“But perhaps…”

The Effectiveness of Probing Questions for Reducing Item Non-response to Financial Questions

Bibliographic Data

ID24217008
AuthorsEvgeniia Shmeleva (0000-0001-8004-3315, HSE University, Moscow, Russia), Daniil Lebedev (0000-0001-7246-8022, HSE University, Moscow, Russia), Ekaterina Bragina (0009-0004-2529-5279, HSE University, Moscow, Russia)
Year2026
Volume32
Issue3
Pages9-29
Publication date2026-09-29
Open AccessYes
TypeARTICLE
VenueSociological Journal (JOURNAL)
Journal identifiersISSN: 1562-2495 • E-ISSN: 1684-1581
PublisherFederal Center of Theoretical and Applied Sociology of the Russian Academy of Sciences (FCTAS RAS) (PUBLISHER)
DOI10.19181/socjour.2026.32.3.1
OpenAlexW7214741992
LanguageRU
References cited24

A high proportion of non-substantive responses to questions about income and expenditures compromises data quality, limits the generalizability of findings, and can lead to sampling bias. The objective of this study is to assess the effectiveness of using follow-up questions to improve the quality of responses to financial questions. The study is based on data from a representative survey of the adult population of Russia, conducted by HSE University as part of the project «Household Economic Behavior Monitoring» (Spring 2025, N = 6 041). The first stage involved identifying predictors of non-substantive responses to the financial questions, while the second stage assessed the effectiveness of follow-up (probing) questions for different respondent categories. Predictors were selected using the Random Forest method with the Boruta algorithm and logistic regressions. The results show that the share of non-substantive responses was 6% for individual income, 17% for household income, and 14% for household expenditures. Follow-up questions converted 36–41% of the initial non-responses into substantive answers, thereby reducing the overall nonresponse rate to 4–10%. The highest risk of non-response was identified among respondents living separately from their spouse (a more than two-fold increase in the odds), those unaware of household expenditures (a 3- to 5-fold increase), those with depressive states (18–32% increase), poor health (50–86% increase), no savings, a short planning horizon, and residents of small towns. Follow-up questions were most effective for respondents with lower education levels, those unaware of expenditures, those without savings, and those from small towns. The method is only partially effective: while it reduces cognitive complexity it does not affect the sensitivity of income-related questions. Based on these findings, recommendations are proposed for improving the design of surveys that include financial questions.

depressive state · financial issues · follow-up questions · health status · household expenditure · Household income · item nonresponse · predictors of non-response · probing questions · Russia · savings · small towns · survey data quality · survey design · Decision-Making and Behavioral Economics · Survey Methodology and Nonresponse · Survey Sampling and Estimation Techniques

Tools

Open DOIOpen Access