Abstract

This study compared the effectiveness of real clinical cases versus AI-generated cases in speech rehabilitation professional education. Using a quasi-experimental design, 45 sophomore students majoring in rehabilitation therapy were randomly divided into real case (n = 23) and AI case (n = 22) groups. The intervention lasted two weeks, with pre-test and post-test assessments measuring students' theoretical knowledge and practical skills across four dimensions: diagnostic, intervention, treatment, and comprehensive abilities. Results showed significant overall improvement in both groups (p < 0.001, d = 0.92). While the real case group demonstrated greater improvement in intervention skills (25.7% vs. 14.6%), therapeutic skills (19.7% vs. 8.8%), and comprehensive abilities (17.4% vs. 5.7%), both groups showed comparable performance in diagnostic skills (13.0% vs. 12.0%). The findings suggest that real cases are more effective in developing advanced professional competencies, while AI-generated cases show promise in basic skill training. This study provides empirical evidence for implementing a staged approach to case-based teaching in speech rehabilitation education, combining the advantages of both real and AI-generated cases.

Keywords: Speech rehabilitation Education, Case-based Teaching, Artificial Intelligence, Professional Competency, Teaching Effectiveness, Educational Innovation, AI-generated cases, Mixed Teaching Method