Structure of An R01: Research Design & Methods

Apr 14 2009 Published by under Grant Review, Grantsmanship, Peer Review

Quite a while back, I began--but never completed--a series on how to structure an effective R01 application. So far we have covered the Specific Aims and Background & Significance sections. Although the next consecutive section of an R01 app is the Preliminary Studies, in response to reader requests, we will skip over that for now and treat the Research Design & Methods. This is the section of the grant where you flesh out your Specific Aims in detail, and provide the specifics of your experimental plans.
(Detailed guidelines are inside the crack.)


The Research Design & Methods section begins with a paragraph or two that essentially summarize the entire Background & Significance and Preliminary Studies sections. This is in keeping with the idea that every section of the grant application should be comprehensible when standing alone. You can use language like this:

In order to blah, blah, blah, it is necessary that yadda, yadda, yadda. Thus, determing the bleh, bleh, bleh and the blargh, blargh, blargh are key issues in the field of blecch. In our Prelminary Studies we have begun to explore novel approaches to determining bleh, bleh, bleh and blargh, blargh, blargh, and observe that ying, yang, yaggity. These Preliminary Studies lead to the following testable hypotheses: (1) whatevs; (2) double whatevs; (3) infinity whatevs. The two (or three) Specific Aims test these hypotheses.

Then, go ahead and list your two (or three) Specific Aims in bold and/or italic typeface. After listing the Aims, write a paragraph with one sentence for each Aim that summarizes the experimental approach for that aim, like this:

Blibbity methods will be used to fling flang flog in blerg cells and measure effects on flub flab flib (Specific Aim #1). Splippity methods will be used to determine the effects of semplenter in blurg cells on flappity fling flong (Specific Aim #2).

Now you get into the nitty gritty for each of your Specific Aims individually. My practice is to break down each Specific Aim into sub-sections where each subsection corresponds to an experimental effort that would likely end up corresponding to a single figure in a manuscript. I do not number or letter these subsections, and definitely do not get into any of this "in Sub-Aim #3(A)(1)(b)(ii) blah, blah, blah". The important things to explicitly treat are not--in my opinion--really ticky tacky shit like concentrations of reagents (unless you're doing an explciti dose-response or something), but to explain the basic experimental design, including negative and positive controls.
I have never included explicit "Expected Outcomes", "Possible Pitfalls", or "Alternative Approaches" sections in my Research Design & Methods. Now that R01s are going to soon be 12 pages, no one is going to be able to include that happy horseshit, and good riddance. I do provide some basic explanation of how the possible outcomes of the experiments relate back to the hypotheses to be tested by the Specific Aims.
After the detailed treatment of the Specific Aims individually, I then conclude with two additional subsections. First, is one paragraph headed "Summary and Conclusions", which should be a completely self-contained summary of the grant application, followed by a forward-looking explanation of what the value of the proposed studies will be for the field:

We have demonstrated in Premiminary Studies using a combination of blah approaches and bleh approaches that yadda yadda is an effective tool for YAHOO! Using this novel tool, we have presented preliminary evidence for HOLY FUCKING SHIT COOL YEAH! These preliminary studies lead to specific hypotheses concerning this, that, and the other HOLY FUCKING SHIT COOL YEAH thing. Using a combination of all sorts of TOTALLY FUCKING COOL NOVEL techniques, we test these hypotheses through the experiments proposed in the Specific Aims. The proposed studies will provide valuable information about TOTALLY FUCKING AWESOME IMPORTANT SHIT YAHOOOOO!!!!!, and thus be of great value to people who care about TOTALLY FUCKING AWESOME IMPORTANT SHIT YAHOOOOO!!!!! These studies will also provide extensive validation of TOTALLY FUCKING COOL NOVEL techniques as a general platform for investigating FUCKTONS OF OTHER TOTALLY FUCKING AWESOME IMPORTANT SHIT!! YAY!

Then you provide a brief timeline that breaks down very broadly what Specific Aims will be pursued in what years of the award, and include some justification for why certain shit will take a long time, while other shit will not. The point here is to justify your request for a full five year budget.
Having said all of this, I will also say that in my own grant reviewing, I first read the Specific Aims, and then the Preliminary Studies/Progress Report. At that point, I have already 90% made up my mind about the score, and I will look at the Research Design & Methods mostly just to see if there are any egregious problems, like impossible experiments or horribly deficient controls.

