Agentic education · Learning science

Actionable companion · Feedback & scaffolding

Learning with agency: The dynamic relationships among AI-assisted formative assessment in music teaching, feedback literacy, and music learning engagement

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.

Read original paperSearch paper agentCC BY 4.0

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 participant population and task before transferring the findings.

Original authors’ abstract, reproduced under the linked source license:

Abstract Artificial intelligence (AI) tools are increasingly used to provide formative feedback in music education, yet the dynamic mechanisms linking AI-supported feedback to student engagement remain insufficiently understood. Drawing on feedback literacy theory and self-regulated learning, this study examined the reciprocal within-person relationships among engagement with AI-generated formative feedback, feedback literacy, and music-learning engagement across one semester. Three-wave panel data were collected from 420 university music-education students and analyzed using a random-intercept cross-lagged panel model to distinguish within-person fluctuations from stable between-person differences. A reliability-corrected model was also estimated to account for measurement error in the composite indicators. The results showed that increases in engagement with AI formative feedback predicted subsequent improvements in feedback literacy, which, in turn, predicted higher levels of music-learning engagement. Feedback literacy therefore emerged as a key mechanism through which AI-supported formative feedback contributed to students' learning engagement. Most reverse effects were not supported, although music-learning engagement predicted later engagement with AI feedback after measurement error was corrected. At the between-person level, epistemic trust in AI feedback accounted for a substantial proportion of the associations among the focal constructs. Overall, the findings indicate a predominantly forward but partially reciprocal dynamic and highlight the importance of feedback literacy, epistemic trust, and measurement-error correction in research on AI-supported music assessment. Keywords: artificial intelligence, feedback literacy, formative assessment, learning engagement, music education, random-intercept cross-lagged panel model

What kind of evidence is this?

Supporting AI learning science · Design requires full-text appraisal

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 analytic sample comprised 420 students, with complete responses at all three waves. Recruitment targeted whole class cohorts to limit self-selection, and participation was voluntary and anonymous. Students were recruited from 22 class cohorts taught by 5 instructors at a single conservatory (approximately 19 students per class on average). The AI-assisted practice tool was a required component of the course curriculum and usage was logged by the system but not separately graded. Individual class and instructor identifiers were not retained in the de-identified analysis file, so a precise cluster-robust correction was not possible, but we use the number and average size of cohorts to bound the likely impact of clustering (§3.4, §5.5).

Limitations and uncertainty · source passage

First, three waves, while sufficient to identify the RI-CLPM, are a floor rather than a comfortable margin. They constrain what can be asked: within-person mediation can be suggested but not firmly established, the autoregressive and cross-lagged structure must be assumed stationary, and a within-person moderation by trust (H7b) cannot be credibly identified. The weak autoregression for feedback literacy also cautions that, once stable differences are removed, its residual within-person process is modest, and larger samples or more waves would estimate it more precisely. Designs with four or more waves, ideally with random slopes, are needed to test reciprocity, mediation, and moderation properly.

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.

  • Give feedback on the learner’s reasoning, then fade hints. Record whether the learner can complete the next task without that scaffold.
  • Use plan–attempt–check–reflect prompts and compare confidence with actual performance. Keep learner choice and reflection visible.
  • Measure engagement separately from achievement. Check whether novelty-driven participation persists and whether it translates to independent performance.
  • Use common rubrics, blind scoring where feasible and a fresh unaided reasoning task. Document permissible assistance and avoid treating AI-detector scores as proof of misconduct.

A 90-day district pilot

  1. 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.
  2. 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.
  3. 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.
  4. 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

He Sun. The dynamic relationships among AI-assisted formative assessment in music teaching, feedback literacy, and music learning engagement. Frontiers in Psychology (2026-08-05). https://doi.org/10.3389/fpsyg.2026.1919875

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.