Effectiveness of Plagiarism Detection by TurnItIn and Copyleaks

Abstract

The objective of this study was to investigate the effectiveness of checking the similarity by TurnItIn and Copyleaks. The research team consists of 31 volunteers, including librarians and library staff or graduate students, who normally provide services for checking plagiarism for students, teachers, and researchers at their own institutions or have experience in using such software. They were from 20 universities located in 25 different areas. All participants received the same set of 14 documents and configured the programs according to the provided guidelines. The results showed that, in overall, TurnItIn had higher accuracy and precision in similarity detection compared to Copyleaks. However, both programs had similar capabilities in checking content from open access journals. Neither program fully supported checking Thai documents, whether they were copied, translated, or generated by AI. Additionally, the functions to exclude quotes and bibliographies/references were not complete. Therefore, these details need to be taken into consideration when using the similarity results.