When Does a Replication Change the Decision?

A result that fails to repeat is easy to label “disproved.” But before changing a research-informed plan, a team still needs to know what differed: the method, the setting, or the people involved.

What would you want checked before treating a failed replication as a reason to revise the decision? For example, would you first compare the original and repeat procedures, look for an independent attempt, or ask whether the finding matters in the team’s actual context?

The answer affects more than confidence in a result. It determines whether the next step is to pause, adjust the plan, or gather one more piece of evidence—and who should own that review.