To Cheat or Not to Cheat: Rationalizing Academic Impropriety

Abstract

Academic cheating and fraud are becoming more prevalent. The Internet removes barriers and opens access to information and increases the opportunities for academic fraud. The incentives to succeed academically also are increasing as higher education continues to grow in importance. A student's rationalization is the last, critical piece needed to spawn an academic fraud. The aim of this study, undertaken in the USA, is to explore and document the association between students' rationalizations and fraudulent academic decisions and behaviours. We use statistical tests and regression analyses to document significant associations between several rationalizations and academic cheating. We find that rationalization only influences students' assessment of the acceptability of actions when there is some ambiguity about whether or not the action is permitted. Further, we find that students are able to justify unacceptable behaviour if they believe their peers have an unfair advantage, they believe they are not getting an unfair advantage, or they feel the instructor does not care about them. Our primary conclusion is that rationalization plays a vital role in students' academic improprieties. Documenting the problem is a first step toward developing solutions. Instructors can take steps to reduce the association between rationalization and fraud. Explicitly clarifying acceptable and unacceptable resources reduces opportunities for ambiguity and fraud. Building personal relationships and connections with students can also limit certain fraudulent academic activities.

Keywords:

• Rationalization
• cheating
• fraud
• incentives
• ethics
• trust

Notes

There are a number of accounting papers, including several recent studies, that use a 100/101-point scale (DeZoort, Hermanson and Houston, 2003; DeZoort and Harrison, 2007; Koonce, Lipe and McAnally, 2008; Holt and DeZoort, 2009; Asare and Wright, 2009; Ganguly and Hammersley, 2010; Rose, Rose and Dibben, 2009; Taylor and Curtis, 2009).

To verify the validity of our results, we considered two alternative specifications. First, we considered a ‘complete’ multivariate regression model that used all four independent rationalization variables along with our control variables. The complete model provides qualitatively similar results to our reported regression results for all our rationalization variables and supports the main findings in our reported regressions. Second, we redefined each of our survey variables that were measured using a 101-point scale as either agreeing (agreement score > 75) or disagreeing. When we replicate our primary regressions, our conclusions are unchanged.