B - 140 Comparison of Performance Validity, Symptom Validity, Cognitive, and Psychological Symptom Report Measures in an Active-Duty Military Sample

Abstract Objective Performance and symptom validity (PVT/SVT) are distinct but related constructs (Nelson et al., 2007; Ruocco et al., 2008). However, studies have rarely examined PVT and SVT via factor analytic techniques, which offer stronger understanding of concurrent performance patterns across tests. Ord et al (2021) conducted one such study and concluded that PVT/SVT each “contribute unique information about examinees’ performance.” This and previous studies, however, have not included cognitive or symptom measures along with SVTs/PVTs, which limits interpretation. Methods In a large sample of Active-Duty (ad) personnel (n = 629), EFA and CFA modeling evaluated underlying patterns between different forms of testing (PVT, SVT, symptom report, and cognitive report). Measures include the MSVT, NV-MSVT, RBANS, and PAI. Results EFA (Maximum Likelihood and Geomin rotation) evidenced shared method variance patterns across PVT, SVT, and both cognitive/symptom scales without good fit (X2 = 4643.0, RMSEA = 0.15, CFI = 0.58). When modeled with CFA, these domains provided good fit, X2(145) = 634.92, CFI = 0.94, RMSEA = 0.07, SRMR = 0.05. Standardized loadings were large for PVT (0.40–0.97), SVT (0.62–0.98), symptom report (0.43–0.91), and cognitive ability (0.48–0.87). Higher order factor structure suggests convergence of PVT and Cognitive Testing (0.70); however, SVT and were distinct from cognitively-focused testing (−0.43). Conclusions Validity testing in neuropsychology has accelerated (Sweet et al., 2021), and understanding relationships between validity and symptom domains is critical to this effort. Our study suggests differentiation between domains, with weaker relationships between validity and clinical symptoms.