Audit Opinion Identification Using Accounting Ratios: Experience Of United Kingdom And Ireland

D. Zdolsek, T. Jagric
Several prior research studies have presented a model for audit opinion identification using a chosen combination of the company's accounting ratios as derived from financial statements. However, from the methodological viewpoint, there is one questionable issue: the appropriate performance evaluation of developed model taking into account auditors' subjective preferences. In the present study, a model for identification of qualified audit opinions using logistic regression is developed and validated using ROC curves and AUC values in combination with misclassification costs as performance criterion. Considering auditors' subjective preferences the model's performance results based on these performance criteria are adequate.
285-299.pdf [1.08 Mb] (Count: 30)