Similarity and other related information

It is important that staff are aware of Brunel's policy on text matching services before analysing similarity in submissions.

Brunel is using Turnitin for similarity detection.

The following links may help with analysing the reports generated by Turnitin.

Understanding the similarity score
Includes an overall view of the Similarity report, similarity percentage ranges, scoring scenarios and a brief description of how Turnitin detects student collusion.
Note: Remember that a high similarity score does not automatically mean that work has been plagiarised, nor is its findings exhaustive. There may be a number of perfectly legitimate factors contributing to a higher score. Guidelines for staff on handling suspected plagiarism can be found on your College's IntraBrunel site (the staff handbook is often a good starting point).
How exclusions-and filters refine the Similarity Report
Downloading and sharing the report
Interpreting the Turnitin Report guide to the interface via UNIwise

Similarity check training

Video of: September 6th 2023 conducted by Turnitin.

Video of: November 29th 2023 conducted by Turnitin.

Video of: March 18th 2024 conducted by Turnitin.

AI detection tools

We are not currently planning to use the Turnitin AI detection tool at Brunel. There are many reasons for this decision, which we will expand on shortly on the APDU pages (https://brightspace.brunel.ac.uk/d2l/home/24715) and institutional AI guidance on the intranet (https://www.staff.brunel.ac.uk/principles-for-ai-in-teaching-and-learning).

However, two primary reasons for not using AI detection are:

Equity – AI detectors are relatively easy to fool, and therefore only weaker students will be caught and punished.
False positives – Most systems admit to at least 1% of false positive results, which is far too high for unfounded allegations of cheating.

The following research conclude that the current available tools (including Turnitin), are not reliable or accurate in detecting AI- generated content. They also advocate rethinking current assessment strategies while embracing AI usage in teaching and learning.

Weber-Wulff, D., Anohina-Naumeca, A., Bjelobaba, S. et al. Testing of detection tools for AI-generated text. Int J Educ Integr 19, 26 (2023). https://doi.org/10.1007/s40979-023-00146-z

Ardito, C.G., 2023. Contra generative AI detection in higher education assessments. arXiv preprint arXiv:2312.05241v2.

Perkins, M., Roe, J., Vu, B.H., Postma, D., Hickerson, D., McGaughran, J., & Khuat, H.Q., 2024. GenAI detection tools, adversarial techniques and implications for inclusivity in higher education. arXiv preprint arXiv:2403.19148.

The findings in the research article below emphasize the importance of academics understanding and exploring how AI responds to their specific assessment questions.
Scarfe, P., Watcham, K., Clarke, A., & Roesch, E. (2024) A real-world test of artificial intelligence infiltration of a university examinations system: A “Turing Test” case study., PLoS ONE, 19(6), e0305354.

Please do not be tempted to upload students work to third party online AI detectors, as there are many IP and GDPR issues to which University would then be exposed.

If you have any questions regarding this policy, please contact the Head of Digital Education (colin.loughlin@brunel.ac.uk).