The General Accounting Office is responsible for auditing and evaluating Government programs. Financial statement work is important, and GAO has staff members who are very skilled at that work, but it represents a relatively small proportion of the total workload of GAO. GAO evaluators do not spend very much of their time setting down or expressing the mathematical value of things. When examining a program, both auditors and evaluators seek answers to basic questions such as: (1) what happened; (2) how does that compare to some standard; and (3) what can be done to improve performance in the future. The same questions are involved whether GAO is reviewing the management of an agency, the efficiency of program operations, or the outcome of a program. The differences between auditing and evaluating processes stem from the intellectual disciplines from which they evolved. Auditing grew out of the accounting discipline, and evaluation grew out of the social sciences. An auditor looks for particular instances of things going wrong. He places great weight on the need for accuracy of administrative records, and is likely to be suspicious of statistical inferences based on aggregate data if he cannot find actual cases confirming the general conclusion. The social science evaluator mistrusts administrative records and is predisposed to seek his own direct observations and a much more generalized view of what happened. He prefers large volumes of data and mistrusts small numbers of specific events because he has little confidence that what was found in one case is typical. In the real world, the administrative records available to the auditor are never complete and never totally reliable. The evaluator, on the other hand, is never able to observe everything that needs to be observed. Each discipline must use the techniques of the other to reach a shared solution of shared problems. Circumstances will dictate which method is more appropriate to a particular review.