Actionable companion · Motivation & engagement
Learning with agency: Understanding secondary school students’ intentions to learn artificial intelligence: a multigroup structural equation modeling analysis
AI-assisted editorial implementation guide, separate from the original research. Proposed actions are not claims 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): Secondary school, K–12 · mixed / unspecified. Verify the participant population and task before transferring the findings.
Original authors’ abstract, reproduced under the linked source license:
Abstract As artificial intelligence (AI) continues to play an expanding role in everyday activities, AI-related knowledge and skills are recognized as essential 21st-century competencies. Although K-12 AI curricula have been prioritized by educational authorities worldwide, limited research has examined the motivational mechanisms that underpin students’ intentions to engage in AI learning. The present study broadens the Theory of Planned Behavior by integrating self-efficacy, attitude toward using AI, subjective norms, perceived usefulness, and AI literacy to examine students’ intention to participate in AI learning at the secondary level. The study also examines whether gender, grade level, school location, and extracurricular AI learning experience shape these relationships. Participants consisted of 632 secondary school students from Zhejiang Province, China. Students’ intention to learn AI was significantly explained by self-efficacy, attitudes toward AI use, and perceived usefulness according to SEM results. Self-efficacy and perceived usefulness also exerted significant positive effects on attitudes toward AI use. By contrast, while AI literacy and subjective norms significantly affected self-efficacy and perceived usefulness, their direct effects on attitudes toward AI use were not significant. The multigroup analysis further revealed that gender did not function as a significant moderator, whereas grade level, school location, and extracurricular AI learning experience did. Practical implications for educators and policymakers emerge from these findings, particularly for the development of inclusive AI learning programs. Keywords: AI learning, intention to learn, K-12 education, multigroup analysis, theory of planned behavior
What kind of evidence is this?
Supporting AI learning science · Design requires full-text appraisal
Stage: Secondary school; K–12 · mixed / unspecified. 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
Participants were drawn from Zhejiang Province, China, using a combination of convenience and snowball sampling methods. The survey was administered through Wenjuanxing, an online questionnaire platform commonly used in China, after obtaining informed consent from students and their parents or guardians. In total, 632 valid questionnaires were collected, and the demographic profile is summarized in Table 1. Among the respondents, 318 students (50.32%) were male and 314 (49.68%) were female. In terms of grade distribution, 182 students (28.80%) were in grade 7, 367 (58.07%) were in grade 8, and 83 (13.13%) were in grade 9. Regarding school location, 304 students (48.10%) attended urban schools, while 328 (51.90%) were from rural schools. In addition, 163 students (25.79%) reported having extracurricular experience in learning AI, whereas 469 students (74.21%) had no such experience.
Limitations and uncertainty · source passage
The interpretation of the results should take several limitations into account. First, all variables were measured using self-reported scales, which may have introduced social desirability bias and common method bias. Future studies could incorporate additional methods, such as interviews and classroom observations, to further validate the findings and enhance the robustness of the conclusions. Second, the sample was drawn exclusively from a single province in China, and snowball and convenience sampling were used, which may constrain the external validity of the findings. Moreover, given China’s collectivist cultural orientation, the functioning of subjective norms in this study may differ from that observed among students in other cultural contexts. Consequently, the generalization of the findings to other provinces, educational levels, and cultural backgrounds should be approached with caution. Future research should consider employing probability sampling or stratified random sampling and conducting cross-group comparisons across multicultural contexts to examine the generalizability of the model. Third, this study primarily used self-efficacy to operationalize perceived behavioral control (PBC). Although this approach is relatively common in educational research, it may not fully capture the controllability dimension of PBC. Future research could adopt a more comprehensive measurement of PBC by incorporating both self-efficacy and controllability, thereby further enhancing the explanatory power of the model. Fourth, this study employed a cross-sectional design, which cannot capture dynamic changes over time. Future research could adopt a longitudinal design to better examine the causal relationships among variables and their developmental trajectories.
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.
- Measure engagement separately from achievement. Check whether novelty-driven participation persists and whether it translates to independent performance.
- Co-design accommodations with learners and specialists. Check access and error burden across groups without inferring subgroup effectiveness from small or unreported samples.
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. Ask learners to attempt a problem unaided, compare an agent hint, explain a revision and verify the final reasoning against a trusted source. Make the permitted use explicit for each assignment. 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. Measure unaided transfer and delayed retention, calibration of confidence and the ability to identify an incorrect AI explanation. Record whether help-seeking becomes answer copying. 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
Yuyu Gu; Bingfu Xiong; Ruxue Li; Zhehao Pan; Yipin Huang; Yingying Pan. Understanding secondary school students’ intentions to learn artificial intelligence: a multigroup structural equation modeling analysis. Frontiers in Psychology (2026-06-23). https://doi.org/10.3389/fpsyg.2026.1833613
Companion prepared 2026-10-30. 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.