3/27/2023

Personality Assessment Inventory:

Recent Developments

Leslie C. Morey, Ph.D.

Abell Professor of Liberal Arts
Department of Psychological and Brain Sciences
Texas A&M University
College Station, TX 77843
morey@tamu.edu

© 1991-2023, L.C. Morey, Ph.D

PAI-Plus: Software
Update
© 1991-2023, L.C. Morey, Ph.D

PAI Plus: software update across

platforms
• Clinical scoring/
interpretive report
offers new scores
and features
• PAI stand-alone
software scoring/
interpretation
software, through
the Interpretive
Explorer, adds
additional features

© 1991-2023, L.C. Morey, Ph.D

1
 3/27/2023

Stand-alone software: report options

• PAI-Plus Score Report
• PAI-Plus Clinical Interpretive Report
• PAI-Plus Interpretive Explorer
– Only offered in the standalone software
– Includes interactive features

© 1991-2023, L.C. Morey, Ph.D

PAI Bibiography on PAR Website

Mendeley option: https://www.mendeley.com/community/personality-assessment-inventory-(pai)/

© 1991-2023, L.C. Morey, Ph.D

Personality Assessment Screener (1997)

• 22 item screener
• Can be completed in 3 to 5
minutes
• Designed to be maximally
sensitive to types of
problems identified by 344-
item PAI
• Assesses overall risk in a
Total Score; 10 element
scores indicate problem
areas to investigate

© 1991-2023, L.C. Morey, Ph.D

2
 3/27/2023

PAS Element Scores

PAS Score Description
Total Score potential for problems
Negative Affect (NA) Element distress, depression, anxiety

Acting Out (AO) Element impulsivity, sensation-seeking

Health Problems (HP) Element somatic concerns

Psychotic Features (PF) Element paranoia, psychotic symptoms

Social Withdrawal (SW) Element social detachment, discomfort

Hostile Control (HC) Element control needs, inflated self-image

Suicidal Thinking (ST) Element thoughts of death, suicide

Alienation (AN) Element lack of support, distrust
Alcohol Problem (AP) Element alcohol use and abuse

Anger Control (AC) Element anger management problems

© 1991-2023, L.C. Morey, Ph.D

PAS Total Score

Predicting PAI Elevations

• Cutting score of 19 yields 84.7% sensitivity, 78.7%

specificity for predicting an elevation on PAI
clinical scales
• Scores above 42 have 82.1% sensitivity, 93.3%
specificity for malingering
• Scores below 13 have 81.4% sensitivity, 71.5%
specificity for defensive responding

© 1991-2023, L.C. Morey, Ph.D

The PAS-O
PAS Observer Report

• Empirically compare informant

report with self report
• Uses PAS items in the third person
• Can be used with either PAS or full
PAI
• Normed on community (n = 504) and
clinical (n = 201) dyad pairs.
• strong cross-informant correlations,
strong correlations with the PAI, and
moderate correlations with measures
of client over- and under-reporting.

© 1991-2023, L.C. Morey, Ph.D

3
 3/27/2023

The PAS-O
Areas of
discrepancy

Elevation

Configuration
© 1991-2023, L.C. Morey, Ph.D

Self/Observer Correlations

© 1991-2023, L.C. Morey, Ph.D

Who should observe?

Observer Similarity and Acquaintanceship

© 1991-2023, L.C. Morey, Ph.D

4
 3/27/2023

PAS-O Total Score

Predicting PAI Elevations

• Cutting scores lower than those for PAS, reflecting

replicated trend for self-report to reflect greater
severity than observer report
• Cutting score of 17 yields 58.6% sensitivity, 80.6%
specificity for predicting an elevation on PAI
clinical scale
• Cutting score of 14 yields 71.7% sensitivity, 71.2%
specificity for predicting an elevation on PAI
clinical scales

© 1991-2023, L.C. Morey, Ph.D

PAS-O discrepancies and PAI validity

PAI Self-report Distortion Indicators Correlation with
PASO Total Raw
Difference1
Positive Distortion Indicators
PIM scale -.32**
Defensiveness Index -.25**
Cashel Discriminant Function .00
Hong Defensiveness Function -.32**
Negative Distortion Indicators
NIM scale .36**
Malingering Index .23**
Rogers Discriminant Function .15**
Hong Malingering Function .34**
Negative Distortion Scale .28**
1 high raw difference scores indicate self reported pathology > observer

© 1991-2023, L.C. Morey, Ph.D

PAS-O Case: Michelle

• 59 year old single woman
• presenting for treatment of problems with alcohol, for
which she had been treated at various points over 20
years
• carried current diagnosis of Alcohol Abuse
• she indicated that she might be experiencing increasing
difficulty in controlling her alcohol use
• accompanied to the evaluation by her younger sister
(age 57), who seemed to be particularly supportive of
the patient's need for treatment
• The younger sister completed the PAS-O as the
observer.

© 1991-2023, L.C. Morey, Ph.D

5
 3/27/2023

PAS-O Case: Michelle

Raw Score comparisons

© 1991-2023, L.C. Morey, Ph.D

PAS-O Case: Michelle

Raw Difference T-scores

© 1991-2023, L.C. Morey, Ph.D

PAS-O Case: Michelle

Absolute Difference T-scores

© 1991-2023, L.C. Morey, Ph.D

6
 3/27/2023

PAS-O Case: Michelle

PAI full scales

© 1991-2023, L.C. Morey, Ph.D

PAS-O Case: Michelle

PAI subscales

© 1991-2023, L.C. Morey, Ph.D

PAS-O Case: Michelle

PAI supplemental indicators

• Although PIM and DEF are elevated,

low CDF suggests that positive
distortion is likely self-deception rather
than effortful underreporting
• PAI profile provides hints of this limited
insight (RXR, DOM)
• Although patient indicates relatively
circumscribed problems with alcohol,
sister describes broad problems with
impulse control, particularly in the
presence of negative emotions, and
that the patient is much more
disconnected and mistrustful of those
around her than she acknowledges
• Given RXR, likely that sister served as a
powerful motivating force for
treatment

© 1991-2023, L.C. Morey, Ph.D

7
 3/27/2023

PAI-Plus: Tabs and

Profile Overlays
© 1991-2023, L.C. Morey, Ph.D

Unpacking NIM/PIM Effects

Predicted and Specific Profiles

• NIM & PIM Predicted Profiles

– regression estimates of other scales using only
NIM/PIM as predictors
• NIM & PIM Specific Profiles
– restandardizing against other profiles falling in
two bands (probable/marked) of distortion

© 1991-2023, L.C. Morey, Ph.D

NIM & PIM Predicted Profiles

100.00
• Regression
90.00 estimates based
80.00
upon clinical sample
(n=1,246)
70.00
B
O
R • Based upon R2
ranging from ~ .04
60.00

