AI-Assisted Autonomous Learning and Reduced Academic Accomplishment in Vocational Higher Education: The Mediating Role of Hardiness

Synthesis: Wang and Zhang (2026) examined how AI-assisted autonomous learning relates to reduced academic accomplishment among 1,264 vocational college students in China, focusing on the mediating role of hardiness (commitment, control, challenge). Using structural equation modeling, they found AI-assisted autonomous learning was negatively associated with hardiness and positively associated with reduced academic accomplishment, with hardiness partially mediating the relationship — the indirect effect accounted for a substantial proportion of the total effect. The study extends research on AI in higher education by suggesting AI-assisted learning has implications beyond efficiency and convenience for students' psychological resources and academic self-evaluation.

Key Findings

• AI-assisted autonomous learning was negatively associated with hardiness and positively associated with reduced academic accomplishment among vocational college students.
• Hardiness was negatively associated with reduced academic accomplishment and partially mediated the AI-learning → reduced-accomplishment relationship, with the indirect effect accounting for a substantial proportion of the total effect.
• The study frames AI-assisted autonomous learning as a double-edged condition: it can support efficiency, information access, and personalized assistance, but may also reduce students' opportunities for independent thinking, effortful problem solving, and the experience of self-generated mastery.
• Reduced academic accomplishment — a core dimension of academic burnout reflecting negative self-evaluation of competence and learning effectiveness — was the focal outcome, particularly salient for vocational students whose confidence is tied to skill development and professional identity.
• The authors call for studying AI beyond technology acceptance and learning efficiency, toward its associations with psychological resources and academic self-perception.

Study Design & Method

The study collected survey data from 1,264 students at a vocational college in China and analyzed the relationships among AI-assisted autonomous learning, hardiness, and reduced academic accomplishment using structural equation modeling (SEM). AI-assisted autonomous learning was operationalized as students' active use of generative AI tools to obtain information, complete tasks, solve problems, and regulate learning. Hardiness (Kobasa's commitment/control/challenge disposition) was treated as the proposed psychological mechanism, and reduced academic accomplishment (a Maslach-derived dimension of academic burnout) as the self-evaluative outcome. The mediation model tested both the direct path and the indirect path through hardiness.

What this means for practice

• Instructors. Keep moderately challenging work in AI-supported courses and require students to explain, reflect, revise, and decide on their own, because AI-assisted autonomous learning was associated with lower hardiness and hardiness carried 53.1% of the association with reduced academic accomplishment (indirect B = 0.111 of a total effect of 0.209).
• Instructors. Use AI to prompt thinking rather than replace it — prompting questions, step-by-step hints, alternative perspectives, and reflective feedback instead of complete solutions — so students in vocational programs, where competence comes from repeated practice, still reach mastery through their own effort.
• Administrators. Treat AI-supported study as an academic-adaptation matter as well as an efficiency one: build challenge orientation, control beliefs, and persistence into advising, counseling, and student-development programs alongside digital competence.
• Administrators. Judge AI-supported courses by students' psychological resources and academic self-evaluation rather than time saved or tasks completed, since the association with reduced accomplishment held with hardiness only partially mediating (direct B = 0.098), and the concern extends to over-reliance and reduced effortful Problem Solving.

Limitations

• The cross-sectional design precludes causal inference: the negative associations observed cannot establish that AI use causes reduced accomplishment.
• All constructs rely on self-report, and the sample is drawn from a single vocational college in China, bounding generalizability.
• Hardiness is treated as a relatively stable disposition, which may understate its plasticity over time.
• The study did not distinguish between different types or intensities of AI use, which could moderate the observed relationships.

Citation

Wang, W., & Zhang, Q. (2026). AI-assisted autonomous learning and reduced academic accomplishment in vocational higher education: The mediating role of hardiness.