In this paper, the author considers certain aspects of the problem of obtaining unbiased information about the merits of a program or product, whether for purposes of decision making or for accountability. The evaluation of personnel, as well as the evaluation of proposals and evaluations, generally involves a different set of problems than those which the author considers here. However, some points made here will apply. Throughout, efforts are made to consider both the credibility and the validity of an evaluation, the former being the audience's estimate of the latter. The author begins by looking at some typical important practical cases of bias in program evaluation. (Contains 4 figures.)