50.00
to .50.
40.00
R Sq Linear = 0.433

30.00

40.00 60.00 80.00 100.00 120.00 140.00

NIM

© 1991-2023, L.C. Morey, Ph.D

8
 3/27/2023

NIM & PIM Specific Profiles

Scale Mean (SD) Community NIM 84t to 91t NIM 92t +
(n=1,000) (n=53) (n=74)

50.00 79.56 80.33

BOR
(10.0) (10.98) (10.45)

• Restandardization against bona fide respondents in “probable”

and “marked” distortion ranges
• NIM-specific sample from clinical standardization sample, PIM-
specific sample from community standardization sample

© 1991-2023, L.C. Morey, Ph.D

Case VU-VA1
• 43 year old Caucasian woman
• Married 5 years, relationship seen as abusive
• "Always had up and down moods," low
frustration tolerance
• Having difficulty sleeping, pain from accident
3 years previous
• Completed PAI as part of a disability
evaluation

© 1991-2021, L.C. Morey, Ph.D

PIM-Predicted Discrepancies

Kurtz, J. E., Henk, C. M., Bupp, L. L., & Dresler, C. M. (2015). The validity of a regression-based procedure for detecting concealed psychopathology in
structured personality assessment. Psychological Assessment, 27 (2), 392–402.

© 1991-2023, L.C. Morey, Ph.D

9
 3/27/2023

PIM Predicted Discrepancy

Standard vs. Role Play Responding

Kurtz, J. E., Henk, C. M., Bupp, L. L., & Dresler, C. M. (2015). The validity of a regression-based procedure for detecting
concealed psychopathology in structured personality assessment. Psychological Assessment, 27 (2), 392–402.

© 1991-2023, L.C. Morey, Ph.D

PIM Predicted
Discrepancy
Standard vs. Role Play
Responding

• Discrepancy
appreciably correlated
with actual score
• Large discrepancies
(>2 SEM) often
associated with “true”
elevations of 65t +

Kurtz, J. E., Henk, C. M., Bupp, L. L., & Dresler, C. M. (2015). The validity of a regression-based procedure for detecting
concealed psychopathology in structured personality assessment. Psychological Assessment, 27 (2), 392–402.

© 1991-2023, L.C. Morey, Ph.D

Use of Context-Specific Norms

PAI-Plus software

• Context norms/mean profile overlays:

– Corsica et al. (2010) bariatric surgery norms
– Hynan (2013) child custody norms (see also Flens data in Morey & McCredie, 2021)
– Karlin et al (2005) chronic pain norms
– Morey et al (2011) deployed military norms
– Sims et al. (2013) egg donor norms
– Cheng et al. (2010) MVA claimants norms
– Lowmaster & Morey (2012) police applicants
– Hurst et al. (2010) kidney donor norms
– Also: Morey (1991) college student norms

• Add-on reports by PAR:

– Roberts (2000) Public Safety applicants norms (including corrections officers)
– Edens & Ruiz (2005) Correctional Norms

© 1991-2023, L.C. Morey, Ph.D

10
 3/27/2023

Police Selection & Performance Project

(Lowmaster & Morey, 2012)
• Pre-employment psychological
evaluations (including PAI) for local city
police department
• Studied performance ratings for 85
applicants who were subsequently
hired and remained employed for at
least 1 year
– Last stage of hiring process (conditional
offers already made), ~ 8% rejected at
this point
• Supervisors completed Officer
Evaluation Form (Benner et al., 2000),
25 item rating form with manual
providing item anchors; median 35
months post-hire
• Factor analyzed this form into three
components:
– Job performance (e.g., job knowledge,
decision-making); alpha = .94, ICC = .78
– Integrity problems (e.g., dishonesty,
unethical behavior); alpha = .82, ICC =
.70
– Disability abuse (e.g., sick leave,
disability status abuse); alpha = .87, ICC
= .49

© 1991-2023, L.C. Morey, Ph.D

Number of Significant Correlations with Outcome

Ratings, Police Sample
(117 possible correlations)

45
40
35
30
25 High Defensive (median +)
20 Low Defensive (< median)
15 Significant Moderation
10
5
0
PIM DEF CDF
Lowmaster, S.E., & Morey, L.C. (2012). Predicting law enforcement officer job performance with the Personality
Assessment Inventory. Journal of Personality Assessment, 94(3), 254-261.

© 1991-2023, L.C. Morey, Ph.D

Case: BrPD
• 21 year old male involved in pre-employment
fitness for duty evaluation with local police
department
• MMPI-2 defensive (K 67t), clinical scales
within normal limits

© 1991-2023, L.C. Morey, Ph.D

11
 3/27/2023

Case BrPD
Follow-up

© 1991-2023, L.C. Morey, Ph.D

Child Custody Litigants

Hynan, D. (2013). Use of the Personality Assessment Inventory in child-custody evaluation. Open Access Journal of Forensic
Psychology, 5, 120–133.

• Sample:
– 250 child custody litigants,
legal parents completing
court-ordered evaluations
– Mothers averaged 37 years
of age (SD = 6.09) and
fathers 40 (SD = 7.51)
– average of 15 years of
education (SD = 2.57)
mean of 1.72 children per
family (SD = .75), most
were Caucasian
• Data from Flens (n=350,
presented in Morey & McCredie,
2021) show similar pattern, but
greater defensiveness in men
litigants (65t vs 59t)

© 1991-2023, L.C. Morey, Ph.D

Case: Mike Casebook

• 32 year old male involved in custody dispute,
each wanting full custody
• obtained a score of 74t on the MMPI-2 L scale
relative to a K scale of 62t; clinical scales
portrayed himself as extroverted and active
(Scale 9 = 59t, Scale 0 = 33t)

© 1991-2023, L.C. Morey, Ph.D

12
 3/27/2023

PAI-Plus: Standard and

Supplemental Indicators

© 1991-2023, L.C. Morey, Ph.D

PAI: Standard Supplemental Indices

Profile Validity Indices

MAL Malingering Index

RDF Rogers Discriminant Function
DEF Defensiveness Index
CDF Cashel Discriminant Function
ALCest ALC Estimated Score
DRGest DRG Estimated Score

Predictive Indices

TPI Treatment Process Index

VPI Violence Potential Index
SPI Suicide Potential Index

© 1991-2023, L.C. Morey, Ph.D

PAI-Plus: Supplemental Validity Indicators

• Non-systematic Distortion Indicators
– Standard indicators: ICN, INF
– “Experimental” Indicators:
• Back Random Responding (based on Morey & Hopwood, 2004)
• Hong Randomness Index (Hong & Kim, 2001)
• Negative Distortion Indicators
– Standard indicators: NIM, MAL, RDF
– “Experimental” Indicators:
• Negative Distortion Scale (Mogge et al., 2010)
• Hong Malingering Index (Hong & Kim, 2001)
• Multiscale Feigning Index (Gaines et al., 2012)
• Malingered Pain-Related Disability Discriminant Function (Hopwood et al., 2010)
• Positive Distortion Indicators
– Standard indicators: PIM, DEF, CDF
– “Experimental” Indicators:
• Positive Distortion Scale (Mogge et al., 2011)
• Hong Defensiveness Index (Hong & Kim, 2001)

