The methods section as a reproducibility artifact
Lesson 1 of 6 · Manuscripts and reviewer responses
1 · Learn the move · Fact-check list
The reader of your methods section is someone trying to reproduce the work - a reviewer probing it, a student in another lab three years from now, or you in eighteen months. Write for that reader and the section becomes a checklist target: software with versions, every parameter and threshold as actually applied (from the filter log and the repository, not from memory), n at every stage with exclusions and the reason for each, reagent sources, instrument settings, statistical tests named with their justification. The model runs the checklist against your draft and marks each item PASS or GAP; every GAP becomes [NEEDS: ...] in the text rather than a plausible default, because a methods section that reads complete but is not is worse than one with visible holes.
Run a reproducibility fact-check on my methods draft: [paste]. Checklist: software + versions / parameters and thresholds as applied / n at every stage with exclusions and reasons / reagent sources / instrument settings / statistical tests named and justified. Mark each PASS or GAP with the line it checks against my filter log: [paste log]. GAPs become [NEEDS: ...] in the text - never fill one with a typical value.
2 · Your turn. You write the prompt
Your methods section was written in three sittings across four months, and the analysis changed twice in between. The threshold in the text says 0.05; you have a memory of relaxing it. The submission checklist wants the section final by Monday, and the only trustworthy record is the filter log in the repository.
Remember: the AI sees only your prompt, not this page. If the situation isn't in your prompt, it doesn't exist.
Optional. These shape the output when you run your prompt below, not your score.