When is Computerized Continuous Auditing Less Effective at Deterring Fraud
George C. Gonzalez · D-Scholarship@Pitt (University of Pittsburgh) · 2012
Companies use a variety of techniques to deter fraudulent behavior. This study focuses on the fraud deterrent effect of computerized continuous auditing systems. Although continuous auditing systems are almost always computerized in the natural environment, this study reports the results of two experiments that separately examine the effects of continuous versus periodic auditing and manual versus computerized fraud detection. It also examines the relative effects of human versus computer-mediated communication of the audit findings. Consistent with theory from criminology, information systems, and psychology, I find that the effectiveness of a continuous audit approach depends on the actual probability of fraud detection, and that at low levels of fraud detection a continuous audit is actually less effective than a periodic audit in reducing the perceived opportunity to commit fraud. I find no differences based on whether fraud is detected by a manual or computerized system, but find moderate support that face-to-face communication of audit findings creates more discomfort in a potential fraud perpetrator than does computer-mediated feedback. Contrary to my predictions, I do not find that individuals’ perceptions translate into corresponding effects on actual fraudulent behavior in my study.