Professional Documents
Culture Documents
Comparing The Validity of MMPI-3 Scores in Prehire Psychological Screenings of Male and Female Police Officer Candidates
Comparing The Validity of MMPI-3 Scores in Prehire Psychological Screenings of Male and Female Police Officer Candidates
To cite this article: Gina M. Talerico, Jasmine J. McCallum, Megan R. Whitman, Anthony M.
Tarescavage, David M. Corey & Yossef S. Ben-Porath (2024) Comparing the Validity of MMPI-3
Scores in Prehire Psychological Screenings of Male and Female Police Officer Candidates,
Journal of Personality Assessment, 106:1, 27-36, DOI: 10.1080/00223891.2023.2191278
Although most police officers perform their duties in a safe, zmunicipal police departments required preemployment psy-
ethical, and competent manner, police officer misconduct chological screening with testing.
is an important concern (Chaney & Robertson, 2013). Past In 2014, the Police Psychological Services Section (PPSS)
research has found that incidences of police misconduct can of the International Association of Chiefs of Police (IACP)
affect citizen safety and negatively impact public perception released preemployment psychological evaluation guidelines
of police officers (Bradford et al., 2009; Carr et al., 2007; aimed at improving these assessments (IACP, 2014/2020).
Mastrofski et al., 2002; Mazerolle et al., 2012; Nagin & Telep, The IACP notes that the testing utilized in these evaluations
2017). These concerns have led to the development of pol- should have documented reliability and validity in pre-
icies aimed at decreasing police misconduct (Nunes, 2015). employment assessments of public safety candidates. The
For example, in an effort to improve the screening process Minnesota Multiphasic Personality Inventory-2-Restructured
of police officers, psychological evaluations and testing have Form (MMPI-2-RF; Ben-Porath & Tellegen, 2008/2011) is
become an integral part of the preemployment selection an example of a test that has well-documented psychometrics
process. for use in police officer screenings (Ben-Porath et al., 2017;
In a recent review, Corey et al. (2022) reported that 37 Corey & Ben-Porath, 2018; Menton et al., 2022; Roberts
states mandate, or conditionally mandate, psychological et al., 2019; Tarescavage, Brewster, et al., 2015; Tarescavage,
screening of police officer candidates, and the most recent Corey, & Ben-Porath, 2015a, 2015b; Tarescavage, Corey,
national Bureau of Justice Statistics (2010) review of local Gupton, et al., 2015; Tarescavage, Fischler, et al., 2015).
law enforcement selection practices indicated that between However, most of these studies included only male partic-
72% to 98% of police departments required a psychological ipants in validation analyses.
evaluation when screening candidates. Cochrane et al. (2003) The MMPI-3 (Ben-Porath & Tellegen, 2020a, 2020b), a
conducted a national survey of municipal police departments revision of the MMPI-2-RF that features an updated, rep-
and found that the majority used a multimethod approach resentative normative sample and new and revised scales,
in their hiring process. Many departments required medical was released in 2020, and subsequent peer-reviewed research
examinations, drug testing, background checks, and poly- on this version of the test has compared validity coefficients
graph testing. At the state level, Corey et al. (2022) reported of test scores across male and female police candidates
that 17 of the 37 states requiring psychological evaluations (Whitman et al., 2021, 2022b). First, Whitman et al. (2021)
specify that written psychological testing should be utilized, compared the criterion validity of MMPI-3 scores in pre-
and Cochrane et al. (2003) reported that 91% of surveyed employment evaluations across male and female public safety
CONTACT Anthony M. Tarescavage [email protected] Cleveland Psychological Testing, 16600 Sprague Road, #190, Middleburg Heights, OH 44130,
USA.
© 2023 Taylor & Francis Group, LLC
28 TALERICO ET AL.
candidates. These authors compared validity coefficients previous MMPI-3 studies (Whitman et al., 2021, 2022b)
using Cohen’s q, which is the difference between two Fisher’s indicated that these test scores are not biased across male
z corrected correlations. The gender differences in validity and female police candidates; however, those investigations
coefficients were minimal for preemployment public safety compared validity coefficients using Cohen’s q, which does
candidates, with few statistically significant differences, most not account for potential intercept bias. The present study
of which were small in magnitude. In another study that sought to expand on Whitman et al. (2021, 2022b) by eval-
also utilized Cohen’s q, Whitman et al. (2022b) compared uating intercept and slope bias more directly, using a
the validity of convergent and discriminant validity of multi-group structural equation modeling approach.
MMPI-3 scores across male and female police candidates.
That study used personal history data (i.e., biodata scales)
Current study
that were collected during the preemployment evaluation as
criterion measures. Similar to Whitman et al. (2021), these The purpose of the current study was to extend previous
authors found minimal gender differences, and all statisti- research by Whitman et al. (2021, 2022b), which compared
cally significant effect sizes were small. the validity of MMPI-3 scores across male and female public
It is important for psychologists to use in their evalua- safety candidates using Cohen’s q effect sizes. Specifically,
tions tests and other sources of evidence that have been we utilized a more direct analytic approach by evaluating
shown to be reliable and valid in the populations to which MMPI-3 scores for potential slope and intercept bias in a
an examinee belongs (American Educational Research structural equation modeling framework. We first correlated
Association & American Psychological Association & MMPI-3 substantive scale scores with conceptually relevant
National Council on Measurement in Education, 2014; composites derived from a preemployment history question-
American Psychological Association Task Force on naire. In cases where these associations yielded at least small
Psychological Assessment & Evaluation Guidelines, 2020). effect sizes (r ≥ |.10|; Cohen, 1988), we used a multi-group
This is particularly true in preemployment psychological structural equation modeling approach in which we regressed
evaluations, as the Civil Rights Act of 1991 prohibits dif- history questionnaire variables on MMPI-3 scores. For each
ferential interpretation of test results and use of different MMPI-3 substantive scale and criterion variable pairing, we
cutoff scores for generating interpretive statements on the compared the fit of freely estimated models for men and
basis of sex and other characteristics. Therefore, psycholo- women against models that were constrained to have slopes
gists are particularly motivated to select tests that have been and intercepts equal across men and women. Comparisons
demonstrated to be equally reliable and valid across demo- that yielded statistically significant decrements in fit indi-
graphic groups, allowing them to accurately interpret results cated the possible presence of slope and/or intercept bias.