Note: see McCredie & Morey, 2018 for cross-validation of many of the experimental indicators
© 1991-2023, L.C. Morey, Ph.D

13
 3/27/2023

PAI-Plus: Supplemental Clinical Indicators

• Standard indicators: SPI, VPI, TPI, ALC-Est & DRG_Est, MCE

• “Experimental” Indicators:
– Inattention (INATTN) Index (Watson & Liljequist, 2018)
– Neuro-Item Sum (Keiski, 2007)
– Violence and Aggression Risk Index (Roche et al., 2017)
– Reactive Aggression Scale (Antonius et al., 2013)
– Instrumental Aggression Scale (Antonius et al., 2013)
– Level of Care Index (Sinclair et al., 2013)
– Chronic Suicide Risk (S_Chron) Index (Sinclair et al., 2016)
– RXR Estimated Score (Morey, 2020)

© 1991-2023, L.C. Morey, Ph.D

PAI-Plus: Supplemental Validity Indicators

• Non-systematic Distortion Indicators
– Standard indicators: ICN, INF
– “Experimental” Indicators:
• Back Random Responding (based on Morey & Hopwood, 2004)
• Hong Randomness Index (Hong & Kim, 2001)
• Negative Distortion Indicators
– Standard indicators: NIM, MAL, RDF
– “Experimental” Indicators:
• Negative Distortion Scale (Mogge et al., 2010)
• Hong Malingering Index (Hong & Kim, 2001)
• Multiscale Feigning Index (Gaines et al., 2012)
• Malingered Pain-Related Disability Discriminant Function (Hopwood et al., 2010)
• Positive Distortion Indicators
– Standard indicators: PIM, DEF, CDF
– “Experimental” Indicators:
• Positive Distortion Scale (Mogge et al., 2011)
• Hong Defensiveness Index (Hong & Kim, 2001)

© 1991-2023, L.C. Morey, Ph.D

Hong Randomness Function

Hong, S. H., & Kim, Y. H. (2001). Detection of random response and impression management in the PAI: II. Detection
indices. Korean Journal of Clinical Psychology, 20, 751–761.

• Scale composition: weighted composite of four PAI scales (ICN,

INF, ALC, and DRG).
• Derivation: discriminant function distinguishing randomly
generated sample from genuine (Korean) responders; 95.5%
correct classification rate
• Distribution: mean of -2.77 (sd=1.36) in US community
standardization, m = -1.62 (sd=1.85) in US clinical
standardization sample; large effect of clinical status suggests
need to use clinical norms, and large loading of ALC and DRG in
function likely problematic in substance abuse settings
• Notable research: distinguished clinical normative sample from
PAI manual random profiles (m = +2.00, sd = -.82, d = 2.71), and
significantly incremented INF (McCredie & Morey, 2018)

© 1991-2023, L.C. Morey, Ph.D

14
 3/27/2023

Identifying “Back Random Responding”

Morey & Hopwood (2004) Psychological Assessment

• Manual suggests Short / Full t-

score comparison
• SUI and ALC scales most
sensitive
• +/- 5 t-score points suggests
BRR
• Interpret 160-item short form if
present
Morey, L.C., & Hopwood, C.J. (2004). Efficiency of a strategy for
detecting back random responding on the Personality Assessment
Inventory. Psychological Assessment, 16, 197-200.

© 1991-2023, L.C. Morey, Ph.D

Identifying “Back Random Responding

Scoring Short-Full Discrepancies

Back Random Responding Analysis

Scale Q15 Q55 Q95 Q135 SUM SF-t FS-t DIFF
> 5t?
ALC
SUI
Q20 Q60 Q100 Q140 For all items, F=0, ST=1, MT=2, VT=3

SF-raw to t 0 1 2 3 4 5 6 7 8 9 10 11 12

ALC 45 50 55 60 65 70 75 81 86 91 96 101 106

SUI 46 52 57 63 69 75 80 86 92 98 104 109 115

© 1991-2023, L.C. Morey, Ph.D

Case EB

• 25 year old male seen in forensic setting

© 1991-2023, L.C. Morey, Ph.D

15
 3/27/2023

Case EB
Follow-Up
• This PAI completed shortly after patient murdered his father
• Interpreted by psychologist as invalid due to confusion,
interpreted by psychiatrist (who had administered the PAI) as
malingering
• Later revealed that psychiatrist had begun by reading and
discussing PAI questions with patient, who seemed to have
trouble understanding the questions*
• This was discontinued after 100 or so items, after which
patient completed test on his own, as recommended

* Note that the test author does not recommend this practice

© 1991-2023, L.C. Morey, Ph.D

PAI-Plus: Supplemental Validity Indicators

• Non-systematic Distortion Indicators
– Standard indicators: ICN, INF
– “Experimental” Indicators:
• Back Random Responding (based on Morey & Hopwood, 2004)
• Hong Randomness Index (Hong & Kim, 2001)
• Negative Distortion Indicators
– Standard indicators: NIM, MAL, RDF
– “Experimental” Indicators:
• Negative Distortion Scale (Mogge et al., 2010)
• Hong Malingering Index (Hong & Kim, 2001)
• Multiscale Feigning Index (Gaines et al., 2012)
• Malingered Pain-Related Disability Discriminant Function (Hopwood et al., 2010)
• Positive Distortion Indicators
– Standard indicators: PIM, DEF, CDF
– “Experimental” Indicators:
• Positive Distortion Scale (Mogge et al., 2011)
• Hong Defensiveness Index (Hong & Kim, 2001)

© 1991-2023, L.C. Morey, Ph.D

Negative Distortion Scale

Mogge, N. L., LePage, J. S., Bell, T., & Ragatz, L. (2010). The negative distortion scale: A new PAI validity scale. Journal of
Forensic Psychiatry & Psychology, 21, 77–90.

