Data Mining Journal Entries for Fraud Detection: A Replication of Debreceny and Gray’s (2010) Techniques
Journal of Forensics and Investigative Accounting, Vol. 8(3) July-December 2016: 501-514
24 Pages Posted: 6 Feb 2016 Last revised: 10 Feb 2017
Date Written: 2016
There is limited published research to detect financial statement fraud using digital analysis to analyse journal entry data. As far as we are aware, Debreceny and Gray’s (2010) study is the first and only such study. In this study, we replicated and extended Debreceny and Gray’s (2010) work by examining generalizability of their techniques beyond subjects from USA. Besides Chi-Square test, we also explored the use of mean absolute deviation method during digital analysis. We found Debreceny and Gray’s (2010) techniques useful in facilitating cross-sectional analysis for journal entry data sets that are based on multiple organizations. Our results confirmed that their techniques offered a comprehensive and systematic way of applying digital analysis on journal entries in a new setting. Our analysis also found that researchers should not rely solely on Benford’s Law during digital analysis because of potential false alarms.
Keywords: fraud; journal entries; data mining; digital analysis; Benford’s Law
Suggested Citation: Suggested Citation