For a scientific, experimental paper, a peer reviewer should consider these issues:
- Are the topic and results appropriate for the journal? If not, there is little point in reviewing other aspects of the paper as it is sure to be rejected.
- Is there a hypothesis?
- Are the methods complete?
- Were appropriate control experiments performed?
- Are there conclusions?
- Are the conclusions motivated appropriately?
- Do the methods support the conclusions?
- Are the graphs/illustrations readable and do they convey useful information?
- Is the data that should be presented as graphs actually in the graphs?
- Is the research ethical?
- Is the research placed in context with the literature?
- Is information taken from references accurate, or at least plausible?
- Are the results likely to be reproducible?
- Is the writing intelligible?
- Do the title, abstract, and introduction describe the paper effectively?
The answers should all be yes.
In my experience many reviewers also identify typos in the paper. I only do that if the typos are in technical information or make the paper incomprehensible. Most journals pay a copyeditor whose job it is to correct ordinary spelling and grammar errors.
The reviewer needs to write a report. The report includes numerical ratings on scales provided by the journal. The report should:
- Summarize the paper in a few sentences, clearly identifying that the reviewer read the correct paper.
- Identify the main strengths of the paper. Sadly this can, at times, be hard.
- Identify the main weaknesses of the paper. If the paper needs revision, then all weaknesses which need correction should be listed.
- Each weakness should be numbered and its location in the paper should be specified. This is to assist the editor and future reviewers in checking to see if the revisions were made.
- Recommend an action the editor can take.
The reviewer is not responsible for telling the authors how to improve their paper.
There is no way to predict how long a review will take until you have read the paper. It varies widely.
If the reviewer has a conflict of interest, this must be disclosed to the editor. Example conflicts of interest are:
- Reviewer works for the same organization as an author.
- Reviewer might make or loose money if the article is published.
- Reviewer is friends with an author.
- Reviewer has published work or applied for funding with an author.
- Reviewer considers an author to be a family member.
- An author has offered the reviewer a bribe.
- Reviewer intends to publish a paper addressing the same hypothesis.