Co-authorship without collaboration: comparison of co-authorship patterns in predatory and non-predatory journals

The study of dishonest co-authorship is a relatively new topic. Researchers have mostly studied the phenomenon of "honorary authorship" when individuals are included as authors of manuscripts or grant applications even if they were not involved in the research effort. In this study, we focused on more explicit manifestation of dubious coauthorship, which is the result of the cooperation of authors in order to secure a place in the article. The false co-authorship is especially evident in the case of interorganizational collaboration resulted from the direct selling of places in the list of authors. Our analysis demonstrates that the co-authorship pattern in questionable journals differs from normal journals including the dominant type of collaboration, number of authors and author’s geographical concentration. The analysis of dubious co-authorships concentrated in predatory journals has the potential to contribute to research on academic dishonesty, in particular, to understanding authors' motives. There is a question of whether the authors are innocent victims of deception or they are rational actors who calculate the risks and do not expect that there will be serious consequences. A study of authors from biomedicine showed that authors who publish in predatory journals have almost no history of previous publications and citations. Kurt showed that 70% of researchers were unaware of the predatory journal concept. However, in another data set, only one author with a predatory publication had no degree, moreover, the number of predatory publications increased with the rank of the author. Authors have not always been deceived, rather, they might deliberately use low (if not zero) rejection, a fast publishing process, and open access. The study of suspicious co-authorships provides more grounds to determine what motives are behind the predatory publication since we can be certain for some types of publications authors bought the place without knowing other collaborators. This understanding of motives is important in order to develop recommendations for misconduct prevention, because the ways to solve the problem should differ depending on the type of authors. Finally, our study shows that the possibility of false co-authorship should be considered in the research projects which rely on co-authorship data. Data on co-authorship is used in many research projects to make conclusion about nature of connection between individuals, journals, research fields, organizations, and countries. However, predatory publishing might create significant bias in the data, especially in case of data coming from developing countries