Why do some journals ask authors to suggest reviewers?
With scientific research becoming increasingly specialized, it can be difficult for journals to find experts to carry out peer review. Asking authors to suggest their own reviewers can therefore be mutually beneficial: journals save time and resources searching for an appropriate reviewer, and the author can support this process by providing contacts from their network.
However, while authors may know who is best qualified to judge their work within their specialist network, is there a risk that they could nominate reviewers more likely to review their paper favorably, thus compromising the objectivity of the peer-review process? It should be remembered that journals are not obliged to use such nominated reviewers (many state this clearly in their Notes for Contributors); but there are obligations on the reviewers, too.
Ethical guidelines treat conflicts of interest for reviewers just as seriously as they do for authors, asking potential reviewers to ‘declare all potential conflicting interests, seeking advice from the journal if they are unsure whether something constitutes a relevant interest’ (COPE, 2013; http://publicationethics.org/files/Peer%20review%20guidelines_0.pdf, accessed 12 May 2016]. Reviewers are expected to decline the reviewing assignment if their impartiality is undermined, but these conflicts of interest are sometimes difficult to avoid. As the PLoS Medicine editors say, ‘in specialized fields, isn't almost everyone a friend, colleague, or competitor?’ (PLoS Medicine Editors, 2008; Making Sense of Non-Financial Competing Interests. PLoS Med 5(9): e199. doi:10.1371/journal.pmed.0050199, accessed 12 May 2016).
There is some evidence that reviewers suggested by authors are more likely to give favorable reviews. A recent study in BMJ Open (BMJ Open, 2015; 5:e008707 doi: 10.1136/bmjopen-2015-00870. http://bmjopen.bmj.com/content/5/9/e008707.full, accessed 12 May 2016) showed a significant disparity between the opinions of author-suggested reviewers and editor-suggested reviewers. The study (which also investigated the operating standards of open review models and single-blind models of peer review) examined 800 reviewer reports for papers submitted to BMC Infectious Diseases, BMC Microbiology, and the Journal of Inflammation, and found that ‘author-suggested reviewers provided reports of comparable quality to non-author-suggested reviewers, but were significantly more likely to recommend acceptance’. Approximately two-thirds of author-suggested reviewers recommended acceptance for papers submitted to BMC Infectious Diseases and BMC Microbiology, compared to 31% and 38% respectively from editor-suggested reviewers.
An earlier study by the Journal of Pediatrics (Helton and Balistreri, 2011; Volume 159, Issue 1, pp 150–151, http://www.jpeds.com/article/S0022-3476%2811%2900160-0/fulltext#back-bib4, accessed 12 May 2016) showed similar figures. Evaluating 178 peer reviewed manuscripts submitted to the journal in 2007, researchers found that 65.3% of author-suggested reviewers recommended acceptance, compared to just 54.2% of editor-suggested reviewers. Editors were more likely to agree with recommendations from the latter; they agreed with 55.2% of accept recommendations from reviewers suggested by editors, and only 49.5% of accept recommendations from reviewers suggested by authors.
Although author-suggested reviewers were more likely to support the acceptance of manuscripts in both studies, it is not possible to determine the reason behind this. Is bias to blame, or are reviewers suggested by authors more likely to have particularly deep expertise in the subject, and therefore more able or more willing to see a paper’s merits?
In any case, the final decision always rests with editors who consider multiple reviews before judging for themselves whether to accept or reject a manuscript.