Exploring the Role of Response Time in Item Response Theory: Rethinking the PISA 2022 Creative Thinking Assessment

Abstract

This article explores the potential of response time-item response theory (RT-IRT) models to enhance creative thinking (CT) measurement in Programme for International Student Assessment (PISA) 2022, which employs traditional IRT models (2PL and generalized partial credit) excluding response time (RT). Given traditional IRT's limitations in capturing cognitive processes, we explored how numerous advanced IRT models, particularly those that incorporate RT as valuable information, have been developed for model testing and comparison for large-scale assessment. In addition, we discuss the critical role of RT as demonstrated in the research literature, linking it to key aspects of creativity. We also explore the possibilities of connecting two types of RTs (i.e., the total time spent on task completion and the time from start to first action spent on each task) to patterns in creative performance across domains and stages, using RT-IRT models. Benefits and types of RT-IRT models (e.g., joint, diffusion, mixture) are further examined as they integrate RT to model speed–accuracy trade-offs, detect aberrant response behaviors, and enhance result interpretability by reflecting engagement and creative processes. Lastly, we propose RT-IRT models to leverage PISA 2022's RT data and provide process-oriented insights, improve ability estimates, and potentially prevent misclassification of spontaneously creative responses as careless ones.