18 responses so far

  • drdrA says:

    Incredibly fortunate timing. I was just digging up your (and Beaker's) excellent advice on Specific aims for a jr colleague!

  • drdrA says:

    One thing to add. After I write that first sort of summary paragraph, I don't write a paragraph for each aim- instead I insert an EASY to follow flow chart here-the key word is easy- that summarizes briefly what will be done in each aim (preferably with pictures), and what the expected result will be, and what is carried on to the next aim (if anything).
    When you are making this flow chart it may seem dumb and obvious to you- but reviewers seem to like it... esp. when they are reading two or three of these babies and you give them something where they can see the whole flow of things in one snap shot. Maybe I should just write a post about flow charts after I turn this grant in today...

  • Beaker says:

    I first read the Specific Aims, and then the Preliminary Studies/Progress Report. At that point, I have already 90% made up my mind about the score, and I will look at the Research Design & Methods mostly just to see if there are any egregious problems, like impossible experiments or horribly deficient controls.
    I was surprised to hear this advice, as some other seasoned reviewers have told me that the Research Design is most of what matters in the whole grant. It's the meat section, with the rest of the application being mere condiments. CPPs advice about the Research Design, if I interpret it correctly, might be boiled down to the following:
    1. Be sure you are addressing important outstanding questions in your field.
    2. Plug your experiments into the general/generic format described by CPP.
    3. Arrange your proposed experiments according to the expected "real world" outcomes: a figure for a paper.
    4. Practice good experimental design, especially the controls.
    5. Don't say anything stoopid, which would create a "fatal flaw" that exposes you Experimental section to damming criticism.
    In addition, I've heard the advice that inclusion of some excessive technical detail may help you because--if used in the appropriate amounts--it "reassures" the reviewers that you are competent and not just glibly making shit up.
    Lastly, I am relieved to hear that CPP does not bother much with the "possible pitfalls" nonsense.
    Thanks CPP. Spot on, as usual.

  • Fodder says:

    I love it when the reviewers use the things I wrote in Potential Pitfalls as things I "didn't address."

  • I was surprised to hear this advice, as some other seasoned reviewers have told me that the Research Design is most of what matters in the whole grant. It's the meat section, with the rest of the application being mere condiments.

    I disagree completely. The Preliminary Studies are the meat, and the rest is condiments. The Research Design & Methods is there to satisfy smalltown grocers on study section. Creative scientists know that if you are capable enough to generate fascinating preliminary data, then you are surely capable of following those data to where they point.

  • Luigi says:

    I think you have to put in lots of detail for skeptical people who don't know you're competent, have methodological sticks up their ass, and love detail. There are lots of people like that. I also think you need to avoid too much detail because people will get distracted by it and nitpick. Plus, there really isn't enough space to go into all the mind-numbing procedures and caveats and potential side-experiments. I think you should include flow charts and stuff to clarify the approach for harried reviewers, but avoid flow charts so that your thoughts don't appear vague and undetailed. You need to include caveats and potential pitfalls, in case the reviewer thinks of them anyway. But you should avoid pointing out caveats and potential pitfalls, in case you end up shooting yourself in the foot. Worry about the methods section a lot, because it is very important. But don't spend too much time on it, because the important parts of the proposal are other parts.
    Get it? Don't miss the forest for the trees. There's no one right way to do it.
    But I defiietely agree with CPP: Build a compelling story in the background & preliminary data so that reviewers are dying to know how it all ends. If you don't do that then it doesn't matter whether you're competent. If you do do that, they'll trust you.
    Other advice (none from me)...
    http://www.anu.edu.au/BoZo/Scott/Studentresources.html
    Almost half the way down this page is a bag of grant advice, in case you want to procrastinate writing the thing some more while pretending you are actually (sort of) working on it.
    I am sure others here know other useful links. Post them.

  • Beaker says:

    The Preliminary Studies are the meat...
    I can think of two reasons to show preliminary data:
    1. "We can do this. See, here is some preliminary evidence."
    2. "We have discovered something really fucking important. See, here is what we have so far--and it is so fascinating that I'm sure the reviewers will agree that more money should be spent to pursue this novel lead further."
    Reason #2 is obviously more compelling, but are there times when "proof-of-principle" alone justifies showing data?

  • Reason #2 is obviously more compelling, but are there times when "proof-of-principle" alone justifies showing data?

