A regression model test of the overall correctness of financial statements

Publisher Website
Google Scholar
Abstract
Focus is placed in this paper on the statistically audited accounts of a financial statement. It is shown that individual accounts can be materially correct and yet the overall financial statement can be materially in error. The measure of overall materiality of error which is proposed consists of a statistical test of the discrepancy between the auditor's estimates and the stated book values. A regression of the auditor's estimated account values is run on the stated book values of the same accounts. The null hypothesis that the intercept is zero and the slope is unity is tested at the ,05 level of significance. Standard linear hypothesis testing procedures are used and the resulting statistical test is the Snedecor F-test. Rejection of the null hypothesis implies materiality of error whereas acceptance of the null hypothesis implies that the overall financial statement is correct. Materiality of error is defined statistically. If the proposed measure of materiality is adopted, the auditor is absolved from subjectively defining a dollar value measure of materiality.

This publication has 0 references indexed in Scilit: