Explaining learning engagement in AI-assisted learning through teacher support and achievement goals: insights from goal-setting theory

Synthesis: Explaining learning engagement in AI-assisted learning — a structural equation modeling study of 758 Chinese university students in AI-assisted English learning. Guided by goal-setting theory, Wang & Wang (2026) show how teacher support drives learning engagement through four achievement goal orientations, extending goal-setting theory to AI contexts.

Wang & Wang (2026) address an under-explored question: not just whether AI helps language learning, but the psychological mechanisms that drive students' engagement in AI-assisted environments. The study integrates teacher support, achievement goals, and engagement into one framework, and tests which goal orientations mediate the teacher-support → engagement link.

Method

• Design: Structural equation modeling of survey data.
• Sample: 758 Chinese university students in AI-assisted English learning.
• Framework: Goal-setting theory, with four achievement goal orientations (mastery-approach, performance-approach, mastery-avoidance, performance-avoidance).

Key Findings

• Teacher support predicted the approach goals (mastery-approach and performance-approach) but not the avoidance goals (mastery-avoidance, performance-avoidance), and directly enhanced engagement.
• Engagement predictors: mastery-approach, performance-approach, and performance-avoidance goals all positively predicted engagement.
• Mediation: only mastery-approach and performance-approach goals served as significant mediators between teacher support and engagement.

What this means for practice

• Instructors. Keep teacher support central in AI-assisted courses: it directly enhanced engagement in this 758-student model, so AI feedback complements rather than replaces the instructor's motivational role.
• Instructors. Frame AI-assisted work with approach goals — mastery-oriented goals about understanding the material and performance-oriented goals about reaching a target standard — because only mastery-approach and performance-approach goals mediated the link between teacher support and engagement.
• Instructors. Check which goals your assessments and tasks reward: performance-avoidance goals also predicted engagement but were not the channel through which teacher support operated, so encourage approach goals rather than relying on students' fear of falling behind.
• Designers. Instrument AI platforms to capture behavioral log data, not only survey responses, so courses can separate the quantity from the quality of engagement and detect superficial, avoidance-driven AI use.
• Administrators. Fund professional development that strengthens teachers' support behaviors in AI-assisted classes, since that is the pathway the study identifies to students' approach goals and engagement.

Limitations

• All data came from self-report questionnaires collected in October 2025 through convenience sampling, with English instructors circulating the survey link to their classes; the authors note the responses may be subject to social desirability bias and may not capture actual behavior with AI tools.
• The cross-sectional design prevents causal inference and cannot show how motivational processes evolve, so the authors call for longitudinal or experimental designs.
• The final sample of 758 came from 813 submitted questionnaires, with 55 cases (6.77%) dropped for implausibly short response times; eligibility required prior experience with AI-assisted English learning, so students with no such exposure are absent.
• Engagement quality was not measured: the authors state the study cannot distinguish quantity from quality of engagement, including whether avoidance-driven AI use produces more superficial learning outcomes.

Citation

Wang, Y., & Wang, Y. (2026). Explaining learning engagement in AI-assisted learning through teacher support and achievement goals: insights from goal-setting theory. Humanities and Social Sciences Communications, 13(1215).