Parallel Session D

Digital tools for detecting and correcting misconduct

Moderator: Elizabeth Wager, Committee on Publication Ethics and Sideview, UK

Several tools can now detect publication misconduct including plagiarism, redundant publication and image manipulation. However, increased sensitivity of detection raises issues about the definitions of misconduct and how journals should respond.

Checking, Correcting and Updating through CrossCheck and CrossMark
Rachael Lammey, CrossRef, UK

CrossCheck (powered by iThenticate) is an initiative started by CrossRef to help its members actively engage in efforts to prevent scholarly and professional plagiarism. It combines a database of current and archival scholarly literature with the iThenticate tool, which compares authored work against the content in the database and highlights matching or similar text for editorial review. It continues to be widely adopted by the scholarly community and has nearly 300 publisher members. CrossMark is a new initiative, launching in April 2012, that should prove to be another useful service. CrossMark allows scholars to easily identify documents that are being actively maintained by their publishers. The appearance of a CrossMark logo on aN HTML, PDF or ePub document indicates that the publisher is taking care of or stewarding it through any updates, corrections, enhancements, retractions, or other changes. Clicking on a CrossMark logo may also provide important publication record information about the document, which might include peer review, publication history, funding sources, location and links to data sources, CrossCheck plagiarism screening or rights. 

Do we need new definitions of plagiarism?
Elizabeth Wager, Committee on Publication Ethics and Sideview, UK

New classification of duplicate publication
Chong-Woo Bae1, Soo Young Kim2, Sun Huh3, Chang-Kok Hahm4
1Department of Pediatrics, Hanyang University, Seoul, Department of 2Family Medicine and3 Parasitology, Hallym University, Chuncheon, and 4Department of Radiology, ChungAng University, Seoul, Korea

In Korea, duplicate publication in medicine has been easy to detect since the creation of the comprehensive nation-based medical literature database, KoreaMed.  We analysed 100 duplicate publications by randomly selecting 5% of papers from KoreaMed for each year. Duplicate publications were determined by three reviewers. These were classified according to a new classification system.  One major category comprised 1) complete copy with different language, 2) complete copy with same language, 3) copy with some modification with different language, and 4)copy with some modification with same language, 5) salami with divided sample number and 6) salami with divided outcome , 7) imalas with extended sample number or extended study periods, 8) imalas with added hypothesis, 9) imalas with extended sample number or extended study periods, and added hypothesis. In others categories,10) reverse imalas and 11) not classified as above were suggested.  This classification is comparable to that of von Elm et al. It may contribute to the easy understanding of the determination of duplicate publication.


How editors use digital tools

Ksenija Baždarić, Mladen Petrovečki, Department of Medical Informatics, University of Rijeka School of Medicine, Rijeka, Croatia

Ana Marušić, Coeditor in Chief, Journal of Global Health; Editor emeriti, Croatian Medical Journal, Department of Research in Biomedicine and Health, University of Split School of Medicine, Split, Croatia

In many journals editors make daily judgments on the presence of plagiarism in submitted manuscripts. According to the best practices, such as COPE Ethics Flowcharts, they often contact authors’ institutions but the plagiarism cases may not be followed by the authorities. One of the reasons for this discrepancy could be different views on the extent of similarity/plagiarism between researchers/academics and journal’s editors.

Our first allegation of misconduct in the Croatian Medical Journal was a high-profile case of duplicate publications, which we resolved with the help of expertise from international editorial organizations and retracted two articles. We were also contacted by the database Déjà vu about a possible plagiarism case, which finally ended in article retraction by author. These experiences and the availability of CrossCheck and other online detection services for plagiarism prompted us to take the preventive action of screening all submissions to the journal. We will present our experience in detecting and dealing with content similarity during 2009 and 2010 in the Croatian Medical Journal.