TurnItIn.com at Dalhousie - An Open Letter

In a news post in early December 2005, I spoke up against TurnItIn.com. I indicated my intentions to make this an on campus issue, to ask for a review and a change to the policy. Ultimately, I want a policy that respects students, respects our intellectual copyright, and respects that many of us work hard at university.

Today, I sent this email to ImproveService@dal.ca as a submission in their recently-ended contest ("Access Dalhousie") soliciting opinions on what policies on campus could do with a little improvement.

The Opening Niceties

Good day,

First, I realize this email comes a few days after the deadline, but like you said - good suggestions are always welcome.

I've been a Dal student for 7 years, two degrees. My history here is replete with circumstances where I felt something was amiss with a Dalhousie policy, and did what I could to change it. By and large, this was successful - administrators responded to reasoned arguments, and things changed for the better. This introduction serves to say that this recent active solitication of new ideas is only a more obvious approach to a process that Dalhousie staff have been willing to undergo for years now, and I appreciate that.

The Problems with TurnItIn.com

With that said, there is one policy that has nagged at me since it was introduced. I've never raised it, never advocated against it, because it has some merits. I refer to Dalhousie's use of TurnItIn.com.

When Dalhousie introduced the service 4 years ago, it was done in the most abrupt, least respectful, most dictatorial process I have seen in 7 years. We were told our professors could use it, and if they chose to, we had to submit, or else. When we tried to ask questions, when we tried to raise our concerns, we were stonewalled, misled, led in circles, or asked to just accept it. After all, what did we have to hide? There was nowhere to go to ask questions, no contact person, and the logical suspects would refer us to other logical suspects in a university-wide game of duck, duck, goose.

Since that time, some efforts have been made. FAQs have been produced, and some legal challenges in the US and at McGill have prompted TurnItIn to be far more forthcoming with information.

But some concerns remain, and though the information available attempts to address these concerns, it falls short.

The first and most obvious is the intellectual property issue. I am in the process of filing two patents - it is not out of the question that students come up with creative ideas they wish to protect. Despite the unsubstantiated claims of TurnItIn that my property is stored "securely", as a technology, security, and privacy expert I simply do not believe that to be true. Over 200 security breaches have been reported by organizations more reputable than TurnItIn since California's breach disclosure law was passed earlier this year. What happens when TurnItIn is next? "We're sorry" won't help. Neither will the disclaimer of liability in their terms and conditions. Even a lawsuit filed to receive compensation will be ground-breaking, and as a student, am I really going to do that? I will simply suffer, comforted only by empty and now worthless promises.

Second, I object to the presumption of guilt. To be required to submit my paper, and to then allow my paper to be used to detect future cases of plagiarism, is demeaning. It implies that not only am I likely to be plagiarizing now, but I am likely to allow others to plagiarize my work. In most cases, a quick look at my paper would be sufficient to determine that I am talking about something that has never been talked about before - a new approach to privacy, a new approach to search engine querying. Why the presumption of guilt?

Third, I doubt the service's efficacy. In a series of test submissions, I found that even very, very minor alterations to text copied from a paper known to be in the database were sufficient to trick the system. TurnItIn.com has known properties. If it is so easily fooled, it loses both its deterrent value and its inherent value.

Additionally, I maintain a website where I write posts, whole pages, and even feature series about my areas of interest. In some cases, I have expanded on this work and passed it in for academic credit - which is not a violation of plagiarism rules. However, I know that the TurnItIn bot used to crawl my website (until I asked them to stop). A quick test revealed that my paper was flagged as being 10% plagiarised, simply because of the similarity between my paper and my website. A rare example? Perhaps, but in this world of blogging and instant online publishing, it may be more common than you think.

More worrisome is the natural tendency of people to trust computer systems as being accurate. The shortcomings of TurnItIn are rarely advertised; this, the student disciplinary committee may accept TurnItIn results as proof positive of plagiarism, as may professors or TAs. Using flawed and tainted evidence against a student is blatantly unfair to them.

