Article brief Example | Topics and Well Written Essays - 750 words

Brief - Article Example The methodology used by the study to come up with a fraud model involved three steps. The first step was to data of a public company from 1995 to 2002. A sample of 100 companies was split into fraud and non-fraud companies (McKee, 2014). The third step was to test 15 predictive variables to determine those that were good predictors of fraud status. The fraud companies were selected from SEC Enforcement Release that provided publicly reported fraud. Non-fraud companies were matched with the fraud companies in terms of three criteria including market value, positive change of 25% in net income, and standard industrial classification (McKee, 2014). The predictive variables were derived from prior research literature. Analysis of the data involved binary logistic regression using fraud status as the variable to be predicted. The variables used when analysing the 15 predictive variables include company size, auditor tenure, and McKee-Lensberg bankruptcy probability. The fraud model was tested with 91 companies which were in the original sample. The other 9 companies not included in testing did not have data for at least one of the three variables used to analyse the predictive variables (McKee, 2014). A company with fraud probability greater than 50 was regarded as a fraud company while a company with less than 50% probability of fraud was regarded as a non-fraud company. The results of the study are that the model predicted 63 of the tested companies’ fraud status correctly. This reflected 69.2% level of accuracy. The model also predicted fraud status of 28 companies incorrectly, reflecting 30.8% rate of error. This model can be compared favourably to the fraud model developed by Alden et al (2012) which showed 75% accuracy of training rates and 64% accuracy of validation. This article is important because it provides a fraud model that can be used by auditors to develop standards