In March, Clarivate Analytics, the company responsible for the Web of Science (WoS) academic database, announced sanctions against some 50 scientific journals included in its extensive selection. The journals in question failed to meet the quality standards required by the company and will lose an endorsement that is essential to attracting new authors—having been excluded from the Journal Citation Reports (JCR), they will no longer be assigned an impact factor, the academic indicator used to measure a journal’s reputation and impact by calculating how many times its articles are cited in other papers.
This removal process occurs every year, but drew particular attention in 2023 because it included 21 titles from two fast-growing open-access publishers. The move also shines a spotlight on a practice widely used by these publishers that was already considered controversial: the publication of special issues edited by guests with no formal link to their staff, which regularly include a huge number of articles and often do not follow the same rigorous level of peer review as regular issues.
Nineteen of the excluded journals are published by Hindawi, which publishes about 250 open-access journals—64 of which were indexed by WoS. The company, founded in Cairo, Egypt, in 1997, is now owned by American publisher John Wiley & Sons. Two other sanctioned publications are published by MDPI, which is headquartered in Basel, Switzerland, and has a collection of 390 journals. One was the International Journal of Environmental Research and Public Health, which published about 17,000 articles in 2022. Its latest impact factor was 4.614, an impressive score for a title with such prolific output.
Clarivate did not provide details about each individual case, but WoS editor in chief and vice president Nandita Quaderi stated that an AI tool was used to detect atypical changes in journal performances, identifying 500 outliers that warranted further investigation. According to Quaderi, Clarivate was able to gather evidence that at least 50 of them were not meeting their quality criteria. “In recent months, we have taken additional proactive steps to counter the increasing threats to the integrity of the scholarly record,” she said in a statement. “Once we determine that a journal no longer meets our quality criteria, we have a responsibility to act.”
At the end of last year, Hindawi announced that it was temporarily suspending special issues after identifying that several of them had been used to publish hundreds of fraudulent articles produced by paper mills—illegal services that pump out manuscripts on demand, generally using fabricated data and images. In October, more than 500 articles from 16 of the publisher’s titles were retracted for peer-review manipulation. Investigations began in April after the editor in chief of one of Hindawi’s journals expressed concern about the content of a special issue. The reviewers’ comments appeared to be duplicated. There were also cases of the same reviewers evaluating many manuscripts and others who submitted their reviews extremely quickly. Hindawi reported that it lost US$9 million as a result of pausing special issues between November and January.
The special issues model was also a factor in the exponential growth of MDPI, founded just 13 years ago and now the fourth largest scientific publisher in the world. The company published around 20,000 articles in its first 15 years of existence, but that figure began to skyrocket in 2015. It published 240,500 papers in 2021, charging an average publication fee of 1,258 Swiss francs per paper (the equivalent of US$1,400). In 2023, its two biggest titles—Sustainability and the International Journal of Molecular Sciences—are expected to publish around 3,500 special issues each, equal to nine a day.
An analysis by Paolo Crosetto of France’s National Research Institute for Agriculture, Food, and the Environment and Pablo Gómez Barreiro of the Royal Botanical Gardens in Kew, London, showed that in 2022 alone, 100 MDPI journals published 17,000 special issues containing a total of 187,000 articles. The pair assessed how long it took for manuscripts to be reviewed, measured as the interval between the date they were submitted and the date they were published. The average was 37 days, compared to more than 200 days for open-access journals in the PLOS collection. “I have no proof that they did anything wrong,” Crosetto told Science. “But it stands to reason that trust is hard when you leave this guest editing to anyone,” he said, referring to documented cases of conflicts of interest and low-quality or even fraudulent peer review in guest-edited special issues. Carlos Peixeira Marques, from the University of Trás-os-Montes and Alto Douro in Vila Real, Portugal, told Science that MDPI has invited him multiple times to be a guest editor of special issues in fields such as agriculture and engineering, but never in his field of business and tourism. “The absolutely insane number of special issues has made it impossible to guarantee minimum peer-review standards,” he said.
In a statement, MDPI attributed its removal to WoS’s “content relevance” criterion. In previous statements, the company has defended its model, arguing that faster peer reviews allow authors to quickly share research results and that being a guest editor helps junior researchers learn about scientific communication processes. Giulia Stefenelli, chair of MDPI’s Board of Scientific Officers, told Times Higher Education that “special issues are initiated by experienced researchers in specific disciplines as an offer to the scientific community.” She stressed that the journals evaluate proposals for special issues made by academics and that the articles are subject to a rigorous peer review, with a manuscript rejection rate “close to the 50% mark.”
Republish