Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 35 - Imputing Race/Ethnicity to Understand Health Care Disparities Among Older Adults
Type: Topic Contributed
Date/Time: Monday, August 3, 2020 : 10:00 AM to 11:50 AM
Sponsor: Social Statistics Section
Abstract #312426
Title: Implicit Bias and Black-White Disparities in Health Care Among Older Adults
Author(s): Madhumita Ghosh-Dastidar*
Companies: RAND Corporation
Keywords: Healthcare disparity; Implicit Bias; Project Implicit; county-level racial bias

Unconscious/implicit biases may play a role in healthcare disparities. We estimate the association between implicit racial bias and county-level healthcare disparities, hypothesizing that counties with more negative attitudes towards Blacks, relative to Whites, have greater healthcare disparities. County-level disparities are estimated using nationally-representative Medicare data. National data on implicit biases were collected online through an Implicit Association Test (IAT), a computer-based measure that relies on differences in response latency to reveal implicit bias. The county-level IAT data suffer from issues of scarcity due to low (county-level) sample sizes, and potential bias stemming from non-representative respondent samples. We applied a combination of nonresponse weighting and model-based shrinkage utilizing socio-demographics (e.g., percent Black, socio-economics) from the American Community Survey, and borrowing information across counties using multi-level modeling. Using this approach, we derived model-based and population-weighted county-level IAT measures (e.g., attitudes towards Blacks versus Whites) and estimated its association with healthcare disparities

Authors who are presenting talks have a * after their name.

Back to the full JSM 2020 program