Authors write a grant or paper which is submitted to a funding agency or journal. It’s then sent out to some other academics who anonymously assess it and come back with comments/criticisms/recommendations. The editor/funder then decides whether to publish the paper or give money to do the research. Reviewing is a major component of being a member of the academic scientific community. Nevertheless, there is huge variation both in the way people go about reviewing and the time they invest in it.
I have very little experience of reviewing grants, so I’ll focus on manuscripts here. I like reviewing manuscripts – I really do. Firstly, it’s an extra opportunity to read a paper in depth, (i.e., not just scout out the information you want or find interesting); it’s a chance to see new approaches to analysing data and new questions in the field; and it’s an opportunity to influence things a tiny bit. I must confess, there is also an ego component to it: “What’s that important editor in the field? You want my opinion?” Sometimes I despair at myself, but there it is.
Much as I like reviewing, it takes a lot of time from my schedule.
I read somewhere (and I can’t remember where – please chip in if you know so I can reference the source) that one should aim to spend two hours on reviewing a manuscript. No. Just no. No-can-do. If I limited myself to two hours, I’d have a superficial knowledge of what the authors were on about and could provide only the weakest of reviews. Likewise, I wouldn’t consider marrying somebody after a speed-date. After submitting a manuscript, it can take 1-3 months for the authors to get the reviewers’ comments back. If one of the comments are “the authors did not consider x” and figure 1 in the paper is “the effect of x on the trait in question”, the authors will, quite rightly, want to do the reviewer a serious physical discourtesy (fists and/or knives may be involved). We all make mistakes, and as a reviewer, I want to minimize the chance of making them. I do this by re-reading everything.
Reviewers are usually asked to submit their comments to the journal within two weeks. I get the sense that a lot of people wait until the end of that period (or even until they are badgered by editors for being late). I’m not judging people for this – some people can write insightful reviews immediately before the deadline. I can’t. I need to do the review, put it away and work on something else for a few days, dig it out and re-read everything. I am someone who very much needs ponder time, and I’m constantly surprised by how much my views can change after this period.
Sometimes, reviewing means I need to learn new stuff.
Another reason why I can’t wait until the deadline to review is that sometimes I need to read up about parts of the paper. This is especially true for statistical analyses sections. If someone is doing an analysis I’m unfamiliar with, I like to see that it makes sense. Selfishly, I find this really useful – there are numerous times I’ve learned new statistical techniques as a result of reviewing. I’ve also had to read up about particular sub-fields, which has given me ideas for my own research – and we all love that feeling, right?