Testing is “heads down” work. A test is worse than useless if it does not yield an accurate result; it would point you in the wrong direction. So test validity is key. But how do you achieve that validity? In Phase I you have access to very little data. All you can depend on is logic, informed by as much evidence as can be gathered.

This is best done via heuristics. Heuristics are simple strategies to form judgments and make decisions in situations where data is scant. Test validity is not binary — it operates on a continuum. Heuristic answers may not rise to the level of statistical reliability, but conclusions drawn from sound heuristic tests will be as accurate as is possible given the phase you are in. And that is enough to make progress.

‍