Prompting Behavior and Human-AI Interaction: Insights Into Learning Dynamics and Critical Engagement in Higher Education



Author Information

Chrysanthi Melanou, Freiburg University of Education, Germany

Abstract

The increasing integration of generative Artificial Intelligence (AI) tools in higher education raises essential questions about how students engage with content, regulate their learning, and develop critical thinking skills in AI-augmented environments. This paper presents results from the second phase of a longitudinal, quasi-experimental study conducted in a dual study program in Germany. The first phase (N = 93) quantitatively examined the impact of AI-supported learning on knowledge gain, motivation, cognitive load, critical thinking, and reflective use across three measurement points (T1, T2, T3) conducted throughout the semester. These findings provided the empirical foundation for the second phase. In phase two, a particular focus was placed on prompting behavior as a potential behavioral indicator of underlying learning processes. At mid-semester (T2), one intervention group (N = 32) was systematically observed with regard to prompt behavior, frequency, preferred AI use-cases, and whether outputs were revised or adopted directly. These variables were descriptively analysed and explored in relation to the findings from phase one. Preliminary patterns suggest that prompting behavior may be meaningfully associated with deeper learning dynamics, including motivation, critical engagement, and over-reliance on AI tools. Students who revised AI outputs more frequently also tended to score higher in critical thinking and reflective use. These findings highlight prompting behavior as a meaningful indicator of student critical engagement with AI and self-regulated learning.


Paper Information

Conference: ACP2026
Stream: Psychology and Education

This paper is part of the ACP2026 Conference Proceedings (View)
Full Paper
View / Download the full paper in a new tab/window


Comments & Feedback

Place a comment using your LinkedIn profile

Comments

Share on activity feed

Powered by WP LinkPress

Share this Research

Posted by James Alexander Gordon