Examining open-ended survey responses between web and face-to-face
In a recent study published in Public Opinion Quarterly, Faculty Associate Christopher Antoun and colleague Stanley Presser analyzed responses from the 2016 American National Election Study. The study compared data gathered through web-based self-administered questionnaires and face-to-face computer assisted personal interviews (CAPI) using independent samples. The focus was on three categories of open-ended questions: identifying the most important national issues, explaining why someone is for or against political candidates, and expressing preferences and criticisms in political parties.
Initial findings from this study indicated that respondents who completed web-based questionnaires tended to provide shorter answers with fewer ideas compared to those interviewed face-to-face with CAPI. Despite these differences, the researchers found that both survey modes yielded similar distributions of substantive answers. Additionally, they found that there was no significant difference in the civility of responses between web-based and interviewer-administered formats. Antoun notes that "these findings call into question the conventional wisdom that open questions are not well suited for self-administration and suggest that researchers should consider more frequent use of open questions in Web surveys."
Antoun and Presser conclude that incorporating open-ended questions more frequently into web-based surveys could potentially enhance the depth of data collected online. Using this approach could contribute to a broader understanding of respondent perspectives and improve the insights obtained through web-based surveys.
Christopher Antoun, Stanley Presser. (2024). "Open Questions Self-Administered on the Web versus Interviewer-Administered in Person: The 2016 American National Election Study Mode Comparison". Public Opinion Quarterly, 88(2), 249–267. https://doi.org/10.1093/poq/nfae012