Generative AI for developing higher-order thinking skills——A review

Abstract

Against the backdrop of rapid advancements in artificial intelligence technology, generative artificial intelligence (GAI) is being increasingly applied in education; however, its impact on students’ higher-order thinking skills remains a subject of significant debate. To address this, this paper employs a systematic review and meta-analysis approach in accordance with the PRISMA guidelines. It selected 62 empirical studies from China and abroad examining how GAI promotes the development of students’ higher-order thinking skills (HOTS). Using a framework that includes learning objectives, GAI platforms and functions, and types of learner HOTS, the study elucidates the underlying mechanisms through which GAI fosters the development of learners’ HOTS. The findings reveal that: (1) Existing research primarily focuses on six major scenarios—including writing, programming, and oral dialogue—to foster seven categories of learners’ higher-order thinking skills; (2) In terms of overall effects, the effect size (Hedges’ g) of GAI on students’ HOTS development was 0.831 ( p= 0.000<0.001), indicating statistical significance. Regarding moderating effects, the moderating effects of subject and experimental duration were significant, while those of learning style and GAI type were not significant.