regardless of a candidate’s demographic background. As described next in the Method section, we quantified the
Practitioners can inform test selection by knowing magnitude of the differences across the models using omega
whether commonly used tests evidence bias. Test bias occurs prime. Based on the results of past MMPI-2-RF investiga-
when there is evidence of differential validity of test scores tions of slope and intercept bias in a bariatric surgery setting
across demographic groups (Kaplan & Saccuzzo, 2018). The (Marek et al., 2015) and forensic inpatient setting (Whitman
first possible manifestation of differential validity is intercept et al., 2019), as well as comparisons of validity coefficients
bias, which is a systematic over-prediction or under-prediction across male and female public safety candidates (Whitman
of a criterion across demographic groups. The second man- et al., 2022a, 2022b), we hypothesized that the current study
ifestation of differential validity is slope bias, which rep- would not show evidence of practically meaningful
resents a difference in the strength of association between cross-gender differential validity in the MMPI-3 scores of
scale scores and criteria across demographic groups. To police candidates.
illustrate the implications of differential validity in this set-
ting, consider the following possibility. Women are under-
represented in law enforcement with about 14% of full-time Method
sworn law enforcement officers in local departments being Participants
female (Goodison, 2022). A measure evidencing intercept
bias could contribute to this hiring discrepancy by, for exam- The current study used a sample of 527 prehire police officer
ple, over-predicting job-relevant problems for female can- candidates (455 males, 72 females) from multiple police
didates. A measure evidencing slope bias could contribute agencies in the Pacific Northwest of the United States. All
to the hiring discrepancy by yielding less valid (and there- candidates produced MMPI-3 protocols that were valid
fore more error prone) test interpretations for women. according to MMPI-3 interpretive guidelines (Cannot Say
Whereas self-selection into police work is likely the pri- [CNS] < 15, Combined Response Inconsistency [CRIN],
mary factor influencing unequal employment rates across Variable Response Inconsistency [VRIN], and True Response
gender, it is important to rule out alternative explanations Inconsistency [TRIN] < 80 T, and Infrequent Responses [F]
to the underrepresentation of women in law enforcement, and Infrequent Psychopathology Responses [Fp] < 100 T;
including test bias in preemployment psychological evalua- Ben-Porath & Tellegen, 2020a). The age of the sample mem-
tions. Moreover, evaluating for test bias can also inform bers was comprised of 20 years or less (0.8%), 21-22 years
psychologists’ test use decisions for the most defensible (12%), 23-24 years (14%), 25-26 years (15%), 27-28 years
selection practices under the Civil Rights Act of 1991. Two (10.2%), to 29 years or more (48%). The sample’s level of
VALIDITY OF MMPI-3 SCORES IN MALE AND FEMALE POLICE OFFICER CANDIDATES 29
education included G.E.D. (2.1%), high school graduate our criteria (see Table 2 for descriptions of item content).
(11%), 1-2 years of college (30.6%), 3-4 years of college with As reported in Table 2, Cronbach’s alpha for these compos-
no bachelor’s degree (16.9%), graduation from a four-year ites ranged from .50 (History of Internalizing [two items])
college with a bachelor’s degree (32.8%), to a postgraduate to .76 (History of Legal Problems [19 items]), which are
degree (6.6%). The sample was composed of candidates who acceptable given the restriction of range expected in police
identified as white/Caucasian (87.1%), Hispanic/Latino candidates being evaluated for psychological suitability after
American (8%), Asian/Pacific Island (5.1%), Native American a conditional offer of employment had been tendered; low
(2.7%), and Black/African American (2.3%). Additionally, a SEM values support this conclusion. The items in the com-
majority of the sample reported their marital status as legally posites were transformed to z scores before taking the aver-
married (51.8%) or single (41.9%). This study was approved age of the items. Therefore, the mean for all composites is 0.
by the [masked] institutional review board (protocol
#[masked]). This sample overlaps with the sample used by
Whitman et al. (2022a) and Whitman et al. (2022b). The Analysis plan
sample was drawn from the same population as the MMPI-3
We first calculated means, SD, Cronbach’s alpha internal
police officer candidate comparison group presented in the
consistency estimates, and SEM for MMPI-3 and PsyQ com-
MMPI-3 Technical Manual (Ben-Porath & Telleegn, 2020b).
posite scores. Given the expected range restriction in police
However, there is no direct comparison of slopes and inter-
candidate samples, Cronbach’s alpha internal consistency
cepts in the test manual or Whitman et al. (2022a, 2022b),
estimates are likely underestimates of the reliability of test
and the present study takes a distinct analytic approach
scores (Schmidt et al., 2006). Therefore, we also calculated
relative to Whitman et al. (2022b), who compared magni-
SEM, which account for both alpha coefficients and sample
tudes of associations between MMPI-3 scores and biodata
variability. When scales with low alpha coefficients also have
scales across male and female police candidates.
low SEM, we can infer that the low internal consistency
estimates are reflective of range restriction rather than scale
Measures heterogeneity or measurement imprecision (Cortina, 1993).
