With the ongoing development of Sino-foreign cooperative programs, all-English and bilingual curricula have become vital avenues for universities to enhance students’ international academic competencies. However, in mathematics courses, the all-English delivery mode still faces challenges such as language barriers, difficulty in grasping concepts, and low levels of classroom engagement. This study takes the Li’ an Innovation Experimental Education Park as a case example and, using learning behavior. data gathered via the Ketangpai classroom platform, constructs a weighted teaching-evaluation model that integrates both formative and summative assessment. This model pays special attention to the objective evaluation of process learning, integrating the completion rates of different teaching categories through weighted methods to ensure that the evaluation is more scientific and fair. . An empirical analysis was conducted in the courses “Probability Theory and Mathematical Statistics” (delivered entirely in English) and “Advanced Mathematics” (bilingual) . Findings reveal that continuous assessment scores exhibit a significant positive correlation with both midterm and final exam results, indicating that formative evaluation can effectively predict students’ academic performance. Moreover, the weighted model based on category-specific completion rates more accurately reflects learners’ true progress than the traditional equal-weight calculation approach, and it offers instructors more targeted feedback. This research provides empirical support and practical insights for optimizing teaching models in Sino-foreign cooperative education.
