Tailoring AI-powered foreign language learning: a Q method analysis of student preferences for intelligent textbook development

Abstract Despite the potential of generative AI to transform foreign language education, existing research on smart textbooks primarily focuses on technical architecture, often overlooking the subjective preferences of learners. This disconnect hinders the development of personalized and effective learning experiences. To address this gap, this study utilized Q methodology to construct a typology of learner viewpoints regarding generative AI intelligent textbooks. Based on the Jobs-to-Be-Done (JTBD) framework, thirty-two foreign language majors (P-set) sorted 48 statements (Q-set) representing various AI functions, content adaptations, and applications. Four distinct learner profiles with divergent orientations have been identified: (1) the AI-empowered Precision Learning Advocates, who are focused on algorithm-driven personalization to optimize efficiency and achieve exam-oriented outcomes; (2) the Proactive AI-Enhanced Contextual Learners, who are committed to real-world relevance and preference-aligned materials for knowledge enrichment and practical engagement; (3) the Comprehensive AI-Integrated Learning Enthusiasts, who are supportive of AI as a flexible academic assistant and dismissive of skepticism toward its pedagogical efficacy; and (4) the Goal-Driven Selective AI Adopters, who are dedicated to leveraging AI exclusively for core knowledge mastery and rejecting extraneous distracting functionalities. Each type demonstrates unique preferences for interaction methods and AI intervention levels. This study contributes to the field by providing an empirical taxonomy of learner demands in the AI era. The findings offer actionable guidelines for intelligent textbook design, suggesting that developers should move beyond a “one-size-fits-all” approach and create adaptive functional modules tailored to these specific learner profiles.