Next, we calculated correlations between MMPI-3 scores
Minnesota multiphasic personality inventory-3 and composites. In the 70 analyses where associations
The MMPI-3 was described in detail earlier. Raw MMPI-3 between MMPI-3 scores and the PsyQ composites yielded
scores were derived from an expanded version of the at least a small effect size (r ≥│.10│), we conducted non-
MMPI-2-RF, the MMPI-2-RF-EX, which was used to develop linear regressions models that used MMPI-3 scores to pre-
the MMPI-3. Hall et al. (2022) demonstrated the compara- dict PsyQ composite scores. Using Mplus 7.3 (Muthén &
bility of MMPI-3 scores derived from the MMPI-2-RF-EX Muthén, 2017), these composites were parameterized using
and MMPI-3 booklets. The MMPI-3 Technical Manual maximum likelihood with robust standard errors (MLM)
reports extensive reliability and validity of evidence of test and a mean-adjusted chi-squared test statistic that is robust
scores across a wide range of settings, including preemploy- to non-normality. To examine for slope bias, we constrained
ment evaluations of police candidates (Ben-Porath & the slopes to be equal across gender (male, female), then
Tellegen, 2020b). Means, standard deviations (SD), Cronbach’s compared the fit of these models with freely estimated mod-
alpha internal consistency estimates, and standard errors of els. Similarly, to examine intercept bias, we constrained the
measurement (SEM) for MMPI-3 scale scores are presented intercepts to be equal across gender, then compared
Table 1. intercept-constrained models with freely estimated models.
Both intercepts and slopes were constrained and compared
Psychological history questionnaire (PsyQ) with slope-constrained models when slope bias was statis-
The study criteria are rationally derived mean scores of tically significant or evidenced by at least a small effect size.
problems endorsed on a standardized personal history ques- This is because slope bias inherently produces differences
tionnaire called the Psychological History Questionnaire in intercepts.
(Johnson et al., 2011). The PsyQ is a psychological and We assessed decrement in statistical fit across the con-
behavioral background questionnaire designed for use with strained and freely estimated models using Satorra-Bentler
law enforcement and other public safety applicants. History adjusted chi-square difference tests (Satorra & Bentler, 1994,
questionnaires are an important component in preemploy- 2010). Any Satorra-Bentler chi-squares associated with p >
ment assessments of police officers (IACP, 2014/2020), and .05 indicated nonsignificant decrement in fit when intercepts
admission of some historical problems can disqualify appli- and/or slopes were constrained to be equal. Put differently,
cants. The questionnaire contains 340 questions covering evidence of statistically significant slope and/or intercept
areas in the following 14 topics: education, employment, bias was indicated by significant Satorra-Bentler chi-square
military experience, law enforcement experience, driving values. We used the Satorra-Bentler chi-square values
record, financial history, legal history, substance use, general (Satorra & Bentler, 1994, 2010) to calculate omega (ω′)
information, developmental history, adult relationships, prime, which was used to quantify effect sizes (.10 small,
parental responsibilities, psychological and evaluation history, .30 medium, .50 large; Cohen, 1988). The Satorra-Bentler
and job relevant sexual history. From the full questionnaire, adjusted chi-square difference test is robust to uneven sam-
we rationally derived seven composite variables to serve as ple sizes (Brace & Savalei, 2017); additionally, due to unequal
30 TALERICO ET AL.
sample sizes, the focus of our analysis was on effect sizes also low (mean SEM = 3.5), indicating that the relatively
rather than statistical significance. low alpha values were due to range restriction rather than
measurement imprecision.
Correlations between the MMPI-3 substantive scales and
Results
the study criteria are provided in Table 3. A total of 70
Table 1 includes descriptive statistics and reliability estimates associations yielded a small effect size or greater (r ≥ .10).
for MMPI-3 scores. Most MMPI-3 mean scores were low We then conducted examinations of slope and intercept bias
relative to the test normative sample which has a mean of for these 70 associations, yielding 140 analyses (70 for slope
50 T (average mean score = 42.5 T). MMPI-3 scores also bias and 70 for intercept bias). These results are presented
had less variability in the present sample relative to the in Table 4. Overall, 24 of these analyses yielded statistically
normative sample, which has a SD of 10 T (average SD = significant findings indicating possible psychometric bias
4.9 T). Cronbach’s alpha internal consistency estimates for (see Table 5 for a summary). However, 23 of the 24 effect
MMPI-3 scores were low (mean α = .4); however, SEM sizes for the statistically significant results were negligible.
values, which account for alpha values and variability, were The only one to reach a small effect size was the Eating
VALIDITY OF MMPI-3 SCORES IN MALE AND FEMALE POLICE OFFICER CANDIDATES 31
Concerns (EAT) scale (ω’ = .21) predicting internalizing bariatric surgery candidate (Marek et al., 2015) and forensic
problems. EAT scores were more strongly associated with psychiatric inpatient (Whitman et al., 2019) samples.
internalizing problems among men than women. Our findings also build on the substantial research lit-
erature on use of the MMPI instruments in prehire screen-
ings (Ben-Porath et al., 2017; Corey & Ben-Porath, 2018;
Discussion
Roberts et al., 2019; Tarescavage, Brewster, et al., 2015;
The purpose of the current study was to examine MMPI-3 Tarescavage, Corey, & Ben-Porath, 2015a, 2015b; Tarescavage,
scores for differential validity across gender (men and Corey, et al., 2015; Tarescavage, Fischler, et al., 2015.
women) in prehire psychological screenings of police offi- Specifically, as shown in Table 3, the MMPI-3 scales were
cers. We found only one instance of at least small-magnitude associated in expected ways with the PsyQ composites (most
statistically significant differential predictive validity for the notably, the externalizing scales, which is consistent with
MMPI-3 scales, and in this case, it demonstrated stronger previous patterns of results [Whitman et al., 2022b]).
