27 May 2019
There has been considerable discussion of the role of bibliometrics in the assessment of research and international ranking systems. It was noted that the current impact factor is based on citations which do not reflect the various journal functions. A document by Paul Wouters, Leiden University, Cassidy R Sugimoto, Indiana University, Vincent Larivière, University of Montreal, Marie E McVeigh, Institute of Scientific Information, Bernd Pulverer, EMBO Journal, Sarah D Rijcke, Leiden University, and Ludo Waltman, Centre for Science and Technology Studies University of Leiden, has proposed the creating of a new set of journal metrics.The comment published in Nature developed from a 2017 workshop in Leiden and argued that metrics should reflect the recognised functions of journals: registration of claims to original research, curation, evaluation, dissemination and archiving.
It was noted that the current impact factor is based on citations which do not reflect the various journal functions. It was suggested that indicators relevant to curation might include the expertise and diversity of the editorial board, acceptance rates, and transparency. The number and diversity of reviewers and their timeliness could be used to assess the function of research evaluation.
The Journal Impact Factor included in Journal Citation Reports, now published by Clarivate Analytics, has become susceptible to abuses such as self-citation and the proliferation of questionable journals.
The document agues that indicators should be justified, contextualized, informed and responsible, and that a governing body should be set up to propose new indicators, develop standards, publish educational materials, and control predatory journals. This will be the task of another meeting in 2020.