Writing a limitation that strengthens your argument.

Students treat the limitations section as a confession, so they write it as vaguely as possible and hope nobody dwells on it. Experienced reviewers read it first, and for the opposite reason: a precise limitation is evidence that the author understands their own design.

A useful limitation does three things. It names the specific threat, says which conclusion it affects, and says what would settle it. Vagueness fails all three.

Rewrite one

Weak: "The sample size was small, which limits generalizability."

This says nothing. Small compared with what, and generalizable to whom?

Rewritten: "All 96 respondents came from a single department, and 71% were in their second year. The relationship between feedback timing and revision may hold only where assignment structure is this consistent; a multi-department sample would show whether it survives across different assessment formats."

Same admission. Now the reader knows exactly how far to carry the finding, and you have handed your successor a study to run.

Rewrite two

Weak: "Self-reported data may be subject to bias."

Every self-report sentence in every paper. Which bias, in which direction?

Rewritten: "Revision effort was self-reported after marks were released, so students who received higher marks may recall having worked harder. This would inflate the association we report. Timestamped draft data would remove the recall problem entirely."

Naming the direction of the bias is the move. It shows you know whether the error works for or against your own argument / and admitting it inflates your result is far more persuasive than pretending it is symmetrical.

Rewrite three

Weak: "Due to time constraints, only 10 interviews were conducted."

Time constraints are not a limitation, they are an excuse. And ten interviews is not automatically too few.

Rewritten: "Ten interviews were enough to reach repetition on the three themes reported here, but not to explore the fourth pattern that appeared in the final two conversations, where students described avoiding feedback entirely. That pattern is reported as an observation rather than a finding, and would need a purposive sample of students who missed submissions."

Notice this limitation increases the value of the paper. It flags a genuinely interesting thread and is honest that the data cannot carry it yet.

What does not belong in there

• Things you could have fixed. A broken question in your survey is an error, not a limitation. Fix it or report it as an error.
• Ritual apology. "This study is not without limitations" is filler. Start with the limitation.
• Limitations that undo the paper. If a threat genuinely destroys your main claim, the fix is to weaken the claim, not to footnote the collapse.

Where limitations really come from

They are decided in the method, months before you write them. When you choose one department because it is the one you can access, you have written your first limitation. Keep a running file from day one / every compromise, dated, one line. By the time you write the section it is already drafted, and you will have caught two or three you would otherwise have forgotten.

---

The test: read your limitation and ask whether a reader could now design the study that answers it. If yes, it is doing real work. If it could be pasted into any paper in your field without editing, delete it and write the specific one.