Looking Over Turnitin's Shoulder

A tech tool for detecting plagiarism has received some low marks from users concerned about privacy and the service's accuracy

Judging from recent research, one of the casualties of the Internet Age may be intellectual honesty. Four in ten college students surveyed said they intentionally cut and paste material from the Internet without citing the source of that information, according to a 2005 survey by Don McCabe, founding president of the Duke University-affiliated Center for Academic Integrity. That's four times as many as in 1999, the research shows.

Even more surprising to McCabe, who's now a professor of management at Rutgers University, about 77% of respondents to the 2005 survey said online plagiarism is "not a serious issue." Many students, McCabe concluded, "struggle to understand what constitutes acceptable use of the Internet."

Stopping the Cheaters

In an effort to help students and teachers achieve that understanding, Oakland (Calif.)-based iParadigms founded a service called Turnitin in 1997. Now being used by more than 9,000 schools, Turnitin uses software to help educators know whether a term paper includes previously published material that has been improperly cited. IParadigms says its membership has doubled each year for the past seven years. "[We're in] a beautiful market position," says Chief Executive John Barrie. He expects the company to sign up an additional 100,000 clients in the next 10 years.

But lately, iParadigms has hit some speed bumps that underscore the difficulty of thwarting plagiarism and other forms of cheating in the digital age. Critics say Turnitin's methods compromise copyright protections and foster a climate of suspicion between students and educators. Some also question its effectiveness in rooting out plagiarism.

Here's how Turnitin works: Students at participating schools and colleges submit most or all of their written take-home assignments to their teacher through the service's Web-connected application. It then compares the work for sentence or phrase matches against three databases: a comprehensive Internet snapshot, a library of published articles, and a pool of millions of previously submitted student papers (about 120,000 papers are submitted daily). The teacher receives a "Similarity Index" for each submission—a measure of what percentage of the work contains plagiarized material—as well as instant access to the sources in question. From there, it's up to the educational institution, which pays about 80¢ a student per year, to mete out any punishment that may be warranted.

Facing a Backlash

Students and parents from at least one school—McLean High School in Fairfax County, Va., which began using Turnitin this academic year—are crying foul. Juniors and seniors formed a committee in September and, with the support of legal counsel and a faculty sponsor, circulated a petition against McLean's use of Turnitin. The students say 70% of their classmates have signed the document. "If it's truly an educational tool, why do they insist on the requirement to archive [student] work?" says Kevin Wade, a McLean High School parent and publisher of the activist website dontturnitin.com. Wade is among parents who back the students' opposition to the use of Turnitin, arguing that a company shouldn't profit from the publication of students' work. Some also fear for students' privacy.

IParadigms' Barrie says the database is private and impenetrable. "Nobody sees a student paper that is submitted to us other than the student and their instructor," he says.

In October school administrators made some concessions, making Turnitin mandatory only for 9th and 10th graders. "Students can run their papers [against the Turnitin database] as they develop them, so the kids are learning as they move along," says Paul Regnier, spokesman for Fairfax County Schools. The county has instituted the service in 17 of its 25 schools since 2003, and pays about $380 annually per school.

Wade says he's not appeased. Among scenarios that cause him concern: A college admissions office turns up instances of copied work in an application essay that was from the same student's archived work.

"Yes, it will identify the source as being the applicant, but when a university is processing tens of thousands of applications, do you want to trust a clerk in an admissions office to make this connection?" he says. He's also concerned that future employers may have access to the Turnitin database and use it to retrieve writing samples that are several years old.

Privacy Concerns

Even more troubling to some parents and students is the possibility that the Turnitin database could be sold off to another business. The service's mandatory privacy policy provides for this very scenario: "As we continue to develop our business, we might sell or buy businesses or assets, or Turnitin might be acquired by another company. In any of those circumstances, personal information in our databases may be included among the transferred assets."

Wade and a team of other concerned parents have spent recent months trying to build a legal case against the Turnitin service, retaining the counsel of intellectual-property specialist Robert Vanderhye of Nixon & Vanderhye in Arlington, Va. The group says it's near filing a lawsuit aimed at forcing the company to discontinue archiving student work and to delete the millions papers in its database. "From an intellectual-property perspective, the copying of student papers for comparison is technically a copyright violation," says Dan Burke, a professor of law at the University of Minnesota.

Dubious Similarities

Some users also raise concerns about the effectiveness of the service. Chris Guy, a professor of physics at Imperial College in London, where Turnitin has been in use for three years, assigned an open-ended paper where students could delve into any physics-related topic of their choosing, as long as they went well beyond Wikipedia-like Internet sources and consulted academic journals of esteem.

Guy became suspicious when he received students' Turnitin reports. Many Similarity Indices were between 20% and 50%, and offered as proof totally unrelated sources (e.g., an Australian car magazine as a reference for a paper on the Big Bang Theory), Internet sources that turned up "404 File Not Found" errors, and papers from students in other parts of the world whom his students vow they have never heard of. He asks, "Are you a plagiarist if you don't reference a source you've never seen?"

Guy abstained from further usage of Turnitin in his class, but he began to submit papers of his own creation to the service as a diagnostic test. Not only did Turnitin make many implausible links when it positively detected plagiarism, but when he purposely plagiarized several prominent academic journals and mainstream publications, they didn't show up on the report at all. An entire word-for-word passage from Scientific American rated a 0% on the Similarity Index.

Guy is urging his school's administration to cease using Turnitin. Last year the University of Kansas and Mount Saint Vincent University in Halifax, Canada, banned use of the service on campus while other schools, including Syracuse University, are considering similar action, according to published news reports.

Guilty Until Innocent?

Critics also express concern that the service promotes a climate of mistrust between students, teachers, and administrators, because it assumes that kids are guilty until proven innocent. Barrie responds to these claims by likening his service to a referee on a football field or a proctor in an exam hall.

The Intellectual Property Caucus of the Conference on College Composition & Communication (CCCC-IP) last September issued a statement addressing plagiarism-detection services and the compromises they pose to the academic environment. The 13-year-old association of college professors warned all teachers and administrators that the use of applications like Turnitin "conflicts with best practices for fostering student engagement and learning," and creates a "hostile environment."

Expanding Its Uses

But the CCCC-IP thinks that if Turnitin is used in the right context, it can be a productive tool—for instance, if students are allowed to play around with the application instead of seeing it as a binding verdict. Michael Day, co-chair of the CCCC committee on computers in composition & communication and director of first-year composition at Northern Illinois University, recommends teachers and students use Turnitin to "critically investigate such notions as ownership, property, originality, and copying intellectual work." He continues in an e-mail, "In an age dominated by Web and Internet writing, wikis, blogs, remixing, and patch-writing, we owe it to our students to interrogate such notions."

Provided that iParadigms and its flagship Turnitin service weather the storm, as Barrie foretells, it also hopes to make waves outside the world of academia. Its other service, iThenticate, is a database of commercial documents, designed to give businesses a reliable way to determine the originality and authenticity of job applications, commissioned white papers, and other areas where plagiarism could hurt organizations financially and otherwise.

This service just got off the ground in the past two years, but already boasts high-profile clients including the U.N., the World Bank, and the World Health Organization. Says Barrie, "We've reached a point where people cannot simply wonder whether something is original or not before they grade it—or pay for it."

Click here to see a slide show of "High-Tech Cheat Sheets."

    Before it's here, it's on the Bloomberg Terminal.