情報処理学会 インタラクション2026

Expressing opinions on product satisfaction, research surveys, and market analysis has become an everyday task in our society. These statistical analyses provide important information on market research and evaluation tools. However, traditional rating systems such as the Likert scale often fail to capture nuanced feedback, as they are limited to binary or ordinal selection. The lack of adaptability can lead to survey fatigue and introduce unwanted agreement or cultural bias into the analysis. The ordinal nature of the Likert scale makes it especially difficult to use on mobile devices. We previously introduced the modified four-quadrant gradient rating scale (mFQS), a visual rating system that enables simultaneous evaluation of two dimensions through spatial interaction. Our comparative study indicated that the mFQS produces broader rating distributions and reduces cultural bias among Japanese participants compared to Likert scales. This work explores possible applications for multidimensional rating scales to broaden the landscape of rating systems in human-computer interaction research.