validity for men than for women. These results indicate Additionally, MMPI-3 mean scores and SD were lower in
that MMPI-3 scores are not differentially valid across men this sample relative to the normative sample, which is con-
and women. Consequently, our findings support MMPI-3 sistent with findings from other public safety samples for
use in preemployment evaluations, for which differential various tests (e.g., Lowmaster & Morey, 2012; Tarescavage,
interpretations of test scores on the basis of gender are Fischler et al., 2015; Whitman et al., 2021, 2022a).
prohibited by law. Aspects of these findings warrant further Importantly, the present findings uniquely support the
discussion. use of the MMPI-3 in preemployment evaluations of police
The results of the current study converge with past candidates with respect to the Civil Rights Act of 1991,
research that has generally demonstrated that the MMPI which prohibits differential interpretation of test scores
instruments manifest negligible evidence of differential valid- across gender, including use of different cutoff scores for
ity. For example, using a very similar design and a sample generating interpretive statements. As a consequence, psy-
of forensic psychiatric inpatients, Whitman et al. (2019) chologists need to select tests to use in preemployment
found that only 5.3% of their analyses yielded statistically evaluations that have been demonstrated to be equally valid
significant evidence of differential predictive validity of across gender in the relevant contexts and population. The
MMPI-2-RF scores across gender and ethnic groups in a present results indicate that MMPI-3 scores are not differ-
forensic psychiatric inpatient setting. There too, the largest entially valid in preemployment evaluations of police can-
differences were of a small magnitude. Marek et al. (2015) didates. No research has directly supported this conclusion
found similar results with the MMPI-2-RF in a sample of through evaluation for slope and intercept bias for any other
bariatric surgery candidates. Moreover, these authors commonly used test in preemployment evaluations, such as
reported that male and female bariatric surgery patients the Personality Assessment Inventory (PAI; Morey, 2007)
produced similar MMPI-2-RF scores, which was also true and California Psychological Inventory-434 (CPI-434; Gough
in the present sample. The present results also converge & Bradley, 2002).
with previous research comparing MMPI-3 validity coeffi- The limitations of this study point to future directions
cients in preemployment evaluations of male and female in research. One limitation in the current study is an inad-
public safety candidates, albeit using a different quantitative equate sample size to test for differential validity across
methodology (Whitman et al., 2021, 2022b). Together, these other demographic characteristics (e.g., race and age). Future
three studies yield strong support for the assertion that research in this area is of interest, although it is important
MMPI-3 scores are not differentially biased across gender to note that past MMPI-2-RF research has yielded results
in preemployment settings, which is similar to findings with similar to the current study for race/ethnicity (Whitman
32 TALERICO ET AL.
Table 3. MMPI-3 substantive scales zero-order associations with PsyQ problem variables.
Victim of Child Childhood
Scale Work Financial Legal Abuse Learning Internalizing Alcohol
Emotional/Internalizing Dysfunction EID .11** .00 .10* .06 .02 .11** .13**
Thought Dysfunction THD .00 .06 .06 .04 .06 –.07 –.02
Behavioral/Externalizing Dysfunction BXD .29** .16** .38** .14** .06 .04 .26**
Demoralization RCd .05 –.02 .05 .03 –.01 .09* .07
Somatic Complaints RC1 .08 .02 .10* .09* .03 .04 .02
Low Positive Emotions RC2 .05 –.04 .03 .00 –.01 .11* .01
Antisocial Behavior RC4 .27** .20** .46** .22** .02 .08 .31**
Ideas of Persecution RC6 .06 .02 .03 –.03 –.01 –.06 .05
Dysfunctional Negative Emotions RC7 .11** –.03 .01 .03 –.04 .08 .06
Aberrant Experiences RC8 .07 .05 .08 .07 .06 –.05 –.01
Hypomanic Activation RC9 .14** .01 .08 .07 .09* –.03 .14**
Malaise MLS .15** .01 .16** .02 .01 .08 .05
Neurological Complaints NUC .04 .01 .10* .03 .03 –.03 .00
Eating Concerns EAT .03 –.04 .06 .03 .22** .11* .14**
Cognitive Complaints COG .09* .11* .03 .07 .12** .04 .05
Suicidal/Death Ideation SUI .11* .00 .02 –.03 –.03 .04 .03
Helplessness/Hopelessness HLP –.03 –.03 –.07 –.03 .04 .03 .01
Self-Doubt SFD .02 –.01 .04 .02 .00 .10* .02
Inefficacy NFC .09* .02 .05 .04 .05 .06 .01
Stress STR .11** .02 .09* .10* .04 .09* .15**
Worry WRY .14** .03 .11** .12** .00 .13** .20**
Compulsivity CMP .07 –.04 .00 .03 –.02 .02 .06
Anxiety-Related Experiences AXR .11* .06 .03 .04 .01 .09* .14**
Anger Proneness ANP .09* .02 .11* .09* –.02 –.02 .11**
Behavior-Restricting Fears BRF .03 –.04 –.01 –.05 –.03 .07 .10*
Family Problems FML .10* .11** .09* .37** .05 .01 .03
Juvenile Conduct Problems JCP .21** .14** .38** .13** .03 .05 .13**
Substance Abuse SUB .20** .12** .27** .05 –.01 .07 .43**
Impulsivity IMP .16** .04 .13** .08 .09* –.04 .15**
Activation ACT .07 –.01 .02 .06 .05 –.01 .06
Aggression AGG .15** .08 .09* .03 –.06 –.05 .04
Cynicism CYN .12** –.03 –.04 .02 .02 –.04 .05
Self-Importance SFI –.06 .04 –.07 –.07 .04 –.14** –.04
Dominance DOM .03 .05 .02 .02 –.08 –.04 .02
Disaffiliativeness DSF .08 –.04 –.04 .04 –.01 .04 –.03
Social Avoidance SAV –.02 –.15** –.01 –.02 –.03 .04 –.08
Shyness SHY .05 –.02 .00 .07 .03 .03 .05
Aggressiveness AGGR .07 .07 .03 .05 –.07 –.04 .04
Psychoticism PSYC .01 .05 .04 .05 .05 –.06 –.03
Disconstraint DISC .28** .18** .46** .16** .07 .05 .37**
Negative Emotionality/Neuroticism NEGE .11* –.01 .02 .10* –.04 .08 .12**
Introversion/Low Positive Emotionality INTR –.01 –.15** –.01 –.02 .00 .05 –.07
Note: *p < .05. **p < .01. Bolded r ≥ │.10│.