    Absolutely! It proves you can do the experiments you are proposing!

  • Pinus says:

    I am putting in my first R01 soon. I have enough preliminary results that show that a) this is pretty cool (for the things that I am well qualified to perform). still working on the proof of principle for the technique thing that is amazingly innovative and exciting. Interesting to hear that people who may be in position to review (cpp) think that prelim results are the real meat, as I have always been told that research design is.

  • whimple says:

    1. "We can do this. See, here is some preliminary evidence."
    I think this is critically important, particularly for a newish investigator without an extensive established independent track record. You really open yourself up for damage when you propose experiments you haven't shown you can personally perform, even if those experiments are "standard in the field and everyone can do this if they try". If that's really the case, then try it one time and show you can do it.
    2. "We have discovered something really fucking important.
    The value of this seems less clear because it is interpretation dependent, especially if the "preliminary" data is unpublished at the time of grant submission. One person's 'important' is another person's 'uninteresting' or 'esoteric' or 'difficult to interpret'. Making too big of a jump forward (if it does indeed turn out to be forward) can be problematic.
    You might think 2) is more compelling, but I'm not sure that is actually so.

  • DrugMonkey says:

    As the resident small town grocer, I should point out that PP is high as a kite and if you follow this advice you* are guaranteeing not to get your grant funded.
    __
    *The first rule of success is to figure out what the specific study section your grants are being reviewed in expect. As I've mentioned repeatedly, I have little doubt that study sections have cultural expectations that persist across long periods of time. These very likely vary across sections and subfields. My study section, for example, has been pretty obsessed with the point by point on interpretation of possible outcomes, discussion of pitfalls, alternate interpretations, etc. Grants that come through are very likely to draw fire if they ignore these sections. get a comment like this and fail to address it in the revision? death knell.
    Yes, these issues are going to change in importance when we go to 12 page apps. But the basic principle is the same. You need to find out what the expectations are for the study sections to which your grants are submitted. Way more important than what happens to work for a single PI like PP. (and as much as I would like to see it, I don't recommend leading off your Introduction to Revised application with "Yeah, but Comrade PhysioProf said....)

  • Pinus says:

    Another naive question from pinus:
    How do you search Crisp by study section?

  • Pinus says:

    disregard my query. figured it out.

  • DrugMonkey says:

    How do you search Crisp by study section?
    It is the box that is titled "IRG". Click in there and start typing (scroll bars in CRISP are nearly useless) the initials for the study section you want.
    A related note: If you click all the way through to the Abstract page for a given grant you will notice the study section in which it was reviewed listed at the bottom next to IRG. Handy when trying to figure out where people doing work similar to yours are getting successfully reviewed.

  • Pinus says:

    Thanks!
    That is some useful information, I wish I had known about this earlier, it somehow slipped under my perception.

  • Luigi says:

    The best rebuttal to a panel's critique that they weren't sure the applicant could do a critical technique was given by my wife for her first revision of her first R01...
    Reviewers questioned whether ... in Aim 1 are feasible. A manuscript, currently In Press at Nature (see appendix), demonstrates that the proposed experiments are feasible and sufficient for addressing the hypotheses outlined in this proposal.
    (or something like that)
    Nothing like answering dopey skepticism with a Nature paper.
    She got the grant, by the way.

  • Pinus says:

    Awesome, I just need to get a paper in at nature for when my grant gets shitcanned! I need to get to work on that!

  • Yes, these issues are going to change in importance when we go to 12 page apps. But the basic principle is the same. You need to find out what the expectations are for the study sections to which your grants are submitted. Way more important than what happens to work for a single PI like PP.

    This is true. But when you're writing your first R01, you gotta start from something that worked for someone. I suppose if you were able to work the angles properly, it would be ideal to get your hands on an R01 that was highly scored by the very study section that is going to review your first R01. This is not usually feasible.
    In my case, I was given a funded R01 from someone working in a pretty different area than I do. In fact, my basic R01 structure still hews pretty closely to the one I was given. And I give out my funded R01s freely to new junior faculty at my institution to use as templates.
    Finally, it should be pointed out that the extent to which "pitfalls and alternative approaches" garbage is required is not only study section dependent but also career-stage dependent. Once you have a track record for getting shit done, you need less of this stuff. (Cue DM's howls of anguish at "track record"!)

Leave a Reply