The classification accuracy of the recognition memory test-words as a performance validity test is affected by gender and education.

Abstract

Objective: Previous research suggested that Recognition Memory Test-Words (RMT-W) scores may be confounded by gender and handedness. This study was designed to examine its classification accuracy as a performance validity test (PVT) and susceptibility to demographic characteristics.

Method: Archival data were collected from a consecutive sequence of 310 diagnostically heterogeneous examinees (181 men; M_Age = 41.4; M_Education = 12.9). The RMT-W's classification accuracy was computed against psychometrically operationalized criterion groups.

Results: Optimal RMT-W cutoffs (≤ 42 to ≤ 40) produced a good combination of sensitivity (.62-.70) and specificity (.90-.96), correctly classifying 85.6%-87.7% of the sample. Women scored 1.5 points higher. RMT-W scores were unrelated to handedness but were correlated with education. A linear relationship emerged between level of education and the cutoff needed to achieve ≥ .90 specificity: ≤ 43 for ≥ 13 years of education, ≤ 41 for 12 years of education, and ≤ 39 for ≤ 11 years of education. RMT-W ≤ 45 had .91 specificity in women with postsecondary education.

Conclusions: Results suggest that overall, the RMT-W remains an effective free-standing PVT. Gender, age, and handedness in isolation had minimal impact on RMT-W scores. However, education had a clinically significant effect. The combined effect of gender and education produced a marked shift in classification accuracy. Systematic research is needed on the relationship between demographics and PVT outcomes to ensure that cutoffs have the same clinical interpretation regardless of patient variables. (PsycInfo Database Record (c) 2025 APA, all rights reserved).