Theory
A result anyone can rerun counts for more than a famous name
When evidence is thin or can't be checked, medicine leans on who said it: the senior author, the famous journal, the expert panel. When anyone can rerun the analysis and test it on new data, the question shifts from who said it to what the evidence keeps showing.
Checkable results move weight from reputation to evidence
Claim: hypothesis
People trust a finding for two kinds of reasons: because of who produced it, and because of what checking it shows. When checking is expensive, the first kind wins by default. A reviewer who can't rerun an analysis judges the authors and the journal. A clinician who can't see how an outcome was defined trusts the guideline panel.
A companion repository makes checking cheap. A reviewer can rerun one result in an afternoon. A clinician can read the exact definition of "treatment response" the study used. A later team can run the method on new patients and publish what it finds. Each of those checks adds weight to the evidence and takes some away from the name on the paper.
I expect this to matter most where a finding drives care: a threshold, a dose, a device setting. Those are the findings where a wrong number reaches a patient, and where being able to check it is worth the most.
What would show this claim is wrong
- Clinicians and guideline writers who can rerun or read the analysis weigh findings no differently from those who can't.
- Findings from well-known groups are rerun and corrected no more often after their code is shared than before.
Related
- Companion Executable Code Repositories
- Versioned releases keep the evidence current while the paper stays fixed