• Scale composition: Fifteen of the most infrequently endorsed

items within a psychiatric inpatient sample
• Derivation: demonstrated a sensitivity of .82 and specificity of
.71, overall correct classification rate of .77 for distinguishing
actual patients from feigning respondents
• Distribution: mean of 4.23 (sd=4.17) in US community
standardization, m = 6.57 (sd=5.31) in US clinical
standardization sample; moderate effect of clinical status,
should use clinical norms to interpret
• Notable research: some studies suggest that the NDS may
outperform the three standard PAI feigning indicators (Rogers
et al., 2013; Thomas et al., 2012); distinguished two feigning
samples from PAI clinical sample (d ~ 3.40), significantly
incremented NIM in both samples (McCredie & Morey, 2018)

© 1991-2023, L.C. Morey, Ph.D

16
 3/27/2023

Negative Distortion Scale

Mogge et al., 2010

• Development strategy
similar to MMPI-2 Fp
• 15 items (from 12
different scales/
subscales) with low
endorsement frequency
in inpatient sample
(n=360)
• Cut score of 13 had
SENS=.82, SPEC=.71
against SIRS (AUC=.76)
Mogge, N.L., Lepage, J.S., Bella T., & Ragatzc, L. (2010). The negative distortion scale: a new PAI validity scale. The
Journal of Forensic Psychiatry & Psychology, 21, 77–90.

© 1991-2023, L.C. Morey, Ph.D

Negative Distortion Scale

Mogge et al., 2010

Mogge, N.L., Lepage, J.S., Bella T., & Ragatzc, L. (2010). The negative distortion scale: a new PAI validity scale. The
Journal of Forensic Psychiatry & Psychology, 21, 77–90.

© 1991-2023, L.C. Morey, Ph.D

Negative Distortion Scale

Applied to existing PAI data

• Feigning group much 40

higher than clinical norm (d 35

= 3.47) 30

• Cut score of 13: 100% 25

sensitivity in Morey & Lanier 20

(1998) Fake Bad group, 88.3% NDS raw score

specificity for Morey (1991) 15

clinical standardization group 10

• Clinical & community norm 5

similar (d = 0.26) 0
Fake Bad Fake Good Clinical Community
Morey & Morey & Morey 1991 Morey 1991
Lanier 1998 Lanier 1998

Morey, L.C., & Lanier, V.W. (1998). Operating characteristics for six response distortion indicators for the Personality
Assessment Inventory. Assessment, 5, 203-214.

© 1991-2023, L.C. Morey, Ph.D

17
 3/27/2023

NDS and other PAI Indicators

Rogers et al., 2012

Rogers, R., Gillard, N.D., Wooley, C.N., & Kelsey, K.R. (2013). Cross-Validation of the PAI Negative Distortion
Scale for Feigned Mental Disorders: A Research Report. Assessment, 20, 36-42.

© 1991-2023, L.C. Morey, Ph.D

Hong Malingering Function

Hong, S. H., & Kim, Y. H. (2001). Detection of random response and impression management in the PAI: II. Detection
indices. Korean Journal of Clinical Psychology, 20, 751–761.

• Scale composition: weighted composite of five PAI scales (ICN,

NIM, ARD, PAR, and WRM)
• Derivation: discriminant function distinguishing negative
impression set from genuine (Korean) responders; 93.7%
correct classification rate
• Distribution: mean of -1.75 (sd=0.88) in US community
standardization, m = -0.87 (sd=1.23) in US clinical
standardization sample; large effect of clinical status suggests
need to use clinical norms for interpretation
• Notable research: distinguished two feigning samples vs. PAI
clinical sample (d ~ 3.10), but only significantly incremented
NIM in one of those samples (McCredie & Morey, 2018);
demonstrated sensitivity of .880, specificity of .872 when
applied to adolescents on PAI-A (Meyer, Hong, & Morey, 2014)

© 1991-2023, L.C. Morey, Ph.D

Multiscale Feigning Index

Gaines, M. V., Giles, C. L., & Morgan, R. D. (2012). The detection of feigning using multiple PAI scale elevations: A new
index. Assessment, 20, 437–447.

• Scale composition: average t-score of seven PAI clinical scales

(SOM, ANX, ARD, DEP, MAN, PAR, and SCZ)
• Derivation: MFI scores of 77 demonstrated a sensitivity of .689
and a specificity of .943 (SIRS as criterion) in the derivation
study; incremented NIM, MAL, and RDF.
• Distribution: mean of 50 (sd=7.53) in US community
standardization, m = 58.93 (sd=11.28) in US clinical
standardization sample; large effect of clinical status suggests
need to use clinical norms for interpretation
• Notable research: distinguished two feigning samples vs. PAI
clinical sample (d ~ 2.75), but only significantly incremented
NIM in one of those samples (McCredie & Morey, 2018)

© 1991-2023, L.C. Morey, Ph.D

18
 3/27/2023

Malingered Pain-Related Disability

discriminant function
Hopwood, C.J., Orlando, M.J., & Clark, T.S. (2010). The detection of malingered pain-related disability with the Personality
Assessment Inventory. Rehabilitation Psychology, 55, 307–310.

• Scale composition: composite of 32 different PAI scales and

subscales
• Derivation: discriminant function distinguishing actual pain vs.
malingered pain disability simulators; 88% correct classification
rate higher than NIM, MAL, RDF for this purpose
• Distribution: mean of +0.36 (sd=1.26) in community
standardization, m = +0.52 (sd=1.36) in clinical standardization
sample; with both means above “cut score” of 0, need caution
in interpretation
• Notable research: not yet cross-validated; mean of +4.72
(sd=1.45) in Morey & Lanier (1998) general malingering data
suggests may not be specific to pain malingering

© 1991-2023, L.C. Morey, Ph.D

Malingered Pain-Related Disability

discriminant function
Hopwood et al., 2010

Hopwood, C.J., Orlando, M.J., & Clark, T.S. (2010). The detection of malingered pain-related disability with the Personality Assessment Inventory.
Rehabilitation Psychology, 55, 307–310.

© 1991-2023, L.C. Morey, Ph.D

Implicit/Explicit Negative
Distortion Continuum
correlation with NIM Clinical Clinical/Fake Bad mix
** **
HongMal .878 .942
** **
MFI .803 .914
** **
NDS .753 .875
** **
MAL .624 .848
** **
FakeADHD .578 .772
** **
MalPainDF .134 .715
** **
RDF .098 .614
Implicit Explicit
NIM HongMal MFI NDS MAL RDF

© 1991-2023, L.C. Morey, Ph.D

19
 3/27/2023

Manual Case #637

• Participant in a “fake bad” study feigning
severe mental disorder

© 1991-2023, L.C. Morey, Ph.D

PAI-Plus: Supplemental Validity Indicators

• Non-systematic Distortion Indicators
– Standard indicators: ICN, INF
– “Experimental” Indicators:
• Back Random Responding (based on Morey & Hopwood, 2004)
• Hong Randomness Index (Hong & Kim, 2001)
• Negative Distortion Indicators
– Standard indicators: NIM, MAL, RDF
– “Experimental” Indicators:
• Negative Distortion Scale (Mogge et al., 2010)
• Hong Malingering Index (Hong & Kim, 2001)
• Multiscale Feigning Index (Gaines et al., 2012)
• Malingered Pain-Related Disability Discriminant Function (Hopwood et al., 2010)
• Positive Distortion Indicators
– Standard indicators: PIM, DEF, CDF
– “Experimental” Indicators:
• Positive Distortion Scale (Mogge et al., 2011)
• Hong Defensiveness Index (Hong & Kim, 2001)

© 1991-2023, L.C. Morey, Ph.D

Hong Defensiveness Function

Hong, S. H., & Kim, Y. H. (2001). Detection of random response and impression management in the PAI: II. Detection
indices. Korean Journal of Clinical Psychology, 20, 751–761.

