B – 154 Evaluating the Boston Naming Test as an Embedded PVT in Veterans

Abstract Objective This study examined the effectiveness of the Boston Naming Test (BNT) as an embedded performance validity test (PVT) in a sample of Veterans who underwent neuropsychological evaluation at a VA medical center. We replicated Abramson et al.’s (2020) analytic approach to evaluate whether BNT scores distinguished between valid and invalid responders. Method Participants were 655 Veterans (88.1% male; 94.0% white; M/SD: age = 57.49/13.97; education = 13.17/2.34). Performance validity was determined by failure of > 2 PVTs. Cognitive impairment was determined by the presence of any neurocognitive disorder diagnosis. Participants were grouped as unimpaired-valid, unimpaired-invalid, impaired-valid, or impaired-invalid. BNT cutoffs were compared across groups using ANOVA. Accuracy was examined with Chi-square and receiver operating characteristic (ROC) analyses. Results BNT performance differed significantly across groups for both raw, F(3, 494) = 40.42, p < .001, and T-scores, F(3, 484) = 17.70, p < .001. A <= 50 raw score cutoff yielded 89.5% specificity and 64.4% sensitivity; a <= 18 T-score cutoff yielded 99.5% specificity and 3.5% sensitivity. ROC analyses showed good discrimination for the BNT raw score (AUC = .821) and fair discrimination for the T-score (AUC = .756). Conclusion These findings support the utility of the BNT—particularly the raw score cutoff—as a practical, embedded PVT with strong classification accuracy for Veterans.