Abstract
This study was designed to replicate an earlier report on the link between low scores on the Grooved Pegboard test (GPB), invalid responding, and elevated self-reported psychiatric symptoms. A fixed battery of neuropsychological tests was administered to 100 consecutively referred outpatients (MAge = 38.3, MEducation = 13.6 years) following traumatic brain injury at a Midwestern academic medical center. Classification accuracy of GPB validity cutoffs was computed against a free-standing PVT and three composite measures of embedded validity indicators. Previously suggested GPB validity cutoffs (T ≤ 29 in either hand) produced good combinations of sensitivity (0.25–0.55) and specificity (0.89–0.98) to psychometrically defined invalid performance. Raising the cutoff to T ≤ 31 resulted in a reasonable trade-off between increased sensitivity (0.36–0.55) and decreased specificity (0.84–0.94). T ≤ 31 in both hands was highly specific (0.93–0.98) to noncredible responding. GPB validity cutoffs were unrelated to psychiatric symptoms or injury severity. Failing PVTs based on forced choice recognition was associated with elevated self-reported depression, somatic concerns, and overall symptomatology. Low scores on the GPB are reliable indicators of noncredible responding. Self-reported emotional distress has a complex relationship with performance validity. Psychogenic interference is a potential mechanism behind PVT failures, and its expression is likely mediated by instrumentation and sampling artifacts. Further research on the topic is clearly needed to advance current understanding of psychogenic interference as a confound in cognitive testing.