Effectiveness of ChatGPT in facilitating learning for students with special educational needs: An empirical study in Saudi Arabia

Study purpose This paper aims to explore the effectiveness of ChatGPT in facilitating learning for medical students with special educational needs (SEN) while acknowledging and addressing the challenges that SEN students may encounter in utilizing this technology. Methods This cross-sectional survey study assessed ChatGPT's efficacy in supporting medical students with SEN across three Saudi Arabian universities. Utilizing purposive and convenience sampling, a questionnaire was administered to 283 SEN students. Statistical analyses, including t-tests and ANOVA, were conducted to evaluate perceptions of ChatGPT's effectiveness, considering demographic factors and impairment types. Results Notable differences were observed in perceptions of ChatGPT's effectiveness by impairment type and education level. Statistically significant differences were observed among the participants with different types of impairments in relation to flexibility in communication (p = .01), scaffolding and guided practice (p = .0435), immediate feedback and reinforcement (p = .0334), visual and audio support (p = .0244), and simplified learning (p = .002) factors. For instance, individuals with communication and interaction impairments rated ChatGPT's support significantly higher for simplified learning (M = 4.39, p = .002) and visual/audio support (M = 4.08, p = .024) compared to other impairments. Education level significantly influenced perceptions across all support factors (p < .05), with diploma holders consistently rating ChatGPT more favorably. Conclusion Although by providing personalized, simplified, and scaffolded learning experiences, along with social and emotional support, ChatGPT demonstrates promising potential in enhancing learning of SEN students; it does not prove to be effective across all types of impairments.