Historic co-authorships speed up editor handling times

Journal editors tend to accept manuscripts written by prior collaborators more quickly.

24 April 2017

Lane Erikson / Alamy Stock Photo

Academic editors at one of the largest multidisciplinary journals tend to accept manuscripts from previous collaborators an average of 19 days faster than other papers, a new study has found.

The analysis, published in Scientometrics, quantifies the impact of social relationships between editors and authors on handling times.

"The worry is that prior authors are getting special treatment just because they are insiders," says Daylian Cain, an expert on conflicts of interest, and associate professor of Marketing & Management at Yale University.

"Many power relationships and conflicts of interest exist in academia. For example, when one needs reference letters from other academics for promotions and awards. Such relationships can cause bias, even without intentional quid pro quo."

Computational social scientist Frank Schweitzer, and colleagues from ETH Zurich in Switzerland, analysed 137,536 articles published between 2007 and 2015 in the multidisciplinary open access journal PLOS ONE.

Schweitzer and colleagues identified 1,067 articles (1.3% of the total) where editors and authors had been co-authors in a paper previously published by PLOS ONE. This occurred more than 20 times more often than expected from randomized models.

Although editors and authors are likely to be part of the same scientific community, author-editor co-authorship within a five-year timeframe is considered a competing interest at PLOS. Despite this, more than 95% of the co-authorships identified in the study occurred within five years of the prior publication, write the authors.

Matt Hodgkinson, head of research integrity at Hindawi, another open access publisher, was a staff editor at PLOS ONE during the period examined in this study. He told Nature Index that editors know these co-authorships occur. "This is something that the editorial team at PLOS and other publishers do address, but it's clear publishers need to be doing more about this," he said.

Since the dataset was limited to articles published in PLOS ONE, the number of prior co-authorships is likely to be an underestimate as authors and editors may have published together elsewhere, says David Garcia, one of the study coauthors.

At PLOS, academic editors are assigned according to expertise in a subject, says David Knutson, public relations manager at PLOS. “So it’s not unexpected that they would handle papers written by previous collaborators. PLOS does however ask editors to avoid handling manuscripts from recent collaborators,” he said.

Faster handling times

The authors found that reduced handling times remained statistically significant when they accounted for proxies of article quality (number of citations and downloads), the prior editorial experience of the handling editor, and the similarity of topic between the editor's and the authors' publications.

The study did not include data on peer-review, such as number of reviewers, the time they took, their expertise, or the number of rounds of review. But the differences in handling times might be related to the review process, says Garcia.

Elizabeth Wager, Publications Consultant at Sideview and Editor-in-Chief at Research Integrity and Peer Review, an open access journal published by BioMed Central, says that one explanation for the faster times might be that reviewers are more prompt when the work is really closely related to their own, or when they know the editor or the authors personally. “But, that of course raises the problem of whether these social links constitute a conflict of interest," she says.

Other factors might also explain the phenomenon, the authors say. For example editors may quickly “desk-reject” low-quality papers of their previous collaborators, authors might send their better work to editors that were their collaborators earlier.

It remains to be tested whether prior co-authorship affects the likelihood of acceptance, says John Ioannidis, an epidemiologist at Stanford University, California, who studies scientific robustness. “Differential likelihood of acceptance and molding of the papers to fit the beliefs of the editor and reviewers are proportionately more serious potential problems.”