Berkeley

A computer-based service to detect student plagiarism is being used by a growing number of university teachers across the United States, and may soon be tested in Britain. The service was developed by a doctoral candidate at the University of California, Berkeley.

Called Plagiarism.org, the service was launched last spring. It allows academics, journal editors — and even students — to rapidly compare articles against thousands of papers available through the Internet. The Internet has itself made plagiarism far easier through simple cuts and pastes.

One German researcher says he has already found the programme so effective that he plans to scrutinize all manuscripts submitted to his recently launched publication, the Journal of Medical Internet Research (see below).

Plagiarism.org was set up by neurophysiology doctoral student John Barrie. According to Doug Zuidema, head of the Office of Student Conduct at the University of California at Berkeley, the university is negotiating a contract with the company.

“We hope to negotiate an agreement for all faculty and students on the campus to have access to the service,” says Zuidema. Plagiarism.org is already negotiating for a pilot study to be carried out at selected British universities through the UK Higher Education Funding Council.

The program originated from Barrie's experience in grading neurobiology papers, when he became suspicious that students were plagiarizing material. In some cases they were taking material from online companies that sell articles over the Internet via websites such as Schoolsucks.com and Cheater.com.

After working as a teaching assistant with psychopharmacologist David Presti, Barrie created the service and enlisted the help of several Berkeley graduates as partners. Last spring, he used the system to check the papers of about 320 juniors and seniors in Presti's upper-division neurobiology class.

When a paper or article is run through the Plagiarism.org vetting system, a printout is generated on which apparently plagiarized material is highlighted. This is done by linking apparently cribbed sections to their sources, typically published articles.

At the beginning of the semester, Presti had told the students that their papers would be checked for plagiarism. Afterwards, Barrie examined the papers and found that 15 per cent of the students had plagiarized material.

One student author from Presti's class, for instance, appeared to have based virtually his whole article on sections of published work lifted from six web addresses.

“Clearly this is a serious problem; we have to do more analysis of it,” says Paul Licht, dean of Berkeley's College of Biological Sciences. “I'm not sure whether I'm more discouraged that they committed plagiarism, or that they continued to do so after they were warned.”

Presti and Berkeley officials are reviewing possible action against the students. They could be given a failing grade for either the paper or the class, or, more seriously, be charged with violating the code of student conduct.