The associations between students' attitudes toward AI and learning engagement: serial mediating roles of perceived autonomy and learning enjoyment.

Introduction: Artificial intelligence (AI) tools are increasingly integrated into higher education to provide real-time feedback and personalized learning support. While previous research has primarily examined the impact of AI attitudes on students' AI usage behavior, less is known about how students' attitudes toward AI influence their psychological experiences and learning behaviors. This study investigates the associations between students' attitudes toward AI and their learning engagement, focusing on the serial mediating roles of perceived autonomy and learning enjoyment. Methods: A questionnaire was administered to 425 university students. All variables were measured with validated Likert-type scales adapted from established instruments. To reduce common method variance, data were collected at two time points and Harman's single-factor test was performed. Regression analyses, bootstrap mediation testing, and propensity score matching were conducted to examine the proposed serial mediation model and address potential self-selection bias. Results: A significant positive correlation exists between students' attitudes toward AI and their learning engagement. Perceived autonomy mediates this relationship, with a notable mediation effect of 0.177. Learning enjoyment also plays a mediating role in linking students' attitudes toward AI and their engagement in learning, with a significant and relatively strong mediating effect of 0.115. A serial mediation effect involving both perceived autonomy and learning enjoyment is observed, with a smaller but still significant effect value of 0.021. Furthermore, the use of propensity score matching helps control for self-selection bias, thereby enhancing the robustness of the findings. Discussion: The findings offer empirical insights into the motivational and emotional mechanisms linking attitudes toward AI and engagement, thereby informing the design of AI-enhanced learning environments to support autonomy, enjoyment, and active participation.