The American Geophysical Union (AGU) conducted a two-journal test of CrossCheck in 2010. The test helped AGU establish parameters for the editorial boards of their 18 journals to follow in their use of the software. Thirty percent was chosen as the Similarity Index (SI) that triggered a look at the CrossCheck report by the editor for most journals (15% for the letters journal). Matching text less than this percentage tends to be innocuous.
All new submissions are checked for wording overlap with CrossCheck before entering the peer-review cycle; one journal also checks revisions using the software. The editorial assistants initially use CrossCheck to check the manuscripts and examine the SI report if the SI exceeds the journal threshold. After review by staff, the manuscript is forwarded to the editor with notes about the matching text. The editor then analyzes the SI to determine the seriousness of the overlap.
AGU editors are not concerned about a high SI until they corroborate problems while examining the matching text. As CrossCheck often points out, matching text is not plagiarism. “Boiler-plate” text was discussed in an EOS article written by the editors-in-chief of the pilot-study journals; the authors of the study dislike the practice of reusing methodology text.1 AGU editors use the CrossCheck results as an opportunity to educate authors about properly citing previous submissions and the dangers of potential self-plagiarism.
The CrossCheck software is a beneficial tool for AGU journal editors. The editors still rely on reviewers to identify manuscripts they are simultaneously reviewing for other journals and submissions published in other languages. However, CrossCheck allows the editors to eliminate a great deal of plagiarized material, saving time for reviewers and preventing copyright infringement of previously published content.
- Kumar P, Calais E. Ensuring integrity in AGU publications and compliance with dual publication policy. EOS, Trans Am Geophy Union 2011 92(13):113 http://www.agu.org/pubs/crossref/2011/2011EO130008.shtml.