• Scale composition: weighted composite of seven PAI scales (INF, PIM,

ANX, NON, RXR, DOM, and WRM).
• Derivation: discriminant function distinguishing positive impression
set from genuine (Korean) responders; 94.7% correct classification
rate
• Distribution: mean of -1.48 (sd=1.20) in US community
standardization, m = -2.51 (sd=1.72) in US clinical standardization
sample; large effect of clinical status suggests need to use community
norms
• Notable research: distinguished four positive impression samples
from PAI community sample (d ranged from 0.70 to 1.74), significantly
incremented PIM in all four samples (McCredie & Morey, 2018);
demonstrated sensitivity of .750, specificity of .723 when applied to
adolescents on PAI-A (Meyer, Hong, & Morey, 2014)

© 1991-2023, L.C. Morey, Ph.D

20
 3/27/2023

Positive Distortion Scale

Mogge, N. L., & LePage, J. S. (2017). The Positive Distortion Scale (PDS): A study of a new PAI validity scale.
Unpublished manuscript.

• Scale composition: comprised of 17 of the most frequently endorsed

PAI items in a psychiatric inpatient sample, reversed scored
• Derivation: demonstrated a sensitivity of .95 and specificity of .79,
overall correct classification rate of .87 for distinguishing actual
patients from positive impression set patients
• Distribution: mean of 30.42 (sd=7.38) in US community
standardization, m = 22.19 (sd=8.75) in US clinical standardization
sample; however, derivation approach suggests that it would be most
useful in clinical settings
• Notable research: distinguished three of four positive impression
samples from PAI community sample (d ranged from 0.64 to 1.23),
significantly incremented PIM in two samples; did not successfully
identify “coached” dissimulators (McCredie & Morey, 2018);

© 1991-2023, L.C. Morey, Ph.D

Self/Other Deception Continuum

correlation with PIM Community Pre-Employment
** **
DEF .629 .537
** **
Cashel .247 .440
** **
HongDef .498 .698
** **
PDS .334 .495

Self Other
PIM DEF PDS HongDef Cashel

© 1991-2023, L.C. Morey, Ph.D

Case: BrPD
• 21 year old male involved in pre-employment
fitness for duty evaluation with local police
department
• MMPI-2 defensive (K 67t), clinical scales
within normal limits

© 1991-2023, L.C. Morey, Ph.D

21
 3/27/2023

PAI-Plus: Supplemental Clinical Indicators

• Standard indicators: SPI, VPI, TPI, ALC-Est & DRG_Est, MCE

• “Experimental” Indicators:
– Inattention (INATTN) Index (Watson & Liljequist, 2018)
– Neuro-Item Sum (Keiski, 2007)
– Violence and Aggression Risk Index (Roche et al., 2017)
– Reactive Aggression Scale (Antonius et al., 2013)
– Instrumental Aggression Scale (Antonius et al., 2013)
– Level of Care Index (Sinclair et al., 2013)
– Chronic Suicide Risk (S_Chron) Index (Sinclair et al., 2016)
– RXR Estimated Score (Morey, 2020)

Note: see McCredie & Morey, 2018 for cross-validation of many of the experimental indicators

© 1991-2023, L.C. Morey, Ph.D

Mean Clinical Elevation

(MCE)

• Calculation: average t-score of the 11 clinical full

scales
• Represents indicator of overall problem severity
• Clinical standardization sample mean = 59.20 (s.d. =
10.05)
• Serves as ipsative reference point for various clinical
indicators, e.g. TPI, SPI, VPI, RXR
• Useful in comparison to NIM
• Related to treatment dropout (d = .42, .49 in 2
studies1,2)
1 Hopwood, C.J., Ambwani, S., & Morey, L.C. (2007). Predicting non-mutual therapy termination with the Personality Assessment Inventory.

Psychotherapy Research. 17(6), 706 – 712

2 Hopwood, C.J., Creech, S., Clark, T.S., Meagher, M.W., & Morey, L.C. (2008). Predicting the completion of an integrative and intensive

outpatient chronic pain treatment. Journal of Personality Assessment, 90, 76 -80.

© 1991-2023, L.C. Morey, Ph.D

Inattention (INATTN) Index

Watson, J. & Liljequist, L. (2018). Using the Personality Assessment Inventory to identify ADHD-like symptoms. Journal of
Attention Disorders, 22(11), 1049–1055.

• Scale composition: Six subscales included (ANT-S, BOR-S, SCZ-T, MAN-G,

MAN-A, & ANX-C); summing number of elevations of 65t+, thus ranging
from 0 to 6.
• Derivation: MANOVA comparing diagnosed ADHD patients with other
patients; a cutoff score of 3 was found to yield a sensitivity of 41% for
individuals who were eventually diagnosed with ADHD, with a specificity
of 79%.
• Distribution: mean of 0.49 (sd=0.92) in community standardization, m =
1.39 (sd=1.42) in clinical standardization sample; large effect of clinical
status
• Notable research: cutscore of 3 SENS = 68%, SPEC = .78, incremental
validity beyond MCE (McCredie & Morey, 2018); increments other ADHD
measures (Lancaster & Liljequist, 2018); correlated .62 with CAARS ADHD
Index but 3 cutscore only had .20 SENS, .94 SPEC (Maffly-Kipp & Morey,
under review)
© 1991-2023, L.C. Morey, Ph.D

22
 3/27/2023

Neuro-item Sum
Keiski, M. A. (2007). Use of the Personality Assessment Inventory (PAI) following traumatic brain injury (Doctoral Dissertation).
University of Windsor, Ontario, Canada. Dissertation Abstracts International DAI-B 68/12.

