How does artificial intelligence improve ophthalmology education outcomes? The mediating role of learning motivation and self-efficacy.

The integration of artificial intelligence (AI) into higher education is rapidly transforming pedagogical approaches, offering unprecedented opportunities for personalized learning and enhanced instructional support. Within the demanding field of medical education, particularly ophthalmology, AI-powered tools such as virtual patient simulators, adaptive learning platforms, and sophisticated image recognition systems hold significant promise for improving learning outcomes. However, despite the rapid advancements in AI, a critical question persists: what are the precise psychological mechanisms through which AI utilization translates into tangible academic achievement in this specialized medical domain?

A recent cross-sectional study conducted at a university in Jilin Province, China, sheds light on this crucial interplay. The research examined 416 undergraduate ophthalmology students across two cohorts in late 2024 and early 2025. The findings reveal a positive association between AI usage and academic performance, with learning motivation and self-efficacy acting as significant mediators in this relationship. Furthermore, the study highlights the important moderating role of AI literacy, particularly in strengthening the link between AI use and learning motivation.

Unpacking the AI Advantage in Ophthalmology Education

Ophthalmology, a discipline that necessitates the synthesis of complex theoretical knowledge with precise practical skills, is particularly well-suited for AI integration. Traditional teaching methods often rely on static images and instructor narration. In contrast, AI-driven systems can provide real-time, multidimensional visualizations of pathological images, simulate disease progression, and algorithmically annotate key features. This dynamic approach allows students to grasp intricate pathophysiological mechanisms more effectively. For instance, AI-assisted retinal image recognition systems can expedite students’ understanding of conditions like macular degeneration and diabetic retinopathy, thereby enhancing their diagnostic accuracy.

Beyond visualization, AI promotes adaptive learning pathways and real-time feedback. Students engaging with AI tools for tasks such as retinal image analysis or diagnostic reasoning receive immediate feedback, enabling continuous refinement of their understanding. AI systems can also generate traceable learning data, providing instructors with valuable insights to identify teaching weaknesses and implement more precise instructional strategies. The advent of AI-driven virtual simulation technology further expands the practical dimension of ophthalmology teaching. Through virtual reality and augmented reality integrated with AI, students can practice ophthalmic procedures and diagnostic decision-making in low-risk virtual environments, accumulating high-intensity clinical skills training.

The Psychological Drivers of AI-Enhanced Learning

The study’s findings underscore that the benefits of AI in ophthalmology education extend beyond mere information delivery. A key insight is that AI usage is positively associated with academic performance (β = 0.247, p < 0.001). This direct effect, supporting Hypothesis 1, suggests that simply increasing engagement with AI tools correlates with better final examination scores. This aligns with the notion that AI can provide personalized support and enhance study efficiency, particularly in visually dense fields like ophthalmology.

More profoundly, the research demonstrates that AI usage influences academic outcomes through two crucial psychological pathways: learning motivation and self-efficacy. The study found significant indirect effects of AI usage on academic performance through these mediators. Specifically, the indirect effect via learning motivation was 0.081 (95% CI [0.059, 0.103]), supporting Hypothesis 2a. This indicates that students who utilize AI more frequently tend to exhibit higher learning motivation, which in turn contributes to improved academic performance. Similarly, the indirect effect through self-efficacy was 0.079 (95% CI [0.017, 0.141]), supporting Hypothesis 2b. This suggests that AI engagement can foster greater confidence in students’ abilities to learn and master ophthalmology concepts, leading to better academic results. These findings are consistent with Social Cognitive Theory, which emphasizes the importance of motivation and self-efficacy in driving learning and achievement.

The Crucial Role of AI Literacy

Adding another layer of complexity, the study investigated the moderating role of AI literacy—an individual’s ability to understand, use, and critically evaluate AI systems. The results revealed a significant moderating effect of AI literacy on the relationship between AI usage and learning motivation (β = 0.183, p < 0.001), supporting Hypothesis 3a. This crucial finding indicates that the positive impact of AI usage on learning motivation is amplified for students with higher levels of AI literacy. In essence, students who are more adept at understanding and utilizing AI tools are better positioned to leverage them for motivational gains. They can navigate AI interfaces more effectively, interpret AI-generated feedback more accurately, and experience less cognitive load, allowing them to focus on the learning content itself, thus fostering greater engagement and interest.

