On January 27th, 2014, the MSU BEACON Center graduate students held a panel on how to
review. The panel was organized by Emily Weigel (@choosy_female) and Jory Schossau,
Bjorn Ostman (@CarnyEvolution), Kristin Parent, Rich Lenski (@RELenski), and Arend Hintze were panel members, together with me.
I put together the following notes, somewhat amended in response to
Rich, Kristin, Bjorn, and Arend's comments, and the students'
questions. I hope you find them useful!
Note that we talked about both grant and paper reviews, and Rich could
provide additional perspective on paper reviews because he is an
editor of PNAS.
Goals as reviewer
Your goals on review should be:
- help editor/panel make good decision
- help authors write a good paper
Titus's approach to reviewing
Skim-read the grant or paper; highlight areas that don't make sense.
Decide if you are pro-, anti-, or on the fence.
If pro, try to write a review that objectively states why. Pay special attention to framing in a larger context, if you can.
If anti, identify major flaws clearly and objectively. Behave as if the authors know who you are and where you live.
If on the fence, look harder for support one way or another; if you can't find, explain that you're on the fence, and why.
I structure my reviews like so:
Broad overview: why this is important, or not.
Big-picture problems ("they failed to convince me of X, Y or Z")
Nits to pick (bad figures; spelling; grammar; etc.)
Note that generally editors want recommendations separate from your text, so that they don't get into arguments with authors about things like "but 2/3 reviewers said accept!"
It's easy to find BS reasons to reject papers; try to find good
Things that I hate to see in grants that will lead me to be grumpy (says
- bad grammar
- bad science, defined as "stuff that is clearly wrong/badly done", not just stuff I disagree with
Do say when you don't know something/aren't an expert. You will be helping the editor!
Things you should try doing:
I have posted other peoples' reviews for one of my papers here: http://ivory.idyll.org/blog/khmer-counting-reviews.html
I post many of my grants on my Web site:
I passed out reviews for three grants; I cannot make the reviews public,
but here are the grants.
- 2011-NSF-CAREER: http://ged.msu.edu/downloads/2011-nsf-career.pdf
- my takeaway: "Good proposal, bad PI, not enough preliminary data"
- 2012-NSF-CAREER: http://ged.msu.edu/downloads/2012-career-nsf-final.pdf
- my takeaway: "OK proposal, OK PI, but too much preliminary data"
- 2012-NIH-BIGDATA: http://ged.msu.edu/downloads/2012-bigdata-nsf.pdf
- my takeaway: "Please fund this person."
Note, 2012 grants were written at same time, but got very different
reviews. Reviews depend heavily on reviewer!
Entertaining #SixWordPeerReview tweets from Rich:
These are absolutely not how you should write reviews. But they
Env. Micro. referee comments:
"It is sad to see so much enthusiasm and effort go into analyzing a dataset
that is just not big enough."
"You call the sample fresh water, this is confusing as it is saline water."
"The biggest problem with this manuscript, which has nearly sucked the will to
live out of me, is the terrible writing style."
Again, try to avoid this kind of thing in real reviews :)
My blog post on why current peer review is broken: http://ivory.idyll.org/blog/angenmap-discussion-on-peer-review.html
Another recent blog post that I found interesting:
Questions? Comments? Ask!
There are comments.