"On Self-Respect" is a popular choice for argument units because students have strong opinions about it and easy access to personal examples. That enthusiasm produces essays that are energetic but frequently unsupported. A student may declare that self-respect means never caring what others think, then illustrate the point with a single story that does not prove it. Teachers grading these papers need a way to reward reasoning over conviction.

The first difficulty is that the essay is itself an argument built from anecdote and observation rather than formal proof. Students sometimes conclude that anecdote is therefore acceptable evidence in every case. A teacher can use this as a teaching moment by asking what distinguishes an illustrative example from a proof and where each belongs in an argument. Making that distinction explicit improves the essays more than any amount of general encouragement.

The second difficulty is that students rarely address a counterargument unless the assignment demands it. Without one, the essay reads as a list of agreements. Requiring a fair statement of the opposing view, followed by a real response, changes the quality of the thinking. It also gives graders a clear place to look for evidence of intellectual honesty.

What to reward in the reasoning

Grade the chain of reasoning, not the position taken. A student who disagrees with Didion but argues carefully should outscore a student who agrees but merely repeats her phrasing. Build the rubric around whether each claim is supported, whether the connections between claims are explained, and whether the conclusion follows from what came before. This keeps the grading fair even when a teacher personally disagrees with a student's stance.

• A thesis that takes a definite position on what self-respect requires
• At least one example that actually tests the claim rather than decorating it
• A counterargument stated fairly and answered with reasons
• Accurate reference to Didion's essay without relying on summary
• A conclusion that follows from the argument instead of introducing a new one

Common scoring problems and how to avoid them

One frequent problem is the halo effect, where a beautifully phrased opening paragraph raises the score of an essay whose later sections fall apart. Another is the opposite, where a shaky introduction depresses the grade of a paper that improves as it goes. Scoring each criterion separately, and returning to the rubric after every paragraph, reduces both errors. AI grading helps here because it applies the criteria evenly from the first line to the last.

A third problem is inconsistency across graders in a shared course. Two teachers may read the same essay and disagree about whether a counterargument is real or token. Shared anchor papers and written descriptors settle most of these disputes. Running a small calibration session before the first major essay saves considerable frustration later.

Writing comments that improve the next draft

Comments on argument essays should identify the weakest link in the reasoning and suggest a way to strengthen it. For example, a note might say that the claim about independence is supported by a story about quitting a job, and ask whether the story shows independence or only impulsiveness. This kind of question pushes the student to examine their own evidence. It is far more useful than telling them to develop the idea further.

Keep the number of comments manageable. Three or four well-chosen observations do more good than fifteen scattered marks, because students can act on a short list. Drafting those observations with an AI first pass and then editing them yourself keeps the quality high while saving time. The final feedback should still sound like you.

Using the essay as a model for the unit

Didion's piece can serve as a mentor text for the argument unit itself. Ask students to identify where she makes a claim, where she offers an example, and where she complicates her own position. Seeing an established writer build an argument out of mixed materials helps them plan their own. It also gives you a shared vocabulary for feedback.

Follow the reading with a short low-stakes writing task before the full essay. A one-paragraph argument scored against the same rubric lets students see where they stand and gives you data on common weaknesses. You can then adjust instruction before the major assignment arrives. That small step tends to raise the quality of the final essays noticeably.