GenAI Dependency Risk: Insights through the lens of mindset, GenAI literacy, and academic stress

Gökçearslan Ş., Aktan M. C., Yılmaz B., Gökçearslan E.

NEW DIRECTIONS FOR CHILD AND ADOLESCENT DEVELOPMENT, cilt.2025, ss.1-15, 2025 (SSCI)

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 2025
  • Basım Tarihi: 2025
  • Dergi Adı: NEW DIRECTIONS FOR CHILD AND ADOLESCENT DEVELOPMENT
  • Derginin Tarandığı İndeksler: Scopus, Social Sciences Citation Index (SSCI), Academic Search Premier, ASSIA, Child Development & Adolescent Studies, Educational research abstracts (ERA), ERIC (Education Resources Information Center), EBSCO Education Source, Linguistics & Language Behavior Abstracts, MEDLINE, Psycinfo, Social services abstracts, Sociological abstracts
  • Sayfa Sayıları: ss.1-15
  • Hacettepe Üniversitesi Adresli: Evet

Özet


Introduction: Generative Artificial Intelligence (GenAI) dependency is an emerging concept highlighting the over-reliance on AI tools, which may hinder students' advanced thinking skills. Academic stress levels and mindset shape the way students utilize these tools.

Objective: This study examines predictors of GenAI dependency, including growth mindset, GenAI literacy, and academic stress, addressing a significant gap in the literature on GenAI dependency in education.

Methods: Using convenience sampling, data analysis, and structural equation modeling, a questionnaire survey was administered to a sample of 276 Turkish university students (67.4% female, 83% undergraduate students), including the GenAI Dependency Scale, GenAI Literacy Scale, Academic Stress Scale, and Mindset Theory Scale.

Results: GenAI literacy increases, GenAI dependency increases (r=.310, p<0.01). As growth mindset increases, GenAI literacy increases (r=.275, p<0.01). Additionally, as academic stress increases, the growth mindset also increases (r = 0.227, p < 0.01). Although these relationships are significant, they are of low strength.

The findings suggest that both GenAI literacy and academic stress contribute to higher levels of GenAI dependency, while a growth mindset appears to ease academic stress and also foster GenAI literacy. Interestingly, male students and those studying at public universities reported greater dependency overall. In addition, students who used GenAI more frequently tended to show stronger signs of dependency.

Conclusion: Universities should introduce training programs and mandatory orientation sessions to help students use GenAI tools ethically and mindfully. To ensure stronger pedagogical and disciplinary relevance, institutions can design learning scenarios and case studies that integrate GenAI support. In addition, personalized feedback systems and self-regulation tools could be embedded into GenAI applications to better guide students’ learning processes. Finally, future research would benefit from a broader methodological approach, including longitudinal designs and the use of secondary data sources.