An Empirical Analysis of Factors Influencing the Effectiveness of Legal Education in Chinese Universities Using Machine Learning

Legal education is essential for improving students’ legal awareness, yet its effectiveness often falls short, and the key influencing factors remain understudied using quantitative methods. This study uses survey data from university students and combines machine learning with traditional statistical approaches to identify the drivers of legal education outcomes. A questionnaire was designed to assess legal cognition and learning attitudes. Using XGBoost and SHAP values, the model revealed that active participation, teacher-student interaction, and real-life content relevance were the strongest predictors of legal cognition. In contrast, background factors such as being a law major or taking legal courses had minimal impact. These findings were validated through PCA, multiple linear regression, logistic regression, and t-tests. Logistic models struggled with class imbalance, resulting in low AUC (∼0.6), which limits their predictive utility. The results underscore that student engagement and positive learning attitudes are more important than formal background in shaping legal understanding. This study provides empirical evidence to guide improvements in legal education and suggests future research directions, including behavioral data integration, class imbalance solutions, and deep learning applications.