1. Home
  2. Quick Answers
  3. Online Assessment and Feedback
  4. Most frequently asked questions about Turnitin

Most frequently asked questions about Turnitin

Written by Dr Mary Davis, Academic Integrity Lead, Oxford Brookes Business School, Oxford Brookes University.

1. Are Turnitin results always accurate?
No.

Scores can have false positives (text matches that are in fact ok or expected eg to the task rubric, cover sheet, a template, commonly used standard phrases, commonly used references or quotations). So high scores can sometimes consist of correctly cited work with many references (though may indicate some over-reliance on source text eg many quotations). High scores can also be due to matched text in appendices or if the submission is set up to exclude bibliography.

Scores can have false negatives (text matching does not indicate problems, but tutors can find problems with use of sources or authorship). So low scores can indicate plagiarism, little or no use of sources, and the use of custom writing services (which know how to avoid matches)

(Weber-Wulff, 2018)
2. Do the matches show the sources students used?
Sometimes. However, it is important to note that as Turnitin has been available since 1997, a vast and exponentially growing number of sources on its database are submissions of student work. Turnitin matches text to the closest (in time) use of the same text. Currently, it is far more likely that matches are made to other students’ use of the same sources and quotes, than to original sources. Therefore, when you see a long list of matches to student work at different universities, it does not mean that these are the sources for the submission, just that the words are the same.

(Weber-Wulff, 2018)
3. Why does Turnitin not show all copying?
Turnitin is a database which is continuously updating. However, it does not have all possible sources on it. In particular, it does not have all books unless in e-versions.

Sometimes it is necessary to use other forms of checking for copying, eg by putting a sentence into Google to look for matches.

(Weber-Wulff, 2018)
4. What is the ideal overall similarity?
This does not exist! The idea of a desirable percentage to achieve on Turnitin, such as 0%, 15% etc, is an absolute myth that has no logic at all. It depends entirely on what the matches are, how other texts are being used and how the Turnitin submission has been set up. The percentages themselves do not indicate plagiarism. They also vary across different textmatching software, depending on the algorithms used, so the overall similarity is NEVER a reliable figure to use.

(Canzonetta, 2018 and Weber-Wulff, 2018)
5. Should we tell students to make the colour go away?
No. Sometimes students interpret any match to mean plagiarism and conclude they must remove all matches to eliminate plagiarism. Some matches are always expected (eg correctly cited quotes, templates, standard expressions). However, if students have any large matches to individual sources, then the recommendation would be to do more paraphrasing, note-taking and summarising in their own words.

(Williams & Davis, 2017)
6. Should we rely on the overall similarity for evidence of academic misconduct?
Definitely not! It depends entirely what the matches consist of. For some time, there have been calls for Turnitin to remove the overall similarity percentage, because it is so frequently misinterpreted or used in an ill-advised way. It is recommended instead that you look carefully at the individual sources matched, and what is matched, and then decide if there is evidence of academic misconduct.

http://stevebentleyhud.wordpress.com/2012/12/04/time-for-turnitin-to-get-rid-of-originality-percentage-scores/

(Davis and Morley, 2018)
7. How do students get around Turnitin matches?
• Use of screenshots onto the document (look out for slightly grey, fuzzy text that cannot be highlighted for comments)

• Use of white letters between words that look like spaces (download, select all, change all to red font)

• Replacing of letters like ‘e’ with a symbol that looks like ‘e’ (look out for inconsistencies in font shape)

• Use of different spelling systems (US/UK or other)

• Fabricating references and quotes (check authenticity of sources)

• Copying from old texts (eg books) with false references (as above)

• Use of translations from other languages/ online paraphrasing tools (check naturalness of language –this one may be hard to judge)
8. Should we respond to TurnitinUK requests to share assignments from OBU students with staff at other universities?
No. These requests are sent by markers at other universities when there is a match to OBU students’ work. Turnitin requests are sent to tutors associated with Turnitin assignments, which means module leaders and co-tutors may receive the same request. Some tutors at other universities automatically request access even when the match is insignificant eg 1-2%. Some requests may be inadvertent as clicking on a match to student work leads to a pop-up message from Turnitin: ‘Because submitted papers remain the intellectual property of their authors, instructors, and respective institutions, we are unable to show the content of this paper at this time. If you would still like to view this paper, please click on the institution name above to submit a permission request to the author’s instructor’.

It is important to understand that the reason why there are multiple matches to other students’ work is usually because students in different institutions are using the same sources, writing about the same topics on similar programmes, and also that many academic phrases are used by students as standard ways of saying something in academic writing. The Turnitin database now has an enormous and continuously growing store of student work, so matches between students are extremely common as a result.

The email requests from Turnitin to share the student work contain the OBU student’s actual copied assignment under the message. Replying to these requests, even to refuse them, involves a further sharing of student work across email. There are thus issues of confidentiality in replying, and even more so in agreeing to share, for example, primary research details.

If there is a very significant match, there could be cause for concern, but rather than reply to the Turnitin request email, it is recommended that you contact the named staff member at their institution through finding their staff email.

References

Canzonetta, J. (2018). Can detection systems be used responsibly? In Student Plagiarism in Higher Education, Eds. Pecorari, D. and Shaw, P. in series Research in Higher Education, Abingdon: Routledge, pp.74-88.

Davis, M. and Morley, J. (2018). ‘How much can you copy?’ In Student Plagiarism in Higher Education, Eds. Pecorari, D. and Shaw, P. in series Research in Higher Education, Abingdon: Routledge, pp.28-46. https://www.routledge.com/Student-Plagiarism-in-Higher-Education-Reflections-on-Teaching-Practice/Pecorari-Shaw/p/book/9781138055162

Weber-Wulff, D. (2018). Why does plagiarism detection software not find all plagiarism? In Student Plagiarism in Higher Education, Eds. Pecorari, D. and Shaw, P. in series Research in Higher Education, Abingdon: Routledge, pp.62-73.

Williams, K., & Davis, M. (2017). Referencing and understanding plagiarism (2nd ed). London: Palgrave Macmillan. E-book available in OBU Library: https://www-dawsonera-com.oxfordbrookes.idm.oclc.org/abstract/9781137530721

Updated on July 3, 2020

Was this article helpful?