| Sumario: | Background: The integration of Generative Artificial Intelligence (GenAI) into higher education is growing rapidly, yet its impact on learning processes remains underexplored. Existing research and theories, such as social cognitive theory (SCT), largely focus on human‐to‐human learning interactions, leaving a gap in understanding how cognitive and motivational mechanisms operate in human–AI contexts. Objectives: This study investigates how GenAI features influence students' critical thinking and self‐efficacy, with a specific focus on the mediating role of cognitive engagement. Methods: Drawing on SCT, we conceptualised GenAI features—playfulness, perceived learning value and output quality—as environmental stimuli influencing student outcomes via cognitive engagement. Survey data were collected from 223 undergraduate and postgraduate students. Structural equation modelling was used to test both direct effects and the mediating role of cognitive engagement. Results and Conclusions: The results indicate that GenAI playfulness and perceived learning value significantly enhance students' cognitive engagement, which then positively affects their critical thinking and self‐efficacy. Cognitive engagement functioned as a key mediator in these relationships. However, output quality did not exhibit a significant effect, suggesting that engagement, rather than content quality alone, is crucial for fostering meaningful cognitive development. This study extends SCT by adapting it to human–AI learning contexts and provides actionable insights for designing GenAI tools that enhance learner engagement and development. Lay Summary: What is currently known about this topic? ○Critical thinking and self‐efficacy are essential learning outcomes.○Cognitive engagement supports academic motivation and deeper learning.○Most studies focus on GenAI's direct effects, not underlying mechanisms.○Current theories like SCT focus on human‐to‐human learning interactions.What does this paper add? ○Shows how GenAI features impact learning via cognitive engagement.○Identifies playfulness and learning value as key drivers of engagement.○Demonstrates cognitive engagement boosts critical thinking and self‐belief.○Extends Social Cognitive Theory to include human–AI interaction.Implications for practice or policy ○Design GenAI tools that actively promote student engagement.○Emphasise learning value and interactivity over output accuracy alone.○Train educators to integrate GenAI in ways that support reflection.○GenAI can act as a co‐learner, peer or teacher, supporting interactive learning experiences.
|
|---|