How to Read a Scientific Paper as a Reviewer
The ability to read a scientific manuscript both critically and efficiently is not just an important skill to have as a reviewer, but can also help you understand how reviewers assess your own work. Although different types of studies may require tailored approaches, the following general guide on how to read research articles as a reviewer can be applied for most article types. Whether you are early in your research career or a seasoned reviewer, practising this approach can help you to make fair and constructive assessments of research articles, supporting authors by helping to ensure that their work is clearly described and robustly reported.
- Start with the abstract
The abstract should provide you with a clear overview of the aims of the study, methodology, key results and conclusions. The abstract should enable you to quickly gain a sense of the overall context of the research, the techniques used, and the key data, before reading the full article.
As a reviewer, you should ensure that the authors have provided an honest summary of their key findings in the abstract without overselling the conclusions or overstating the generalizability of the data.
- Initial ‘high-level’ readthrough of the manuscript
As a reviewer, you will most likely need to read through the manuscript multiple times before submitting your report. Before reading the manuscript in detail however, it can be helpful to have an initial ‘skim read’ of the entire manuscript. At this stage, don’t worry about spending time trying to understand every detail as if there are any sections that you find hard to understand you can return to these later. This first-pass readthrough will help you to orientate your reading and prepare you to assess the main strengths and weaknesses of the study.
If at this stage you believe that the manuscript topic or methodology is out of your area of expertise, you can contact the Editor or Editorial Office to explain and they will be able to guide you. Very often Editors invite reviewers to review a specific aspect of a manuscript and they may explain to you which aspects of the study they were hoping for your expertise on.
As you read, keep notes of any limitations or errors that you notice, as often some of the most important issues to highlight to authors will become apparent during this first read. Make a note of any obvious issues you notice with the data and study design, such as insufficient power, missing controls, or statistical errors. Identify any aspects of the study that could be strengthened through additional data collection or analysis.
You should also check the figures, which should clearly display the results of the study without requiring a deeper read of the text to understand them. Remember that any conclusions drawn by the authors must be supported by the data presented.
- Detailed readthroughs and assessment
After your initial readthrough, you should have already gained a general impression of the work and may have highlighted some important issues that the authors will need to address. A more detailed readthrough can now be carried out to focus on specific issues that will require revision. If any sections were unclear in your first readthrough, this is the time to look at these areas in more detail.
The following guidance provides specific advice on how to assess each section of the manuscript, and are applicable for most research papers:
Introduction:
The introduction of the manuscript should provide context for the study grounded in previously published work, and demonstrate a clear rationale for the research. The authors should use the introduction to present the main aims and hypothesis of the study.
As a reviewer, you should check to ensure that the introduction includes sufficient and appropriate citations of primary research in the field, without demonstrating bias towards a particular school of thought, journal or geography. Make a note of any key studies related to the work that have been omitted. Authors should avoid citing too many review articles in lieu of the primary research article, and the literature cited should be up-to-date. It is acceptable to ask authors to include important relevant citations if there are key studies that have been missed, but this request must be clearly justified to the authors in the review report and you should not ask authors to include your own citations (for more information see the COPE Ethical Guidelines for Peer Reviewers).
Methods:
The methodology used in the study should be clearly described, reproducible and transparent. Ensure that the authors have used the most appropriate techniques to address their hypothesis and have included suitable controls. Make a note of any methodological weaknesses and limitations that the authors will need to address. It is worth asking yourself whether alternative or additional techniques and analysis could more accurately address the study aims. You should also ensure that the statistical analyses are appropriate, that any tests chosen are correct for the research question and data collected, that multiple comparisions have been controlled for, and that any potentially confounding variables have been captured and accounted for.
Ensure that the study design adheres to international standards in ethical practice, that relevant ethical approvals and participant consent have been obtained, and that the authors have followed the appropriate reporting guidelines for the study type. A full list of reporting guidelines for most study types can be found on the EQUATOR network website, and it can be helpful to check the manuscript against the relevant checklists to make sure that the authors have adhered to best practises.
