Learning to review

Yesterday I said how my work is all log-jammed by an overdue review. I feel bad that I haven't finished it, yet I can't get myself to get it done. I think the reason is that I feel unsure about my comments. Although I'm still green enough to feel sort of honored to be a reviewer, I hate doing it because I'm not sure if I know what I'm doing. Chuck recently wrote a nice post (with some good advice) about how learning to peer review is a major gap in grad student training. I agree.

This is not my first review. Occasionally, Research Advisor will get asked to review a paper that looks straightforward but she doesn't have time to read. If she thinks I can handle the content, she'll suggest me as a reviewer, letting the editor know that I am her student and that she will assist me if needed. So I read the paper and write the review, asking her questions as necessary. I also rely on her vast experience to help me determine how serious any errors are. I know that grad students are notoriously harsh reviewers and I have to admit I don't always have the judgment to decide if the fact that the authors used technique x instead of y is enough to kill the paper*.

This system works, I guess. It gives me an opportunity to cut my teeth on reviewing, a job I'm going to have to learn sooner or later. It's great to have a safety net. I think I'd be paralyzed with nerves if I had to submit the review without confirmation that it was sensible.

On the other hand, I would have really liked a little more training for this. If feels weird to me to write reviews when I have not received any for my own work. Sure, I've seen the reviews from a few of my colleagues' papers, but that's different from reading comments on your own work, which you know much more intimately (I hadn't even read the versions of my colleagues' manuscripts that were reviewed). I think the best scenario would be -- with the editor's blessing -- for RA and I both to review the same paper, with her having responsibility for it. We'd both write comments, discuss them, and then she would submit the review but I would get to read what she wrote. Perhaps we could do it that way a couple of times and then switch roles so that I would submit the final comments. After that, I could do my own reviews with the opportunity to check things with her as needed. I know that creates a lot more work**, but this is such an important part of the way we do science that I think it's worth it to take the time to properly train new students how to do it.

*I know the editor makes the final decision and all, but reviewers are still asked to give an overall judgment of the paper.
** I haven't asked her to do it this way because usually if she passes a review to me, she's totally swamped and I feel bad asking for her time. Plus, by the time it occurred to me to do it this way, I had done enough that I think she probably thinks I'm doing just fine.