Audit quality measurement model using factor analysis technique, structural equations and decision trees C5.0-C & R

The quality of audit institutions is always one of the most important things in auditing. Researchers in this field believe that the most effective variable in the discussion of audit quality is audit institutions .The purpose of this study is to present a new method for predicting and ranking the quality of audit institutions affiliated to Iranian Institute of Certified Public Accountants (IACPA). and its innovation is to achieve a highly accurate forecasting model. For this purpose ,we used data records of the quality control and status control of IACPA which had already been gathered through questionnaires from 2012 to 2017. We selected 1555 pieces of information records. After screening and deleting incomplete data, 1367 pieces of information were studied. Then, to reduce the data dimensions and find the internal and optimal pattern of variables set. Factor analysis and structural equation techniques were applied. The analysis revealed three categories paly ing a major role in quality of audit. These categories, constructs, were labelled process indicators , environmental or contextual indicators and input indicators. Finally, we used decision tree algorithm from data mining techniques, namely, C&R, and C5.0 algorithms. The results showed that C5.0 algorithm with 92% accuracy and sensitivity, and 97% detection power is the best performance for predicting audit quality and The C&R algorithm is able to predict the qualitative ranking of auditing institutions with 83% accuracy and sensitivity and 93% detection power