Abstract The article focuses on the problems in experimenting with the application of statistical techniques in auditing. Before statistical sampling techniques can be applied to any area, a clear statement of the purpose of sampling has to be developed. This statement of purpose is essential for the determination of the alternative decisions from which is to be made, of the criteria according to which the appropriate decision is to be chosen, and of the relevant statistical model for the population under study. Problems which arise in this initial stage may be extremely difficult ones. This paper discusses some of the problems of this initial stage in introducing statistical techniques in auditing, which must be solved before the more technical statistical problems, such as determination of the optimum sampling procedure, necessary sample size, and the like can be studied. Sampling is used extensively in auditing, for instance, in verifying transactions and in confirming balances of accounts receivable. It is important to distinguish between sampling in auditing and sampling of accounting records in general. In the latter area, some interesting results of the usefulness of statistical sampling techniques have been reported.
John Neter (Fri,) studied this question.