et al., 2019). Second, the present Pacific Northwest sample on differential predictive validity using prospective designs
consisted predominately of white participants (87.1%), which and post-hire performance data is needed. For example,
raises questions about the generalizability of these findings validation studies might use supervisor ratings of post-hire
to other police candidate samples. Indeed, the subsample of probationary period performance (Tarescavage, Fischler,
white participants in the present study is disproportionately et al., 2015), field performance ratings (Tarescavage, Corey
high even relative to other police and sheriff departments; & Ben-Porath, 2015a, 2015b), internal affairs or civilian
for example, Goodison (2022) conducted a survey of U.S. review authority complaints (Sellbom et al., 2007), or invol-
local police department personnel and reported that in 2020, untary or adverse departure (Tarescavage, Brewster, et al.,
69% of full-time sworn officers identified as non-Hispanic 2015) as criteria. Fifth, PsyQ composites were relatively
white. Future research with more diversity is needed. Third, lacking in the area of thought dysfunction, and the use of
the present sample included only individuals who identified criteria with greater construct coverage is needed to better
as male or female, which is a dichotomization of a fluid examine for differential validity in thought dysfunction
and dimensional construct (Bass et al., 2018). Oversampling scores, which have been demonstrated to be predictive of
for candidates who identify as nonbinary or another gender poor post-hire performance of police officers (Sellbom et al.,
identity, as well as using multiple methodologies to evaluate 2007; Tarescavage, Fischler, et al., 2015). Lastly, future
for differential validity, may help improve the external valid- research should evaluate other components of the selection
ity of the present conclusions. process for differential validity, including background inves-
Fourth, this study utilized self-reported background infor- tigation results and interviews.
mation as criteria. Although this is an informative source Notwithstanding these limitations, this study indicates
of information in preemployment evaluations, future research that the MMPI-3 does not evidence meaningful differential
Table 4. Differential validity analyses.
Slope Intercept
Scale Criterion χ2 p Ω' M W χ2 p Ω' M W
EID Work Problems .00 .947 .00 – – .32 .571 .01 – –
BXD Work Problems 1.01 .316 .02 – – 1.29 .256 .02 – –
RC4 Work Problems .30 .586 .01 – – .63 .427 .01 – –
RC7 Work Problems .33 .563 .01 – – .68 .409 .01 – –
RC9 Work Problems .65 .420 .01 – – .17 .682 .01 – –
MLS Work Problems 2.25 .134 .03 – – .38 .537 .01 – –
SUI Work Problems 1.83 .177 .02 – – .43 .510 .01 – –
STR Work Problems .53 .465 .01 – – .00 .970 .00 – –
WRY Work Problems 4.52 .034 .04 .18** .02 4.43 .109 .04 – –
AXR Work Problems .02 .896 .00 – – .53 .465 .01 – –
FML Work Problems .21 .649 .01 – – .08 .773 .00 – –
JCP Work Problems 9.36 .002 .05 .18** .44** 7.87 .020 .05 .13* .31*
SUB Work Problems 1.91 .168 .02 – – .09 .760 .01 – –
IMP Work Problems 3.04 .081 .03 – – .22 .640 .01 – –
AGG Work Problems 10.03 .002 .05 .10* .48** 10.53 .005 .05 .05 .31*
CYN Work Problems 1.06 .304 .02 – – .88 .349 .02 – –
DISC Work Problems .09 .759 .01 – – .08 .779 .00 – –
NEGE Work Problems .29 .590 .01 – – .18 .669 .01 – –
BXD Financial Problems .31 .581 .01 – – .09 .761 .01 – –
RC4 Financial Problems .24 .625 .01 – – .08 .778 .00 – –
COG Financial Problems 1.08 .298 .02 – – .15 .696 .01 – –
FML Financial Problems 0.26 .608 .01 – – .30 .582 .01 – –
JCP Financial Problems 3.45 .063 .03 – – .11 .745 .01 – –
SUB Financial Problems .01 .937 .00 – – .02 .898 .00 – –
SAV Financial Problems .10 .748 .01 – – .01 .943 .00 – –
DISC Financial Problems .02 .893 .00 – – .05 .827 .00 – –
INTR Financial Problems .35 .555 .01 – – .10 .747 .01 – –
EID Legal Problems 3.94 .047 .03 .18 -.08 4.67 .097 .04 – –
BXD Legal Problems .49 .484 .01 – – .19 .662 .01 – –
RC1 Legal Problems 2.65 .104 .03 – – .03 .867 .00 – –
RC4 Legal Problems .60 .438 .01 – – .14 .708 .01 – –
MLS Legal Problems 1.49 .223 .02 – – .08 .783 .00 – –
NUC Legal Problems .96 .328 .02 – – .04 .851 .00 – –
WRY Legal Problems 8.61 .003 .05 .16* -.11* 9.59 .008 .05 -.09 -.12
ANP Legal Problems .04 .848 .00 – – 1.17 .281 .02 – –
JCP Legal Problems .37 .545 .01 – – .31 .579 .01 – –
SUB Legal Problems .03 .874 .00 – – 1.81 .179 .02 – –
IMP Legal Problems .05 .816 .00 – – .46 .496 .01 – –
DISC Legal Problems .64 .423 .01 – – .73 .393 .01 – –
BXD Victim of Child Abuse .68 .410 .01 – – 5.00 .025 .04 .34** .23
RC4 Victim of Child Abuse .80 .370 .02 – – 7.70 .006 .05 .37** .19
STR Victim of Child Abuse .72 .397 .01 – – 1.15 .284 .02 – –
WRY Victim of Child Abuse 1.68 .195 .02 – – 3.03 .082 .03 – –
FML Victim of Child Abuse 3.98 .046 .03 .33** .48** 7.90 .019 .05 .30** -.18
JCP Victim of Child Abuse 7.30 .007 .05 .21** -.12 14.47 .001 .06 .22** .33*
DISC Victim of Child Abuse .90 .343 .02 – – 6.55 .011 .04 .36** .22
NEGE Victim of Child Abuse 2.86 .091 .03 – – .69 .405 .01 – –
EAT Childhood Learning Problems .35 .555 .01 – – .03 .874 .00 – –
COG Childhood Learning Problems 4.90 .027 .04 .17* -.03 4.78 .092 .04 – –
EID Internalizing Problems .60 .437 .01 – – .60 .440 .01 – –
VALIDITY OF MMPI-3 SCORES IN MALE AND FEMALE POLICE OFFICER CANDIDATES
WRY (Worry). BRF (Behavior-Restricting Fears). JCP (Juvenile Conduct Problems). AGG (Aggression). INTR (Introversion/Low Positive Emotionality). SUI (Suicidal/Death Ideation). EID (Emotional/Internalizing Dysfunction).
