To perform the analysis, confirmation respondents were classified into two groups: 1) accounting clerks (accountants) performing accounts payable duties for business; and 2) consumers responding to personal confirmation inquiries.

Performance rating in the confirmation process is obtained by examining error rate, error amount, and error type. An error is defined as the difference between the book value and the audited value of an account.

Sample Data

To compare respondent behavior, data was obtained from audits of accounts receivable populations. Only positive confirmations were used because it is generally recognized that positive confirmations are more reliable in the detection of errors.

The sample data was obtained from the working papers of two regional and seven international CPA firms. Table 1 shows the number of errors, confirmations, and audits for each class of respondent.

In two-thirds of the audits studied, confirmations were returned by accountants. These confirmations comprised less than half of those sent and yet detected almost three-fourths of the errors.

Error Rate

The most striking and possibly the most important differences noted were the error rates. The error rate was obtained by dividing the total number of errors reported by the total number of confirmations sent. As indicated in Table 1, for the accountants, 207 errors were reported from 4,258 confirmations for an error rate of 4.86%; for the consumers, 86 errors were reported from 5,130 confirmations for an error rate of 1.68%. The proportion of errors reported by accountants was 2.9 times larger than those reported by consumers.

These results support general knowledge that accountants are more likely than consumers to detect errors. This difference is attributable to the fact that accountants perform account analyses regularly and are generally more adept at detecting and reporting errors.

Error Amount

To reduce the impact of larger account balances, the data sets for accountants and

TABULAR DATA OMITTED

consumers were matched by client size as measured by total accounts receivable. Audits were retained in the analysis if the total accounts receivable for one respondent class were within approximately $1 million of the other respondent class. The matching had the effect of reducing the number of audits in the accountant analysis from 93 to 85 with the number of errors going from 207 to 130. The number of audits in the consumer analysis went from 47 to 45 with the number of errors going from 86 to 83.

The results of the matched data analysis are more conclusive. The evidence indicates that the average error amount detected by accountants does not differ significantly from those detected by consumers. This finding is not surprising because error amount would seem to be more a function of the system error-generation process rather than respondent ability, with the possible exception that accountants would tend to work with larger account balances.

Error Type

Confirmations can detect only certain error types. For example, they can detect whether a credit sale transaction was properly valued but cannot detect whether credit was approved. Eight types of errors can be detected by confirmations. A listing of error types and frequency of occurrence by class of respondent is contained in Table 2. Some errors would not be identified in the audit working papers as to error type, and thus were included in the table as "error type 9, unidentified errors." The analysis of type 9 errors is excluded from this discussion.

From an examination of Table 2, it can be seen that the first three error types were the most commonly detected. The infrequent occurrence of the remaining error types prohibits meaningful analysis.

It appears that consumers are better able to detect unrecorded collections than are accountants and are marginally better at detecting unrecorded returns and allowances. Accountants appear to detect errors of sales valuation better than consumers. This suggests that consumers are more concerned with decreases to their accounts while accountants are more concerned with increases to their company's accounts.

Some possible explanations for the above are as follows: 1) consumers are more aware of the bills that they pay and notice when a payment is omitted because it is their money and they personally write the check, while an accountant does not have the personal involvement because it is the company's money and the checks are usually prepared through an automated system; 2) using the same logic, consumers are more concerned that sales returns and allowances are recorded; 3) accountants typically utilize a more sophisticated system for controlling purchases, which facilitates detection of sales valuation errors.

TABULAR DATA OMITTED

James G. Swearingen, PhD, CPA, Professor of Accounting, Weber State University; Floyd A. Wilkes, PhD, CDP, Associate Professor, Weber State University; and Sandra L. Swearingen, MPA, CPA, Instructor of Accounting, Weber State University.