Abstract
Objective: To test embedded symptom validity scales of the Neurobehavioral Symptom Inventory (NSI) as predictors of performance validity.
Setting: A Veterans Affairs Level II TBI/Polytrauma outpatient care unit in the Midwestern United States.
Participants: Veterans with a history of mild traumatic brain injury undergoing neuropsychological assessment as part of their routine care within the TBI/Polytrauma clinic.
Design: Retrospective analysis of the existing clinical data.
Main Measures: The NSI, the b Test, Test of Memory Malingering, Reliable Digit Span, California Verbal Learning Test-II Forced Choice.
Results: Embedded NSI validity scales were positively correlated with number of performance validity test failures. Participants identified as invalid responders scored higher on embedded NSI validity scales than participants identified as valid responders. Using receiver operating characteristic analysis, the embedded NSI validity scales showed poor sensitivity and specificity for invalid responding using previously published cutoff scores. Only 1 scale differentiated valid from invalid responders better than chance.
Conclusion: The embedded NSI validity scales' usefulness in predicting invalid neuropsychological performance validity was limited in this sample. Continued measurement of both symptom and performance validity in clinical settings involving traumatic brain injury treatment is recommended, as the present results support the existing research suggesting symptom validity tests and performance validity tests tap into related but ultimately distinct constructs.