Fostering Inclusiveness in Big Data: Changing the Normative Process for Selecting Covariates in Statistical Models

Oregon Convention Center 

In observational and/or epidemiological research we often adjust for a host of socio-demographic variables to clarify the influence of a predictor on an outcome. However, we rarely utilize All possible covariates, instead we utilize what is available or the normative variables we have been taught to include. If we want to be more inclusive of underrepresented communities, we have to start making the invisible visible by showing representation in our statistical analyses. Thus, in this presentation I will present ways to be more inclusive in the ways in which we select and utilize covariates in statistical analyses.