No abstract

The politics of "bean counting" often trump the methods of counting. As a result, program fate is sometimes decided independent of the count's results. In The Potpourri case, our evaluator friend faces the question: What if the count's results are actually going to be used to make program decisions? Should I weave the best possible scenario I can with the available data? Of course, positive results that appear inflated can call into question the evaluator's competence and, consequently, do more harm than good to the program.

Editors' notes

No abstract