Further Validation of the Multiracial Identity Integration Scale (MIIS): Factor Structure, Validity, Reliability, and Measurement Invariance

The purpose of this dissertation was to perform a comprehensive psychometric evaluation of the Multiracial Identity Integration Scale (MIIS), the most cited measure of racial identity for multiracial individuals. Despite the prominence of this measure in informing theories of how multiracials understand their different racial backgrounds, the MIIS has demonstrated limited psychometric validity outside of the original scale development study. To address these gaps, across three studies I examined the factor structure, validity, reliability, and measurement invariance of the MIIS using the largest sample (N = 1,592) of multiracial American adults to date. An 8-item orthogonal, two- factor model with correlated errors to account for method effects was supported by exploratory and confirmatory factor analyses. Consistent with the original scale development study, Racial Distance (4 items) referred to multiracial individuals’ perceptions of separation among their different racial backgrounds, whereas Racial Conflict (4 items) referred to perceptions of incompatibility and tension among their multiple racial groups. Multigroup confirmatory factor analyses indicated that the MIIS demonstrated configural, metric, scalar, and residual invariance across gender (male vs. female), nativity (U.S.-born vs. foreign-born), multiracial status (biracial vs. multiracial), multiracial generation (first-generation vs. multigenerational multiracial), and White, Black/African American, Latinx, and American Indian racial backgrounds. Moreover, results suggested at least partial scalar invariance across age (emergent vs. established adulthood), annual income (<$19,000 vs. $20,000 to $39,000 vs. $40,000 to $59,000 vs. >$60,000), multiracial groups (Asian-White vs. Black-White vs. Latinx-White), and Asian American racial background. Evidence of convergent validity was partially supported with MIIS subscales correlating with lack of family acceptance, challenges with racial identity, multiracial pride, and self-esteem. Nonsignificant correlations between MIIS subscales and social desirability bias demonstrated discriminant validity. In support of criterion-related validity, subscale means varied across multiracials who endorsed different primary racial typologies. Moreover, incremental validity was supported with MIIS subscales predicting whether multiracials identified as monoracial or multiracial, above the contributions of a measure of multiracial racial identity exploration. Despite promising psychometric evidence, several important caveats concerning reliability and how to appropriately use this optimally weighted scale given the presence of method effects are discussed.