When University of Virginia Professor Lou Bloomfield first heard that students were recycling term papers in his popular physics class, he spent a night designing a computer program to check for plagiarism.
His class, called “How Things Work,” was billed as physics for non-scientists, and drew a crowd of 500 students each semester. To simplify communication and grading, Bloomfield had students submit all their term papers electronically beginning in the spring of 1999. Because of the way he ran his class, he had an exact copy of every paper submitted over the course of four semesters.
On Friday, April 6, 2001, the first version of his controversial program began churning through thousands of papers looking for cheating and unoriginal work. About 100 hours later, the results came back much worse than Bloomfield had expected.
Nine months later, the physics teacher still finds himself at the center of a plagiarism controversy that involves 154 honor council prosecutions, seven expelled students, 23 pending trials, 44 cases that have yet to be investigated and 10 students who dropped out of school rather than face a jury of their peers.
“The plagiarism primarily was students reusing other students’ papers,” Bloomfield said. “That sort of behavior has happened here and elsewhere since papers were first written.”
“What has changed is my ability to find it,” he said.
The Need to verify
Plagiarism has always been a concern in education, but most experts agree that the Internet has made information more accessible and plagiarism easier. In response, several individuals and companies have developed software to electronically analyze papers and compare them to a vast array of available resources.
Simple programs, such as Bloomfield’s, compare papers to other papers in a database. Bloomfield’s searches for occurrences in which six or more words in a row are replicated and warns the user when it finds occurrences of dishonesty. Bloomfield’s program is available free on the web, and he says it has been downloaded by almost 4,000 users.
More sophisticated software is for sale and literally deconstructs papers into complex mathematical pieces. These programs claim the ability to not only search for blatant copying but also check for paraphrasing and the duplication of “intellectual property.” Some, like the software offered by Turnitin.com, claim to compare submitted papers to both an extensive database and 1.5 billion webpages.
In what Georgetown University administrators call a move to protect and educate students, the University will sign a contract this semester with Turnitin.com. This semester, members of the school’s Honor Council have approached faculty to begin using the software in their classes on a pilot basis. If all goes well, the software will be encouraged university-wide.
Georgetown University Mathematics Professor James Sandefur has been helping solicit interested faculty members to try the software.
“We have a half a dozen professors in a range of departments,” Sandefur said. “The recommendation is that the professor inform students at the beginning of the course, and then when a paper is due students will turn a copy into the professor and also submit a copy of their paper electronically.
Once a paper is submitted online, Turnitin.com takes over. The software, currently in use by dozens of schools and the entire University of California school system, breaks the paper down into sections, scans for synonyms and begins looking for plagiarism. The software compares the paper to a database of hundreds of thousands of stored papers and then uses search engines to compare the paper to online material.
The software’s creators claim that Turnitin.com is capable of finding all kinds of plagiarism. Both exact copies and paraphrasing should be caught by its search engines. The company promises that a paper will be fully processed within 24 hours. At that point, a professor can check an “originality report” online which links plagiarized material to its respective web sources. Each paper is ranked on a scale of one to five, with one showing the least plagiarized content and five showing the most.
Georgetown Perspective
Schools have different policies on how the program is used. Some schools require all papers in certain departments to be run through the website. Other schools use it only for suspected plagiarists.
According to Sonia Jacobson, assistant for Academic Affairs and an adviser to the Honor Council, Georgetown still has not worked out exactly how the software will be used at Georgetown.
“We look at Turnitin as a deterrent,” she said. “Do we only use it when we’re suspicious or for ever paper that comes through?”
Jacobson said, that for her, the most important issue is one of equity. Both she and Sandefur agree that any program that is not applied to all students runs the risk of seeming unfair.
“That’s the kind of stuff that faculty and students need to talk about,” she said. “Faculty who have been burned [through plagiarism] are eager not to have to worry about the work that comes in and whether it has been plagiarized.”
Chuck Weiss, Distinguished Professor in the School of Foreign Service, agreed. Weiss has been using Turnitin.com to check his students’ papers since last semester.
“From a professor’s point of view, serious plagiarism is a kick in the teeth,” he said. “You feel terrible when you find it, and it’s a betrayal of the professor’s effort and the ideals of the University.”
Last year, an assistant informed Weiss that one of his students had plagiarized a large section of a paper for his class. He had already given the student a good grade on the paper and thought it was well written.
“I thought it was very clear,” he said, “and the reason it was very clear is that it was copied.”
Weiss believes that Turnitin.com offered three conveniences: It reminded students to cite sources correctly, it acted as a deterrent, and it was a way of catching students. He said that he found no serious plagiarism during his semester trial.
“The things I ran across this year were not intentional evasions of academic responsibility,” Weiss said. “They were much more a student’s feeling that they could not equal the clarity of a professional account.”
Weiss said he doesn’t know if plagiarism is increasing because of the Internet, but that the Internet certainly makes cheating easier.
“Pressures on a student are very great and the temptation [to cheat] must be very great, and everybody’s human,” Weiss said. “If students know they are going to have to jump this particular hurdle, I’m hoping it will nudge them back to the straight and narrow.”
Weiss said he expects to use the system in all of his classes.
“I would recommend it,” he said.
Debating how far
But not everyone agrees with Weiss. Don McCabe, a professor at Rutgers University and the founder of the Center for Academic Integrity at Duke University, is one of the country’s most respected authorities on plagiarism. He thinks the focus of universities must be on honor codes, and he starts speaking faster, noticeably irritated, on the subject of anti-plagiarism software.
“I’m very much in favor of promoting academic integrity instead of policing it,” McCabe said. “If a faculty member has a reason to believe a specific students is plagiarizing, fine. But I do have a problem with running everyone through the system.”
