- One paper, given by a team from Texas Tech University, described an experiment in which they ran 400 papers (apparently of unknown plagiarism degree) through turnitin and SafeAssign. They discovered that the systems returned different values for degree of plagiarism (no suprise to me, as we almost never see identical values in our tests) and flagged many innocent bits as being potential plagiarisms. I cannot find a link to their paper, but the team is listed on the conference site as having given the session "Plagiarism Detection Technology: From the roundtable to the grail—deconstructing Turnitin and Safe Assignment" by Susan Lang, Laura Palmer, Monica Norris, and Kathleen Gillis.
- Much more troubling was the discussion in this article about how iParadigms, the company that sells turnitin, paid for the travel of some researchers who were giving (positive) talks about turnitin. Shouldn't a researcher, who perhaps has no travel money, at least tell the audience listening who paid for the travel? I realize this is rather like the company purchasing an ad on the first slide, but isn't it more transparent than a speaker saying nothing?
Saturday, March 14, 2009
Inside HigherEd reports on a plagiarism detection track at a large writing conference in San Francisco. Two important points from the article