Measurement Error in American Community Survey Paradata and 2014 Redesign of the Contact History Instrument

Written by:
RSM2016-01

Introduction

In recent years, many studies have addressed the potential of using paradata to reduce total survey error (Kreuter & Casas-Cordero, 2010; Nelson et al., 2014). There is less research, however, on the quality of paradata itself and how it may be prone to measurement error. This includes missing, incomplete, or inaccurate paradata, which may bias estimates of total survey error (Kreuter & Casas-Cordero, 2010). In this research, we analyze paradata from the U.S. Census Bureau’s American Community Survey (ACS) to see if modifications to the U.S. Census Bureau’s Contact History Instrument (CHI) reduce measurement error, and thus improve paradata quality.

Page Last Revised - October 8, 2021