Design and assessment of AI-based learning tools in higher education: a systematic review

Abstract

Artificial intelligence (AI)-based learning tools are increasingly integrated in higher education, offering benefits such as personalized learning experiences, real-time feedback, and increased flexibility. However, effective design and implementation strategies for these tools are not well established. This study addresses this gap through a systematic literature review with two main objectives: (1) to summarize the design features of AI-based learning tools currently employed in higher education, focusing on aspects such as algorithm types, training datasets, modes of information presentation, and their roles in the learning process; and (2) to assess their impacts on college students’ cognitive, skill-based, and affective learning outcomes. Our review encompasses 63 peer-reviewed articles published between January 2014 and April 2024. Notably, approximately half of the reviewed studies employ publicly available AI systems for instructional purposes (n = 32), while the other half develop proprietary AI-based learning tools (n = 31). 26 studies use AI techniques to generate and deliver multimodal learning materials. Moreover, we identify three primary roles of AI in higher education: assessment and evaluation (n = 45), personalized feedback and recommendations (n = 46), and intelligent tutoring (n = 26). In terms of learning outcomes, although AI tools generally improve cognitive knowledge acquisition and affective outcomes, their effectiveness in developing cognitive process and skills varies significantly. Lastly, we provide recommendations for optimizing the design and implementation of AI-based learning tools in higher education and outline promising directions for future research in this field.

Ähnliche Arbeiten