Assessing How a Household Survey is Perceived by Respondents

Daniel K Yang

Abstract

Survey modifications could impact respondent burden. Increased burden could potentially lead to refusals in the following waves or inadequate answers to questions, which could in turn induce bias affecting overall data quality. Census has studied survey participant impressions of data security and privacy. Burden measurement would allow us to identify where interventions may be needed
to offset the impact of respondents perception of burden and to mitigate burden-induced bias on data quality. During the 2012 and 2017 Consumer Expenditure Surveys (CE) Quarterly Interview, answers on perceived burden were collected at the end of the final interview wave. In this study, we will introducing a composite burden index score using a multivariate technique. We studied respondent burden proxy indicators by using the nonparametric recursive partitioning model under
a complex survey design for the 2012 CE Quarterly Interview Survey. We will also present the results from the 2017 CE newly revised respondent burden questions.