Imprimir Republish

Good practices

Progress on shaky ground

University ranking uses indicators linked to scientific misconduct to classify institutions

bgwalker q / Getty Images

An unusual university ranking has listed 1,500 institutions, not for excellence in teaching or research output, but based on indicators related to ethical misconduct. The Research Integrity Risk Index, described in a paper not yet peer reviewed but shared on the ArXiv preprint server, considers two negative parameters of a university’s research output: the proportion of its scientific articles that have been retracted (withdrawn from publication due to errors or evidence of misconduct) and the proportion published in journals that have been removed from the Scopus and Web of Science databases for engaging in anomalous publishing practices or committing ethical violations in editorial processes.

“Universities want to be seen as rising stars, but it is not always clear whether they are rising on solid ground or statistical quicksand,” the index’s creator, Lebanese information scientist Lokman Meho, head librarian at the American University of Beirut, told Nature. He emphasized that while his classification uses objective data, it is not designed to estimate the scale of ethical misconduct committed at universities. Rather, it may help reveal institutions with structural vulnerabilities overlooked by traditional rankings, such as those that pressure their researchers to artificially inflate performance metrics or that are lenient toward questionable publishing practices.

The index classifies institutions into five categories: 1) red flag; 2) high risk; 3) watch list; 4) normal variation; and 5) low risk. The rating depends on the proportion of articles published in 2022 and 2023 that were later retracted and on the share of research output from 2023 to 2024 published in journals subsequently removed from Scopus and the Web of Science. The “red flag” group includes 124 institutions with “extreme anomalies and systemic integrity risks.” One quarter of them are in India, 17% in Saudi Arabia, and 15% in China. Indian universities occupy the top nine positions.

A notable example is Anna University in Chennai. The public institution, focused on engineering, technology, and architecture, saw retractions of 372 papers published by its researchers in 2022 and 2023. Other institutions on the red flag list have also been involved in scandals recently, such as the Dental College at the Saveetha Institute of Medical and Technical Sciences in Chennai, which boosted its impact indicators by encouraging students to write 1,500-word essays describing scientific activities they did the previous year and then submit them to journals and conference proceedings (see Pesquisa FAPESP issue n° 329). Many were accepted for short communications or correspondence sections, and because they cited numerous papers by authors from their own college, they artificially inflated the institution’s citation impact factor. Of the papers published in 2022 and 2023, 177 were retracted for ethical violations.

Most Brazilian universities appear in the lower categories—normal variation or low risk—including the three major universities in the state of São Paulo: the University of São Paulo (USP), University of Campinas (UNICAMP), and São Paulo State University (UNESP). There are, however, two federal universities in the higher tiers: the Federal University of Uberlândia (UFU), in Minas Gerais, was ranked 115th, and the Federal University of Goiás (UFG), was ranked 151st, with 27 retractions each. The retracted papers linked to these universities relate to a controversial case involving biologist Guilherme Malafaia Pinto, from the Goiano Federal Institute (IF-Goiano), who was corresponding author of 45 articles published in Science of the Total Environment (STOTEN) that were later retracted due to manipulation of the peer-review process. When submitting papers, Malafaia suggested reviewers, but gave fake email addresses for three of them. Editors at STOTEN sent dozens of his submissions to the fake email addresses and received detailed reports in return—though it remains unclear who actually wrote them (see Pesquisa FAPESP issue n° 349).

“Many of the retracted papers listed coauthors from the Federal Universities of Goiás and Uberlândia, which explains why the institutions appear in the ranking,” explained Edilson Damasio, a librarian at the State University of Maringá who studies the retractions of articles by Brazilian authors. Because the index only included institutions with large research outputs, IF-Goiano itself was not included, unlike the larger universities where the coauthors were based.

Carlos Ueira Vieira, research director at UFU, said he does not recognize the number of retractions attributed to the university and has requested clarification from Meho. According to Vieira, three UFU researchers collaborated with Malafaia, but the records show they only coauthored one of the retracted STOTEN papers. In some cases, the UFU-affiliated author was Malafaia himself, who cited his links to the university’s graduate program in ecology, conservation, and biodiversity. Vieira said Malafaia has since been dismissed from the program and no longer has any affiliation with the institution.

Fabíola Souza Fiaccadori, a professor and research director at UFG’s office of research and innovation, explained that Malafaia completed his PhD at UFG and frequently collaborated and coauthored papers with researchers from the university. She stressed that the coauthors had no involvement in the irregularities, and the retractions were related solely to misconduct in the review process, rather than the scientific content of the articles. “Us being named in this ranking is the result of a specific snapshot in time, since historically our researchers have had a very low volume of retractions,” she said. Fiaccadori added that UFG has been working to promote a culture of integrity, including by creating a guide to good academic practices and offering postgraduates a course on research integrity.

Republish