Confessions of a User

I've been wanting to respond to some of the Turnitin debates over at Kairosnews all week. I just haven't had time. I've also been considering submitting a proposal about Turnitin best and worst practices to a local technology conference. In light of that, here are some thoughts I've jotted down this morning.


Turnitin Problems

Turnitin does not distinguish between things that have been properly quoted and documented and things that have not. The percentage matched number cannot be used to determine whether a student has cheated.

Turnitin is inconsistent. In one student's essay, he had several quotes from the Bible. Some were marked as unoriginal; others were not. We know that the entire Bible is available online in all of its various translations. Anything quoted from it should have shown up as unoriginal.

Turnitin even marks the headings on the papers as unoriginal. We know that a certain amount of the percentage matched number can be accounted for simply in the student spelling the instructor's name correctly.

One student had several sentences marked as unoriginal in an essay about Harry Potter. The link provided by Turnitin for where the sentences had been copied took me to a message board where all of the users were posting under anonymous screen names. The “unoriginal sentences” were not all together in one place. They were scattered throughout various threads and topics on the board. For all I know, the student was only quoting herself.

If essays are submitted to Turnitin in rough draft form, revisions of those same essays cannot be submitted without showing up as entirely plagiarized. This discourages teachers from having students submit rough drafts and using the originality reports as a teaching tool.

One student wrote an essay about her senior project in high school. The portions that showed up as matched were linked to an article in an educational journal about possible assignments for senior projects. No whole sentences were matched, only phrases. It is my thinking that the student had internalized the instructions for her assignment after spending so much time working on it. It is also my suspicion that the high school teacher, on the other hand, had lifted the assignment straight from a journal without changing a thing.

Turnitin only catches one form of cheating. If a student gets another student to write a paper, Turnitin can only catch that if more than one student has submitted the same paper to Turnitin.

Turnitin Worst Practices

** Accepting Turnitin's word as law (or even as true)--anything identified as plagiarized by Turnitin should be carefully reviewed by the instructor to determine why it was marked.
**Using the percentage matched number as an indicator of how many points should be deducted from an assignment for plagiarism—even if a student has incorrectly copied some sentences from sources, a portion of that number can be accounted for by things that do not constitute plagiarism.
**Submitting student work to Turnitin without informing the student that this will happen.
**Having students submit their own work to Turnitin without explaining to them exactly what it is or how it works.
**Submitting only final drafts to Turnitin without working with the student on identifying and correcting possible instances of plagiarism first.

Turnitin Better Practices

**Submitting rough drafts only to Turnitin and using the originality reports as opportunities for student revision rather than as proof of crimes.
**Recognizing that not all forms of “unorginality” are plagiarism.
**Using the originality reports as an opportunity to revise for lack of original thinking and/or original voice in writing as well as for outright plagiarism.
**Recognizing that some sharing of information and some internalizing of information is likely to show up in these reports in ways that do not constitute cheating.
**Having students submit their own papers to Turnitin and giving them open access to their own originality reports.
**Recognizing that students have to be taught how to avoid plagiarism before they can be expected to demonstrate expertise in avoiding it.
**Recognizing that information sharing is so prevalent in our common business and educational practices that students have good reason to expect that some degree of “borrowing” is acceptable in their writing unless they are told otherwise.

Like grammar checkers, plagiarism detection services are unreliable. They should not be the only way that plagiarism is dealt with in student writing, and they should not be used as a means of creating less work for the instructor. Used correctly, plagiarism detection services do require time and attention on the part of the instructor.

It is my thinking that people trained as writing instructors are the most likely people to understand how to avoid improper use of plagiarism detection services. Like grammar checkers, however, it is the people who understand how to use them who are the least likely to need them.

Though I have been using Turnitin as part of my department's adoption of the service, I agree with Becky Howard and others who have spoken out against it. I also agree that compositionists should be educating themselves on exactly how Turnitin works and exactly how it is being used so as to be better prepared to educate others in its possible pitfalls.