Agentic education · Learning science

Actionable companion · Feedback & scaffolding

Learning with agency: Does artificial intelligence-supported rhythm-enhanced phonological awareness training improve early reading in children with attention-deficit/hyperactivity disorder? A randomised controlled intervention study

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): Early childhood · preschool, Children · stage unspecified. Verify the participant population and task before transferring the findings.

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

Abstract Introduction Children with attention-deficit/hyperactivity disorder (ADHD) are at increased risk of reading-related difficulties, and phonological awareness (PA) plays a vital role in early reading development. Rhythm may serve as a temporal scaffold for PA, yet whether rhythm-enhanced PA training produces additional benefits beyond typical PA training for children with ADHD remains unclear. This study examined whether rhythm-enhanced PA training, delivered through an artificial intelligence (AI)-supported and research-assistant-facilitated interactive teaching model, could improve early reading-related abilities and rhythm discrimination in Mandarin-speaking kindergarten children with ADHD. Methods A total of 165 eligible kindergarten children with ADHD were stratified by gender and randomly allocated in a 1:1:1 ratio to the rhythm-based phonological awareness (RPA) intervention group, the phonological-awareness-only (PAO) intervention group, or the control group. Of these, 146 children completed the intervention and were included in the final analysis. Both intervention groups received AI-supported instruction for 8 weeks, with two 30-min sessions per week, while the control group received AI-supported oral storybook activities. Outcome measures included character reading, four PA measures (syllable, rhyme, tone, and phoneme-level awareness), and rhythm discrimination. Results Significant time × group interactions were found for character reading, all four PA measures, and rhythm discrimination (all ps < 0.001, ηp2 = 0.11–0.53). Bonferroni-adjusted comparisons showed that both intervention groups outperformed the control group on syllable, rhyme, tone, and phoneme-level awareness (ps < 0.05). The RPA group further outperformed the PAO group in character reading (mean difference = 6.92, p = 0.003), syllable awareness (mean difference = 1.72, p = 0.004), rhyme awareness (mean difference = 1.54, p < 0.001), and rhythm discrimination (mean difference = 3.11, p < 0.001). Discussion Within the shared AI-supported and human-facilitated interactive teaching model, PA intervention improved PA-related outcomes, while rhythm enhancement provided additional benefits for character reading, syllable awareness, rhyme awareness, and rhythm discrimination. This study further extends the Temporal Sampling Framework by demonstrating that rhythm may serve as a temporal scaffold for PA and early reading development in children with ADHD. Keywords: ADHD, AI, phonological awareness, reading, rhythm

What kind of evidence is this?

Supporting AI learning science · Randomized study reported

Stage: Early childhood · preschool; Children · stage 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 recruited using convenience sampling from 27 kindergartens in one city in China that were accessible to the research team. All 27 kindergartens approached by the research team agreed to participate. Online recruitment posters were distributed through the participating kindergartens’ parent communication platforms to the parents or guardians of all 4,651 enrolled children. A total of 227 parents or guardians responded and reported that their children showed possible signs of ADHD. These children were subsequently invited to undergo formal diagnostic and eligibility assessment conducted by a licensed psychologist. Children were excluded if they had an identified comorbid psychiatric, neurodevelopmental, neurological disorder that could independently affect their participation in or response to the intervention. The excluded conditions included eating disorders, autism spectrum disorder, anxiety disorders, and neurological disorders. Comorbidity status was determined on the basis of a licensed psychologist’s clinical assessment and parent-reported medical and developmental history.

Limitations and uncertainty · source passage

This study also has several implications. First, the results suggest that AI-supported delivery can be feasibly incorporated into structured PA intervention when trained human facilitation is available, especially when the target group (i.e., children in this case) require repeated practice, consistent prompts and immediate feedback. Second, RPA activities may be more useful than typical PAO training when the intervention goal is to improve not only PA skills but also character reading and rhythm discrimination. This is based on the fact that the RPA group in this study showed greater gains than the PAO group in character reading, syllable awareness, rhyme awareness and rhythm discrimination. From a practical perspective, kindergarten-based PA intervention for children with ADHD may benefit from embedding rhythmic chanting, clapping, tapping and syllable-to-beat mapping into AI-supported PA activities. In addition, the study results showed that RPA training may provide temporal scaffolding for children’s processing of syllabic and rhyming structures. At the same time, the findings suggest that rhythm effects are domain-specific, as RPA produced additional benefits for syllable awareness and rhyme awareness, but not for tone awareness or phoneme-level awareness. More broadly, this study extends AI-supported language intervention research by showing the feasibility of using AI as a standardised delivery component within a RA-facilitated interactive teaching intervention model.

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.
  • Limit simultaneous instructions and unnecessary dialogue. Compare task performance and a suitable mental-effort measure; easier-feeling work is not automatically better learning.

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. Use a brief adult-mediated story, turn-taking or language activity. An educator controls every interaction; the agent never becomes an unsupervised companion. Preserve hands-on play and human conversation. 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. Measure independent naming, explanation or turn-taking in a human-led activity after the tool is removed. Record frustration, overstimulation and the amount of adult help. 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

Linzi Zhang; Fangya Huang; Panpan Li. Does artificial intelligence-supported rhythm-enhanced phonological awareness training improve early reading in children with attention-deficit/hyperactivity disorder? A randomised controlled intervention study. Frontiers in Psychology (2026-09-03). https://doi.org/10.3389/fpsyg.2026.1906514

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.