Ethnic identity: Factor structure and measurement invariance across ethnic groups.

Abstract

Considering a historically diversified (and growing) population in the United States, one's ethnic identification is often an important psychological-as well as social and political-construct because it can serve as a hindrance to interpersonal interaction. Despite the importance of ethnic identity in psychological research, the most widely developed ethnic identity measurement tool, the Multigroup Ethnic Identity Measure (MEIM; Phinney, 1992), lacks consensus regarding its psychometric properties. The purpose of this article is to identify the factor structure of this measure and identify whether it exhibits measurement equivalence/invariance (ME/I) across ethnicities. The current findings offer several contributions to the state of the literature. First, our data suggests a two-factor model, including affirmation/commitment and exploration factors, is the most appropriate structure when considering fit and parsimony indices via confirmatory factor analysis. Second, configural and metric measurement equivalence was found across Caucasian and non-Caucasian participants. Interestingly, partial scalar invariance was established when comparing Caucasians with the minority groups with the exception of the Hispanic subgroup, which exhibited no scalar invariance. Third, differences in ethnic identity factor means were found, especially across Caucasians and African Americans. In conclusion, the use of the two-factor model of the MEIM is recommended, and results suggest that the MEIM is an appropriate measure of ethnic identity in most ethnic groups. Limitations and future research are also discussed. (PsycINFO Database Record