• Scale composition: Sum of item scores from 11 items (from SOM, DEP,
SCZ) thus ranging from 0 to 33.
• Derivation: Items initially selected by expert raters and then found able to
differentiate between patients with severe closed head injury and controls
• Distribution: higher in moderate closed head injury (m=20.57) than severe
(m=15.63); mean of 7.27 (sd=5.69) in community standardization, m =
13.00 (sd=8.09) in clinical standardization sample, large effect -> reference
against MCE
• Notable research: very limited cross-validation; found to distinguish
patients with organic disorder clinical diagnosis with a moderate effect
size but did not increment SOM for this purpose (McCredie & Morey,
2018); elevated (m = 17.33) in those diagnosed with ADHD, even more so
in feigned ADHD (m = 25.49; Maffly-Kipp & Morey, under review)

© 1991-2023, L.C. Morey, Ph.D

Violence and Aggression Risk Index

Roche, M. J., Sinclair, S. J., Denckla, C., Chung, W. J., Stein, M., & Blais, M. (2017). The empirically derived Violence and
Aggression Risk Index from the Personality Assessment Inventory: Development, validation, and application in general
psychiatric settings. Bulletin of the Menninger Clinic, 81, 213–232

• Scale composition: uses t-scores of five PAI clinical scales (NON, PAR-
H, ANT-A, AGG-A, and AGG-P); for each scale a value of 0 was used to
denote t scores less than 50; 1 between 50 and 59; 2 between 60 and
69; 3 between 70 and 79; 4 between 80 and 89; 5 between 90 and 99;
6 for t scores 100 or above.
• Derivation: VARI distinguished patients with history of violence (d
effect size = 1.33) and modestly incremented both AGG-P and the VPI
in the postdiction of violence history on cross-validation
• Distribution: mean of 3.32 (sd=3.10) in US community
standardization, m = 6.02 (sd=4.55) in US clinical standardization
sample; large effect of clinical status suggests need to use clinical
norms for interpretation
• Notable research: distinguished two assault history samples (d ~ .36,
.85), but only significantly incremented AGG in one of those samples
(McCredie & Morey, 2018)

© 1991-2023, L.C. Morey, Ph.D

Reactive/Instrumental Aggression Scales

Antonius, D., Sinclair, S. J., Shiva, A. A., Messinger, J. W., Maile, J., Siefert, C. J., . . . Blais, M. A. (2013). Assessing the heterogeneity of
aggressive behavior traits: Exploratory and confirmatory analyses of the reactive and instrumental aggression Personality Assessment
Inventory (PAI) scales. Violence and Victims, 28, 587–601.

• Scale composition: Reactive Aggression: 30 items indicating

impulsivity and hastiness in aggression; Instrumental Aggression: 24
items indicating premeditation and manipulation in aggression.
• Derivation: factor analyses indicate reasonable model fit, but no
external validity correlates
• Distribution: Reactive Aggression: m = 19.38 (sd=7.88) in US
community standardization, m = 26.39 (sd=10.79) in US clinical
standardization sample; Instrumental Aggression: m = 18.23 (sd=6.31)
in US community standardization, m = 19.85 (sd=6.94); large effect of
clinical status for Reactive Aggression suggests need to use clinical
norms for interpretation
• Notable research: distinguished two assault history samples (d ~ .21
to .53), but only Reactive Aggression significantly incremented AGG in
those samples (McCredie & Morey, 2018)

© 1991-2023, L.C. Morey, Ph.D

23
 3/27/2023

Level of Care Index (LOCI)

Sinclair, S. J., Slavin-Mulford, J., Antonius, D., Stein, M. B., Siefert, C. J., Haggerty, G., . . . Blais, M. A. (2013). Development
and preliminary validation of the Level of Care Index (LOCI) from the Personality Assessment Inventory (PAI) in a psychiatric
sample. Psychological Assessment, 25, 606–617.

• Scale composition: uses t-scores of five PAI scales (NIM, SUI, DEP-A, PAR-P,
and ANT-S); for four scales a value of 0 was used to denote t scores less
than 50; 1 between 50 and 59; 2 between 60 and 69; 3 between 70 and
79; 4 between 80 and 89; 5 between 90 and 99; 6 for t scores 100 or
above. For NIM, T scores less than 40 were assigned an index score of 0;
between 40 and 49 = 1; between 50 and 59 = 2; between 60 and 69 = 3;
between 70 and 79 = 4; between 80 and 89 = 5; 90 or above = 6.
• Derivation: Selection of scales intended to differentiate clients requiring
inpatient and outpatient; cutoffs between 15 and 18 demonstrated 73%
sensitivity for inpatient care, incrementing MCE and other indicators
• Distribution: mean of 4.04 (sd=3.58) in US community standardization, m
= 7.51 (sd=5.30) in US clinical standardization sample; large effect of
clinical status and nature of index suggests need to use clinical norms for
interpretation
• Notable research: distinguished inpatient from outpatient samples in
cross-validation (d = .56), incremented MCE (McCredie & Morey, 2018)
© 1991-2023, L.C. Morey, Ph.D

S_Chron Index
Sinclair, S. J., Roche, M. J., Temes, C., Massey, C., Chung, W. J., Stein, M., . . . Blais, M. (2016). Evaluating chronic suicide risk
with the Personality Assessment Inventory: Development and initial validation of the Chronic Suicide Risk Index (S_Chron).
Psychiatry Research, 245, 443–450.

• Scale composition: Chronic Suicide Risk Index (S_Chron) uses t-scores of

six PAI scales (NIM, STR, MAN-G, BOR-N, BOR-S, and ANT-A); for five
scales a value of 0 was used to denote t scores less than 50; 1 between 50
and 59; 2 between 60 and 69; 3 between 70 and 79; 4 between 80 and
89; 5 between 90 and 99; 6 for t scores 100 or above. For MAN-G; a value
of 5 for T-scores between 31 and 40; 4 between 41 and 50; 3 between 51
and 60; 2 between 61 and 70; 1 between 71 and 80; 0 for 81 or greater.
• Derivation: Selection of scales intended to differentiate clients with a
history of multiple previous suicide attempts from clients with only one
attempt or no attempts, incrementing SUI and SPI in derivation sample
• Distribution: mean of 6.75 (sd=3.40) in US community standardization, m
= 10.39 (sd=4.76) in US clinical standardization sample; large effect of
clinical status suggests need to use clinical norms for interpretation
• Notable research: distinguished two suicide risk samples (d ~ .27, .37),
but only significantly incremented SUI in current precautions group
(McCredie & Morey, 2018)

© 1991-2023, L.C. Morey, Ph.D

Supplemental Clinical Indicators

Global Profile Influences

Correlate NIM MCE

** **
VPI .665 .833
** **
SPI .728 .902
** **
TPI .623 .822
** **
Neuro-sum .583 .650
** **
LOCI .800 .812
** **
INATTN .652 .758
** **
ReactAgg .582 .759
** **
InstruAgg .188 .365
** **
VARI .541 .742
** **
S_CHRON .661 .780

© 1991-2023, L.C. Morey, Ph.D

24
 3/27/2023

Estimated RXR (RXR-Est)

