Catching Copy Cats
The fall semester opened this year with unprecedented concern over the scope of plagiarism in higher education. A virtual epidemic of cheating, or perhaps just a new awareness, has spread across the academic world. A web search for "plagiarism" reveals numerous articles published this past summer alone in the higher education press.
Reports of book authors and journalists caught copying the works of others without attribution are also frequent. A common thread through these events is the idea that the web and internet make cheating easy. Yet many of these instances of plagiarism are detected by sharp-eyed users of the networks.
Estimates of the frequency of plagiarism by students vary, with rates of 70 percent or more often asserted. The Center for Academic Integrity (www.academicintegrity.org) at Duke University (N.C.), for example, surveyed more than 50,000 students and found that 40 percent admitted using at least few unacknowledged borrowings in their papers.
Most observers concede that plagiarism is not new but wonder really how widespread it has become. The temptation to cut and paste is undeniable; the vastness of the web and internet gives an illusion of safety from detection.
Cheaters exhibit varying degrees of deception. Some, desperate in the face of deadlines or simply lazy, lift a few sentences, weaving them into their own writing. Or perhaps they appropriate whole passages from sources they think nobody else will find. But others make more deliberate efforts to cover their tracks, substituting synonyms or embroidering new sentences to alter copied paragraphs.
Commonly used web search engines are an obvious and easily available means for trying to counter plagiarism. If casual or incautious copiers find materials in places discoverable by ordinary searching, then faculty should be able to trace suspicious wording by the same means. But a growing multitude of software packages and search services is springing up to offer help in detecting plagiarism.
The University of Virginia website, for one, offers a free program called Wcopyfind that compares submitted files to search for shared phrases. It searches files on local and locally networked drives but is not able to search the web or internet.
The Essay Verification Engine, or EVE2, searches the web to find pages from which a writer might have plagiarized and returns links to the suspected sites, highlighting in red all the passages in the submitted paper that appear to have been copied from the detected sources. Compatible with Microsoft Word, Corel Word Perfect, or plain text formats, EVE2 sells for $30, with no recurring fees.
Plagiarism-Finder from Mediaphor Software AG, meanwhile, conducts web searches starting from texts in several formats, including Adobe PDF. The publisher suggests that authors and journalists could also use the program to check whether their texts are being plagiarized on the web. Plagiarism-Finder costs $125 and is available for a free 30-day trial.
A different approach is taken by the Glatt Plagiarism Screening Program. This program challenges writers to recreate the texts they claim to have created. It masks every fifth word of a submitted text and then measures accuracy and elapsed time while the author of that paper is challenged to fill in the blanks correctly.
The software is based on a principle that every writer has a unique style and will be highly successful at remembering written passages, even those with missing words. The Glatt Plagiarism Screening Program sells for $250.
Considered by many to be the leading anti-plagiarism product, Turnitin searches the web and some proprietary databases and a collection of already-submitted papers. A product of iParadigms, Turnitin creates what the company terms a "fingerprint" of a submitted paper, which it then compares to its three sources of information. Patterns, distributions, and profiles of language use are computed statistically and combined to yield ratings of probability of plagiarism.
Additionally, each paper submitted for testing is added to the company's database, augmenting the stock of existing papers available for each new search. Turnitin is available for campus licenses, with pricing based on institutional enrollment.
Within the past month, iParadigms has teamed with LexisNexis to offer a new service called CopyGuard. It searches more than six billion documents within the LexisNexis collection, using pattern-matching techniques to identify possible sources for submitted texts. While the service was not designed specifically with higher ed use in mind, it could certainly be used at schools.
Software programs and online services vary in their claims of certainty about identification of plagiarism. Some herald the end of stolen texts, while others are careful to say they report degrees of probability. But all represent themselves as tools of detection and offer themselves as specialized means to combat the epidemic of plagiarism.
Many institutions of higher education have established websites to promote student awareness of the dangers of plagiarism and ways to avoid the practice. The Student Judicial Affairs Office of the University of California, Davis has a web page entitled "Avoiding Plagiarism: Mastering the Art of Scholarship" (sja.ucdavis.edu/avoid.htm). It gives guidelines on how to cite sources, as well as guidelines--with examples--for avoiding inadvertent plagiarism. The site recommends Diana Hacker's A Writer's Reference (St. Martin's Press, 1995) as an authority on proper references.
A comparable web resource at Indiana University, "Plagiarism: What It is and How to Recognize and Avoid It" (www.indiana.edu/~wts/pamphlets/plagiarism.shtml) illustrates acceptable and unacceptable paraphrases and advises on strategy and mechanics to avoid plagiarism.
computed statistically and combined to yield ratings of
probability of plagiarism.
The Writing Center of Claremont McKenna College (Calif.) has published a study of software designed to teach students about the dangers of plagiarism (writing.claremontmckenna.edu/plagiarism.asp). Students were surveyed on their assessments of the effectiveness of six software tools. Their findings were that while some of the tested products rated better than others, none were better than "marginally acceptable."
A web page at the University of Michigan gives instructors background and pedagogical advice for preventing plagiarism in their assignments and cheating on tests (www.lib.umich.edu/acadintegrity/instructors). Various articles and letters from faculty seek to explain why students plagiarize and what faculty can do to lessen the likelihood that work they assign will result in infractions.
General awareness-raising is widely suggested as a counter to the lure of improper copying. The most frequently suggested ways to discourage plagiarism through pedagogical adjustments are for faculty to: give unique assignments (topics not likely to be represented in the universe of papers offered for sale); assign some specific sources students must use; and require students to submit outlines, project plans, or drafts in advance of the final product. Faculty are also commonly urged to be alert for anomalies in style, format, diction, and voice in student papers--as those might be signs that passages have been plagiarized.
While no one is able to say with certainty whether plagiarism is more rampant than in the past or just more detectable now, scholarship and publication are undergoing a crisis of credibility.
Students are not alone in being implicated in improper appropriation of the work of others: Journalists and scholars have been tainted also (see End Note). To some extent, faculty and editors have had to scramble to become familiar with the means for testing suspicious passages as students and writers have gotten ahead of them.
A new market for aids to search out evidence of plagiarism is developing. These tools just might join the array of reference resources that shape the conduct of scholarship and policing of intellectual property.
Tom Warger is a consulting principal for Edutech International (www.edutech-int.com).