Negative Emotions). RC9 (Hypomanic Activation). AXR (Anxiety-Related Experiences). IMP (Impulsivity). CYN (Cynicism). SFI (Self-Importance). MLS (Malaise). NUC (Neurological Complaints). SFD (Self-Doubt). ANP
EAT (Eating Concerns). COG (Cognitive Complaints). FML (Family Problems). BXD (Behavioral/Externalizing Dysfunction). RC1 (Somatic Complaints). RC2 (Low Positive Emotions). RC4 (Antisocial Behavior). RC7 (Dysfunctional
Scale Criterion Bias ω' Effect Size Interpretation
-.03
W
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
WRY Work Slope .04 Negligible Men > Women
Problems
JCP Work Slope .05 Negligible Men < Women
Problems
−.11*
AGG
M
–
–
–
–
–
–
–
–
Problems
JCP Work Intercept .05 Negligible Men > Women
Problems
AGG Work Intercept .05 Negligible Men > Women
.03
.01
.01
.03
.01
.01
.04
.06
.03
.00
.01
.01
.00
.00
.00
.03
Ω'
Problems
EID Legal Slope .03 Negligible Men > Women
Intercept
Problems
EAT Legal Slope .07 Negligible Men < Women
Problems
.100
.595
.712
.121
.569
.715
.095
.001
.122
.926
.765
.422
.861
.906
.769
.139
WRY Legal Slope .05 Negligible Men > Women
p
Problems
WRY Legal Intercept .05 Negligible Men > Women
Problems
FML Victim of Slope .03 Negligible Men < Women
Child
Abuse
JCP Victim of Slope .05 Negligible Men > Women
2.70
.28
.14
2.40
.32
.13
4.70
13.68
2.39
.01
.09
.65
.03
.01
.09
3.94
χ2
Child
(Anger Proneness). SUB (Substance Abuse). SAV (Social Avoidance). DISC (Disconstraint). STR (Stress). NEGE (Negative Emotionality/Neuroticism).
Abuse
BXD Victim of Intercept .04 Negligible Men < Women
Child
Abuse
RC4 Victim of Intercept .05 Negligible Men < Women
Child
-.03
-.04
-.02
Abuse
W
–
–
–
–
–
–
–
–
–
–
–
–
–
.14*
Child
M
–
–
–
–
–
–
–
–
–
–
–
–
–
Abuse
JCP Victim of Intercept .06 Negligible Men < Women
Child
Abuse
.03
.02
.01
.02
.02
.03
.04
.06
.02
.02
.01
.02
.02
.01
.01
.03
Child
Abuse
Slope
the latter sample). For cases of slope bias, > (the associations between the
Criterion
scale score and the criterion were stronger among the prior sample than
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
Alcohol Problems
among the latter sample). WRY (Worry). JCP (Juvenile Conduct Problems).
AGG (Aggression). EID (Emotional/Internalizing Dysfunction). EAT (Eating
Concerns). COG (Cognitive Complaints). FML (Family Problems). BXD
(Behavioral/Externalizing Dysfunction). RC4 (Antisocial Behavior). RC6 (Ideas
Note: *p < .05. **p < .01.
DISC
WRY
ANP
BXD
AXR
SUB
RC4
RC9
BRF
IMP
STR
EAT
JCP
EID
SFI
indicate an absence of bias when using the MMPI-3 in Civil Rights Act of 1991 § 105, 42 U.S.C. § 2000 et seq. (1991). https://
preemployment evaluations of female police candidates. www.eeoc.gov/civil-rights-act-1991-original-text
Cochrane, R. E., Tett, R. P., & Vandecreek, L. (2003). Psychological
testing and the selection of police officers: A national survey.
Criminal Justice and Behavior, 30(5), 511–537. https://doi.
Disclosure statement org/10.1177/009385480325724
Regarding conflicts of interest, Anthony Tarescavage, David Corey, and Cohen, J. (1988). Statistical power analysis for the behavioral sciences
Yossef Ben-Porath receive research funding from the MMPI-3 test (2nd ed.). Erlbaum.
Corey, D. M., & Ben-Porath, Y. S. (2018). Assessing police and other
publisher, the University of Minnesota Press. Yossef Ben-Porath is a
public safety personnel using the MMPI-2-RF: A practical guide.
paid consultant to the MMPI-3 publisher, the University of Minnesota
University of Minnesota Press.
