The Transformation of Legal Teaching Models Under the Obe Concept: Based on AI Multi-Modal Technology in a Chinese University

Abstract

This study examined how three levels of multimodal artificial intelligence (AI) support—primary, advanced, and integrated—were associated with undergraduate law students’ professional knowledge, practical skills, and professional development within an Outcome-Based Education (OBE) framework. A quantitative cross-sectional correlational and comparative design was employed. Data were obtained from 268 undergraduate law students in selected Chinese universities through a structured questionnaire. The measurement and structural models were evaluated using structural equation modeling (SEM), and instructional-context scores were compared using an independent-samples t-test. The measures demonstrated satisfactory reliability and convergent validity (Cronbach’s α = .84–.92; composite reliability = .86–.93; average variance extracted = .58–.71). The structural model showed an acceptable fit (χ²/df = 2.31, CFI = .94, TLI = .92, RMSEA = .056, and SRMR = .048) and explained 62% of the variance in professional knowledge, 68% in practical skills, and 59% in professional development. All nine hypothesized paths were positive and statistically significant. Integrated AI support had the strongest association with each competency, particularly practical skills (β = .45, p < .001). Students in integrated AI-supported contexts also reported higher overall competency than those in traditional teaching contexts (M = 4.12 vs. 3.54, t = 6.87, p < .001). These findings support systematic alignment of multimodal AI activities and assessments with explicitly stated legal learning outcomes. Because the design was cross-sectional, the results indicate associations rather than causal effects.