Now Reading
Another Case Against Science’s Objectivity Myth: Nepotism in Publishing

Another Case Against Science’s Objectivity Myth: Nepotism in Publishing

Photo: Anne Nygård/Unsplash

  • The scientific literature is expected to be an honest reflection of the work scientists are doing, but a new study is the latest to question this assumption.
  • An international team has attempted to quantify the extent to which nepotism – the practice of favouring one’s friends or relatives – undermines the biomedical literature.
  • The study’s findings “could have an impact on productivity-based metrics and suggests a risk of instrumentalisation, if not corruption, of the scientific enterprise.”

Scientific publishing is an important part of doing modern science. In the simplest sense, it’s an offshoot of scientists’ studies, experiments and discussions, and also helps disseminate scientific findings within the larger community.

But in practice, it has become much more than that – principally because the numbers generated by the publishing industry are often used to ‘measure’ the quality of a scientist’s work, and the extent of their contributions to science. And this exercise is frequently flawed, biased or both.

Against this backdrop, a new study by an international team has attempted to quantify the extent to which nepotism – the practice of favouring one’s friends or relatives – undermines the biomedical literature.

Specifically, the team found that a significant number of journals published many papers authored or coauthored by the same persons. These journals often also responded to submissions from these persons sooner, published other work that cited these persons’ papers, and allowed them to be a member of the journals’ editorial boards.

On the face of it, the scientific literature – consisting of all the scientific papers published on all topics – is expected to be an honest reflection of the work scientists are doing around the world. But the current study is the latest in a long line of such studies to have questioned this assumption.

As the paper states, the study’s findings “could have an impact on productivity-based metrics and suggests a risk of instrumentalisation, if not corruption, of the scientific enterprise.”

Skew by the numbers

Journals stacked on a shelf. Photo: yeaki/Flickr, CC BY 2.0

The research team, from Canada, France, Italy and the UK, focused on biomedical journals catalogued by the US National Library of Medicine. They zoomed in on 5,468 journals, in 152 biomedical categories, and the papers they published from 2015 to 2019.

Based on the papers, they identified the authors who had published the most articles in each journal, whom they called the “most prolific author”. Then they calculated the ratio of the number of articles this author had published in the journal to its total number of articles – both in the five year period.

The study found that the most prolific authors published about 14 articles each in the five years, on average. This amounts to 3% of all articles published in this period. And in 5% of the journals, the most prolific author had more than 10% of all their papers.

The numbers imply a possibility that these 5% of the journals could have favoured a few authors over others.

To make sure, the researchers calculated the Gini index, a number widely used in economics to measure income inequality. It varies from 0 (income equality) to 1 (maximum inequality). The researchers calculated the index based on the number of publications and all the authors of the 5,468 journals.

Specifically, they plotted the total number of publications up to a predetermined percentage versus the percentage of authors listed as authors in those articles. If x% of authors had authored x% of the papers, there would be perfect equality.

Then the researchers calculated the ratio of the skewed percentage against the total number of publications. The Gini index effectively quantifies the deviation from perfect equality in the number of publications, against the number of authors.

They arrived at an average index value of 0.183. This means the distribution of authorship deviates by around 18% from a scenario in which authorship is equally distributed. This observation further cemented their suspicion that publications they had suspected of nepotism appeared to be favouring a few authors over others.

In about half of the journals, which reported the article submission and acceptance dates, the researchers found that the time to acceptance for the most prolific authors was on average 85 days – versus 107 days for articles published by others. This period also became shorter the more the most prolific authors published.

According to the researchers’ paper, this suggests that “for certain journals, which are outliers on papers by the most prolific author and/or the Gini index, peer reviews may have been absent or only superficial for prolific authors.”

And because the publication lag decreases with the number of articles published, “our description of these outliers based on papers by the most prolific author could be only the tip of the iceberg, capturing solely the most extreme cases of hyperprolific publication in a given journal.”

Compromised editorial boards

It is not always virtuous for scientists to publish a very large number of papers. In fact, independent experts have unearthed many potential cases of scientific misconduct or fraud because a scientist was publishing a suspiciously large number of papers.

A recent example was the case of Didier Raoult, the French microbiologist whose name appears as a coauthor on at least 2,300 papers. Raoult has come under fire – from government authorities and scientists – in the last two years: for promoting hydroxychloroquine as a treatment for COVID-19 sans evidence, for initiating legal proceedings against Elisabeth Bik after she flagged concerns about images in some of his papers, and for conducting an illegal tuberculosis trial.

