Objective: The purpose of this study was to evaluate the measurement invariance of a neuropsychological battery across race/ethnicity by sex/gender subgroups over repeated measurements.
Method: Participants were 6,057 non-Hispanic White (NHW), Black, and Hispanic men and women in the Washington/Hamilton Heights Inwood Columbia Aging Project (WHICAP) who were administered neuropsychological tests of memory, language, and visuospatial abilities at 18 to 24-month intervals for up to 25 years. Invariance analyses were conducted on the three-factor model across sex/gender, racial/ethnic, and sex/gender by racial/ethnic subgroups, as well as across five assessment timepoints.
Results: The three-factor model demonstrated full measurement invariance across sex/gender groups and over repeated measurements. However, partial measurement invariance (invariant factor structure and factor loadings but nonequivalent observed score intercepts) for the language domain was exhibited across racial/ethnic and sex/gender by racial/ethnic subgroups.
Conclusion: Establishing measurement invariance is essential for valid interpretation of group differences in cognitive test performance. Findings from the current study highlight the need for continued examination of sex/gender by racial/ethnic differences in measurement properties of assessment tools, as well as expanded research on sex/gender variability across other understudied racial/ethnic groups. (PsycINFO Database Record (c) 2020 APA, all rights reserved).