Abstract
The newly issued Statement on Auditing Standards No. 56—Analytical Procedures [AICPA 1988] has placed greater emphasis on analytical procedures by requiring their use in the audit of financial statements. Regression analysis may be used as an analytical procedure and has been shown through research to be an effective audit tool.
The purpose of this paper is to empirically investigate whether different levels of dispersion of the data around the regression line affects auditor decisions when using regression models as analytical procedures.
Monthly data from 15 companies in various industries used in analytical reviews by a CPA firm were used to determine the effect of data base dispersion on auditor's decisions.
The decision model used in this study is the statistical technique for analytical review (STAR) approach used in the audit practice of Deloitte, Haskin, & Sells. This study reveals data bases with larger dispersions are more likely to have incorrect rejections than those data bases with smaller dispersions (an incorrect rejection is the signaling of a month for investigation when no error exists in that month). This study also develops a rule of thumb model for the auditor as follows:
then dispersion may be too large, and results should be carefully interpreted. Regression analysis has proven to be a useful technique for auditors in the analytical review phase of the audit; however, care should be exercised to identify individual data sets which may not be suitable for application of this type of model.
Get full access to this article
View all access options for this article.