Evaluation, particularly of complex or multi-component community prevention projects, is a struggle. For a start, there are so many ways to avoid doing a really good evaluation that will tell us something about a program.
Those who are enthusiastic about evaluation may put together a bevy of fancy instruments that we have always wanted the opportunity to try, and/or that look good up front to the funder, but then struggle to be able to sort out something statistically meaningful from the tangle of results. At worst we may have little skill at evaluation, but not want anyone to know.
So where to from here? Well first we want to argue that we have an obligation to evaluate. Then we want to walk you through two programs that have demanded careful thought, yet still involve us in a struggle. Finally we propose an overarching schema, drawing on literature from other fields of human endeavour.