May EDI Corner - Data Disaggregation
Data disaggregation refers to separating data into smaller units to help identify inequities.1 Disaggregating addresses health disparities by uncovering disparate patterns of disease and access masked by full population estimates. Data disaggregation can be accomplished by various dimensions such as age, sex, geographic area, education, race, ethnicity, or income.1
Data should be disaggregated using the following variables at minimum.2
|
However, in disaggregating data, as the analyst or researcher, one should know the reason for disaggregating data. If data are disaggregated by “race” because that is typically how data are disaggregated, this could be problematic and could “perpetuate theories of biological inferiority and discriminatory behavior.” Assuming that there is a reason to disaggregate data by “race,” it is important to appreciate the degree to which current race/ethnicity categories may be incomplete. To support the inclusion of the diversity of ethnic ideas, the Census Bureau will include in the 2020 decennial census a classification for people of Middle Eastern and North African Descent.3
The lack of adequately disaggregated data fails to meet the needs of some minoritized populations by rendering them invisible in policy making, resource allocation, and program development.4 To this degree, policies, practices, and norms supported by those incomplete data may maintain White supremacy.4 As an example, during the COVID pandemic, American Indian/Alaska Native communities were not included in surveillance which challenged interventions to support the communities with COVID resources.5 Epidemiologist Abigail Echohawk has named the problem of “data genocide” and has advocated for better data collection, a decolonization of data, data sovereignty, and other protections to prevent erasure.
|
|
