Journals need ‘diversity factor’ to improve representation

Journals should be given a “diversity factor” to help improve representation across research, according to a new paper.

The current methods of evaluating a journal’s impact rely heavily on the fragile impact factor, researchers said, which does not capture the contribution a piece of research can make to furthering health outcomes for different communities.

The study – led by Jack Gallifant, a postdoctoral researcher at the Laboratory for Computational Physiology at MIT – said there is no objective measure for tracking progress towards inclusivity in science or for evaluating who contributes to health research.

Published in the Public Library of Science, the paper says the impact factor has transformed into a proxy for the quality of individual articles even though highly cited papers skew calculations, and leading journals have “gamed the system”.

Campus resource: What are the barriers to success for early career researchers in STEM fields?

“A shift from a single citation-based metric to using several different metrics that provide a more complete perspective on factors aligned with scientific excellence based on contribution to advancing diversity and inclusion, improving health outcomes, and achieving equity is therefore necessary,” the authors write.

Researchers used the OpenAIex database to examine the metadata of all papers published between 2000 and August 2022, and published a dashboard displaying the proposed diversity factor of major titles.

They found “significant underrepresentation” of female authors and, despite signs that things are improving, warned that it is unclear how many journals will reach gender parity in the next five years.

The study also found that those from low- or middle-income countries are most at risk of being “left behind” – with no positive correlation to the journal’s impact factor.

“Academia is becoming more and more centralised especially in high-income countries,” the study says. “As it does so, it becomes more difficult to penetrate, especially when considering the trends of self-citation.”

The paper proposes that the next steps towards implementing a diversity factor would be the availability of dataset characteristics, detailed funding sources, patents and downstream policy impact, and citation mapping – which would allow for a better understanding of who is impacted and who is causing the impact.

The study “uncovers a bleak reality unseen by citation-based metrics”, the researchers said, because current academia and the healthcare system that it shapes cannot provide for all if it is not reflective of the wider population.

Instead, they see the diversity factor as a “call to action” for improved representation and to prevent the perpetuation of biases against certain subgroups.

“It reminds journals and authors to assess how thoughts and data reach the manuscript and whether they consider all perspectives, not just those available at hand.

“Otherwise, we will continue learning and practising medicine in an echo chamber created by the few ivory tower academics with access to the resources and data required to advance the field left in the hands of a select few institutions.”

patrick.jack@timeshighereducation.com