Identifying and measuring information content of financial statements in audit reporting adjustment

Document Type : Research Paper


1 Department of Accounting, Damavand Branch, Islamic Azad University, Damavand, Iran

2 Department of Statistics, Firoozkooh Branch, Islamic Azad University, Firoozkooh, Iran


Making economic decisions and allocating resources optimally without the presence of valid and reliable data is not possible. Capital flows towards superior economic activities when capitalist decisions rely on timely, relevant and reliable information. In this regard, auditing plays a vital role in determining the validity of information; in other words, given the accountability requiring the presence of valid and reliable data, it can be stated that auditing is one of the fundamental accountability processes. In the current research, an optimum prediction method for independent auditor's report types is selected and two approaches of the J48 algorithm and random forest are compared. This research has been conducted on 84 corporates during 2008-2017. In order to train, test and investigate the research variables, Weka software was used. The dependent variable is the auditor's report type. Results indicated that the accuracy of the J48 algorithm has been 72.61% and 60.42% in training and test sections, respectively and the accuracy of the random forest has been 94.57% and 63.09% in training and test sections, respectively; so, the random forest model is more effective.


[1] G. Asadi, Investigation of a relationship between profitability and liquidity in corporates and its effect on dividends, Manag. Message J. 28 (2008), 133–155.
[2] M.A. Bagherpor Valashani, M.J. Saie, A. Moshkani and M. Bagheri, Predicting independent auditor report in Iran: data mining approach, J. Account. Audit. Res. 5 (2013), no. 19.
[3] B. Barkhordarian, S.A. Hashemi and S.M. Hosseini, Predicting conditional auditing comments using multilayer neural perceptron network and decision tree, M.A. Thesis, Isfahan University, 2011.
[4] P. Dechow and I. Dichev, The quality of accruals and earnings: The role of accruals in estimation errors, Account. Rev. 77 (2002), 35–59.
[5] K. Efstathios, C. Spathis, A. Nanopoulos and Y. Manolopoulos, Identifying qualified auditors’ opinions: A data mining approach, J. Emerg. Technol. Account. 4 (2007), no. 1, 183–197.
[6] C. Gaganis, F. Pasiouras and M. Doumpos, Probabilistic neural networks for the identification of qualified audit opinions, Expert Syst. Appl. 32 (2007), 114–124.
[7] C. Gaganis, F. Pasiouras, C. Spathis and C. Zopounidis, A comparison of nearest neighbors, discriminant and logit models for auditing decisions, Intell. Syst. Account. Finance Manag. 15 (2007), 23–40.
[8] Y. Hasas Yegane, M.T. Taghavifard and F. Mohammadpor, Using probable neural networks to identify auditor comment type, Theory Practice J. 1 (2014), no. 1, 131–159.
[9] M. Tabatabaee, D. Javid and M.R. Moazami Godarzi, Investigating j48 algorithm capability in prediction and comparison of bankruptcy with neural networks and Logit and Probit Binary, M.A. Thesis, Islamic Azad University, 2014.
[10] D. Zdolsek and T. Jagric, Audit opinion identification using accounting ratios: Experience of United Kingdom and Ireland, Actual Prob. Econ. 115 (2011), no. 1, 285–310.
Volume 14, Issue 8
August 2023
Pages 119-127
  • Receive Date: 02 August 2022
  • Revise Date: 02 October 2022
  • Accept Date: 15 October 2022