Unpacking GenAI-enabled deep learning engagement: role perceptions, human-GenAI synergy strategies, and underlying...

Area of Science:

• Educational Technology
• Cognitive Science
• Artificial Intelligence in Education

Background:

• Generative AI (GenAI) is reshaping student learning experiences.
• Individual impacts of GenAI on deep learning engagement are highly variable.
• This study explores how students' perceptions of GenAI influence their deep learning.

Purpose of the Study:

• To investigate the relationship between students' role perceptions of GenAI, their synergy strategies with GenAI, and their deep learning engagement.
• To identify mechanisms that mediate the impact of GenAI on student learning.
• To uncover learner profiles and potential paradoxes in the digital learning era.

Main Methods:

• A mixed-methods approach was utilized with 181 undergraduate students in a blended course.
• Qualitative coding of reflection reports was integrated with quantitative path analysis.
• Latent profile analysis was employed to identify distinct learner profiles.

Main Results:

• Human-GenAI synergy strategies fully mediate the link between role perception and deep learning.
• A cognitive threshold for metacognitive engagement was identified when GenAI is perceived as a collaborator.
• Six learner profiles revealed a skill-cognition decoupling, and an assessment paradox emerged where exploratory learners scored lower.

Conclusions:

• Deep learning in the digital age necessitates pedagogical shifts.
• Focus should move from technical skills to reshaping GenAI role perceptions.
• Assessment paradigms must evolve to value metacognitive synergy between humans and GenAI.