Measurement Equivalence Testing of the American Identity Questionnaire Across Black, Latino, and White Adolescents

Objectives: The current study examined the psychometric properties of the American Identity Questionnaire (Phinney & Devich-Navarro, Journal of Research on Adolescence, 1997, 7, 3). American identity has been associated with societal and personal benefits for ethno-racially diverse populations, but limited research has assessed whether American identity measures function equivalently across members of different groups. Thus, the current study examined the measurement equivalence and construct validity of the American Identity Questionnaire among Black, Latino, and White adolescents. Method: Using a cross-sectional design, adolescents completed self-administered surveys during regular school time. The current study included U.S.-born adolescents (N = 1,326; Mage = 16.16 years; SD = 1.12; 53% female) who self-identified as either Black (n = 315), Latino (n = 345), orWhite (n = 666). Results: Multigroup confirmatory factor analysis was carried out using ethnic-racial group membership as the grouping variable. Findings suggested that the American Identity Questionnaire demonstrated configural (equivalent factor structures) and metric (equivalent factor loadings) invariance across the three groups. Partial scalar invariance was supported after allowing one item intercept to be freely estimated among Latino youth. Regarding construct validity, American identity was positively associated with self-esteem and personal identity, and negatively associated with depressive symptoms across the three groups. Conclusions: Findings suggest that the American Identity Questionnaire can be used to assess associations between American identity and other constructs with samples of Black, Latino, and White adolescents. Mean-level comparisons across the three groups may also be possible. Construct validity results indicated that American identity was positively associated with adolescents' psychosocial adjustment.