The Development of the Inventory of Problems–29: A Brief Self-Administered Measure for Discriminating Bona Fide From Feigned Psychiatric and Cognitive Complaints

ABSTRACT

This article describes the development of the Inventory of Problems–29 (IOP–29), a new, short, paper-and-pencil, self-administered measure of feigned mental and cognitive disorders. Four clinical comparison simulation studies were conducted. Study 1 (n = 451) selected the items and produced an index of potential feigning. Study 2 (n = 331) scaled this index to produce a probability score, and examined its psychometric properties. Study 3 tested the generalizability of Study 2's findings with 2 additional samples (ns = 128 and 90). Results supported the utility of the IOP–29 for discriminating bona fide from feigned psychiatric and cognitive complaints. Validity was demonstrated in feigning mild traumatic brain injury, psychosis, posttraumatic stress disorder, and depression. Within the independent samples of Studies 2 and 3, the brief IOP–29 performed similarly to the MMPI–2 and Personality Assessment Inventory, and perhaps better than the Test of Memory Malingering. Classifications within these samples with base rates of .5 produced sensitivity, specificity, positive predictive power, and negative predictive power statistics of about .80. Further research is needed testing the IOP–29 in ecologically valid field studies.

Notes

¹ Consistent with Rogers and Bender (2013), here we refer to malingering to indicate the “deliberate fabrication or gross exaggeration of psychological or physical symptoms of the fulfillment of an external goal,” and feigning to indicate the “deliberate fabrication or gross exaggeration of psychological or physical symptoms (Rogers & Vitacco, 2002) without any assumptions about its goals” (p. 518).

² The word simulators is used here to characterize research participants in simulation studies feigning any psychiatric disorders or cognitive impairments; in contrast, the word malingerers is used to describe individuals presenting disorders or impairments in real life.

³ In practice, of course, this decision is complicated by both multiple, related response styles, and the fact that motivation (i.e., malingering) cannot be deduced from test findings. Also, findings from individual tests are often indeterminate regarding the likelihood of feigning (Rogers & Bender, 2013).

⁴ Diagnostic information was incomplete and reported differently in the dissertations and data files, so we estimated frequencies of diagnosis by category.

⁵ Accordingly, both the DV2 and DV3 versions of the IOP contain the IOP–29 items. Both also included other items being evaluated for a longer, more complex test of feigning.

⁶ The base rate for malingering in Study 3 is .503.

⁷ As reported in Table 3, AUC was .845 (SE = .042) for the psychotic subsample and .898 (SE = .037) for the depression subsample. For the PTSD subsample, AUC was .903 (SE = .037); for the MTBI subsample AUC was .840 (SE = .048).