Fourth, I still object to the inadequacy of the information provided to students. For those who are willing to look, there is a FAQ on the Dal website that is still very much in the tone of "Because we said so. Deal with it." I find it curt and dismissive - no mean feat when you remember that it is only a web page. And try going to the turnItIn.com website for information - you will find no fewer than four policy documents with varying applicability to different jurisdictions, and that appear to sometimes conflict. Students don't know their rights, do not know the implications, and no one is going out of their way to inform them. Rather, the approach is often "because we said so."

Fifth, the privacy implications are important. Although the site recommends not to submit any papers with personal information, in some cases this may be unavoidable. Even as a computer science student, I have used real-world stories to illustrate my point in my papers. I am sure the same could be said for social work students, for example.

Sixth, professors and TAs may very well grow complacent in the face of the program - only looking for plagiarism in papers with a 50% score or higher, say. This reliance on a system that is not accurate to begin with (see point #3) will not decrease the occurences of plagiarism - in fact, it will do quite the opposite.

Seventh, it is quite possible to configure the service so that students never know what their "plagiarism score" was. This is completely and utterly unfair to students. They have a right to know the case against them. It can also be construed as their personal information, or information generated from their personal information, which they definitely have a right to access. Barring access is a recipe for trouble.

Proposing Solutions

Of course, I recognize the merits of the system. Certainly plagiarism is a growing problem. Detecting plagiarism in 100 first-year papers is a daunting task for humans, and some technological help could be beneficial. I understand that it is up to the individual professors to choose whether or not to use the service. It may even be an effective deterrent (are incidents of plagiarism increased or decreased over the last 4 years?).

But as it stands now, the policy of "if the professor chooses to use it, so also do everyone of his students" is unfair and inadequate to students worried about their rights, especially their intellectual property rights.

I do not come to you empty handed - I propose real alternatives. Some or all of these policies would be a marked improvement on the existing policy.

1. Students may, by way of a letter or email to their professor before the add/drop date of the course, ask to be excused from the TurnItIn program.
   - rather than being a "get out of jail free" card, this may invite extra human scrutiny to the paper - scrutiny which is almost certainly more accurate than computer scrutiny. No habitual plagiariser will opt-out of a system so easily fooled in favour of human attention to the paper.
   - note that at McGill, it was ruled that students cannot be required to use the service. This policy is an acknowledgement of that.
2. Professors may decide to use the service, but they must not compel all of their students to submit their papers to TurnItIn.com. Rather, they may only submit the paper if they believe there are reasonable grounds to suspect plagiarism. Students will still be required to submit an electronic copy to facillitate this. This allows humans to come up with reasonable grounds, rather than the computer.
3. Under the contract with TurnItIn, no Dalhousie student papers will be retained by the system for longer than 3 weeks, nor will any be added to the indexes for any reason.
4. Put the whole service to a referendum vote - let the students decide if they see the value in this service.
   “a mandated threat of a less-than-effective technological solution is not an effective long-term solution to plagiarism”
5. Recognize that a mandated threat of a less-than-effective technological solution is not an effective long-term solution to plagiarism. Efforts on the front of education, writing workshops (even mandatory ones!), increased human resources and increased human vigilance, a streamlining of the process to bring justice to plagiarism cases... all of these measures will be more effective then throwing time and money at the white elephant that is TurnItIn.com.

I hope that this suggestion is informative and complete. I welcome further comments, suggestions, and criticisms. This is an open letter; I intended to publish it myself, and I encourage you to do the same.

I look forward to speaking with you, and working with you, as I work to bring this issue to the center of debate and to the attention of students and student leaders on campus.

Thank you once again for the effort of soliciting ideas from students on campus. I understand that you have received several hundred responses thus far. I am sure the the respondents appreciate, as do I, the opportunity to work with you to make Dalhousie a better place for our students now, and for our students yet to come.

Best Regards,

Mike Smit

The Opening Salvo

It starts now.