On the faculty development Listerv POD, there have been a couple of excellent posts by Ludy Goodson of Georgia Southern University.
In the posts she highlights the problems of relying solely on plagiarism detection programs (there are a number of them out there, but Turnitin.com may be the best known). Among the points she raises are the following:
- The software gives a false sense of security in that it doesn't catch several forms of cheating and plagiarism. It finds "matches" but not lots of other kinds of cheating (such as putting in false citations, improper paraphrasing, making up information, etc.). Students who wish to cheat eventually discover the workarounds to avoid detection.
- Most of these programs also do not check any proprietary databases (for example, Turnitin.com only checks those databases owned by ProQuest), or from non-digital sources e.g. books not in the database.
- Reports of high percentages of plagiarism are often false-positives because these tools only match text. An honest person who quotes a source will trigger a plagiarism report from the tool.
- Students accused of plagiarism often have not been taught the process of good research and the correct way to use resource materials. There are a number of different strategies that instructors can use to limit the possibilities of plagiarism. Some of these are listed at this site by Jaime McKenzie and a site maintained by Marylynne Diggs at Clarke provides an excellent example of guiding students through a writing assignment process, with checklists to guide them, separate grading sheets for each level of assignment in the "Resource Packet," other links explaining her policy, and resources to support the students in their development process.
Ludy Goodson has also developed a useful and very comprehensive set of resources about student plagiarism and it is available here.