Malingering on the Social Security Disability Consultative Exam: Predictors and Base Rates

Abstract

The Disability Determinations Service (DDS) obtains evidence about an applicant from sources that may include the Psychological Consultative Examination (PCE), which usually includes a mental status examination and Wechsler Scale findings. Although evidence for good effort is necessary to validate the findings, psychologists have been officially discouraged from determining effort by the use of formal tests. In Chafetz, Abrahams, and Kohlmaier (2007), indicators of effort within the PCE in WAIS-age (adults) and WISC-age (children) claimants were determined, and a PCE Malingering Rating Scale was developed. In the present study, this new scale was used along with established symptom validity tests (SVTs) to determine base rates and additional predictors of malingering within the PCE. Claimants were apportioned into separate effort groups in a “dose–response” manner: Definite Malingerers, Chance-Level, Fail Both (SVT and Rating Scale), Fail One, Fail Indicators, and Not Fail. Data from separate studies using the Test of Memory Malingering (TOMM), and the Medical Symptom Validity Test (MSVT) were used to assign claimants into separate effort groups. An evidence-based assessment approach utilized these base-rates to calculate post-test odds. Other predictors of effort within the PCE include the presence of conduct disorder symptoms, and the number of other family members on Disability.

Keywords:

• Disability
• Evidence-based assessment
• Malingering
• Mental retardation
• Social security

ACKNOWLEDGMENTS

I wish to thank Dr. Thomas Frazier for his help with evidence-based assessment, and for his informative workshop at NAN, 2006. I thank Dr. Scott Millis for helping with likelihood ratio discussions. I appreciate the discussions and interest in this work shown by Drs. Kevin Bianchini, Kevin Greve, Thomas Merten, Paul Green, Kyle Boone, and Manfred Greiffenstein. Moreover, I am grateful to Drs. Joy Kohlmaier and Joel Abrahams for their reviews, commentary, and encouragement. I also appreciated the helpful review of the manuscript during the submission process.

Notes

†M (SD). B = Black; W = White; O = Other; M = male; F = female.

*p<.01; Bold*: p<.001. Format: Mean (SD).

Val=valid; Fail 2=fail SVT and Rating Scale Total; Fail 1=fail either SVT or Rating Scale Total; Fail Ind.=fail indicators but not SVT or Rating Scale; Tot=total; RDS=Reliable Digit Span; Miss Start=# Wechsler Subtests in which items missed before start point; Err=Errors; df=within groups degrees of freedom ranging from 109–112 depending on missing data. Eta² is the proportion of variance in the dependent variable explained by the differences among groups. R² is the proportion of variance explained in the test for linearity (R) when the groups are ordered.

*p < .01; Bold*: p < .001. Format: Mean (SD).

The Fail 2 category in the TOMM Child sample had only one subject and was combined into the Fail 1 group to make the Fail 1 or 2 group.

Val = valid; Fail 2 = fail SVT and Rating Scale; Fail 1 = fail either SVT or Rating Scale Total; Fail Ind. = fail indicators but not SVT or Rating Scale; Tot = total; RDS = Reliable Digit Span; Miss Start = # Wechsler Subtests in which items missed before start point; Err = Errors; df = within groups degrees of freedom ranging from 70–75 depending on missing data. Eta² is the proportion of variance in the dependent variable explained by the differences among groups. R² is the proportion of variance explained in the test for linearity (R) when the groups are ordered.

*p < .01; Bold*: p < .001. Format: Mean (SD).

Val = valid; Fail 2 = fail SVT and Rating Scale; Fail 1 = fail either SVT or Rating Scale Total; Fail Ind. = fail indicators but not SVT or Rating Scale; IR = Immediate Recognition; DR = Delayed Recognition; PAR = Paired Associate Recall; FR = Free Recall; Tot = total; RDS = Reliable Digit Span; Miss Start = # Wechsler Subtests in which items missed before start point; Err = Errors; df = within groups degrees of freedom ranging from 49–51 depending on missing data. Eta² is the proportion of variance in the dependent variable explained by the differences among groups. R² is the proportion of variance explained in the test for linearity (R) when the groups are ordered.

*p ≤ .01; Bold*: p ≤ .001. Format: Mean (SD).

Val = valid; Chance Bel. = chance or below; Fail 1 or 2 = fail either SVT or Rating Scale Total or both; Fail Ind. = fail indicators but not SVT or Rating Scale; IR = Immediate Recognition; DR = Delayed Recognition; PAR = Paired Associate Recall (p = .016); FR = Free Recall; Tot = total; RDS = Reliable Digit Span; Miss Start = # Wechsler Subtests in which items missed before start point; Err = Errors; Eta² is the proportion of variance in the dependent variable explained by the differences among groups. R² is the proportion of variance explained in the test for linearity (R) when the groups are ordered.

Presence of conduct disorder frequencies for Adult (A) and Child (B) samples. Order violations for Medical Symptom Validity Test Adults (A) are shown for straight order violations and ORDER20 violations.

% = presence of conduct disorder evidence within each effort category. For order violation results, % is relative to all order violators. For ORDER20 violators, % is relative to all persons within each malingering category. *p < .02; Bold*: p < .001.