Detection of Response Bias and Noncredible Performance in Attention-Deficit/Hyperactivity Disorder

The development and utilization of indices to detect noncredible assessment performance is an area of increasing importance as recent prevalence estimates of malingering regularly exceed 30% in specific samples (Larrabee, 2007). Assessment of Attention-Deficit/Hyperactivity Disorder (ADHD) in adults presents clinicians with a variety of challenges given changes in the symptoms across ages and limitations in the diagnostic criteria and clinical measures. In an effort to address clinical symptoms, adults diagnosed with ADHD are frequently prescribed stimulant medication or receive a variety of accommodations at work or school. These benefits serve as potent incentives for individuals to feign ADHD. The purpose of the proposed investigation is to determine the predictive validity of MMPI-2 validity scales in detecting and accurately classifying individuals attempting to feign ADHD. Review of archival data from university psychological assessment clinic identified 36 adults diagnosed with ADHD and administered the MMPI-2 and ADHD Current and Childhood Symptoms Scales - Self-Report Forms (Barkley & Murphy, 2006) who served as the ADHD group (AC). Additionally, a sample of undergraduate students (N = 96) were randomly assigned to either a normal control group or analog group instructed to feign ADHD symptoms (prescribed malingering, PM). Both groups were administered MMPI-2 and ADHD scales in addition to a background history questionnaire. Results indicated that the ADHD behavior rating scales were unable to differentiate the AC group from the PM group, while both groups endorsed significantly higher symptoms than the control group. Correlations revealed that utilized MMPI-2 validity scales (i.e., P, L, K, Pb, Pp, RBS, PBS, HHI) were significantly correlated with the number and severity of endorsed current and childhood symptoms. Subsequent AN OVA and post-hoc tests found that the response patterns of the PM group resulted in significantly higher scores on P, Pb, Pp, RBS, PBS, and HHI as compared to the AC and control groups. Multinomial logistic regression revealed that Pp best differentiated the PM group from the AC group; however, Pp did not have incremental validity over the collective variance accounted for by other significant scales. Results also indicate that recommended cutoffs for HHI, RBS, and PBS display inadequate sensitivity and specificity. The current investigation supports employment of indices of effort or response bias in ADHD evaluations. Consistent with prior findings (Harrison et al., 2007; Jachimowicz & Geiselman, 2004; Quinn, 2003), individuals attempting to feign ADHD symptoms are able to respond in a convincing manner of self-report behavioral measures. Current results indicate that the MMPI-2 offers a number of validity indices that may assist in detecting individuals attempting to feign ADHD