Preprint
Machine Learning

Unlocking Potential: Key Factors Shaping Undergraduate Self-Directed Learning in AI-Enhanced Educational Environments

Di Wu(Hubei University), Shuling Zhang(Hubei University), Zhiyuan Ma(Zhongnan University of Economics and Law), Xiao‐Guang Yue(European University Cyprus), Rebecca Kechen Dong(University of Technology Sydney)
August 29, 2024Systems87 citations
PDFDOI

87

Citations

4

Influential Citations

Systems

Venue

2024

Year

Abstract

This study investigates the factors influencing undergraduate students’ self-directed learning (SDL) abilities in generative Artificial Intelligence (AI)-driven interactive learning environments. The advent of generative AI has revolutionized interactive learning environments, offering unprecedented opportunities for personalized and adaptive education. Generative AI supports teachers in delivering smart education, enhancing students’ acceptance of technology, and providing personalized, adaptive learning experiences. Nevertheless, the application of generative AI in higher education is underexplored. This study explores how these AI-driven platforms impact undergraduate students’ self-directed learning (SDL) abilities, focusing on the key factors of teacher support, learning strategies, and technology acceptance. Through a quantitative approach involving surveys of 306 undergraduates, we identified the key factors of motivation, technological familiarity, and the quality of AI interaction. The findings reveal the mediating roles of self-efficacy and learning motivation. Also, the findings confirmed that improvements in teacher support and learning strategies within generative AI-enhanced learning environments contribute to increasing students’ self-efficacy, technology acceptance, and learning motivation. This study contributes to uncovering the influencing factors that can inform the design of more effective educational technologies and strategies to enhance student autonomy and learning outcomes. Our theoretical model and research findings deepen the understanding of applying generative AI in higher education while offering important research contributions and managerial implications.

Analysis

Why This Paper Matters

As generative AI tools like ChatGPT and adaptive tutoring systems become ubiquitous in higher education, understanding how they affect students' ability to learn independently is critical. This paper addresses a gap in the literature by moving beyond technology acceptance models to examine the interplay of teacher support, learning strategies, and AI interaction quality in shaping self-directed learning (SDL). The study is timely because many institutions are rapidly adopting AI without a clear evidence base for how to optimize these environments for student autonomy.

The research is significant for AI practitioners because it provides a validated framework for designing AI educational tools that not only deliver content but also cultivate metacognitive skills. By identifying motivation and technological familiarity as key factors, the paper offers actionable insights for product designers and curriculum developers.

Technical Contributions

  • Comprehensive factor model: Integrates teacher support, learning strategies, and technology acceptance as antecedents of SDL, with self-efficacy and motivation as mediators.
  • Empirical validation: Uses structural equation modeling on survey data from 306 undergraduates to test the proposed model.
  • Context-specific insights: Focuses on generative AI environments, distinguishing from prior work on general e-learning or rule-based AI.
  • Practical implications: Provides a framework for educators to enhance SDL through targeted support and strategy training.

Results

The study found that teacher support and learning strategies significantly increase students' self-efficacy, technology acceptance, and learning motivation. These three mediators then positively influence SDL abilities. Direct factors—motivation, technological familiarity, and quality of AI interaction—also emerged as significant. The model explains a substantial portion of variance in SDL, though exact R-squared values are not reported in the abstract. The findings confirm that improvements in teacher support and learning strategies within generative AI environments contribute to higher student autonomy.

Significance

This research bridges educational psychology and AI system design, offering a theoretical basis for creating AI tools that support self-regulated learning. For AI practitioners, it highlights the importance of designing interfaces that are intuitive (technological familiarity) and responsive (quality of interaction), while also considering the human elements of teacher involvement and strategy instruction. The study's managerial implications suggest that institutions should invest in both AI infrastructure and teacher training to maximize the benefits of generative AI in education.

generative-aiself-directed-learninghigher-educationtechnology-acceptanceteacher-supportlearning-strategiesself-efficacymotivation

Related Papers