Validating the California Verbal Learning Test-Second Edition-Short Form Forced Choice Trial as an Embedded...

This study validated the California Verbal Learning Test-Second Edition-Short Form Forced Choice Recognition Trial (CVLT-2-SF-FC) against an independent battery of criterion performance validity tests (PVTs).

Ninety-eight veterans (mean age = 69.74 years) completed comprehensive outpatient neuropsychological evaluations that included the CVLT-2-SF-FC, Advanced Clinical Solutions Word Choice Test, Test of Memory Malingering-Trial 1, and Reliable Digit Span from the Wechsler Adult Intelligence Test-Fourth Edition Digit Span subtest. Participants were diagnosed with cognitive impairment based on the comprehensive evaluations. Two criterion group approaches (Single Failure and Two Failure) were used to classify participants into invalid, valid with no neurocognitive disorder (NND), and valid with neurocognitive disorder (ND) groups. Receiver operator characteristic (ROC) analyses tested the classification accuracy of the CVLT-2-SF-FC.

Using both grouping approaches, cut scores of ≤7 or ≤ 8 were associated with adequate specificities (≥0.90) and accuracy for identifying invalid performance in the total sample (AUCs = 0.71-0.81), albeit with modest sensitivities (0.41-0.50). Although the CVLT-2-SF-FC exhibited acceptable accuracy for differentiating between the invalid and NND groups with a cut-score of ≤8 (AUCs = 0.73-0.83), overall accuracy was reduced when differentiating between the invalid and ND groups (AUCs = 0.68-0.79), particularly when using the Single Failure classification (AUC = 0.68).

These findings add to the current knowledge of the CVLT-2-SF-FC as an embedded PVT, highlight the need to tailor cut-scores to different clinical groups, and support using this metric alongside other PVTs. Future studies should examine the CVLT-2-SF-FC's utility among larger samples that are more diverse with regards to age, sex, veteran status, and cognitive functioning.