Being a referee is somewhat of a thankless task. It means being a good citizen and putting the interests of the scientific community before your own. Furthermore, if you are conscientious and provide a helpful report in a timely manner, the journal editors will "reward" you by sending you more and more papers to referee. Postdocs can easily sink a lot of time into refereeing papers, to little personal gain. Someone once told me a helpful rule for being a good scientific citizen: you should referee approximately the same number of papers per year as you submit [normalised to the number of co-authors]. I encourage my postdocs to follow this guideline.
Some good reasons to referee a paper:
- You have some constructive feedback that will improve the paper.
- The paper has serious flaws.
- It is good experience to learn how referee's may respond to your papers.
Some good reasons to not referee a paper:
- You are too busy and can't provide a report in a timely manner.
- You can't be objective because of a conflict of interest.
- You already referee more than your share of papers [see the rule above].
- You already have one paper to referee.
- You really don't have the expertise to evaluate the validity and/or importance of the paper.
- The paper is so poorly written it is difficult to follow.
Some bad reasons to referee a paper:
- You just can't say no.
- You mistakenly think that if you say no the journal editors won't be sympathetic to your next submission.
- The paper is from a competitor and you want to stop it getting published.
- You mistakenly think that being a referee looks particularly good on your CV.
- The paper cites a least one of your papers and so you want it to be published.