Acceptance of AI-Assisted English Language Learning Tools in Higher Education: Psychological Correlates Across Disciplinary and Proficiency Groups

Synthesis: Wu et al. (2026) examined how learning motivation, self-efficacy, anxiety, and risk perception relate to acceptance of AI-assisted English language learning in a Chinese higher-education context, building on the Technology Acceptance Model (TAM). Drawing on survey data from 210 undergraduates (STEM = 91, Humanities = 119; English proficiency Low = 77, Intermediate = 103, High = 30), they found that learning motivation and self-efficacy were consistently and positively associated with acceptance-related indicators (perceived usefulness, perceived ease of use, behavioral intention, and satisfaction), while anxiety and risk perception showed more nuanced, generally weaker patterns that must be interpreted cautiously. The study provides a learner-centered account emphasizing psychological correlates and descriptive group-level differences rather than causal claims.

Key Findings

Learning motivation and self-efficacy were consistently and positively associated with acceptance-related indicators across the sample.

Anxiety did not show a uniformly negative pattern — its positive associations with acceptance outcomes were exploratory concurrent patterns and should not be read as evidence that anxiety is beneficial.

Risk perception was positively but generally more weakly associated with acceptance outcomes, and the authors caution against over-interpreting this association.

Mean-level comparisons showed descriptive differences in several acceptance outcomes across academic discipline and English proficiency groups, with exploratory subgroup-specific regressions describing within-group patterns (not formal between-group tests).

Analyses used SPSS 27.0 and Stata 18.0 with robust-standard-error adjusted linear regressions, mean-level comparisons, post hoc tests, and exploratory subgroup regressions; findings are associative, not causal, given the cross-sectional design.

Study Design & Method

The study surveyed 210 undergraduates from a Chinese university, with balanced disciplinary (STEM vs. Humanities) and proficiency (Low/Intermediate/High English) groups. Instruments measured AI-assisted English learning acceptance outcomes (perceived usefulness, perceived ease of use, behavioral intention, satisfaction) alongside learning motivation, self-efficacy, anxiety, and risk perception. Analyses included descriptive statistics, reliability and convergent validity, full-sample adjusted linear regressions with robust standard errors, mean-level group comparisons, post hoc comparisons, and exploratory subgroup-specific regressions. The authors were explicit that subgroup regressions describe within-group patterns rather than testing between-group coefficient differences, and that the cross-sectional design supports associative rather than causal interpretation.

Implications for AI in Education

The findings argue that acceptance of AI-assisted Language Learning tools is not purely a matter of perceived usefulness or ease of use — it is shaped by learners' psychological resources (Motivation, self-efficacy), affective states (anxiety), and evaluative judgments (risk perception), which vary descriptively across disciplines and proficiency levels. For practitioners, this supports designing AI-assisted English learning that builds learner confidence and motivation rather than merely optimizing usability, and tailoring support to disciplinary norms and proficiency groups. It also cautions against simplistic assumptions that anxiety always blocks or risk perception always deters AI adoption. The study grounds a learner-centered extension of TAM for Generative AI English learning, complementing system-centered technology acceptance accounts.

Limitations

The cross-sectional design precludes causal inference, and all constructs rely on self-report. The sample (N = 210) is a single-institution Chinese higher-education sample, bounding generalizability to other contexts and populations. Subgroup analyses were exploratory and descriptive, not formal tests of between-group differences. The counterintuitive associations of anxiety and risk perception with acceptance underscore the need for replication and for cautious interpretation of the descriptive patterns observed.

Connected Concepts

Language Learning

Generative AI

Motivation

Self Regulated Learning

Higher Ed

Personalized Learning

Teacher Role

Connected Articles

GenAI Pronunciation Feedback Wtc 2026 — Associations Between Generative AI-Based Pronunciation Feedback and Willingness to Communicate

AI Guided Learning Audiovideo 2026 — AI-Guided Learning: Research on Knowledge and Skill Acquisition

GenAI Feedback Design Multisite Experiment — Human-Centered GenAI Feedback Design in Higher Education

AI Tools Arab English Classrooms — AI Tools in Arab University English Classrooms

Will Skill Not Tool Chinese University Students Acceptance Of Generative AI For — Will, Skill, Not Tool: Chinese University Students' Acceptance of Generative AI

Citation

Wu, J., Wang, Y., He, Y., Yin, X., Chen, F., & Wan, B. (2026). Acceptance of AI-assisted English language learning tools in higher education: Psychological correlates across disciplinary and proficiency groups.