administered personality inventory. To identify
those participants who misrepresent themselves on the MMPI-2, a series of validity and detection
devices exist. A significant body of literature exists concerning the identification of participants
who malinger, or fake bad, on the test. However, conflict exists regarding which, if any, of the
Criminal Personality

Page 9 of 17

standard malingering detection devices effectively define those individuals attempting to
exaggerate symptoms. Effectively identifying those participants who malinger is a critical
concern to forensic psychologists, counselors in corrections and mental health settings, and
clinicians, because these individuals are often seeking preferential treatment and may
unnecessarily exhaust limited resources. The standard malingering detection devices for
identifying malingering on the MMPI-2 are the F Scale (one of the three validity scales including
the L and K Scales) and the F K Index. In addition to these commonly used approaches, there
are a wide variety of other malingering detection devices including analysis of subscales, review
of specific item sets, and simple profile review. A number of researchers have considered the
validity of these various approaches (Pelfrey, 2004).


Evaluation about Norm Process of MMPI

Although the F Scale, K Scale, and F K Index are the most commonly employed malingering
detection devices, a number of alternative strategies are available with varying degrees of
success. Based on the notion that some MMPI-2 questions very clearly tap pathology and others
do not, Weiner and Harmon developed the Obvious and Subtle Item Indexes. The Subtle-
Obvious Scales detect malingering attempts at no better than chance. Despite this assertion,
several studies report the utility of this indicator. The general strategy underlying this technique
is to subtract the total number of standardized subtle scores from the total number of
standardized obvious scores. AT score cutoff of 160 is commonly used with 79% and a 53%
effectiveness rate utilizing this technique. A secondary strategy employing the Subtle-Obvious
Scales is to subtract the subtle T score from the total number of obvious responses.
Criminal Personality

Page 10 of 17

Consequently, generating 81% accuracy with this approach in evaluating clinical populations
known to be malingering (McCusker, 2007).

Summary

Although there is a number of malingering detection devices, there is limited consensus
regarding their veracity and utility. There is a wide range of accuracy among the various devices,
and most of the research considering the effectiveness of these devices fails to consider those
factors that contribute to successfully malingering. A review of the malingering and MMPI-2
literature suggests several viable research questions. The extant literature indicates that several
factors, particularly knowledge of the test and general intelligence, predict successfully escaping
detection as a malingerer. One may reasonably ask to what degree these two variables contribute
to escaping detection as a malingerer. It should be noted that, heretofore, studies involving
coached participants or participants knowledgeable in the MMPI-2 have not measured the degree
of that knowledge. The following research methodology describes how these research questions
may be addressed.

Evaluation of the Use of MMPI for Diagnosis of Criminality

According to Folsom & Atkinson (2007), the science of predicting criminal recidivism has
improved dramatically in the past few decades, and there is far less emphasis on unstructured
clinical judgment. There are now several actuarial instruments in use that predict the risk of an
offender committing a new offense of any kind and a new violent and/or sexual offense in
particular. Research has supported the use of these instruments with men of various ages and
races. Their use has even been supported with men diagnosed with a personality or other mental
Criminal Personality

Page 11 of 17

disorder. Despite there being too few to count, incarcerated women have recidivism rates
varying from 22% to 48% for follow-up periods of 2-5 years. In a study of 67 Canadian female
offenders serving sentences of more than 2 years, found that 43% were convicted of a new
offense within 2 years of their release. Further, 48% of a sample of 50 Canadian female
offenders serving sentences of more than 2 years reoffended within 3 years of release. Using a
longer follow-up period and a broader definition of recidivism (i.e., readmission to custody for
any reason, including technical violations of their release conditions without a new offense), and
that 22% of a sample of first-time Canadian female offenders serving sentences of more than 2
years violated their release conditions or reoffended. Thus, recidivism among female offenders is
a significant problem.
The utility of risk assessment instruments that have been developed for male offenders, for
female offenders, is currently unclear. The General Statistical Information on Recidivism
Scale, widely used in Canada with male offenders, has been successfully revalidated on men
with an overall correlation of .42 with recidivism. However, it does not fare as well in predicting
recidivism for female offenders. It yields lower correlations (r = .25) and an irregular pattern;
that is, the Good Risk Category yielded the highest rates of recidivism (Folsom, & Atkinson,
2007).
Because a self-report version of the MMPI has been shown to be effective in predicting
recidivism for some groups of female offenders, it was decided to test its usefulness on a sample
of women who were serving sentences of over 2 years. These women would be more serious
offenders than those previously studied. It was hypothesized, therefore, that the MMPI would
continue to be useful in the prediction of their recidivism, although the absolute scores would be
Criminal Personality