• Scale composition and derivation: designed to provides
an estimate of the expected RXR score as determined by
the nature of the problems reported on other PAI scales.
This is a regression-based estimate derived from scores on
the 11 PAI clinical scales
• Distribution: 77% of patients obtained an RXR t-score that
was within 10t of their observed RXR T score.
• Notable research: removes most associations of
defensiveness indicators with RXR in clinical sample;
estimate appears to be 10t+ lower than observed in both
fake good and fake bad samples

© 1991-2023, L.C. Morey, Ph.D

Treatment motivation = f (Treatment need)

Predicting RXR from the clinical scales

• RXR = 76.84 - (.623 * MCE)

• RXR = 73.237102 + (SOM * .045996 ) - (ANX
*.046582) - (ARD * .019073) - (DEP *
.163310) + (MAN * .004796) + (PAR *
.119282) - (SCZ * .007849) - (BOR * .326771)
+ (ANT * .008794) - (ALC * .111612 ) - (DRG *
.020836)

© 1991-2023, L.C. Morey, Ph.D

PAI Case 301

• 42 year old male Vietnam combat veteran
• Inpatient admitting diagnosis: polysubstance abuse
• Provisional diagnosis: antisocial personality
• Third marriage recently deteriorating
• Suicide attempt one week after brief 3 day stay (his decision) in
inpatient VA substance abuse program (included this PAI)
• Second hospitalization in psychiatric unit revealed physical
abuse of all 3 wives

© 1991-2023, L.C. Morey, Ph.D

25
 3/27/2023

DSM-5 Alternative Model for

Personality Disorders
(e.g. Skodol et al, 2011)

• Severity/Level of Personality Pathology

• Problematic Personality Traits
• Personality Disorder Types

Skodol, A.E., Clark, L.A., Bender, D.S., Krueger, R.F., Livesley, W.J., Morey, L.C., Verheul, R., Alarcon, R.D., & Bell, C.C. (2011). Proposed
changes in personality and personality disorder assessment and diagnosis for DSM-5. Part I: Description and rationale. Personality
Disorders: Theory, Research and Treatment, 2(1), 4-22.

© 1991-2023, L.C. Morey, Ph.D

Criterion A
DSM-5 PD Severity
• Impairment in Self/
Interpersonal Functioning
– Characteristics of individuals at
different levels described in
the “Level of Personality
Functioning” scale
– Implicit in describing these
dysfunctions is that these
reflect failures in
developmental and
maturational processes

© 1991-2023, L.C. Morey, Ph.D

Criterion B: Problematic Personality Traits

Five broad domains (25 lower order facets)

• Negative Emotionality (emotional lability,

anxiousness, submissiveness, separation insecurity,
perseveration, depressivity*, suspiciousness*)
• Detachment (withdrawal, restricted affectivity*,
anhedonia, intimacy avoidance)
• Antagonism (callousness, manipulativeness, grandiosity,
attention-seeking, hostility*, deceitfulness)
• Disinhibition (impulsivity, distractibility, risk-taking,
irresponsibility) vs. Compulsivity (rigid perfectionism)
• Psychoticism (unusual beliefs and experiences,
eccentricity, cognitive and perceptual dysregulation)

© 1991-2023, L.C. Morey, Ph.D

26
 3/27/2023

Criterion B
DSM-5 Trait Model
• Trait definitions for higher
order domains and lower
order facets are provided
• Implicit in describing these
traits is that these reflect
extensions of normal
temperamental,
biobehavioral individual
differences
© 1991-2023, L.C. Morey, Ph.D

Personality Disorder Types

Borderline Personality Example

• Disorder types
reflect “hybrid” of
global PD pathology
(Criterion A) and
problematic traits
(Criterion B)

© 1991-2023, L.C. Morey, Ph.D

The PAI and the DSM-5 Alternative Model

for Personality Disorders

• PAI scales for the five AMPD pathological trait

domains developed by Ruiz et al. (2018)
• Regression functions to score level of
personality functioning & pathological trait
facets developed by Busch et al. (2017)
Busch, A.J., Morey, L.C., & Hopwood, C.J. (2017). Exploring the assessment of the DSM-5 Alternative Model for Personality
Disorders with the Personality Assessment Inventory. Journal of Personality Assessment, 99(2), 211-218.

Ruiz, M.A., Hopwood, C.J., Edens, J.F., Morey, L.C., & Cox, J.M. (2018). Initial development of pathological personality trait
domain measures using the Personality Assessment Inventory (PAI). Personality Disorders Theory, Research and Treatment, in
press. http://dx.doi.org/10.1037/per0000286

© 1991-2023, L.C. Morey, Ph.D

27
3/27/2023

28
4 of 7 required
PAI Case 301 & AMPD: Borderline Type
Psychoticism

Psychoticism
Unusual beliefs Unusual beliefs
Busch et al. PAI/AMPD Regression Equations

Eccentricity Eccentricity
Unusual beliefs Unusual beliefs
Rigid perfectionism (lack) Rigid perfectionism (lack)
Risk taking Risk taking

Disinhibition

Disinhibition
Distractibility Distractibility
Impulsivity Impulsivity

PAI Case 301 & AMPD

Irresponsibility Irresponsibility

Busch et al (2017) Functions

Busch et al (2017) Functions

Callousness Callousness

• …

Antagonism

Antagonism
Attention seeking Attention seeking
Grandiosity Grandiosity
Deceitfulness Deceitfulness
Manipulativeness Manipulativeness
Suspiciousness Suspiciousness
Restricted affectivity Restricted affectivity

Detachment

Detachment
Depressivity Depressivity
Anhedonia Anhedonia
Intimacy avoidance Intimacy avoidance
Withdrawal Withdrawal
Restricted affectivity (lack) Restricted affectivity (lack)

Negative Affectivity

Negative Affectivity
Perseveration Perseveration
Hostility Hostility

© 1991-2023, L.C. Morey, Ph.D

© 1991-2023, L.C. Morey, Ph.D

© 1991-2023, L.C. Morey, Ph.D

Submissiveness Submissiveness
Separation insecurity Separation insecurity
Anxiousness Anxiousness
Emotional lability Emotional lability
Personality disorder severity Personality disorder severity

80

70

60

50

40

30

20

10

80

70

60

50

40

30

20

10
 3/27/2023

PAI Case 301 & AMPD: Antisocial Type

Busch et al (2017) Functions
80

70

6 of 7 required
60

50

40

30

20

10
Personality disorder severity

Emotional lability
Anxiousness

Separation insecurity
Submissiveness
Hostility
Perseveration
Restricted affectivity (lack)

Withdrawal
Intimacy avoidance
Anhedonia
Depressivity
Restricted affectivity

Suspiciousness

Manipulativeness
Deceitfulness
Grandiosity
Attention seeking

Callousness

Irresponsibility
Impulsivity
Distractibility
Risk taking

Rigid perfectionism (lack)

Unusual beliefs
Eccentricity
Unusual beliefs
Negative Affectivity Detachment Antagonism Disinhibition Psychoticism
© 1991-2023, L.C. Morey, Ph.D

PAI: Recent New

Directions
© 1991-2023, L.C. Morey, Ph.D

Autism Spectrum Disorder-DF

Harrison, K. B., McCredie, M. N., Reddy, M. K., Krishnan, A., Engstrom, A., Posey, Y. S., ... & Loveland, K. A. (2020). Assessing autism
spectrum disorder in intellectually able adults with the Personality Assessment Inventory: Normative data and a novel
supplemental indicator. Journal of Autism and Developmental Disorders, 50, 3935-3943.

