How to be a good peer reviewer

Peer review underpins most phases of the research cycle: applying for research funding, obtaining ethics committee approval, vetting research plans and evaluating scholarly manuscripts (before and after publication). By undertaking editorial peer review, journals offer authors certification and validation of their research results while providing readers with a reliable source of filtered and curated content.

Peer review
Despite ongoing debate about its weaknesses, peer review is generally recognised as an important part of the scientific and publishing process. In a 2015 survey, 75% of respondents agreed or strongly agreed that “Scholarly communication is greatly helped by peer review of published journal papers”, and 82% agreed or strongly agreed that “Without peer review there is no control in scientific communication”. A 2009 Sense about Science survey confirmed its importance to individuals: 91% of respondents said that their last manuscript was improved through peer review. On the flipside, being a reviewer can help researchers keep abreast of the latest literature, expand their knowledge, establish their expertise and place in the community, and develop their analytical skills.

Despite the importance of robust peer review, few researchers are taught how to be good reviewers. This lack of training has unfortunate consequences. Bad peer reviews waste time (the authors’, editor’s and reviewer’s), delay the review process and erode confidence in its overall usefulness.

It is thus in the scientific community’s best interests to help researchers become good reviewers. We dissect below some of the essential elements of best-practice peer review.

Participating in peer review

Your ethical obligations as a peer reviewer begin as soon as you accept an editor’s invitation to assess a manuscript. (The basic principles of these obligations are expertly outlined in the Committee on Publication Ethics’ Ethical Guidelines for Peer Reviewers).

Reviewers should maintain the confidentiality of manuscripts. Don’t discuss any details of the manuscript with anyone not involved in its peer review (including the authors), and never use any information from an unpublished manuscript for your own or anyone else’s benefit.

If you think there might be a conflict of interest that could prevent you from providing an impartial review, let the journal’s editor or Editorial Office know as soon as possible. Potential conflicts include a personal or professional relationship with an author, a financial interest in the work or its publication, working in competition with the authors, or having seen earlier versions of the manuscript. The editor or Editorial Office is best placed to determine whether you should be recused from your role.

Writing your review

The purpose of the review is two-fold: to provide authors with feedback on their manuscript; and to help the editor decide whether the manuscript (in its current or revised form) is suitable for publication in the journal. Together, these factors help to improve not only the manuscript but also the broader scientific process.

There are many types of peer review, but they tend to fall into two broad categories: blind, in which the identities of at least one group of participants (such as authors or reviewers) are hidden from the other groups; or open, in which all participants know each others’ identities. Whether you are asked to provide a blind or open review may have some bearing on how your review is formatted. For example, in single- or double-blind peer review, reviewers may be asked to separate their feedback on a manuscript into ‘Comments to the authors’ and ‘Comments to the editor’. Check the journal’s instructions to ascertain which type of peer review it uses.

The time to undertake a review varies – if a manuscript is of good quality, it may only take 1-2 hours. A manuscript that requires major revisions may take 4-5 hours or even more. A manuscript that is impenetrable or unpublishable may only take a short while to review. The principles outlined below are relevant to each case.

Comments to the authors
Good reviews are clear, concise and constructive. They summarise the manuscript, its findings, and its contribution to the field and the journal. Good reviews are fair and impartial; they scrutinise the research, not the researcher. Where there are limitations to the design or data, good reviews describe how the underlying scientific work might be improved. They refrain from unsubstantiated comments, unsupported assertions and unnecessarily harsh criticisms.

Journals’ requirements vary, but most editors appreciate the reviewer’s opinion of the manuscript’s relevance to the journal’s readership; its novelty and significance; the quality of its scientific design, methods, results and analysis; the validity of its conclusions; and the quality of its presentation.

A good review starts with a general overview of the manuscript’s strengths and weaknesses, and then list specific comments or criticisms (ideally numbered). Think about each section of the manuscript and and what it contributes:

  • Title and Abstract: Do they reflect the importance and relevance of the work and its main findings?
  • Introduction: Does it place the work into context and present hypotheses clearly? Are there any errors of fact?
  • Methods: Are the methods (including statistical analyses) appropriate and used correctly? Have the authors performed adequate replication? Are the methods explained in sufficient detail to allow replication by others? Have the authors adhered to established codes of practice and ethics when conducting their research?
  • Results: Are they clearly explained? Are any caveats or limitations identified and/or mitigated?
  • Conclusions: Are all the conclusions supported by the results? Have the authors considered any alternative explanations? Have the authors published any of the data or conclusions (in whole or in part) previously?
  • Tables and figures: Are they clear and easy to interpret? Do they follow basic principles of information presentation?
  • Data: Do the authors list the sources of all publicly available data? Have the authors adhered to the journal’s policy (if relevant) on data availability and sharing?
  • References: Has all the relevant literature been cited? Do the authors rely primarily on their own publications or do they cite others’ work?

Consider also the manuscript’s presentation. Is it clearly written or would it benefit from language editing by a native speaker? Have the authors followed ethical publishing practices when preparing their manuscript?

In blinded peer review, the editor may prefer that you don’t share your recommendation on publication with the authors, as it could be confusing or contradictory depending on the other reviewer’s and editor’s feedback. Some manuscript tracking systems ask reviewers to rate the suitability of the manuscript for publication by selecting from a list of options. Otherwise you can convey your opinion to the editor.

Comments to the editor
Some (but not all) journals allow reviewers to provide confidential comments to the editor as part of their review. These comments are not shared with the authors.

There is no need to repeat your comments to the author in this section, but it can be helpful to the editor if you summarise them. State your recommendation regarding the manuscript’s suitability for the journal, with clear arguments for or against publication. If the manuscript has raised any concerns about potential breaches of ethical publishing practices, you should provide sufficient information for the editor to investigate further.

Submitting your review

Follow the journal’s instructions for submitting your review online (the invitation email usually provides a link). If you’re uploading any additional files, such as a marked-up manuscript, make sure they don’t reveal your identity (e.g. in the file’s properties or tracked changes).

Try to submit your review on time, ideally before the deadline. If you can’t meet the deadline, a quick email to let the editor or Editorial Office know your estimated date of completion will be much appreciated.

What happens next will depend on the journal. Some journals send reviewers the final decision on the manuscript (often including all reviewers’ comments), which can be useful for seeing how your assessment compares with others.

Peer review takes time and effort – for you as a reviewer to perform it, and for journals and publishers to collate, use and communicate the results. Improving peer review is everyone’s responsibility and contributes to ensuring that the community in your field of expertise – and science and scholarship more broadly – benefit from your expertise, knowledge and hard work.

Authors: Caroline Hadley and Dugald McGlashan