Page 12 of 17

higher. The finding that criminal history variables that define early onset (age at first arrest) and
persistent antisocial behavior (number of convictions) are the strongest predictors of recidivism
is not new. This is the same conclusion reached in the male offender literature. Several
instruments have been found to work well in the prediction of recidivism among men, and it is
those with items pertaining to antisocial behavior that fare the best. It is not surprising, therefore,
to find that the same theoretically relevant items perform best when predicting the future
antisocial behavior of women.
Given that the single self-reported history items outperformed both actuarial instruments, it is
worth questioning the need for such instruments, especially the longer pen and pencil types or
those based upon structured interviews. Across several samples, however, a scale with more
items would be expected to outperform these single items due to the increased reliability
afforded by multiple (overlapping) items. For psychometric reasons, therefore, actuarial
instruments made up of several of the variables known to be related to future offending would be
expected to be more stable and robust predictors of recidivism than any single variable alone
This is indeed the case in the well-established literature on the prediction of recidivism among
men (Folsom, & Atkinson, 2007).
Historically, risk assessment instruments have been comprised of static information available on
file (e.g., Statistical Information on RecidivismRevised scale) or that require corroboration from
collaterals or files, or information that is received from the offender in an interview. Pen-and-
paper self-report measures have been considered unreliable because of their vulnerability to
deception. The self-report format may be a limitation in that more accurate information may
have been obtained from corroborating information available in files. However, recent evidence
Criminal Personality

Page 13 of 17

supports the validity of self-report questionnaires in the prediction of recidivism for male
offenders (Kroner & Loza, 2003).
Accurate assessment of criminal attributions can assist in explaining crime and identifying
offender treatment targets. 300 participants including incarcerated offenders, released offenders,
sex offenders, and university students participated in this study. Results indicate minimal
relationships between socially desirable responding and the CRAI. In a fake-good testing
situation, the faking index had appropriate sensitivity, specificity, and positive and negative
predictive power in identifying fake-good responses. The CRAI's instructional set of general
observation as opposed to self-representation appears to limit socially desirable responding and
also allows the CRAI to be administered to non-offender populations (Kroner, et al, 2004).

Evidences of test Misuses and Biases

To apply the proactivereactive scheme to criminal thinking, Walters (2006) constructed
composite scales from the 80-item Psychological Inventory of Criminal Thinking Styles
(PICTS). The Proactive (P) and Reactive (R) composite scales are constructed from key PICTS
thinking style, factor, and content scales and are designed to measure proactive and reactive
criminal thinking, respectively. Research indicates that the two composite scales correlate
differentially with measures of positive outcome expectancies for crime and hostile attribution
biases in a manner consistent with research conducted on juvenile subjects.




Criminal Personality

Page 14 of 17



Possible strategies of reducing biases

Arrests are properly classified as count data in that they are nonnegative integers that ordinarily
form a positively skewed pattern that more closely approximates the Poisson distribution than
the normal distribution. Poisson regression is consequently considered the benchmark model for
count data, although if the distribution is over-dispersed (i.e., model variance exceeds the mean)
negative binomial regression is the model of choice. There are three questions that need to be
answered before a specific Poisson class procedure can be selected: over-dispersion,
truncation/censoring, and zero inflation. Over-dispersion occurs in least squares dispersion test.
Truncation occurs when a value, like zero, is missing from the distribution, whereas censoring
occurs when a range of values are collapsed into a single value (e.g., three or higher). Zero
inflation is marked by an excess number of zeros in ones distribution. This can be tested with
the statistic for non-nested models. Values greater than +1.96 favor the zero-altered model,
whereas values on V below -1.96 reject the zero-altered model (Walters, et al, 2007).
Protections. The American psychological Association's ethical standards require that
psychologists use only techniques or procedures that lie within their competence.
In addition, the purchase of Testing Materials is generally restricted by the publisher to
individuals or institutions that can demonstrate their competence in administering,
scoring, and interpreting tests.
The question of privacy. The examinee must be given only test relevant to the purpose
of the evaluation.
The question of confidentiality. Information revealed to psychiatrist and clinical
psychologist is typically regarded as a privilege, there are continuing assaults on the right
to withhold such information. For example, the Tarasoff decision of the plant California
Supreme Court makes it clear that information provided by a patient in the course of
therapy cannot remain privileged if that information indicates that the patient may be
dangerous.
The question of discrimination. Within psychology, attacks have centered on ways in
which test discriminate against minorities. It is often charge that most psychological tests
Criminal Personality