Press, and distributor, Pearson Assessments. Yossef Ben-Porath and
Corey, D. M., McElroy, H. K., & Ben-Porath, Y. S. (2022). Statewide
David Corey receive royalties on sales of the MMPI-3 Police Candidate psychological screening mandates for police candidates in the United
Interpretive Report and derivative materials. Yossef Ben-Porath receives States: A review and comparison to the standard of practice.
royalties on MMPI-3 sales. Professional Psychology: Research and Practice, Advance online pub-
lication. https://doi.org/10.1037/pro0000487
Cortina, J. M. (1993). What is coefficient alpha? An examination of
ORCID theory and applications. Journal of Applied Psychology, 78(1), 98–104.
https://doi.org/10.1037/0021-9010.78.1.98
Megan R. Whitman http://orcid.org/0000-0002-0150-453X Goodison, S. E. (2022). Local police departments personnel, 2020. U.S.
Anthony M. Tarescavage http://orcid.org/0000-0001-9608-2968 Department of Justice Bureau of Justice Statistics. https://bjs.ojp.
David M. Corey http://orcid.org/0000-0001-6512-4512 gov/sites/g/files/xyckuh236/files/media/document/lpdp20.pdf
Yossef S. Ben-Porath http://orcid.org/0000-0002-4597-6280 Gough, H. G., & Bradley, P. (2002). California Psychological Inventory
manual. Consulting Psychologists Press.
Hall, J. T., Menton, W. H., & Ben-Porath, Y. S. (2022). Examining the
References psychometric equivalency of MMPI-3 scale scores derived from the
MMPI-3 and the MMPI-2-RF-EX. Assessment, 29(4), 842–853.
American Educational Research Association, American Psychological https://doi.org/10.1177/1073191121991921
Association & National Council on Measurement in Education. International Association of Chiefs of Police. (2014/2020).
(2014). Standards for educational and psychological testing. American Preemployment psychological evaluation guidelines: 2020 revisions.
Educational Research Association. https://www.theiacp.org/sites/default/files/Preemployment%20
American Psychological Association Task Force on Psychological Psychological%20Evaluation%20Guidelines%202014.pdf
Assessment and Evaluation Guidelines. (2020). APA guidelines for Johnson, Roberts, & Associates, Inc. (2011). Psychological History
psychological assessment and evaluation. American Psychological Questionnaire (PsyQ) for law enforcement and other public safety
Association. https://www.apa.org/about/policy/guidelines-psychologica applicants. Johnson, Roberts, & Associates, Inc.
l-assessment-evaluation.pdf Kaplan, R. M., & Saccuzzo, D. P. (2018). Psychological testing: Principles,
Bass, M., Gonzalez, L. J., Colip, L., Sharon, N., & Conklin, J. (2018). applications, and issues. Cengage Learning.
Rethinking gender: The nonbinary approach. American Journal of Lowmaster, S. E., & Morey, L. C. (2012). Predicting law enforcement
Health-System Pharmacy, 75(22), 1821–1823. https://doi.org/10.2146/ officer job performance with the Personality Assessment Inventory.
ajhp180236 Journal of Personality Assessment, 94(3), 254–261. https://doi.org/1
Ben-Porath, Y. S., & Tellegen, A. (2008/2011). MMPI-2-RF: Manual 0.1080/00223891.2011.648295
for administration, scoring and interpretation. University of Minnesota Marek, R. J., Ben-Porath, Y. S., Sellbom, M., McNulty, J. L., & Heinberg,
Press. L. J. (2015). Validity of minnesota multiphasic personality
Ben-Porath, Y. S., & Tellegen, A. (2020a). Minnesota multiphasic per- inventory-2-restructured form (MMPI-2-RF) scores as a function
sonality inventory-3 (MMPI-3): Manual for administration, scoring, of gender, ethnicity, and age of bariatric surgery candidates. Surgery
and interpretation. University of Minnesota Press. for Obesity and Related Diseases, 11(3), 627–634. https://doi.
Ben-Porath, Y. S., & Tellegen, A. (2020b). Minnesota multiphasic per- org/10.1016/j.soard.2014.10.005
sonality inventory-3 (MMPI-3): Technical manual. University of Mastrofski, S. D., Reisig, M. D., & McCluskey, J. D. (2002). Police
Minnesota Press. disrespect toward the public: An encounter-based analysis.
Ben-Porath, Y. S., Corey, D. M., & Tarescavage, A. M. (2017). Using Criminology, 40(3), 519–552. https://doi.org/10.1111/j.1745-9125.2002.
the MMPI-2-RF in preemployment evaluations of police officer tb00965.x
candidates. In Police psychology and its growing impact on modern Mazerolle, L., Bennett, S., Antrobus, E., & Eggins, E. (2012). Procedural
law enforcement (pp. 51–78). IGI Global. https://doi. justice, routine encounters and citizen perceptions of police: Main
org/10.4018/978-1-5225-0813-7.ch003 findings from the Queensland Community Engagement Trial
Brace, J. C., & Savalei, V. (2017). Type I error rates and power of (QCET). Journal of Experimental Criminology, 8(4), 343–367. https://
several versions of scaled chi-square difference tests in investigations doi.org/10.1007/s11292-012-9160-1
of measurement invariance. Psychological Methods, 22(3), 467–485. Menton, W. M., Corey, D. M., & Ben-Porath, Y. S. (2022). Evidence
https://doi.org/10.1037/met0000097 for the comparability of local and remote administrations of the
Bradford, B., Jackson, J., & Stanko, E. A. (2009). Contact and confi- MMPI-2-RF in police candidate evaluations. Psychological Assessment,
dence: Revisiting the impact of public encounters with the police. 34(1), 98–104. https://doi.org/10.1037/pas0001088
Policing and Society, 19(1), 20–46. https://doi.org/10.1080/ Morey, L. C. (2007). The Personality Assessment Inventory professional
10439460802457594 manual (2nd ed.). Psychological Assessment Resources.
