From Perception to Precision: Applying Multimodal Data Analytics to Assess Ideological Outcomes in Vocational College English Courses

Vollständiger Abstract

Worum geht es in dieser Arbeit?

The integration of ideological and political education (IPE) into foreign language teaching at higher vocational colleges presents a fundamental assessment challenge: how can subjective, internalized values such as cultural confidence, professional ethics, and patriotism be systematically measured and evaluated? Traditional approaches relying on teacher observation and summative testing fail to capture the developmental, multidimensional nature of ideological outcomes. This empirical study addresses this gap by developing and validating a multimodal data analytics framework for assessing IPE outcomes in English courses at a higher vocational college in China. Drawing on learning analytics theory and curriculum IPE principles, the study constructs a technical pathway for collecting and analyzing diverse data streams—including classroom interactions, online discussion posts, project-based assignments, speech recordings, and behavioral engagement metrics—over a 16-week teaching experiment involving 186 students across four classes. Using natural language processing, speech recognition, social network analysis, and behavioral analytics, the study generates individual and class-level “digital competence profiles” that visualize students’ ideological development across cognitive, emotional, and behavioral dimensions. Results indicate that data-driven assessment significantly enhances diagnostic accuracy, reduces teacher evaluation bias, and supports targeted pedagogical interventions. Student engagement scores increased by 12.4%, while the correlation between evaluation results and observed behavioral outcomes improved from r=0.43 to r=0.78 compared to traditional assessment methods. The findings contribute to the emerging field of data-driven values education assessment, offering both theoretical insights and practical guidelines for implementing multimodal analytics in vocational education contexts.