Understanding the Impact of Enhanced Weighting Tools on Key Health Survey Indicators in the California Health Interview Survey
Submission ID: 5596

Date: Thursday, 8:00 AM to 9:30 AM 
Session: Session C: T8:00 - 9:30 AM

Primary Presenter

Jiangzhou Fu, UCLA Center for Health Policy Research

Additional Authors or Round Table Presenters

Todd Hughes, UCLA Center for Health Policy Research ,

Royce Park, UCLA Center for Health Policy Research ,

YuChing Yang, UCLA Center for Health Policy Research ,

Ninez Ponce, UCLA Center for Health Policy Research ,

Arina Goyle, SSRS ,

, ,

, ,

, ,

Abstract

The California Health Interview Survey (CHIS) has implemented several methodological innovations in recent years. Since 2019, CHIS has employed address-based sampling (ABS) frame with a mail push-to-web survey followed by telephone nonresponse follow-up as the respondent recruitment approach, which has been proven efficient to increase response rate and reduce survey cost (Todd Hughes et. al., 2020). Apart from advancements in survey sampling and data collection strategies, CHIS 2020 post data collection adjustment approaches have also been improved, mainly regarding California population projection estimates and weighting calibration variables. To be aligned with CHIS 2020 revised weighting tools, CHIS 2019 estimates were also re-weighted utilizing the updated weighting tools.

This paper will first discuss the motivations of enhanced weighting tools and impacts on CHIS 2019-2020 sample compositions, especially focusing on reducing the underrepresentation of Asian subgroups. Additionally, we will explore the discrepancies between original and re-weighted CHIS 2019 estimates. CHIS 2020 estimates of crucial indicators monitoring health behavior, health care, health insurance will be presented and examined, and differences by key subgroups (age group, race/ethnicity, citizenship, and poverty level) will be analyzed. This paper will also scrutinize multiple indicators with salient variation in 2020 compared with previous years, potentially due to COVID-19 pandemic.

Keywords: health surveys, survey re-weighting, minority representation