#694: Encouraging Deeper Learning with GenAI: A Review of Pedagogical Constraints on GenAI and Their Impact in Higher Education

This paper reviews how generative AI (GenAI) can be pedagogically constrained to promote deeper learning in higher education. Analyzing 15 intervention studies comparing constrained GenAI use with free or no GenAI use, we found three delivery methods (process, activity, and tool levels), operationalized through five mechanisms. Results show primarily positive impacts on performance, motivation, and higher order thinking, though nearly half rely on self-reports. The findings highlight the role of constraints in GenAI use, and invite further research explicitly linking constraints to outcomes.