Before analysis:

  1. Fix what must stay constant.
    • Quantitative: estimand, population, time window, support/inconclusive rule.
    • Evaluation/ranking: decisive criteria, exclusions, tie-breakers.
    • Research synthesis: claim and source/inclusion criteria.
  2. Identify the 2-4 defensible forks most likely to change the answer.

After the primary analysis:

  1. For each important fork, name the choice made and reasonable alternatives.
  2. Re-run alternatives that could plausibly change the conclusion.
  3. Report:
    • STABLE: conclusion survives.
    • SENSITIVE: magnitude/ranking changes but action does not.
    • FLIPS: conclusion/action changes; name the fork responsible.
  4. Report the range/distribution where meaningful, not merely the preferred run.
  5. Distinguish computational reproducibility from analytical stability.

stability-check