What's the deal with preliminary data?

Dec 19 2012 Published by under Uncategorized

A few weeks ago, I attended a fancy pants conference for the first time. I was very nervous about going--worried that I would only know two or three people, feel impossibly out of my league, and retreat into a corner and cry for the extent of the meeting. As it turned out, I knew something closer to 20 or 30 people there, and it was one of the most fun, stimulating, and rewarding meetings I've ever been to. One of the people I reconnected with was someone my grad program had tried to recruit, but who'd ultimately decided on somewhere warmer. He's now an assistant professor like me, and so we bonded/commiserated on numerous assistant professor talking points, which naturally included the NIH grants system.

What struck me in this conversation was something he said when the topic of R21s came up. "The problem with R21s," he bemoaned, "is that you need preliminary data." Now, I found this to be a surprising thing to say, since my R21 was funded without any preliminary data--for that particular project. I did include a few figures to show that my post-doctoral work had taught me the techniques I proposed to use in that project. But it made me wonder whether different people have different ideas about what "preliminary data" actually means, and how necessary they think it is when evaluating grant proposals.

It's been said that when you submit an application, you basically need to have half the thing done already. Do you think this is true? If so, why? Where did this attitude come from? Do you think about R21s vs R01s differently?

Please take my handy poll! And please expand on your answers in the comments. I love comments!

15 responses so far

  • Dr. O says:

    Not absolutely sure it was intended, but I'm totally replaying the title of this post in my head as Seinfeld's voice.

    Regarding the poll, I voted the third because it's what I've been told - for both R01s and R21s. I'm very curious to see what the blogosphere commentariat has to say, though.

  • DrLizzyMoore says:

    I may have voted twice...the last two options, because I think it depends on the proposal.

    If the hypothesis is solid, straightforward, yet innovative, then having preliminary data to show that you are competent with the techniques is enough.

    If your hypothesis is complex, then I think that you need substantial data demonstrating that it is a 'reasonable' project.....just my opinion.....

  • Dr24hours says:

    I think it depends of the proposal too much for me to vote. I also think the idea of "preliminary data" is expandable to "preliminary work". For example, in my work, we often don't need any preliminary real-world data at all. Just evidence that the simulation can be built, or that it already has been, and the real world data will inform the output of the simulation.

    So there: very little or no prelim data. Massive prelim work.

  • bashir says:

    Ok I'm glad I'm not the only one thinking Seinfeld.

    I don't review grants so this is more what advice I've been given. Basically more prelim data is always good. Necessary? No, but highly correlated with getting funded.

  • drugmonkey says:

    1) It is demonstrably possible to get funded with mere "technical feasibility" data.

    2) It is also true that a very common set of StockCritiques has to do with the quality/quantity of Preliminary Data.

    2a) This comes both from the Fairness Codicil and the Confirmational Exercise Toolbox.

    3) There is no such thing as a Perfect Grant that will bludgeon all and sundry reviewers into a 4%ile score.

    Given 3, it is my firm belief that applicants should trust more in 1 and fear 2 less.

  • drugmonkey says:

    And oh yes, there is no actual requirement for preliminary data in the R01 application.

    • ecologist says:

      I think you mean in the R21 application. At least, the last time I checked, it was quite explicit that for R21, no preliminary data were required.

      • Dr Becca says:

        I think DM means an R01. You're right, it's explicitly un-needed for the R21 (although that doesn't stop reviewers from dinging an app for not having it), but it's also not explicitly stated as necessary for an R01. However, it's become generally assumed that all R01s do need at least some, and this is the phenomenon I'm interested in probing a little here.

        As a reviewer, what do you look for in preliminary data?

        • drugmonkey says:

          For an R01. There is no formal requirement and the de facto requirement is an emergent property of the peer review process. "We" have decided that it is necessary. We can therefore choose to decide that it is not necessary....either wholesale or on an individual application basis.

  • biochembelle says:

    I wonder if, to an extent, this is dependent on institute/study section and/or program officer. As DM points out, the prelim data thing is a stock critique, but if reviewing for a mechanism intended to promote new lines of research (eg no req for substantial prelim data), someone in the room has responsibility to point out the irrelevance of the critique.

  • eeke says:

    I'm not sure what you were asking - whether preliminary data are EXPECTED by a study section and the applicant should comply, or whether we, as reviewers, would prefer to see preliminary data. In my ideal world, preliminary data shouldn't be necessary. A grant proposal is just that - a proposal, and you're asking for money to try stuff. If the fuckin project is already finished (which I have been told it must be to get a decent score on an R01 application), I think the whole thing is a sham. In my eyes, if the project is nearly finished, I wouldn't know what the money would be paying for. But if it is not finished (or even started), then the proposal would seem more genuine and, even though there is a risk of the project not working, the point of the review is to evaluate whether the application is a risk worth taking. That's my how-I-wish-things-worked fantasy land.

  • One reviewer tore into my R21 saying that I had no preliminary data to prove that my hypothesis was on the right track. The grant did not get funded but we did the study anyway and my hypothesis was proved to be correct. Bring on the R01.

  • Thanks for sharing your info. I truly appreciate your efforts and I am waiting for your next post
    thank you once again.

  • Though it can be irritating that may suggest that you will want to down load
    an uninstaller to end off the occupation for you. ll lose all files stored on
    computer, some of which may be precious for you. It has a spider
    that will allow you to download whole webpages or pieces of a webpage (why anyone would want to do this is beyond me, but if you
    want to, you can).

Leave a Reply