Aug. 01 2022

New Journal Seeks to Reduce Bias in Scientific Publishing

Bias, be it racial, ideological, or scientific, can find its way into academic publications. Editors or peer reviewers who often work behind closed doors can reject work that is submitted by an author whose name does not sound like their own (usually European), an article that does not align with their political views, or a topic that produces null findings.

Qeios, an innovative new journal on which Columbia Mailman School Professor Peter Muennig is a board member, seeks to eliminate bias in scientific publishing while also making the content free. Since 2018, Qeios (pronounced “chaos”) has been operating as a pre-print platform for young scientists to quickly release findings and generate feedback through crowd-sourced peer review prior to publication. It is well on its way to becoming what may be the world’s first academic journal that relies on the scientific public, rather than editors, for publication decisions.

During the pandemic, the use of pre-print servers exploded when timely information about the outbreak was in short supply. Traditional peer review takes many weeks because submissions are first vetted by human editors then sent out to reviewers with their feedback addressed by authors prior to publication. Qeios speeds up that process. An AI-based system finds peer reviewers suited to each submission. Reviews are then published alongside the article as they come in. Authors have the chance to respond and update their article, as needed. The reader sees the whole history of comments and changes. The entire process of vetting an article is transparent—and fast. Qeios receives a median of 10 peer reviews per article and the time from submission to publication is roughly 10 days.

“The real question was whether the Qeios model encourages transparency and rigor by generating a public dialogue,” says Muennig, a professor of health policy and management. “The journal deems an article ‘published’ once it has received three favorable peer reviews. The hope is that authors will fear reputational damage if other scientists feel that they have not adequately responded to the peer reviews and that the reviews themselves will be of higher quality when the public is looking over their shoulder at the writer’s thoughts.”

Qeios may soon get recognition as a proper journal, which means when academics publish there it counts toward tenure and promotion—giving it another advantage over preprint servers. It is already indexed on Google Scholar and is archived in Portico and the British Library. The journal is now applying to be indexed by PubMed and Web of Science—the two biggest indices.

It remains to be seen whether crowd-sourced peer review can eliminate other sources of bias, such as bias against authors whose names do not sound familiar, difficulties in publishing null findings, or civil discourse when articles published by controversial authors are uploaded into the system. “The hope is that scientists will see value in professional, intellectually challenging discourse rather than entrenching within their scientific or ideological camp,” says Muennig.