Agentic education · Learning science

Actionable companion · Feedback & scaffolding

Learning with agency: When Do Undergraduate Students Prefer AI? Insights into AI Scoring and Feedback

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 The integration of artificial intelligence (AI) into higher education assessment has prompted growing interest in how students perceive and prefer AI involvement in scoring and feedback. While prior research has largely focused on technical performance and accuracy, this study aims to fill a gap in the literature by examining students’ preferences and perceptions regarding AI scoring and feedback, with particular attention to context, assignment stakes, and post-evaluation reflections. Ninety-three undergraduate students completed a survey consisting of Likert-type items, scenario-based questions, and an activity in which they generated AI-based scoring and feedback using ChatGPT. Results showed that students preferred structured, moderately detailed AI feedback, particularly for grammar and organization, but generally favoured human evaluation, especially for subjective tasks. While AI was seen as useful in lower-stakes contexts, concerns remained about its ability to assess more complex aspects of writing. Participants expressed a strong preference for hybrid approaches in which AI augments rather than replaces human judgment, along with a need for transparency and opportunities for human review. Collectively, these findings highlight that undergraduate students’ preferences are highly context-sensitive and role-specific, underscoring the importance of student-centred implementation strategies of AI in higher education. Keywords: artificial intelligence, student preferences, automated essay scoring, feedback, ChatGPT

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

This study employed a convenience sampling method, recruiting undergraduate students from the university’s research participation pool, who participated voluntarily in exchange for one percent extra course credit. Ethical approval for this study was obtained from the University of Alberta Research Ethics Board. Data collection was conducted at Concordia University of Edmonton. Informed consent and survey administration were conducted online through Qualtrics in Fall 2025, with participants completing the survey individually in a self-administered format. Participation was voluntary, with the option to withdraw at any time without penalty.

Limitations and uncertainty · source passage

Several limitations should be considered. The study relied on a convenience sample from a single university, with a large proportion of psychology students and female participants, which may limit generalizability across disciplines and institutional contexts. Therefore, the findings should be interpreted as reflecting the perceptions of this specific student population and may not generalize to students from other disciplines, institutions, or demographic backgrounds. Future research should extend this study to additional institutions to improve the generalisability of the findings. In addition, future research should examine whether preferences differ across fields in which assessment norms vary substantially, such as engineering, the fine arts, or professional programs. It also focused primarily on self-reported preferences and perceptions rather than long-term behavioural outcomes. Future research could examine how students actually use AI-generated feedback across a semester and whether preferences change with sustained exposure rather than a single interaction.

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.
  • 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

Seyma N Yildirim-Erbasli; Munevver Ilgun Dibek; Mackenzie L Thomas; Nicolya Lesoway. When Do Undergraduate Students Prefer AI? Insights into AI Scoring and Feedback. Behavioral Sciences (2026-07-15). https://doi.org/10.3390/bs16071196

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.