RMP.3:2 - Problem

A selected design leaves room for operational choices, but not every remaining choice is harmless. Replacing a missing comparison reach, changing a construct’s measure or recruiting only convenient cases can change the question the study is capable of answering. Teams often discover the problem after collecting material and then describe the actual arrangement as though it were the intended design.

The opposite response freezes every detail before the first encounter with data. That can be appropriate for a particular confirmatory commitment, but can obstruct legitimate qualitative, exploratory or source-critical development. A protocol must distinguish useful advance commitments from choices whose controlled evolution serves the inquiry.

How can a practitioner make the design executable, preserve its criticism and allow appropriate adaptation without either hiding a defeated premise or imposing a universal research checklist?