However, the study did not find a significant moderating effect of AI literacy on the relationship between AI usage and self-efficacy (p = 0.364), leading to the rejection of Hypothesis 3b. This suggests that while AI literacy enhances the motivational benefits of AI, it may not directly influence the development of general self-efficacy. One interpretation is that general self-efficacy, as measured in the study, is a more stable, trait-like construct influenced primarily by mastery experiences and generalized feedback, rather than by specific technological competencies. Future research may explore domain-specific self-efficacy (e.g., self-efficacy in interpreting fundus images) to potentially uncover a stronger moderating influence of AI literacy.

Study Methodology and Findings

The cross-sectional study involved 416 undergraduate ophthalmology students from a university in Jilin Province, China, surveyed in November 2024 and March 2025, corresponding to the autumn and spring academic cohorts. Participants completed surveys assessing AI usage, learning motivation, self-efficacy, and AI literacy. Final examination scores served as the objective measure of academic performance. Structural equation modeling was employed to analyze the data, utilizing bootstrapping for mediation and moderation analyses.

To ensure the robustness of the findings, common method bias was assessed using Harman’s single-factor test and a confirmatory factor analysis (CFA). Both methods indicated minimal bias and confirmed the discriminant validity of the constructs. Descriptive statistics revealed positive correlations between AI usage and learning motivation (r = 0.22), self-efficacy (r = 0.41), and academic performance (r = 0.38). Learning motivation also showed positive associations with self-efficacy (r = 0.30) and academic performance (r = 0.28).

The structural equation modeling results further elaborated these relationships:

• Direct Effect: AI usage directly and positively predicted academic performance (β = 0.247, p < 0.001).
• Mediating Effects:
  • AI usage positively predicted learning motivation (β = 0.053, p < 0.001), which in turn positively predicted academic performance (β = 0.159, p < 0.001). The indirect effect via learning motivation was significant (0.081).
  • AI usage positively predicted self-efficacy (β = 0.663, p < 0.001), which in turn positively predicted academic performance (β = 0.186, p < 0.001). The indirect effect via self-efficacy was significant (0.079).
• Moderating Effect: AI literacy significantly moderated the path from AI usage to learning motivation (β = 0.183, p < 0.001). Specifically, higher AI literacy amplified the positive effect of AI usage on learning motivation. However, AI literacy did not significantly moderate the path from AI usage to self-efficacy (β = 0.039, p = 0.364).

Implications for Ophthalmology Education

The findings of this study offer several critical implications for the future of ophthalmology education. Firstly, they underscore the importance of not only introducing AI tools but also ensuring students possess the necessary AI literacy to benefit fully from them. Educators should consider integrating explicit instruction on AI literacy alongside the use of AI-powered learning resources.

Secondly, the mediating roles of learning motivation and self-efficacy highlight that AI’s effectiveness is deeply intertwined with students’ psychological states. AI tools should be designed and implemented in ways that actively foster these internal drivers. This could involve incorporating features that provide personalized encouragement, celebrate incremental progress, and offer clear pathways for skill development, thereby enhancing students’ belief in their ability to succeed and their intrinsic drive to learn.

Thirdly, the differential moderating effect of AI literacy on motivation, but not self-efficacy, suggests a nuanced approach is required. While AI literacy might be crucial for maximizing the motivational benefits of AI, the development of general self-efficacy might be more closely tied to the actual learning experiences and successful outcomes achieved, regardless of the specific technological proficiency. Therefore, the focus for self-efficacy development should remain on providing ample opportunities for mastery and effective feedback.

Limitations and Future Directions

While this study provides valuable insights, its cross-sectional design limits the ability to establish definitive causal relationships. Future longitudinal or experimental studies are needed to confirm these findings and explore causal pathways more rigorously. Additionally, reliance on self-reported data for AI usage, motivation, and self-efficacy could introduce bias. Incorporating objective metrics, such as AI platform usage analytics or performance-based assessments, would strengthen future research. The use of a general self-efficacy scale may also have masked more specific effects; future studies could benefit from employing domain-specific self-efficacy measures. Finally, the study’s focus on a single university and the ophthalmology discipline necessitates replication in diverse settings and across various medical specialties to enhance generalizability.

Conclusion

In conclusion, this research elucidates the multifaceted ways in which artificial intelligence can enhance academic performance in ophthalmology education. The findings indicate that AI usage not only has a direct positive impact but also operates indirectly through the mediating roles of learning motivation and self-efficacy. Crucially, AI literacy emerges as a significant moderator, amplifying the positive influence of AI on learning motivation, thereby optimizing the learning experience for students. As AI continues to evolve, educators must strategically integrate these technologies, coupled with robust AI literacy training, to cultivate motivated and confident learners, ultimately driving improved academic outcomes in the specialized field of ophthalmology.