Panel Conditioning in the U.S. Consumer Expenditure Survey

Abstract

The U.S. Consumer Expenditure Interview Survey asks many filter questions to identify the items that households purchase. Each reported purchase triggers follow-up questions about the amount spent and other details. We test the hypothesis that respondents learn how the questionnaire is structured and underreport purchases in later waves to reduce the length of the interview. We analyze data from 10,416 four-wave respondents over two years of data collection. We find no evidence of decreasing data quality over time; instead, panel respondents tend to give higher quality responses in later waves. The results also hold for a larger set of two-wave respondents.

Panel Conditioning in A Probability-based Longitudinal study: A Comparison of Respondents with Different Levels of Survey Experience

“…Comparing those results to previous research using the Dutch probability-based LISS Panel (Schonlau and Toepoel 2015), our study did not replicate their findings of higher levels of nondifferentiation among more experienced respondents. Our findings on motivated misreporting, however, support some of the previous research that did not provide evidence for an increase in misreporting to filter questions with higher panel experience (Silber et al 2019;Bach and Eckman 2020;Eckman and Bach 2021).…”

“…Within the context of repeated interviewing, motivated misreporting among respondents with an advanced knowledge of the questionnaire has been documented for unemployment (Bailar 1975;Halpern-Manners and Warren 2012), party membership (Warren and Halpern-Manners 2012), everyday hygiene product use (Nancarrow and Cartwright 2007), and functional limitations (Mathiowetz and Lair 1994). However, other studies analyzing reports of household purchases and social contacts did not find any evidence of an increase in motivated misreporting among experienced respondents (Silber et al 2019;Bach and Eckman 2020;Eckman and Bach 2021).…”

Panel Conditioning in a German Probability-Based Longitudinal Study: A Comparison of Respondents with Different Levels of Survey Experience

“…When comparing those results to previous research using the Dutch probability-based LISS Panel (Schonlau & Toepoel, 2015), our study did not replicate their findings of higher levels of non-differentiation among more experienced respondents. Our findings on motivated misreporting, however, support some of the previous research, which did not provide evidence for an increase in misreporting to filter questions with higher panel experience Eckman & Bach, 2021;Silber et al, 2019).…”

“…Within the context of repeated interviewing, motivated misreporting among experienced respondents (with an advanced knowledge of the questionnaire structure) has been documented for unemployment (Bailar, 1975;, party membership , everyday hygiene product use (Nancarrow & Cartwright, 2007), and functional limitations (Mathiowetz & Lair, 1994). However, other studies analyzing reports of household purchases and social contacts did not find any evidence of an increase in motivated misreporting among experienced survey respondents Eckman & Bach, 2021;Silber et al, 2019).…”

Data trust and data privacy in the COVID-19 period

“…The recruitment rate for the replenishment was 12.1%. The replenishment also allows us to gain some insight into potential panel conditioning (Eckman and Bach, 2021). We find no significant differences in May 2020 responses between those who completed surveys in 2018/2019 and those who were newly recruited to the panel.…”