Abstract This article is a report on an experiment directed at determining the extent of nonsampling errors in accounts receivable confirmation procedures. In an ordinary audit situation it is not practical to measure the extent of improper responses. Auditors generally accept a confirmation at its face value without questioning the respondent further, however, under various circumstances, the auditor will look beyond the confirmation for satisfaction. The proportion of nonresponses to confirmations which ask the recipients to confirm the information provided on the confirmations is equal to the proportion of nonresponses to confirmations which ask the recipients to provide the information concerning their accounts. The acceptance of the hypothesis might lead to a solution to the problem of improper responses. Unless the recipient contacts the office being audited, the data on the confirmation would be obtained from his records. It is proposed here that the immediate solution to the problem of nonsampling errors is for the auditor to use the blank confirmation form and circularize a number sufficiently large to yield the number of responses needed to satisfy him.
Eugene Sauls (Sat,) studied this question.