The claim-evidence-reasoning framework is a staple of middle and high school writing instruction, and Stockton's "The Lady, or the Tiger?" fits it almost perfectly. Students must commit to a claim about the princess's choice, find evidence in the story, and then explain why that evidence leads where they say it does. The third step, reasoning, is where most student essays are thinnest, and it is also where feedback is most valuable.

Providing detailed feedback on reasoning takes time that many teachers simply do not have across a hundred or more essays. A student might write that the princess is jealous and therefore chooses the tiger, but never explain why jealousy would outweigh her love. Pointing out that gap in every essay is repetitive work, and it is the kind of work AI feedback can handle consistently.

The strongest use of AI in this setting is as a first reader that responds to the structure of the argument. It can note when a claim is missing a supporting detail, when a detail appears without explanation, or when a paragraph drifts away from the thesis. Teachers then add the human judgment about voice, nuance, and the particular student's growth.

What good feedback on a claim looks like

Feedback on a claim should help the student sharpen the position rather than simply approve it. A claim such as "the princess chose a door" is not arguable, so the feedback should push the student toward a specific stance about which door and why. Prompting students to include a reason within the claim itself, such as her fear of losing him to another woman, produces thesis statements that guide the whole essay.

• Flag claims that restate the prompt instead of taking a position.
• Point out thesis statements that cannot be supported with evidence from the text.
• Ask students to name the princess's motive within their claim.
• Suggest where a counterargument could be introduced without weakening the thesis.
• Highlight claims that are strong so students recognize what works.

Pushing students past quote dropping

A common weakness in short story essays is the quote that sits alone, with no explanation of what it shows. A student might quote the line about the princess's hand moving to the right and move on, assuming the meaning is obvious. Feedback that asks what this gesture reveals about her state of mind pushes the writer to interpret rather than report.

Effective feedback models the missing move without writing it for the student. For example, a comment might suggest adding one sentence that connects the gesture to the princess's earlier jealousy, leaving the exact wording to the writer. This keeps the thinking in the student's hands while making the expectation unmistakable.

Keeping teachers in control of the final voice

AI feedback works best when teachers review and adjust it before students see it. A teacher can quickly delete a comment that misreads a student's intent, add encouragement for a risky interpretation, or tighten the language to match classroom vocabulary. This review step typically takes a fraction of the time that writing comments from scratch would require.

Students also benefit when teachers explain how the feedback is produced and why it is anchored to the rubric. Transparency reduces suspicion and helps students treat comments as guidance rather than as a mysterious verdict. Over time, they begin to internalize the claim, evidence, and reasoning checks and apply them to their own drafts before submitting.

Turning feedback into measurable growth

Growth in argument writing becomes visible when feedback uses the same categories from one assignment to the next. A student who scored low on reasoning for the Stockton essay can be tracked on the same row for the next literary analysis, showing whether the targeted feedback made a difference. This continuity turns isolated assignments into a longer story of improvement.

Teachers can also use aggregated feedback patterns to plan instruction. If most of a class struggles to explain evidence, a mini-lesson on analysis sentence starters may do more than individual comments. Using feedback data in this way connects grading directly to teaching decisions.