Does Respondent Self-Commitment Improve Online Survey Data Quality?

• In Event: Contemporary Issues in Public Opinion Research Methods

Abstract

Online surveys are increasingly the norm in public opinion research. Self-administered internet surveys bear opportunities, but also specific risks for data quality. In search of solutions, scholars are returning to self-commitment as one way of encouraging participants to provide accurate responses. This article provides the most comprehensive study yet of self-commitment as a potential method for improving online survey data quality. Theoretically, the effectiveness of self-commitment in motivating good conduct depends on respondents’ material and immaterial cost-benefit calculations. In four survey experiments in 2021 on more than 32,000 respondents from the general populations of ten countries worldwide, I find that self-commitments make participants less likely to cheat on knowledge questions, but are ineffective or even counterproductive with respect to respondents’ attentiveness, satisficing, and skipping. Researchers should thus use self-commitments carefully and selectively. We need additional methods to seize the opportunities of online surveys, while mitigating their risks.

Author

• Farsan Ghassim, University of Oxford