Disclosure of sensitive behaviors across self-administered survey modes: a meta-analysis

• Published in: Behavior research methods Vol. 47; no. 4; pp. 1237 - 1259
• Main Authors: Gnambs, Timo, Kaspar, Kai
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.12.2015; Springer Nature B.V
• Subjects: Acoustic Stimulation; Behavior; Behavioral Science and Psychology; Cognitive Psychology; Computers; Disclosure; Effect Modifier, Epidemiologic; Experiments; Humans; Meta-analysis; Motivation; Polls & surveys; Psychology; Reproducibility of Results; Social Behavior; Surveys and Questionnaires; Systematic review; Computer; Paper-and-pencil; Survey; Self-disclosure; Sensitive question
• ISSN: 1554-3528; 1554-3528
• DOI: 10.3758/s13428-014-0533-4

In surveys, individuals tend to misreport behaviors that are in contrast to prevalent social norms or regulations. Several design features of the survey procedure have been suggested to counteract this problem; particularly, computerized surveys are supposed to elicit more truthful responding. This assumption was tested in a meta-analysis of survey experiments reporting 460 effect sizes (total N =125,672). Self-reported prevalence rates of several sensitive behaviors for which motivated misreporting has been frequently observed were compared across self-administered paper-and-pencil versus computerized surveys. The results revealed that computerized surveys led to significantly more reporting of socially undesirable behaviors than comparable surveys administered on paper. This effect was strongest for highly sensitive behaviors and surveys administered individually to respondents. Moderator analyses did not identify interviewer effects or benefits of audio-enhanced computer surveys. The meta-analysis highlighted the advantages of computerized survey modes for the assessment of sensitive topics.