What do you think about plagiarism detection software? This is a common question among academics. It is really a question that begs a conversation that is much broader and deeper than you might first expect. As with most tools and technologies, there are always both affordances and limitations. With that in mind, I offer a few thoughts as a way to introduce some to the conversation, and to promote further dialogue among those who are already interested and engaged with the topic. As I get started, please know that my comments are not directed at any specific plagiarism detection tool. Instead, these are general comments about various plagiarism detection tools, as well as the broader topic of academic dishonesty.
- Accountability – Dwight L. Moody is quoted as having said that, “Character is who you are in the dark.” While this quote has good wisdom in it, the reality is that most of us are a little better behaved when we know that others are watching us. In fact, there is some evidence that even, “the illusion of being watched can make you a better person.” With that in mind, it is likely true that plagiarism detection tools add a measure of accountability. If one knows that the paper will be reviewed with such a tool, it might challenge the person to think twice before intentionally plagiarizing.
- Blatant Plagiarism – It catches many cases of extensive and blatant plagiarism, and helps to mitigate against some (but not all) of the services that sell papers.
- Improved Writing – Some students and instructors note that using such plagiarism detection services help students develop better writing skills. This is especially true as some of the services not only check for plagiarism, but also problems with a student’s writing style. Turnitin.com, for example, boasts of tools that improve writing by facilitating peer review of papers.
- Accidental Plagiarism – It is a helpful tool for a student who wants to catch accidental or unintentional plagiarism before submitting a final version to the teacher. The one downside that some note is that submitting the same paper to the service multiple times can create “false positives.”
- False Security – It can give faculty a false sense of security. Many types of plagiarism and academic dishonesty are not noticed by the software. This is not a critique of the tools as much as a critique of over-dependence upon it for things that it can’t or was not designed to do. Many of the following limitations fit into this category.
- Metaphors & Stories – Suppose a student takes a story, metaphor, or illustration from a source; but puts it in their own words and fails to cite it. That will likely not get caught as plagiarism.
- Sources out of Reach – There are many sources that are not in the plagiarism detection databases. There is still a good measure of content that is not digitized or available to check against.This is even more true as learners begin to leverage the vast amount of content available on the web in video and/or audio format (TedTalks, iTunes U, Academic, Academic Earth, Khan Academy, lectures in EdX or Coursera courses, etc.). Perhaps the student got the idea from a documentary, older paper texts, unpublished works, through interviews, etc. Again, this is not a critique of the software. I’m just pointing out that none of these are likely to be detected. The software can’t be the entire plan to promote academic integrity.
- Source Deception – Most of the time, these tools will not pick up what I refer to as “source deception”…getting a general idea from one source, but citing it as coming from another. Or, it will notice it, but the student has it quoted and cited, so it looks alright. If you actually looked for the reference in the cited text, you would not find it. This often happens when the student finds a relevant idea from a non-academic or unacceptable source, so they use it but claim that it came from elsewhere. Unless the professor looks up every citation, this will usually not get caught.
- Many Types of Plagiarism – This is really just an expansion of point #2 above. There are many types of plagiarism and most of these tools target certain types, but not others. There are several good sources on the web about this, like this article from Valdosta State University that describes five distinct types.
Plagiarism detection tools are helpful, but not adequate. They do not replace the hard work of cultivating a class culture of academic integrity and honesty. They do not replace the importance of instructors and students becoming informed about the broad topic of academic integrity, developing a robust vocabulary that allows them to understand the many nuanced forms of academic dishonesty. When used with this in mind, plagiarism detection tools play a helpful supporting role. Even in these situations, there is more to writing and submitting papers than getting caught. If we are not careful, we can frame the entire endeavor in the negative and not consider the larger picture of the intended learning experience. That is why some, like the Composition Program at the University of Louisville, have a carefully constructed a policiy against the use of plagiarism detection software. In the case of the University of Louisville, the authors of the policy support their position with six thought-provoking points. Whatever your opinion on the matter, this is a subject that warrants ongoing thought, attention, and discussion. As you have interest, feel free to share a few of your own ideas in a comment, or you are welcome to move the conversation over to Twitter using the hashtag #cheatmooc in anticipation of the following.
You are invited! – If you are reading this before May 5 of 2013, there is still time and room to sign up for a MOOC that I am hosting at Canvas.net on Understanding Academic Cheating on Online Learning Environments. If you are interested, please don’t wait too long to register, as we are capping enrollment at 1000. There is an online learning focus to the course, but I suspect that anyone interested in academic integrity will find it rich with good resources and conversations.