• Scale composition: Discriminant function including 18 scales/subscales

• Derivation: Function derived by comparing higher-functioning ASD
patients to psychiatric inpatients; article also presents mean ASD
scale/subscale profile
• Distribution: ASD patients higher (m = 1.91) than inpatients (m = -0.21);
community standardization actually somewhat higher (m = 0.16, sd = .88)
than clinical standardization sample (m = -0.24, sd = .99) so largely
independent of profile elevation
• Notable research: cross-validation in original article but not yet
independently; virtually uncorrelated with NIM, MCE in clinical
standardization sample

© 1991-2023, L.C. Morey, Ph.D

29
 3/27/2023

Case AUT
• 22 year old Caucasian woman, reasonably well functioning in school (with
accommodations)
• Describes difficulties with social skills, social interaction
• Also describes sensory sensitivities, some stereotype-like behavior
observed in interview
• Possible family history of “Asperger’s” in father, depression in mother
• Autism Diagnostic Interview with mother indicates patient meets criteria,
Social Communication Questionnaire suggests impairment as reported by
mother but not father
• Remarkably poor performance on Reading the Mind in the Eyes task,
taking 10 times longer than normal and accuracy (33%) not much above
chance
• PAI ASD-DF score of 1.79

© 1991-2023, L.C. Morey, Ph.D

Calmenson PAI ADHD Index

Calmenson, N.E. A New Subscale for the Personality Assessment Inventory (PAI) to Screen Adults for Attention-Deficit/Hyperactivity
Disorder (ADHD), dissertation, August 2021; Denton, Texas. University of North Texas Libraries Digital Library

• Scale composition: 12 items, selected mainly from MAN, SCZ,

DEP and BOR
• Derivation: Experts rated items thought likely to identify ADHD,
then refined to a 12-item scale through factor analysis (alpha =
.84); correlated .76 with CAARS, distinguished ADHD patients
from LD (AUC = .82)
• Distribution: mean of 9.15 (sd=5.09) in community
standardization, m = 13.60 (sd=7.34) in clinical standardization
sample; large effect -> reference against MCE (correlates .805);
college students also somewhat elevated (m=10.65, sd=5.32)
• Notable research: large effect (AUC = .789, d = 1.28)
distinguishing ADHD (m=21.67, sd=8.78) from college student
controls, much larger than INATTN index (Maffly-Kipp & Morey,
under review); also was markedly elevated in ADHD feighners,
correlates .630 with Cognitive Bias Scale

© 1991-2023, L.C. Morey, Ph.D

Feigned Adult ADHD Index

Aita, S. L., Sofko, C. A., Hill, B. D., Musso, M. W., & Boettcher, A. C. (2017). Utility of the personality assessment inventory in
detecting feigned attention-deficit/hyperactivity disorder (ADHD): The feigned adult ADHD index.
Archives of Clinical Neuropsychology, 1-13.

• Scale composition: two versions: weighted composite of 4 PAI

scales/24 specific item scores (scale version less sensitive but seemed
to cross-validate better in original study)
• Derivation: regression function distinguishing actual ADHD vs.
simulated ADHD, scores expressed as probability of feigning; 85%
sensitivity, 97% specificity malingered ADHD simulators
• Distribution: mean of .27 (sd=.36) in community standardization, m =
.29 (sd=.37) in clinical standardization sample; similarity supports use
of community norms, independence from clinical status for item
version only (.44 vs. .25 for scale version)
• Notable research: weak AUCs of .64-.65 in PVT/SVT failures (Harrison
et al., 2022); item version (AUC = .72) much better than scale (AUC =
.56) for feigners in Maffly-Kipp & Morey (under review); mean of .82
(sd=.30) in Morey & Lanier (1998) general malingering data suggests
may not be specific to ADHD malingering;

© 1991-2023, L.C. Morey, Ph.D

30
 3/27/2023

Cognitive Bias Scale (CBS)

Gaasedelen, O. J., Whiteside, D. M., Altmaier, E., Welch, C., & Basso, M. R. (2019). The construction and the initial validation of the
Cognitive Bias Scale for the Personality Assessment Inventory. The Clinical Neuropsychologist, 33(8), 1467-1484.

• Scale composition: ten items (from 7 different scales, only one from NIM)
• Derivation: IRT refinement of initial set of 40 items found to discriminate
between neuropsychological patients who passed or failed PVTs; cutscore
16+ .37 SENS, .90 SPEC
• Distribution: derivation sample m = 13.92 in failed PVT sample; mean of
5.76 (sd=3.45) in community standardization, m = 9.39 (sd=5.44) in clinical
standardization sample, large effect -> reference against MCE
• Notable research: A number of recent cross validations, cutoff 16+ .22-.27
SENS, .93+ SPEC against PVT failure, d = .70 (Armistead-Jehle et al., 2020;
Ingram et al, 2023) although Tylicki et al. (2021) found SENS=.63,
SPEC=.71; Shura et al. (2022) found CBS more related to SVT (M-FAST)
failure (d = 1.27) than PVT (WMT; d = 0.55); huge elevations (m = 22.7) in
feigned ADHD but also elevated (m = 14.5) in those diagnosed with ADHD
(Maffly-Kipp & Morey, under review); correlates .79 with NEURO-SUM in
clinical standardization sample

© 1991-2023, L.C. Morey, Ph.D

Case ADD
• 22 year old Caucasian male, reasonably well functioning in school (with
accomodations)
• Describes initial diagnosis of ADHD at age 12
• WAIS FSIQ 117, PSI 97; CPT performance generally average but with some
comm
• Both self and other (mother) report on CAARS show ADHD symptoms total
> 85t
• TOMM performance within normal limits
• PAI markers of interest:
– INATTN Index: 77t
– Calmenson PAI-ADHD: 87t
– Aita et al. Feigned ADHD: item probability .00, scale probability .15
– NEURO-ITEM sum: 67t
– Cognitive Bias Scale: 45t

© 1991-2023, L.C. Morey, Ph.D

31