Students who are aware that their work will be checked by plagiarism-detection software are just as likely to cheat as those who are not.
Conventional wisdom - and intuition - suggests that the threat of discovery, and subsequent punishment, is an effective deterrent against plagiarism - indeed, one of the comments on the article points to another study that suggested that students’ awareness of the use of Turnitin on a course significantly reduced plagiarism.
It’s not always clear whether plagiarism is an intentional and cynical attempt to deceive, the result of bad time management and poor writing or referencing skills, or due to genuine lack of understanding of the concept of plagiarism or differing cultural norms around it. The first category of student is the category most likely to resort to essay mills as a safer alternative where it’s made clear that plagiarism detection is in use, which suggests that the majority of students ‘caught’ by Turnitin and other text matching techniques when their use is advertised as a supposed deterrent are those whose main problem is not a desire to cheat but academic or personal factors.
Findings like this seem to strengthen the arguments in favour of using Turnitin formatively, as part of a student’s academic development and the essay writing process, rather than as a way of detecting problems once it’s too late to do anything about them and the student has entered the disciplinary process. The use of plagiarism detection only after submission seems to be based on the assumption that plagiarism only occurs through a deliberate desire to cheat, and as I’ve argued before, positions all students as potential cheats rather than as developing academics who may be in need of guidance and support to achieve their potential.