Actionable companion · Self-regulation & metacognition
From AI evidence to responsible practice: How generative AI enhances self-regulated learning in EFL learners: a chain mediation model of “intention to use” and “learning engagement”
AI-assisted editorial implementation guide, separate from the original research. Proposed actions are not findings or recommendations attributed to the source authors. No independent peer review of this guide is asserted.
Evidence and context
Title and abstract screening; stage labels describe reported settings, not verified numerical ages. Design labels require full-text appraisal. Reported stage(s): Higher education. Verify the participants, intervention and comparator before applying this study in a new setting.
Original authors’ abstract, reproduced under the linked source license:
Generative Artificial Intelligence is transforming English as a Foreign Language writing instruction. While much research has focused on technology acceptance, limited attention has been paid to how AI influences learners’ psychological and behavioral processes, thereby affecting self-regulated learning. This study constructs and tests a theoretical model that examines the chain mediation of “intention to use → learning engagement” to explore how generative AI impacts self-regulated learning in EFL learners. Involving 386 Chinese EFL learners from universities, the study uses a questionnaire survey and structural equation modeling for data analysis. The findings show that the perceived ease of use, usefulness, and interactivity of generative AI significantly predict both intention to use and self-regulated learning. Intention to use is related to learning engagement, which in turn fosters self-regulated learning. Moreover, generative AI enhances self-regulated learning through intention to use and learning engagement, with these factors serving as key mediators. This study illuminates the pathway from “technology perception” to “behavioral intention,” “immersive engagement,” and “regulatory process,” providing novel insights into AI-assisted second language writing learning and offering empirical evidence for designing effective human-computer collaborative EFL writing practices.
What kind of evidence is this?
Supporting AI learning science · Survey / observational study
Stage: Higher education. Screening labels describe reported context and study-design language. They are not a risk-of-bias rating or proof of effectiveness for an age band.
Study methods · source passage
The descriptive statistical analysis results (see Table 1) indicate that the mean values of the key variables ranged from 3.47 to 3.76, suggesting that participants’ perceptions of AI tools and their overall learning states were generally at a moderately high level. Specifically, participants reported higher scores for the perceived interactivity of generative AI (M = 3.71, SD = 0.91), perceived ease of use (M = 3.66, SD = 0.91), and perceived usefulness (M = 3.56, SD = 0.96), reflecting the strengths of current AI tools in terms of interactivity, ease of use, and usefulness. Additionally, intention to use (M = 3.76, SD = 0.90) was relatively high, indicating that learners were generally willing to engage in AI-assisted learning activities. In contrast, learning engagement had a slightly lower mean (M = 3.61, SD = 0.90), and self-regulated learning had the lowest mean (M = 3.47, SD = 0.92), suggesting that respondents exhibited some degree of learning engagement and self-regulated learning activities, albeit at relatively lower levels.
Limitations and uncertainty · source passage
This study developed and validated a chain mediation model, which empirically examined how the perceived characteristics of generative AI (ease of use, usefulness, and interactivity) influence self-regulated learning in EFL learners. The key findings are as follows: First, the study confirms that the perceived ease of use, perceived usefulness, and perceived interactivity of generative AI significantly and positively predict learners’ intention to use AI, supporting the applicability of the TAM in this context. Second, the structural equation modeling (SEM) analysis suggests a theoretically specified sequential relationship: the perceived characteristics of AI tools not only have a significant direct positive effect on learners’ self-regulated learning (with perceived usefulness having the strongest direct effect), but they also exert an indirect influence on self-regulated learning through the “intention to use →learning engagement” pathway. When learners perceive AI tools as easy to use, genuinely useful, and highly interactive, they are more likely to adopt them. This strong intention to use, in turn, positively relates to deeper cognitive, emotional, and behavioral engagement in writing tasks, and sustained high-level engagement provides the necessary context and motivation for the application and internalization of self-regulation strategies. Based on these findings, several concrete pedagogical and design implications emerge. For educators, we recommend designing writing tasks that require multiple rounds of interaction with AI—such as iterative drafting, feedback-seeking, and revision cycles—to sustain engagement and foster self-regulation. For technology developers, enhancing AI functions that provide explicit scaffolding for writing strategies—for instance, prompting learners to set goals, offering metacognitive questions, and visualizing revision progress—could further support learners’ regulatory development. These actionable recommendations provide practical guidance for maximizing the educational potential of generative AI in language learning contexts.
Design the system before the pilot
Use this study to inform a learning mechanism or implementation question. It is not direct evidence that autonomous agents improve learning. Keep the proposed agent behavior separate from what the source actually tested.
- Use plan–attempt–check–reflect prompts and compare confidence with actual performance. Keep learner choice and reflection visible.
- Limit simultaneous instructions and unnecessary dialogue. Compare task performance and a suitable mental-effort measure; easier-feeling work is not automatically better learning.
- Measure engagement separately from achievement. Check whether novelty-driven participation persists and whether it translates to independent performance.
A 90-day district pilot
- Days 1–15 · Appraise. Have the curriculum lead and a teacher read the methods and limitations. Record the study design, sample, comparator, outcomes and evidence gaps. Define a local question that the study can help inform.
- Days 16–30 · Design. Choose one disciplinary task. Require a first attempt, a documented AI consultation, a critique of the output and an independent explanation. Align the comparison group and scoring rubric with the same objective. Name an accountable owner and set the baseline, success threshold and budget before implementation.
- Days 31–60 · Test. Start with a small voluntary cohort, preserve an accessible non-AI route and require teacher review. Use synthetic or de-identified examples; approve data handling and tool terms through district processes.
- Days 61–90 · Decide. Blind-score reasoning and accuracy on a new task without AI. Distinguish model examination scores, student perceptions and actual learner gains; only the last directly tests educational benefit. Compare with baseline, document uncertainty and stop or revise if learning worsens, access is unequal or critical errors persist.
The board decision brief
Report the learning problem, evidence quality, participating population, total cost including review time, measured benefit, errors, access differences and remaining uncertainties. Make a dated recommendation to stop, revise or expand. A positive perception survey alone does not establish learning improvement.
Attribution and reuse
Jin Xu. How generative AI enhances self-regulated learning in EFL learners: a chain mediation model of “intention to use” and “learning engagement”. Frontiers in Psychology (2026-04-01). https://doi.org/10.3389/fpsyg.2026.1808183
Companion prepared 2026-11-01. Original authorship and DOI are preserved. The abstract is an attributed source extract; the implementation plan is new AI-assisted editorial material, offered under the same CC BY 4.0 license. This guide has not undergone independent peer review.