In the ever-evolving landscape of academic research, the role of artificial intelligence has undergone a dramatic transformation. Not too long ago, the…
Peer-reviewed research articles are how researchers communicate their work to the scientific community. In recent years, and particularly since the start of the Covid-19 pandemic, mainstream media has judged the trustworthiness of scientific findings by whether they are peer-reviewed. But peer review is constantly evolving; innovation is reducing biases within the system and increasing diversity. To explore issues in peer review, I attended an online Quality and Peer Review workshop on Friday 19 June 2020, organised by Sense about Science.
Sense about Science (SaS) is a charity that champions the public interest in sound science and evidence. They organise workshops to discuss the systems, challenges and future of peer review with early career researchers (ECRs) from the Voice of Young Science (VoYS) network. The workshops are part of SaS’s Quality and Peer Review programme, partnered in 2020 by PLOS, and supported by SaS’s free guide, Peer Review: the nuts and bolts, written by and for ECRs.
On the workshop panel were Dr Amara Anyogu, a lecturer in microbiology at the University of Westminster; Paul Whaley, systematic reviews editor for Environment International; and Dr Diana Marshall, head of reviewer programmes at Taylor & Francis. “Peer review is an integral part of being a researcher”, said Amara. (The general steps are described in the flowchart.)
There are three main types of peer review: single-blind, double-blind and open. Single-blind is the most common, where reviewers see the identity of the authors but the authors do not know the identity of the reviewers. Double-blind peer review is thought to reduce bias by concealing authors’ and reviewers’ identities, but in small research fields it may be possible to work it out. Open peer review is also thought to reduce bias by improved transparency. Post-publication review takes open review one step further by publishing articles prior to peer review and publishing the reviewer comments.
There are, however, limitations to peer review. “There is a lack of diversity of reviewers,” said Diana. “There is an imbalance between the community of scientists publishing papers and the scientists doing peer reviews.” Researchers are often not recognised for reviewing when applying for jobs and grants. Journals often pick reviewers they have previously worked with, meaning the pool is limited. These factors perpetuate the lack of diversity.
“Funders can drive change towards review diversity,” said Amara, “as they have done by enforcing open access and data availability on the research they fund.” New innovations in open science and raw data availability may lead to development of better tools (such as AI technology) to differentiate between fraud and honest mistakes.
“Registered reports are fundamental to improving peer review,” said Paul. Registered reports involve a two-stage review process: researchers submit their introduction and methods to a journal, prior to carrying out the research. The journal peer reviews this submission and can accept the paper in principle for publication. There is a second review stage following completion of the research that ensures methods were adhered to and results were clearly written up. Champions of this new approach say it reduces publication bias against negative results – the paper is accepted on the importance of the research question and the rigour of the scientific method.
There are multiple ways to get involved with peer review:
- publishers run courses to teach researchers the fundamentals of peer review, which allow them to build up a relationship with that journal
- journals often allow ECRs to co-review a paper with their supervisor or mentor
- participate in SaS workshops by joining the VoYS network.
“Choose to work with journals you have a constructive relationship with,” said Paul. ECRs can reach out to journals directly with their CV, but it is important to communicate expertise online, by published papers and social media such as LinkedIn and Twitter.
Many ECRs struggle with imposter syndrome when it comes to peer review, but they can get involved as PhD students. “Everyone starts out as a novice and builds expertise,” said Amara. “You are making a great contribution to the science and you are learning.”
The system is designed so that if there are sections of a manuscript a reviewer cannot comment on, they can be covered by another reviewer. Editors are there to support reviewers and peer-review should be from peer-to-peer, not expert-to-novice.
New models of peer-review are putting emphasis back on good science; innovation is increasing diversity and reducing bias. Increasing diversity can be led by ECRs getting involved with the process. Amara finished by saying: “People are at the heart of peer-review and publication, there is imperfection, but it is always getting better.”