Stop Trusting the Questionnaire. Check the Track Record.

Confidence: hypothesis. The failures below are drawn from published industry data and widely reported patterns, not firsthand deployment. The pattern at the end is an argument, not a shipped system. Argue with it.

Ask someone to rate themselves and they will rate themselves well. This is not dishonesty. It is human. Which is exactly why a decision built on self-description is a decision built on the weakest evidence in the room. And yet self-description is what most selection processes run on, because it is easy to collect and it arrives in a tidy form.

Choosing which hospitals and clinics will run a clinical trial is a clear case. The main input is a questionnaire the site fills out about itself, and the questionnaire is built from hope.

The clinical why

A trial has to pick its sites: the institutions that get contracts, budgets, drug shipments, and training, on the bet that they will turn their patients into enrollment. The usual basis for that bet is the site's own questionnaire. How many eligible patients do you see. How fast can you start. How good is your team. The site answers optimistically, because it wants the study, and because a coordinator recalling a few patients who might qualify genuinely believes the number they write down.