Abstract
This study examined the effect of memory impairment on the Word Memory Test (WMT) and investigated whether multiple alternate interpretative algorithms, including the previously validated genuine memory impairment profile (GMIP) and a novel primary validity indices mean (PVIM) composite score, improved its accuracy as a performance validity test (PVT) among examinees with genuine memory impairment. Data from 91 veterans administered the WMT, Hopkins Verbal Learning Test-Revised (HVLT-R), and Brief Visuospatial Memory Test-Revised (BVMT-R) during clinical evaluations were examined. Validity groups were established by four independent criterion PVTs (57 valid/34 invalid). Learning/memory impairment was determined by HVLT-R and BVMT-R. WMT primary validity subtests accurately differentiated valid from invalid groups overall and when the sample was subdivided by memory impairment; however, the optimal cut scores were lower than the WMT manual-recommended cutoff to maintain adequate specificity among those with memory impairment, which substantially decreased sensitivity. The GMIP significantly improved specificity (97%), but sensitivity remained weak (53%) when strict quantitative criteria were used. By contrast, use of the full GMIP criteria (quantitative criteria plus clinical history) resulted in 88% sensitivity/95% specificity. The novel PVIM had 85% sensitivity/88% specificity at optimal cutoff, supporting its utility as an alternate algorithm. Strictly adhering to the WMT manual-recommended cut score for patients with significant learning/memory impairment increases the risk of misclassifying these examinees as invalid performers. Alternate interpretive algorithms, such as the GMIP and PVIM, adequately mitigate this issue and enhance the accuracy of the WMT as a PVT, but should only be applied in clinical conditions empirically associated with significant neuropsychological impairment.