Improving analytical procedures: A case of using disaggregate multilocation data

According to SAS No. 56, Analytical Procedures, the use of disaggregate, in dividual location data can improve the effectiveness of analytical procedur es used in multilocation audits. Using a case-study approach, we investigat e whether improvements in the accuracy and precision of account balance exp ectations can be obtained by using disaggregate, individual location data i n a large, multilocation company. Specifically, we examine two issues: (1) whether the summation of individual location expectations generates more ac curate and precise expectations of company-wide account balances than expec tations based on company-wide data only and (2) whether the accuracy and pr ecision of analytical procedures is enhanced by including peer location obs ervations of the account balance in individual location expectation models. We find that for the multilocation company examined in this case study the summation of individual location account balance expectations is not more a ccurate or precise than an expectation derived from aggregate models unless the individual location models include peer location observations of the a ccount balance. When the individual location models include the same accoun t observations from other peer locations within the company, the company-wi de account balance expectations developed from disaggregate models are more accurate and precise (less variable) than expectations developed using agg regate, company-wide data only. The results from this case study indicate t hat when auditors are generating expectations of company-wide balances, dis aggregate models incorporating peer location account observations provide a ccount balance expectations that are both more accurate and more precise th an company-wide, aggregate models. Given the limitations of a case-study ap proach, future research should be directed at establishing the generalizabi lity of these findings.