When it’s not plagiarism
Posted February 14, 2017on:
Every semester I provide formative feedback on written work submitted by graduate students. Before I do this, the students submit their assignments to Turnitin via an institutional LMS to determine the extent to which their work matches other work in the database.
Every semester I get at least one email from a concerned student worrying about the matching score. The worry is good in that the plagiarism talks they attend have an impact. However, the worry is bad because they misunderstand what plagiarism is and how tools like Turnitin work.
Turnitin runs on formulae and algorithms. It has a huge database of references and previously submitted work. Any new student work is compared against this content. The extent to which the new content matches with the existing work is a percentage that I call the matching score.
Some students seem to think that the matching score is the same as plagiarism. This is not necessarily the case.
If a student uses a template provided by curriculum committee or tutor, the headers and helping text will match. If another student correctly and ethically cites common quotations and lists reference, these will match with other existing work. All these means that the matching scores go up, but this does not mean the students have plagiarised.
In 2009, I provided examples on how the scores alone are not valid or reliable indications of plagiarism. A low score could hide plagiarism while a high score could actually come from the work of conscientious student with lots of correctly cited references.
Both the students and I have the benefit of not just the quantative matching scores, but also the qualitative highlights of matching texts. The latter should allay fears of plagiarism or highlight what is potential plagiarism. The student can take remedial action and I can determine if a score actually indicates plagiarism.
The problem with the system is the human element. Grading teams, administrators, librarians, advisers, and supervisors often arbitrarily set ranges of matching scores to mean no plagiarism, possible plagiarism, or definite plagiarism. The numbers are an easy shortcut because they take out human decision-making. The reports with highlighted text require reading and evaluation and thus mean a bit more work.
Both faculty and students need to be unschooled in focusing on numbers and playing only the numbers game. Life is not just about what can be quantified. Neither is the quality of a student’s assignment and their mindset on attribution.