In response to persistent limitations in college English instruction in non-English-speaking contexts, this study develops an integrated pedagogical framework combining Task-Based Language Teaching (TBLT), the ARCS motivational model (Attention, Relevance, Confidence, and Satisfaction), Moodle-supported learning activities, and selected intelligent-computing tools for instructional analysis. A Chinese vocational university is used as the instructional context. The framework is designed to connect authentic or quasi-authentic language tasks with motivational scaffolding, structured online resources, learner feedback, and data-informed evaluation. The study examines learner engagement, perceived comprehensibility, English-skill performance, motivation-related indicators, score dispersion, and a structural model of the proposed teaching framework. The results presented in the source materials indicate higher levels of reported participation and classroom attention in the ARCS-supported condition than in the comparison condition. Skill-level comparisons also show gains in several reading- and discourse-oriented indicators, while some listening and speaking outcomes remain comparatively resistant to improvement. The structural model suggests that attention, confidence, satisfaction, intrinsic motivation, and perceived relevance are interrelated, although the reported relevance-to-motivation path is not statistically significant (\(p=0.218\)). The findings are interpreted descriptively because the supplied material does not provide a complete participant register, raw dataset, validated measurement protocol, or full inferential-statistical specification. The study therefore supports the pedagogical value of integrating task design, motivational support, and digital learning resources while calling for more rigorous future evaluation of AI-assisted adaptive learning and data-driven instructional optimization.
