Evaluating health promotion: a tale of three errors.

The main purpose of this article is to question the relevance of the Randomised Controlled Trial for the evaluation of health promotion programmes. In its concern to manage Type 1 error, the RCT underestimates or virtually ignores Type 2 and 3 errors. Because of the peculiar complexities of health promotion programmes and the importance of gaining insights into the effect of interventions--rather than merely recording whether or not they achieve their goals--a new kind of validity is needed. The central assertion here is that we should adopt a principle of "judicial review" which is based on a broad spectrum of triangulated evidence.