Didier Raoult. Photo: Twitter/raoult_didier

Scientists have identified at least two reasons for the size of Raoult’s output. First, Raoult attached his name to every paper published by scientists at the institute where he works. Second, the vast majority of his papers were published in a small number of journals whose editorial management included his close collaborators.

The researchers in the current nepotism study also found that the most prolific authors were members of the editorial board in more than half the journals they published in.

The researchers also selected 100 journals from the pool of the 5% – in which the names of the most prolific authors were listed on at least 10% of the papers – at random. Two of them didn’t reveal the composition of their editorial boards, just the editors-in-chief, who were not the most prolific authors on the pages. But in the other 98, the most prolific author was a part of the editorial board. In 25 of them, they were the editors-in-chief.

Aside from possibly boosting the prospects of their own papers, the researchers reported some evidence that the implicated journals also published many articles that cited previous work by the same authors.

Citations are popular in the world of science as a proxy measure of one’s contributions to science. But they are also ‘hackable’. For example, based on a study published in 2019, Nature News reported that 94% of the citations that Sundarapandian Vaidyanathan, a Chennai-based computer scientist, has received are from his own papers or those of his coauthors.

Peer reviewers have also been known to demand the authors of the paper they’re reviewing to cite the reviewers’ work – as a short-cut to boosting their CV without necessarily contributing to science.

In the limited context of journals, citations carry a lot of weight because they contribute to the value of a parameter called the impact factor (IF). It’s equal to the average number of citations per year that a journal has received in the last two years. The higher a journal’s IF is, the more it is said to be important.

However, citation counts can be hacked, which means the IF can be hacked too.

The nepotism study’s researchers found that the journals suspected of nepotistic editorial practices also had a higher IF than those in which the editor-in-chief wasn’t also the most prolific author.

In academia, a scientist’s citations and IF of the journals they have published in help secure academic positions and promotions.

Also read: Science Journals Might Be ‘More International’ but Not Their Editorial Boards

Potential consequences

“Easy publishing in high-impact journals brings easy recognition and better profiles to put forward on job applications,” said Kaushal Kumar Bhati, a postdoctoral researcher at the Louvain Institute of Biomolecular Science, Belgium. But “graduate students from less famous laboratories struggle a lot to publish [in these journals] and later find jobs.”

The researchers speculate that at least some of these journals and their most prolific authors could indeed have been extremely productive or could be part of a small community of scientists carving out a niche. Both considerations are, however, limited by the number of papers a scientist can humanly produce. (Raoult’s paper count defies this limit.)

Another possible explanation is that a particular researcher’s name is added to most articles in a particular field of work because of their seniority instead of active participation. “If high profile researchers aren’t leading the study or taking active participation, they get the authorship thanks to their academic power and influences,” as Bhati said.

The researchers ultimately suggest that others use their method to identify potential journals with nepotistic behaviour. They also write that such journals should be investigated further, with careful attention to their publication processes.

Sylvia Karpagam, a public health doctor and researcher, said scientific studies are often funded by private companies, and many of them set out to prove an idea instead of simply validating it. If there are influential scientists on the editorial board who steer clear of honest and critical peer-review, “this can be devastating for policymaking because the credibility of the journal then carries forward these opinions that come in with an agenda other than scientific research.”

Deepthi Lavu, a researcher at the Algorithm for Comorbidities, Associations, Length of stay and Mortality Study Unit, UK, said policy interventions alone may not be able to eliminate nepotism in journals. “Any more layers added to the system will make it more bureaucratic and possibly more expensive as well,” she said.

Instead, Karpagam said the editorial team should be open to public scrutiny, and assessments of journals based on the study’s recommendations should be placed in the public domain.

Shweta Ramdas, a postdoctoral researcher in human genomics at the University of Pennsylvania, added that scientific publishing as such requires a better alternative to peer-reviews, which are currently and typically opaque.

Bhati echoed her: “I believe the best practice to control this trend is to publish the review reports along with the accepted manuscript.”

Debdutta Paul has a PhD from the department of astronomy and astrophysics at the Tata institute of Fundamental Research, Mumbai. He is currently a freelance science communicator and journalist. He tweets at @dbdttpl.

Scroll To Top