Too tired to think? Survey fatigue and the quality of financial literacy responses

How to cite

Weinel, Jette; Sakshaug, Joseph. 2025. "Too tired to think? Survey fatigue and the quality of financial literacy responses." AEA RCT Registry. AEARCTR-0017425. https://doi.org/10.1257/rct.17425-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Jette Weinel* , Joseph Sakshaug

* Primary investigator

Registry ID
AEARCTR-0017425
DOI
https://doi.org/10.1257/rct.17425-1.0
First registered
Last updated
Status
Completed
Trial dates
2025-09-30 – 2025-11-30
Countries
Albania (), Bosnia and Herzegovina (), Bulgaria (), Czechia (), Hungary (), Macedonia (the former Yugoslav Republic of) (), Poland (), Romania (), Serbia ()
JEL codes
C83, D14, G53, C93
Registry record
http://www.socialscienceregistry.org/trials/17425

Abstract

This study investigates the impact of survey design on the measurement of financial literacy in cross-national surveys. Using a randomized controlled trial (RCT) embedded in the 2025 wave of the OeNB Euro Survey across nine Central, Eastern, and Southeastern European countries, we examine whether the placement of the financial literacy module within the questionnaire influences response quality and reliability. Specifically, we analyze four research questions related to survey fatigue, gender-specific treatment effects, interviewer-induced variance, and order effects on subsequent modules. Respondents are randomly assigned to either an early-placement treatment group or a control group where the module appears at the end of the survey. The financial literacy module includes the widely used “Big Three” questions and an additional item on exchange rates. Our empirical strategy employs multinomial logistic models and multilevel specifications to assess response validity, heterogeneity by gender, and interviewer variance, complemented by tests for priming effects on subsequent questions. By addressing methodological challenges such as fatigue and interviewer bias, this study aims to improve the robustness and comparability of financial literacy measures in surveys.

Intervention

To assess whether the placement of the financial literacy module within the questionnaire influences response quality and measurement reliability, we implement a randomized controlled design in a face-to-face survey in nine CESEE countries. Respondents assigned to the treatment group receive the financial literacy module at the earliest feasible position in the survey, whereas those in the control group encounter the identical module at its conventional position near the end of the questionnaire.

Experimental design

The study employs a randomized controlled trial (RCT) embedded in the 2025 wave of the OeNB Euro Survey across nine CESEE countries. Respondents are randomly assigned to one of two groups with equal probability. The Treatment Group (Group A) receives the financial literacy module at the earliest feasible position in the questionnaire. The Control Group (Group B) receives the identical module at its conventional position near the end of the survey. The intervention involves only the placement of the module, not its content or wording. The module consists of four questions: the widely used “Big Three” items (interest rates, inflation, risk diversification) and an additional question on exchange rates.

Design details

Randomization method
We use the standard randomisation procedure established in the OeNB Euro Survey. See also Beckmann, E., M. Koch, V. Leins, S. Bergmann, K. Allinger, T. Scheiber, and J. Weinel (2025). Oenb euro survey methodology.
Randomization unit
Individuals aged 18+
Clusters
9 CESEE countries.
Observations
9,000 respondents (1000 per country)
Arms
4,500 per treatment arm (500 per country)
Primary outcomes
Survey Fatigue: Share of correct, incorrect, `don't know', and `no answer' responses to each financial literacy item Gender Gap: Difference between male and female respondents on financial literacy responses. Interviewer variance: Interviewer-level variance in responses to financial literacy questions. Measured via intraclass correlation coefficients (ICC) in multilevel models.
Secondary outcomes
Order Effect: Changes in response patterns in later modules, especially the savings module, due to cognitive priming.

Cited by

No citation data available for this registration yet.

Keywords