"There are many approaches to evaluating public health communication programs, all of them struggling to resolve the tension between making strong inferences and making sure that an intervention has gotten a fair test. There will always be some way to question the inferences made or the generality of the results to other contexts. That does not take away from the legitimacy of the evaluations. The fair question for them is whether they have gone reasonably down the path toward reducing uncertainty. A valuable study is one that can usefully inform the policy community about whether the intervention approach is worthy of support, without promising that there is no risk of a mistake. A study is valuable if future judgments about programs are better made taking this information into account than remaining ignorant of it."