A Writing Study Can Find Better Scores and Weaker Motivation
A question to investigate
Can automated feedback help a student revise as well as a teacher’s comments? A randomized study published on 30 October 2026 offers one answer, but its results show why “as well as” needs more than one definition. The researchers compared revised-essay quality and motivation among undergraduates receiving AI feedback, teacher feedback, or no substantive feedback. Their findings support a focused claim: in this setting, both kinds of feedback improved essay quality relative to the control, while teacher feedback was associated with stronger intrinsic motivation than AI feedback.
Name the study design
The study assigned 312 undergraduates to three conditions; 287 participants were included in the analyses. Random assignment helps make a comparison more informative because it is designed to distribute pre-existing differences across groups. Still, readers need to know what outcome was measured and when. Here the article reports a single feedback-and-revision cycle at one institution. That scope matters. The result concerns revised essays within this study, not every kind of writing, every student, or long-term development across a school year.
In the margins Why this passage works +
“That scope matters.”
This short transition reminds the reader that sample and setting limit the reach of a result.
Separate the outcomes
The authors report that both the AI-feedback and teacher-feedback groups produced higher-quality revised essays than the group without substantive feedback. They also report that the AI and teacher groups were statistically equivalent within a prespecified margin for writing performance. “Equivalent” here describes a defined statistical comparison, not proof that every comment was equally accurate or useful. The essay should keep the outcome attached to the method: this trial compared revised-essay quality under its conditions; it did not settle every question about feedback.
Keep motivation distinct
Performance and motivation tell different parts of the story. In the same trial, teacher feedback elicited higher intrinsic motivation than AI feedback, although AI feedback also exceeded the no-feedback condition on that measure. A writer who reports only the essay-quality result would miss a separate outcome. A writer who says the study proves teachers are always more motivating would go too far in the other direction. The evidence supports a comparison among these groups in this study, not a universal ranking of feedback sources.
Respect the analysis
The authors discuss feedback engagement and perceived relatedness to the feedback source as possible pathways associated with outcomes. Their analysis says the indirect effects are associational rather than causal. That wording is important: even in a randomized experiment, not every relationship examined inside the study has automatically been randomized. A reader should not convert “associated with” into “caused by.” This is a useful model for writing about research: identify which part of the design supports a causal comparison and which analysis remains correlational.
In the margins Why this passage works +
“A reader should not convert “associated with” into “caused by.””
The sentence makes the study’s own evidence-status distinction explicit.
State the study’s limits
The authors call for more research across institutions and repeated feedback cycles. A strong conclusion should preserve that limit while stating what the experiment contributes. One defensible summary is that, during a single revision cycle at one institution, AI and teacher feedback produced comparable essay-quality results within the study’s equivalence test, while motivation favored teacher feedback. That statement is informative without turning the finding into a policy rule. It also points to a next question: whether combining rapid automated suggestions with human explanation changes either outcome over time.
Build a paragraph around one outcome
Use a topic sentence that states one result, follow it with the exact evidence and comparison, then explain what the comparison means. If a study measures multiple outcomes, give each outcome its own paragraph or clearly signal the shift. Keep the study’s sample and setting near the claim so a reader can judge its scope.
Write two bounded summaries
Write one sentence about the essay-quality result and a second about motivation. In each sentence, name the comparison and setting. Then add a sentence identifying a limitation. A sound response distinguishes the trial’s performance comparison from its motivation finding and avoids a claim about all students or all feedback situations.
Try the idea.
Write two bounded summaries
See the self-check +
Write one sentence about the essay-quality result and a second about motivation. In each sentence, name the comparison and setting. Then add a sentence identifying a limitation. A sound response distinguishes the trial’s performance comparison from its motivation finding and avoids a claim about all students or all feedback situations.
Sources and evidence
- As effective as the teacher, but less motivating: a randomized controlled comparison of AI-generated and teacher feedback on students’ writing performance and intrinsic motivation · Frontiers in Psychology
- Anatomy of a Body Paragraph · Harvard College Writing Center