Generative Artificial Intelligence in higher education: A systematic review of student use and learning outcomes

Abstract

This systematic literature review of 39 peer-reviewed empirical articles outlines a usage framework comprising seven ways higher education students have utilized Generative Artificial Intelligence in their learning tasks since January 2018 and examines the resulting actual and perceived learning outcomes. Findings indicate that students’ actual learning outcomes achieved through Generative Artificial Intelligence were predominantly successful, while perceived outcomes vary, presenting a mixed picture of success and challenges. Specifically, when students used Generative Artificial Intelligence as a translator, refiner, navigator, evaluator, dialoguer, or self-regulatory supporter, they perceived higher success-to-challenge ratios in learning effectiveness, learning efficiency, interactivity, self-regulation, and personalized learning. By contrast, using Generative Artificial Intelligence as a creator resulted in an approximately equal proportion of successes and challenges. Notably, Generative Artificial Intelligence as a self-regulatory supporter resulted in the lowest success-to-challenge ratio among all usage types, with the few challenges attributed to insufficient holistic integrated self-regulated learning skills. The findings suggest that students need to enhance their comprehensive self-regulated learning capabilities to optimize Generative Artificial Intelligence use in learning task.