In addition to helping authors identify potential errors in their study, one of the most important roles of the review is to help Editors identify manuscripts that are unsuitable for publication because of methodological flaws or ethical concerns. On rare occasions you may identify concerns about how the study was designed, conducted or analysed. If you have ethical concerns about a manuscript you are reviewing, check the journal policy about who you should report this concern to. Typically, this information can be found in the Reviewer Guidelines, but if this information is not readily available, you can contact the Editor or the Editorial Office directly and they will be able to support you.
Results and figures:
The results section should provide a descriptive account of the data shown in the supporting figures, and the figures should be clear, accurate, and easily understandable. The figures should include relevant datapoints, suitable error bars and statistical information. Check whether there are any outliers in the data, and whether these have been discussed. Remember to check the supplementary figures as well as the main figures. As a reviewer, you should ensure that the data have been accurately interpreted and support the conclusions drawn in the manuscript.
Read through the figure legends to ensure that these contain sufficient information, including detail of statistical tests if appropriate, sample size or number of repeated experiments. If the authors have provided access to raw data or source code, e.g. via a data repository or in supplementary files, ensure that these are complete.
Discussion and conclusion:
Whereas the results section should describe the data that has been collected, the discussion and conclusion sections enable authors to discuss potential interpretations of their data and highlight new findings in the context of previous research. As a reviewer, it is important to assess whether the author’s interpretations can be supported by the data presented in the manuscript, or whether the conclusions may be overstated or overgeneralized.
The discussion section should also be used to highlight limitations of the study, and propose future directions for research. Ensure that the authors have sufficiently discussed alternative interpretations of the data, and that they have clearly acknowledged the limitations of their study. Transparent communication about sources of uncertainty in the data and conclusions can ultimately build trust in the authors’ work and, as a reviewer, you help them by providing an external perspective on the clarity of their communication.
Reference list:
You should check the reference list to ensure that the authors have cited appropriate research. Check to ensure that the authors have included recent papers, and that primary research articles, not just review articles, have been included. The journal will ensure that the reference formatting adheres to their guidelines, but if you notice any major errors or issues with bias or relevance of the citations you should highlight this in your review.
- General tips
- As you read through the manuscript, it is a good idea to keep detailed notes that you can refer back to when writing your reviewer report.
- Remember that as a reviewer, your role is to assess the scientific quality of the manuscript rather than act as a proof reader. If the writing quality of the manuscript prevents understanding, you can highlight the need for improvements in your report but you do not need to provide a line-by-line list of typographical or grammatical errors.
- If you do not feel confident or qualified to assess certain aspects of the manuscript, make sure that you inform the Editor as they may need to invite an additional specialist reviewer, such as a statistician.
- When writing your final report, remember to maintain a professional and respectful tone and avoid being too personal in your comments (e.g. use “the authors…” rather than “your…”).
- You must not upload any of the peer review information or materials to an AI tool as this would be a violation of the confidentiality of peer review and of the authors’ copyright. You can check the journal’s reviewer guidelines for more information on appropriate use of AI in peer review.
- If you have any ethical concerns regarding a submission, make sure that you highlight these to the Editor or Editorial Office.
Summary
As with most skills, the more you practise reading and critiquing scientific manuscripts, the easier the process becomes and the more confident you will be at identifying issues. You can also practise these skills when reading published manuscripts, by participating in journal clubs, or by commenting on pre-prints. If you are early in your career and would like to gain more experience as a reviewer, we also have opportunities for you to review manuscripts in partnership with a more senior colleague or supervisor as a co-reviewer.
If you would like further information on how to become a reviewer for Karger journals and access to our online courses for reviewers, please visit our reviewer information pages here.
(Featured image declaration: photo by cottonbro studio from Pexels: https://www.pexels.com/photo/light-man-coffee-cup-6830875/)