An Integrated Difference-in-Differences and Text Mining Approach for Evaluating MOOCs Improvement

Evaluating the implementation effectiveness of the first-class course policy has important theoretical and practical implications for improving online course quality. This study adopts a learner-centered perspective, using online review data to construct a difference-in-differences model for assessing the policy's impact on course quality. The results show that the policy has effectively promoted the overall quality of first-class online courses, with significant improvements in content quality and instructor explanation, but no noticeable impact on supporting services. Further analysis reveals that the policy's impact on course quality varies significantly across disciplines. For natural science courses, improvements were observed across all policy-related quality attributes, including overall quality, content quality, instructor explanation, and supporting services. However, for humanities and social science courses, only modest improvements were observed in content quality and instructor explanation, with no significant enhancement in supporting services or overall quality.