The purpose of evaluation is not always clear to those who have watched surveys conducted, data entered, and then reports written which are promptly filed away never to be seen again. The only thing to show for the whole exercise is that money, which could have been used to expand the program, has now disappeared into this evaluation and is now no longer available. This story is most common when evaluations are imposed by others.

If, on the other hand, those responsible for making decisions about how to design their program, or those who decide which programs to implement, have critical questions, evaluations can help them find answers. An evaluation is most useful when the program heads or policymakers are driving the discussion of what should be evaluated. Further discussion can be found in the Why Evaluate section.

Perhaps as frustrating as an evaluation that asks the wrong questions is one that asks the right questions but produces unreliable answers. Significant amounts of money, time, thought, and effort will go into figuring out what the right questions are. It is not too much to ask for accurate answers. In many cases, if the wrong evaluation methodology is used, even the fanciest statistical techniques cannot correct its mistakes. A randomized design can help ensure that the answers are reliable.