Multidimensional Measurement of Race: Exploring Racial Identity across Self-Identification, Social Perception, Skin Tone, and Self-Conceptualization

Introduction: Race is widely used in social and health research, yet its measurement remains conceptually and methodologically contested. Most survey-based studies rely on fixed categories aligned with standards such as those from the U.S. Office of Management and Budget (OMB), which support comparability in federal data, but often fail to capture the complexity of how individuals understand and experience race. Prior research has highlighted important limitations of these approaches, underscoring the need to account for dimensions such as self-identification, external perceptions, and lived experience. Single standard or improperly coupled measures of race with outcomes could obfuscate the differential ways race can be experienced and imposed, with significant implications for how we understand how social problems are produced and who is most affected by them. In response, there has been a growing shift toward more nuanced approaches that treat race as socially constructed and context-dependent. This study examines divergence across four measures of race to assess how different operationalizations capture distinct dimensions of race.

Methods: Using survey data from a sample of Philadelphia-based residents (N=1,480), this study examined four measures of race: 1) self-identified race using standard OMB categories, 2) socially perceived race, 3) skin tone (1 (lightest) to 10 (darkest)), and 4) open-ended self-identification. Descriptive and comparative analyses were conducted to assess concordance between self-identified and socially perceived race, as well as variation in skin tone across and within racial categories. Content analysis of open-ended responses examined how respondents conceptualized their racial identity beyond the structured categories.

Results: Agreement between self-identified and socially perceived race was high for White (93%) and Black (90%) respondents, but lower for Asian (73%), Other (45%), and American Indian and Alaska Native (AI/AN) (5%) respondents, with AI/AN and Other respondents frequently being perceived as Latino/a/e, Black, or multiracial rather than their self-identified category. Skin tone followed a clear gradient across perceived race (White: M = 1.64, SD = 0.82; Black: M = 6.12, SD = 2.04) and aligned more closely with perceived race than self-identification, suggesting that skin tone, a visible marker of race, shapes how race is socially assigned, even with substantial within-group variation. Open-ended responses showed that respondents described their racial identity in multidimensional terms, incorporating nativity, ancestry, religion, gender, and personal characteristics and values. These forms of self-conceptualization were not captured by either self-identified or perceived race categories, indicating that structured measures may obscure meaningful dimensions of identity.

Conclusions: These findings highlight that race is not a singular construct but operates across multiple, partially overlapping domains. Each measure captures distinct aspects of racialized experience, and reliance on a single operationalization may lead to incomplete or misleading conclusions. In survey research, this suggests that standard race questions may be insufficient for capturing racialized experience, underscoring the need for multidimensional measurement approaches, particularly when examining social and health disparities and inequities.