Web-based surveys are widely used in the social sciences, including psychology. However, such surveys often include inattentive respondents—individuals who answer without carefully reading the questions—which can distort analytical results. Although various detection methods have been proposed, many rely on Likert-scale items. Because some surveys do not include Likert scales—and because such items may introduce directive or unnatural content—there is a need to develop alternative detection methods that do not depend on Likert formats. This study examined inattentive-respondent detection across multiple-choice, free-response, and ranking formats while preserving natural survey content. Study I demonstrated that the proposed detection methods were ineffective for free-response and ranking formats; therefore, subsequent analyses focused exclusively on multiple-choice items. Within multiple-choice items, however, a “choose X from N" format showed promising detection performance but identified an excessively large proportion of respondents as inattentive. In Study II, the items were refined into behavior/fact-based questions, resulting in a method that maintained strong detection performance without over-identification. Thus, the primary contribution of this study is the development and validation of a detection method suitable for multiple-choice surveys that do not rely on Likert scales. The results further indicate that applying the proposed method improves the validity of psychological measures, such as the Big Five scale. Future research should examine its applicability in broader psychological survey contexts. • Proposed a new inattentive respondent detection method. • Method preserves respondent comfort and works in surveys without Likert-scale items. • Behavior/fact-based items reduce inattentive response rates and cognitive burden.
Ozaki et al. (Tue,) studied this question.