Measuring Microaggression: Assessing Differences in Factor Structure for Black, Hispanic/Latino, & Asian Young Adults

Background: Microaggression refers to small, often subtle, acts of discrimination experienced in the daily lives of oppressed populations. Subtle bigotry like microaggression may have particularly harmful impacts. Empirical evidence regarding microaggression and its impacts has relied so far almost exclusively on qualitative research methods. The Racial and Ethnic Microaggression Scale (REMS) (Nadal, 2011) was recently developed to address the need for measurement of this complex construct. Validation in the initial development of the instrument did not use optimal factor analysis methodology. Additionally, the instrument is intended to be used across racial and ethnic groups. Since there are likely to be differences in the manifestations and conceptualization of microaggression for different racial and ethnic groups, it is essential to determine if the construct of the REMS were being understood in the same way across groups. We conducted a series of EFAs to evaluate the validity of the REMS, with particular attention to assessing the factor structure across different racial and ethnic groups. To increase the utility of the measure, the study also aimed to establish a shortened version of the REMS.

 Methods: Participants were 18 to 35 year old Black (N = 82), Hispanic/Latino (N = 103) and Asian (N = 101) undergraduate students enrolled at an urban public college in the United States. Students who were selected through a stratified, randomization methodology, were administered the REMS through an anonymous online survey. We conducted an EFA first with the whole sample of participants using the 45-item REMs. We then applied a resulting 28-item scale to conduct EFAs with each of the racial groups separately. We also assessed internal consistency of the initial and revised instruments and subscales.

 Results: Findings support the reliability of the instrument but suggest that a 5-factor model was a better fit with the data than the 6-factor solution of the initial analysis. Results also reveal differences between how the racial and ethnic groups conceptualize microaggression. While the Black participant data loaded very similarly in factor structure to the overall sample, the Hispanic/Latino and Asian groups found some substantial differences. The Hispanic/Latino data fit a 4-factor model better, with 2 of the subscales aligning together in entirety. There were also 2 items that loaded together but not with any other factor. The Asian data found 2 items from one factor loading with one of the other factors, and 3 items in another factor did not load at all. Interpretation of the results suggests that the differences reflect current stereotypes and discriminatory attitudes that are specific to each group’s experience.

 Implications:  It is likely that the access to and outcomes of social work interventions are impeded by  microaggression. Clinicians who are not informed about microaggressions may perpetrate them unintentionally. Agency contexts might be experienced as unwelcoming, and therapeutic relationships less than safe. On the other hand, research and education about microaggression holds immense promise for enhancing cultural competence. Clearly optimal measurement that is sensitive to variation in racial and ethnic groups will be a crucial component toward these advancements.