Bureau of Justice Statistics. (2010). Local police departments, 2007. Muthén, L. K., & Muthén, B. O. (2017). Mplus: Statistical analysis with
Bureau of Justice Statistics. latent variables: User’s guide (Version 8).
Carr, P. J., Napolitano, L., & Keating, J. (2007). We never call the cops Nagin, D. S., & Telep, C. W. (2017). Procedural justice and legal
and here is why: A qualitative examination of legal cynicism in compliance. Annual Review of Law and Social Science, 13(1), 5–28.
three Philadelphia neighborhoods. Criminology, 45(2), 445–480. https://doi.org/10.1146/annurev-lawsocsci-110316-113310
https://doi.org/10.1111/j.1745-9125.2007.00084.x Nunes, I. S. (2015). Hands up, don’t shoot: Police misconduct and the
Chaney, C., & Robertson, R. V. (2013). Racism and police brutality need for body cameras. Florida Law Review, 67(5), 1811–1844.
in America. Journal of African American Studies, 17(4), 480–505. Roberts, R. M., Tarescavage, A. M., Ben-Porath, Y. S., & Roberts, M.
https://doi.org/10.1007/s12111-013-9246-5 D. (2019). Predicting postprobationary job performance of police
36 TALERICO ET AL.
officers using CPI and MMPI–2–RF test data obtained during pre- cal utility in the PERSEREC police integrity study sample. Assessment,
employment psychological screening. Journal of Personality 23(2), 173–190. https://doi.org/10.1177/1073191115575070
Assessment, 101(5), 544–555. https://doi.org/10.1080/00223891. Tarescavage, A. M., Corey, D. M., Gupton, H. M., & Ben-Porath, Y.
2018.1423990 S. (2015). Criterion validity and practical utility of the Minnesota
Satorra, A., & Bentler, P. M. (1994). Corrections to test statistics and multiphasic personality inventory–2–restructured form (MMPI–2–
standard errors in covariance structure analyses. In A. von Eye & RF) in assessments of police officer candidates. Journal of Personality
C. C. Clogg (Eds.), Latent variable analysis: Applications for devel- Assessment, 97(4), 382–394. https://doi.org/10.1080/00223891.2014.
opmental research (pp. 399–419). Sage Publications. 995800
Satorra, A., & Bentler, P. M. (2010). Ensuring positiveness of the scaled Tarescavage, A. M., Fischler, G. L., Cappo, B. M., Hill, D. O., Corey, D.
difference chi-square test statistic. Psychometrika, 75(2), 243–248. M., & Ben-Porath, Y. S. (2015). Minnesota Multiphasic Personality
https://doi.org/10.1007/s11336-009-9135-y Inventory-2-Restructured Form (MMPI-2-RF) predictors of police of-
Schmidt, F. L., Oh, I.-S., & Le, H. (2006). Increasing the accuracy of ficer problem behavior and collateral self-report test scores. Psychological
corrections for range restriction: Implications for selection procedure Assessment, 27(1), 125–137. https://doi.org/10.1037/pas0000041
validities and other research results. Personnel Psychology, 59(2), Whitman, M. R., Corey, D. M., & Ben-Porath, Y. S. (2022a). Does
281–305. https://doi.org/10.1111/j.1744-6570.2006.00065.x prior law enforcement experience affect scores on preemployment
Sellbom, M., Fischler, G. L., & Ben-Porath, Y. S. (2007). Identifying psychological testing? An investigation using the MMPI-3.
MMPI-2 predictors of police officer integrity and misconduct. Psychological Services. https://doi.org/10.1037/ser0000679
Criminal Justice and Behavior, 34(8), 985–1004. https://doi. Whitman, M. R., Corey, D. M., & Ben-Porath, Y. S. (2022b).
org/10.1177/0093854807301224 Associations between MMPI-3 and psychosocial history findings
Tarescavage, A. M., Brewster, J., Corey, D. M., & Ben-Porath, Y. S. obtained in preemployment evaluations of public safety candidates.
(2015). Use of prehire Minnesota Multiphasic Personality Assessment. Advance online publication. https://doi.org/10.1177/
Inventory-2-Restructured Form (MMPI-2-RF) police candidate 10731911221138931
scores to predict supervisor ratings of posthire performance. Whitman, M. R., Elias, L. S., Cappo, B. M., & Ben-Porath, Y. S. (2021).
Assessment, 22(4), 411–428. https://doi.org/10.1177/1073191114548445 Criterion validity of MMPI-3 scores in preemployment evaluations
Tarescavage, A. M., Corey, D. M., & Ben-Porath, Y. S. (2015a). of public safety candidates. Psychological Assessment, 33(12), 1169–
Minnesota multiphasic personality inventory-2-restructured form 1180. https://doi.org/10.1037/pas0001042
(MMPI-2-RF) predictors of police officer problem behavior. Whitman, M., Tarescavage, A., Glassmire, D., Burchett, D., & Sellbom,
Assessment, 22(1), 116–132. https://doi.org/10.1177/1073191114534885 M. (2019). Examination of differential validity of MMPI-2-RF scores
Tarescavage, A. M., Corey, D. M., & Ben-Porath, Y. S. (2015b). A by gender and ethnicity in predicting future suicidal and violent
prorating method for estimating MMPI-2-RF scores from MMPI behaviors in a forensic sample. Psychological Assessment, 31(3),
responses: Examination of score fidelity and illustration of empiri- 404–409. https://doi.org/10.1037/pas0000677