“It’s hard to develop a bond of trust between students and faculty if students feel the faculty will check everything they do,” he said.
A recent survey bolsters McCabe’s position. A study, released in February of last year and first published by the Los Angeles Times, showed schools that focused heavily on honor codes had fewer incidences of plagiarism than schools that spent a lot of time prosecuting it.
Policing every paper that comes through the academic system doesn’t sit comfortably with some professors and ethicists. At worst, the policy conjures up images of a 1984-like existence where a person’s every move is watched. At best, the policy sets up a system reminiscent of the Cold War, where deterrence is the best policy.
Bloomfield, at the University of Virginia, disagrees with those characterizations. The University of Virginia has one of the country’s strictest honor policies. One incidence of cheating and a student is expelled.
“Is there something wrong with looking for plagiarism?” he asked. “Students who are just innocently doing their work should fear nothing; nobody plagiarizes by accident. If you follow the rules, you’ll never see [a charge of plagiarism.]”
But following the rules doesn’t always guarantee that a student is kept out of the adjudication process. Bloomfield acknowledged that a significant number of the 154 cases of plagiarism found in his class were false alarms, where one person’s paper had been copied without the author’s knowledge. Bloomfield doesn’t think that hurts his image though.
“I’m doing this because it is my duty as a member of the university community to uphold the honor system,” he said. “Part of that duty is making sure the people who violate the community are not part of it.”
How much is too much?
One of the largest questions still to be satisfactorily answered is whether incidences of plagiarism have grown in recent years. Members of the Georgetown Honor Council believe they have. Bloomfield believes that they haven’t. Sandefur and Weiss aren’t sure. McCabe, however, thinks he knows the answer.
McCabe has been doing plagiarism surveys for years. His most recent study, which surveyed 25 schools and 4,500 students, indicated that the number of college students who plagiarize hasn’t increased in recent years. He said that about five percent of college students admitted blatantly plagiarizing from the Internet. About another 10 percent admitted to cutting and pasting or weaving information into their papers without properly citing it. He said these numbers are actually less than those who have cheated using paper sources.
“The most interesting thing, particularly at school like Georgetown, is that the problem is really on the way,” McCabe said.
He described what some have come to call the “MP3 kids,” a whole generation of junior high school and high school kids who have grown up with free information on the Internet and free music. For these students, intellectual property has never been an issue. Information has always been free and without an author.
For current junior high and high school students, the plagiarism numbers are many times that of current college students. Fifteen percent admit downloading a paper from a website and 52 percent have cut and pasted text from websites.
“These students are using information in a clearly inappropriate way,” McCabe said, “But they are refusing to acknowledge it as inappropriate.”
Jacobson said that this lack of clarity is probably the most common reason for Internet plagiarism. “I’m not sure if there has been an increase [in plagiarism], but there has been the perception of an increase,” she said. “It’s not clear that students are prepared to cite and use Internet information in the proper way.”
Jacobson does not believe that most students intend to plagiarize from the Internet, but a lack of familiarity with Internet citation methods and the stress of paper deadlines create situation where student inadvertently cheat.
As a result, Georgetown will be implementing the new software, not because it expects to catch a lot of students but to better educate them about the proper use of intellectual property.
But will the software work?
Does it work?
The Voice submitted 15 papers to the company in an attempt to judge the strengths and weaknesses of the software. Some of the submitted papers were actual term papers from college classes. Some were online news articles from www.georgetownvoice.com and www.thehoya.com. Some submitted articles were deliberately plagiarized from a wide variety of online sources from commercial sites to news outlets. Some articles were paraphrases of online articles. One paper contained extremely common phrases and song lyrics. Finally, two papers were resubmitted to Turnitin.com a second time to test whether the software accurately checked papers against its database.
Approximately half of the submitted material came back as entirely or partially plagiarized.
Term papers, which to the authors’ knowledge had no plagiarized content, came back with a clean bill of health. No false plagiarism was detected in those cases.
The software did an adequate job of detecting online source material. Approximately half of the deliberately plagiarized works came back as plagiarized. Turnitin.com actually had better success detecting plagiarism from rarely used websites, the kind that most students would be lifting from.
With one exception, the software failed to detect news articles lifted from the online versions of various newspapers, such as The Washington Post and The New York Times. No articles from Lexis-Nexis came back plagiarized. Articles from the Hoya and the Voice, even though they had been posted online for months, came back clean.
In one ironic twist, several quotes from a Voice cover story that were properly referenced came back as plagiarized, and the website that Turnitin.com linked to was itself a work of plagiarism.
Turnitin.com failed to catch all but one instance of paraphrasing. The case it did catch was caught only because six words in a row were left unchanged from the original.
Finally, the two papers resubmitted to the site came back clean. Sandefur, however, said that in his test a resubmitted paper came back as plagiarized. Turnitin.com did not return a phone call asking how long it took before a submitted paper was added to the database.
Although the test was by no means scientific, it did show that Turnitin.com will go a long way to stopping direct plagiarism of Internet sources. A student may get lucky and avoid detection once, but the chances are good that if the plagiarism is of the cutting and pasting kind, once will be all.
In Sum
The legal implications of the software are nearly ironed out. Student papers will not become the property of Turnitin.com and papers that are plagiarized from other papers will not directly link to the source material the way they do with websites.
Georgetown is still finalizing the details of its contract with Turnitin.com, but Jacobson said the final contract is almost ready.
Assistant University Counsel Kathryn Baerwald would not comment on any specific issues, but she said that Georgetown has been working closely with the company to resolve any potential legal issues.
Jacobson believes that everything will run smoothly with the pilot program. She said the honor council is undertaking “low-level” marketing to emphasize the importance of academic integrity and that the real mission will be increasing the awareness of the honor code. After the pilot program ends, the University will decide how the software will be implemented on a larger level.
