Plagiarism detection software: Promises, pitfalls, and practices

Abstract

An increasing number of students at universities around the world seem to be submitting plagiarized texts to their instructors for credit, although no exact figures are available, either for how much was plagiarized in the past or how much is plagiarized now. Instructors, overwhelmed with an ever-increasing workload, wish for a simple method ‘ rather like a litmus test ‘ to quickly sort out the plagiarized works, so that they can concentrate their efforts on the rest of the students. The good news is that software can be used to identify some text parallels that could constitute plagiarism. The bad news is that the reports are often not easy to interpret correctly, software can flag correctly referenced material as non-original content, and there are cases in which systems report no problems at all for heavily plagiarized texts, as studies conducted by the author in 2004, 2007, 2008, 2010, 2011, 2012 and 2013 have repeatedly shown. Different systems have also been shown in these studies to report various amounts of plagiarism for identical texts, as they use individual, often proprietary, algorithms and sometimes only examine samples of the text under investigation. This chapter will examine the promises and pitfalls of technology-based plagiarism detection and look at some good practices for using such software in a university setting.