Plagiarism-Detection Software Stems Students' Use Of "Paper Mills"

However, a new study – the first of its kind – reveals how professors can best deter their students from using online “paper mills” and from plagiarizing.

In their study of college students, Brian Gaines, a professor of political science at the University of Illinois, and Bear Braumoeller, a professor of government at Harvard University, found that even stern warnings not to plagiarize seem “to have no discernable deterrent effect.”

But, revealing to the students that their papers would be run through plagiarism-detection software proved to be a remarkably strong deterrent, or, put another way, “seemed to concentrate minds wonderfully,” Gaines and Braumoeller wrote in a recent issue of PS: Political Science and Politics.

While the professors concede that existing plagiarism-detection software is not perfect, “its success rate is high enough to merit use in a wide range of classroom situations.”

The professors focused on two groups of students and the papers they were assigned to write. One group was given a written and a strong verbal warning about plagiarism for the first paper; the other was not. The assignment was intentionally broad. “Our purpose was not to encourage plagiarism, but rather to remove impediments to it in order to assess student behavior when topical constraints are few.”

On the second assignment, all students were told their papers would be checked by plagiarism-detection software. After the students deposited their papers electronically, the professors used EVE (Essay Verification Engine), a program to test for plagiarism. Other findings and observations:

* Ironically, paper mills may in the long run make plagiarism more difficult, the professors said. For one thing, paper mills have “created a niche for plagiarism-detection software.” Also, what is available online is “of middling quality at best; students may reach the same conclusion.” And, with the spread of printed matter now being scanned and put online, plagiarism-detection programs are increasingly capable of catching passages taken from printed sources.

* Only about one out of eight papers turned up “problematic” because of either casual or blatant plagiarism. “While we cannot with confidence establish an upper bound on percentage of papers demonstrating plagiarism, one-eighth serves as a fairly solid lower bound.”

* While a few students engage in intentional academic dishonesty, “far more were unclear on the rules about plagiarism, but paradoxically, had received enough lectures on it that they simply ‘tuned out’ any warnings.” “The challenge for the educator is to deter the first group and to motivate the second to pay closer attention. Plagiarism-detection software seems to serve both functions quite well.”