Modeling Motivated Misreports to Sensitive Survey Questions

Ulf Böckenholt*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

Asking sensitive or personal questions in surveys or experimental studies can both lower response rates and increase item non-response and misreports. Although non-response is easily diagnosed, misreports are not. However, misreports cannot be ignored because they give rise to systematic bias. The purpose of this paper is to present a modeling approach that identifies misreports and corrects for them. Misreports are conceptualized as a motivated process under which respondents edit their answers before they report them. For example, systematic bias introduced by overreports of socially desirable behaviors or underreports of less socially desirable ones can be modeled, leading to more-valid inferences. The proposed approach is applied to a large-scale experimental study and shows that respondents who feel powerful tend to overclaim their knowledge.

Original languageEnglish (US)
Pages (from-to)515-537
Number of pages23
JournalPsychometrika
Volume79
Issue number3
DOIs
StatePublished - Jul 1 2014

Keywords

  • item response models
  • response set
  • self-deceptive enhancement
  • socially desirable responding
  • survey research

ASJC Scopus subject areas

  • Psychology(all)
  • Applied Mathematics

Fingerprint Dive into the research topics of 'Modeling Motivated Misreports to Sensitive Survey Questions'. Together they form a unique fingerprint.

Cite this