Analysis-plan to output consistency checklist

A compact review path for finding mismatches between the planned question and the result being reported.

Check the planned and analyzed population

Confirm inclusion rules, exclusions, duplicates, follow-up windows, and whether the final sample represents the planned estimand.

Check variable and model consistency

Verify coding, transformations, reference groups, outcomes, adjustment sets, interactions, dependence, and missing-data handling.

Check claim and artifact consistency

Ensure narrative values match tables and figures, diagnostics correspond to the final model, and limitations reflect any deviations.

Failure modes to check

Choosing a method because it is familiar rather than because it matches the estimand and data structure.
Treating assumptions as a generic pass/fail checklist instead of evidence about a specific model.
Changing the workflow after seeing results without recording the decision or its effect on uncertainty.
Reporting software output without tracing the claim to final code, diagnostics, and limitations.

Continue the decision path

Related decision guides

Methodological note

This guide provides general educational information. The appropriate decision depends on the study question, design, data, evidence, ethics, and institutional requirements.