With the National Science Foundation (NSF) ending the middle of its process of revising merit review criteria, it finally occurred to me to see how it’s done over at the National Institutes of Health (NIH). My ignorance of the NIH process, now that I’ve seen the official description of it, may well be shared by the folks involved in revising the NSF merit review criteria. The guidance issued with the new criteria may prove me wrong in that assumption, and I’d willingly eat my words at that point.
(Let’s also note the greater resources available to the NIH, which may make their more involved process feasible and out of reach for NSF.)
There are two stages to the NIH review process. The first stage focuses mostly on the scientific merits of the proposed research, with the reviewers chosen based on their knowledge and expertise in relevant scientific fields. The criteria used in this stage by the NIH maps closely to the intellectual merit criteria used by NSF.
The second stage is a review conducted at the Advisory Council/Board level for the Institute or Center. NIH staff gather relevant information to inform this stage fo the process. The Council/Board considers the needs of the center as well as the results of the first stage before deciding what gets funded. The Council/Board consists of “scientists from the extramural research community and public representatives” that are selected by the Institute or Center. There really is no comparable stage of review at the NSF. Staff review the first stage results and make final recommendations to the Division Director, who makes the final decision.
Given the different foci of the two merit review criteria, a two stage process similar to that of NIH may make sense (though Roger Pielke, Jr. doesn’t seem supportive of a broader impacts criterion – more on that later – he does seem to like separating that review from considering intellectual merit). While logistics and resources may make it impractical for NSF, an excellent question is whether this has been, or will be, considered as an option for implementing the revised broader impacts criterion.