Page 15 of 17

are really designed for white middle-class population and that other groups are being
tested with the devices that are inappropriate for them.
Test bias. It is important to remember that significant differences between man scores
on a test of different groups do not in and of themselves indicate test bias or
discrimination. Rather, test bias or discrimination is a validity issue. That is, if it can be
demonstrated that the validity of a tests varies significantly across groups, then a case can
be made that the test is biased for that purpose



Conclusion

The findings of this research merit the interest of forensic examiners for several reasons. First,
knowledge of the MMPI-2 (either through education, discussions with a defense attorney, or
simply repeated exposure to the test) may influence a participants ability to escape detection as
a malingerer. Second, the finding that intelligence is significantly related to the ability to escape
detection should serve as one of the cautionary signals to forensic examiners. Third, this study
addresses a deficit in the MMPI-2 malingering detection literature by considering exogenous
factors related to the ability to escape detection. In addition to considering individual scores on
the clinical scales in treatment planning, it is important to consider their overall configuration,
that is the pattern of scores among them, known as the MMPI-2 code-type. An examination of
the code-type of an individual MMPI-2 profile often yields additional information about a
clients needs and possible responsiveness to various approaches to psychotherapy. Code-types
are ways of classifying an MMPI-2 profile that take into account more than one clinical scale at a
time. Code-types can include one scale (the highest scale), two scales (the two highest scales),
three scales (the three highest scales), and to a lesser extent, four scales (the four highest scales).
Although there is some debate about the most appropriate approach to determining code-types,
research has suggested that a 5-point difference between the lowest scale in the code-type and
Criminal Personality

Page 16 of 17

the next highest clinical scale is essential for minimizing the effects of measurement error. The
number of possible code-types prohibits here an exploration of their potential role in treatment
planning and evaluation.





















Criminal Personality

Page 17 of 17

References
Forbey, J., & Ben-Porath, Y. (2002). Use of the MMPI-2 in the treatment offenders.
International Journal of Offender Therapy and Comparative Criminology. London: Vol. 46, Iss.
3; pg. 308, 11 pgs

Foley, P., Hartman, B., Dunn, A., Smith, J., & Goldberg, D. (2002).The utility of the state-trait
anger expression inventory with offenders. International Journal of Offender Therapy and
Comparative Criminology. London: Vol. 46, Iss. 3; pg. 364, 15 pgs

Folsom, J. & Atkinson, J. (2007). Criminal Justice and Behavior. U.S. Department of Justice.
(n.d.). Bureau of Justice statistics. Retrieved March 3, 2009, from http://www.ojp.usdoj.gov/bjs

Gacono, C. (2002).Why there is a need for the personality assessment of offenders. International
Journal of Offender Therapy and Comparative Criminology. London: Vol. 46, Iss. 3; pg. 271, 3
pgs

Kroner, D., Mills, J., Yessine, A., & Hemmati, T. (2004). The generalized instructional set of the
criminal attribution inventory (CRAI): Socially desirable responding and faking. International
Journal of Offender Therapy and Comparative Criminology. London: Vol. 48, Iss. 3; pg. 360

McCusker, P. (2007). Issues Regarding the Clinical Use of the Classification of Violence Risk
(COVR) Assessment Instrument. International Journal of Offender Therapy and Comparative
Criminology. London. Vol. 51, Iss. 6; pg. 676

Mills, J. F., Loza, W., & Kroner, D. G. (2003). Predictive validity despite social desirability:
Evidence for the robustness of self-report among offenders. Criminal Behavior and Mental
Health, 13, 144-154.

Morey, L., & Quigley, B. (2002). The use of the personality assessment inventory (PAI) in
assessing offenders. International Journal of Offender Therapy and Comparative Criminology.
London: Vol. 46, Iss. 3; pg. 333, 17 pgs
Pelfrey, Jr, W. (2004). The relationship between malingerers' intelligence and MMPI-2
knowledge and their ability to avoid detection. International Journal of Offender Therapy and
Comparative Criminology. London:Vol. 48, Iss. 6; pg. 649

Walters, G., Frederick, A., & Schlauch, C. (2007). Postdicting Arrests for Proactive and
Reactive Aggression with the PICTS Proactive and Reactive Composite Scales. Journal of
Interpersonal Violence. Beverly Hills: Vol. 22, Iss. 11; pg. 1415

Walters, G. (2006). Use of the Psychological Inventory of Criminal Thinking Styles to Predict
Disciplinary Adjustment in Male Inmate Program Participants. International Journal of Offender
Therapy and Comparative Criminology. London: Vol. 50, Iss. 2; pg. 1