Innovative learning design based on P3 task taxonomy and AI to enhance students' critical thinking

Abstract

Critical thinking (CT) and the limitations of Artificial Intelligence (AI)-based learning design have become major challenges in contemporary education. This study aims to identify students' critical-thinking challenges and to develop an AI-supported instructional framework based on the P3 Task Taxonomy. This study employed the ADDIE instructional design model and was limited to the Analysis and Design phases. The participants consisted of 42 pre-service physics teachers. Data were collected through open-ended CT tasks and analyzed descriptively based on students' cognitive response patterns. The findings revealed that 76.2% of the students predominantly exhibited System 1 thinking, characterized by intuitive responses with minimal elaboration, whereas only 20.3% demonstrated analytical System 2 reasoning. Based on these findings, an AI-supported P3 Task Taxonomy was developed that integrates Practice, Problem, and Project tasks with adaptive feedback and scaffolding mechanisms tailored to students' cognitive characteristics. Practically, the framework guides instructors in designing AI-assisted learning environments that foster CT development. Theoretically, this study extends current understanding of the relationships among cognitive systems, instructional task design, and AI support in the development of CT skills.