Analysis Of Compliance Constructs Using Behavioral Indicators From Online Surveys

Abstract

In the United States, job fraud committed by an employee of an organization remains a serious, ongoing predicament. In order for a company’s compliance program to be effective and detect violations, they must first understand the motivations of maleficence. In our experiment, we aimed to determine if responses and behavioral indicators from an online compliance survey distributed to working individuals can predict malicious acts and explore the level of financial gain that motivates cheating and fraud. We examined this research question by analyzing the results of a compliance survey and card game task and developed four models for detecting fraudsters. We found that reducing the number of features improved the predictive model but adding mobile data did not. In future work, researchers can continue to analyze mobile data, such as mouse movements, to better understand fraud perpetrators, which we believe could help improve the effectiveness of compliance programs in the coming years.