B - 128 Examining Word Choice (WC) and Test of Memory Malingering Trial 1 (T1) Cutoffs by Ethnicity, Language, and Education

Abstract Objective Optimal cutoffs on performance validity tests can vary by examinee characteristics including ethnicity, language, and education. We examined WC and T1 cutoffs in Hispanic/Latino (H/L) and non-H/L participants with consideration of language and education. Method Local clinical data were retrospectively examined. Cases with WC and T1 data were included if there was no diagnosis of dementia, intellectual disability or psychosis yielding a sample (N = 580) in which 41% of participants identified as H/L and 26% of participants with language data (n = 530) reported bilingualism. Outcome measures included WC and T1 scores and dichotomous cutoffs. Results Education showed significant associations with WC and T1 (r = 0.12–0.20) and varied by ethnicity and bilingualism (pH/L) with a small effect (d = 0.21). No language differences were observed on WC or T1. Logistic regression models predicting dichotomous WC and T1 with education, ethnicity, and language as predictors were significant overall, but ethnicity and language were not significant. Grouping with independent validity indicators, WC false positives (5–6%) did not vary significantly by ethnicity or language; T1 false positives were not significantly different but were slightly higher in H/L and bilingual (14–15%) participants compared to non-H/L and monolingual (9–10%) participants. Conclusions WC and T1 cutoffs from prior research in mixed clinical and forensic samples appear appropriate for use with H/L and bilingual examinees. Lower educational levels remain an area in need of further validity measure research.