Blog 1: Debunking 10 Myths about Turnitin

I’ve been a researcher of Turnitin, source use and academic integrity since 2005. Over the years, various myths about Turnitin have proliferated, in particular about the overall similarity percentage. I’m writing this blog to try to debunk ten of these myths and encourage more accurate interpretations of Turnitin results.

Myth #1. The overall similarity result shows the amount of plagiarism in the student’s submission.

Overall similarity is not the same as overall plagiarism. Many similarities are not plagiarism at all; they could be completely expected similarity, such as cover sheets, templates, given sources, quotations, references lists, appendices. They could also be good practice (eg re-use of standard formulaic sequences (phrases such as ‘this report will be organized into four main parts’).

Myth #2. There is an ideal overall similarity or an overall percentage above which is plagiarism, and below which is OK.

An ideal overall similarity or a cut off point above which is problematic and under which is fine, cannot be set. It depends entirely what the match is to. Some matches are acceptable, others are not. There are many, many factors that influence the score such as how the assignment submission has been set up, the task requirements, the use of quotes, the popularity of the topic, the way the student writes etc. No judgement should rely on the overall similarity.

Myth #3. Students should always be told to reduce the overall similarity match.

Students might be recommended to do this at a draft stage if there appears to be problematic copying. However, merely trying to get a lower similarity score should not be seen as the main way to improve a text. It is more important for students to look at how they are using sources, how they are expressing themselves, how they are following the task criteria etc.

Myth #4 Zero overall similarity is the best result.

Definitely not! If a student gets a zero match, it may be seen as a suspicious result, achieved by circumventing Turnitin or even cheating through use of essay writing companies which offer ‘zero plagiarism’ in the form of a zero similarity on Turnitin. However, some zero matches are not problematic as they are due to assignment settings, less popular topics or another factor.

Myth #5 Staff can rely on the overall similarity to decide if plagiarism occurred.

As explained in 1 and 2, the overall similarity only indicates similarity, not plagiarism. Overall similarity results should never be relied upon to decide plagiarism. The marker needs to check the work and the sources to decide if there is plagiarism. Turnitin is not a plagiarism checker – the marker is!

Myth #6 A high overall similarity always indicates plagiarism.

There may be many ‘false positives’ in the similarity results, such as matches to correctly cited work, or to ungraded appendices. Other possible glitches occur from excluding bibliography in the settings and then a submission with the word ‘bibliography’ on the first page resulting in the whole assignment being excluded.

Myth #7 The matches show the sources that the student used.

Turnitin is not set up to show the same sources that students used, only to find similarity between the student’s submission and other texts. Sometimes it shows the actual sources the student used, but the majority of the time it matches to other students’ work who have used the same texts.

Myth #8 Matches to other students’ work show that the student has colluded.

As point 7, Turnitin is set up to show similarity, not identical sources. Turnitin has a vast amount of student work on its database and a relatively smaller amount of texts from original sources. It is very likely that where a student has used the same source text as another student, there will be a match to another student’s work.

Myth #9 Giving students draft opportunities to submit to Turnitin is a problem because they match to the final submission and make the similarity report inaccurate.

When draft submissions are given, select ‘no repository’ instead of ‘institute repository’ in the settings. In this way, no matches will occur to the draft when the final submission is set up. Draft submissions on Turnitin are highly recommended to provide useful learning opportunities for students.

Myth #10 Turnitin shows every instance of copied text in a student submission.

Even though it is continuously updated, Turnitin does not have all source texts to check against. Sometimes students may have amended the text to stop copying showing up, or there may be a glitch. If you suspect copying but cannot see it in the Turnitin report, try Googling a phrase, check the references used or any source material related to the topic.