Wednesday, November 26, 2014

Grant writing tips

I have been asked to speak at a grant writing workshop for the School of Mathematics and Physics at UQ.
Here are a few preliminary thoughts.

Consider not applying.
Seriously. Consider the opportunity cost. An application requires a lot of time and energy. The chances of success are slim. Would you be better off spending the time writing a paper and waiting to apply next year? Or, would it be best to write one rather than two applications? You do have a choice.

Don't listen to me.
It is just one opinion. Some of my colleagues will give you the opposite advice. I have never been on a grant selection committee. My last 3 grant applications failed. Postmortems of failed applications are just speculation. What does and does not get funded remains a mystery to me.

Take comfort from the "randomness" of the system.
You have a chance. Don't stress the details. Recycle old unsuccessful applications. Don't take it personally when you fail.

Who is your actual audience? Write with only them in mind.
For the Australian Research Council it is probably not your international colleagues but rather the members of the College of Experts. It needs to be written in terms they can understand and be impressed by.

Why should they give YOU a grant?
I find many people sweat about the details of the research project or think if they have brilliant cutting edge project they will get funded. I doubt it. Track record, and particularly track record relevant to the proposed project is crucial.

Not all pages are equal.
Unfortunately, the application will be 60-100 pages. Don't kid yourself that reviewers will carefully read and digest every page. Some are much more important than others. You should focus on those.

The first page of the project description is the most important. Polish it.
I sometimes read this and I have no idea what the person is planning to do. I quickly lose interest.

"Contributions to the field" and "Research accomplishments" means scientific knowledge generation not career advancement or hyperactivity.

Choose your co-investigators carefully.
They may lift you up or weigh you down. I am usually skeptical of people who have "big name" co-investigators they have never actually published with before. The more investigators the larger the application, and the more material available for criticism. Junior investigators need to realise that the senior people will usually get all the credit for the grant, even if they contributed little to the application. This is the Matthew effect.

Trim the budget.
The larger the budget the greater the scrutiny. It is better to get a small grant than no grant at all. Ridiculously large requests will strain your credibility.

Moderate the hype, both about yourself and technological applications.
There are reviewers like me who will not take you seriously and be more critical of the application.

Be discerning about what publication metrics (citations, journal impact factors) to include.
Impact factors have no impact on me. I don't see the point or value of short term citations.

Writing IS hard work, even for the experienced.
See Tips in the writing struggle.
Get started early. Get feedback.
Write. Edit. Polish. Rewrite. Polish. Polish.

My criteria for research quality.

Responding to feedback from administrators in the university Research Office.
Put in the application early. They can give very helpful feedback about compliance issues, formatting, page limits. Take with "a grain of salt" advice/exhortations about selling you and the science.

I welcome comments and suggestions.
What advice and suggestions have you received that were helpful or not helpful?


  1. To me the devil is in the details, particularly how to separate details from the generalities.
    Too much "background and significance", including a broad view of the impact the proposed work might (!) have in the field, and comments come back saying it's not practical enough.
    Too much detail, proposed experiments, parameter space explored and hypotheses formulated for that space, and specific details on how to measure what you propose, and it looks like a vision is missing.

    So, the separation of vision/b&s/... from the details that you need to convey in order to make it possible to be judged on feasibility and understanding of the issues, is what I find hard.
    It takes away the linearity of the story (from broad to detail), as writing a vision, and then a part on how to implement this in practical work, necessitates overlap.

    So, summarizing the relative fraction of B&S/vision and practical details, and most importantly how the former is distributed to make the latter not look like a list of "to do's" is what I find hard.

    1. I agree finding the right balance between generalities and details is a real challenge. Furthermore, the referees/assessors can be quite frustrating. One will say there is not enough technical detail and another will say there is too much technical detail.

  2. oh, and the most important thing is that you should be able to communicate the proposal in 3 sentences. That includes 1 sentence of why (i.e. background and vision), and 2 sentences of how.
    If you can't do that, don't start writing: you're not ready.

  3. As you asked for comments here are a few - they're not very well thought out.

    1) Impact factors. One of your objections is that they encode things you know e.g. PRL~JACS>PRB~JCP - but that might not be known to everyone reading your grant (I bet it's not obvious to many physicist that JCP > CPL and do chemists know the relative rankings of PRB and JPCM?) So this information has some value (eg in the lists of publications)

    2) I think you should ignore the ARC's structure somewhat. This is hard to do well - but the rules re headings can get in the way of making a compelling case. I think one can benefit from bending the rules - but it is very important not to break the rules. (This may be hard to communicate to a large audience.)

    3) You mentioned this - but I think tailoring the track record to emphasise how this has positioned you to do this grant is important. I think you give some really good advice re this after my first, unsuccessful, application for a future fellowship. The track record there focused on what I thought were the best things I had done. In my second, successful, application I wrote about why I was ready to do this project.

    4) Remember you need to explain why the taxpayer should give _you_ money to do _this_ science. The two need to tie up. This goes double for fellowships. Why are you the best person to do this science?

    5) I find it useful to explain the science case to someone outside of my sub-field. Before I start writing. The figures you sketch on the whiteboard need to go into the grant.

    6) Get lots of feedback from different people - but make sure to send then the entire proposal, not just the science case. Often some of the most useful feedback does not concern the science case.

    7) Get a grant writing mentor - this person should be your supervisor!

    8) Write the boring bits first. This is helpful psychologically as you're less likely to get writers block and it feels like you're getting somewhere (50 pages, written just 10 pages of science to go!)

    1. Hi Ben,
      Thanks for the helpful comments.
      It is good to hear a complementary perspective.

      One minor point that is not clear to me is your point 7.
      Why should your grant writing mentor and supervisor be the same person?
      I think there could be some value from having different people for the different roles.

  4. Oops - there should have been a NOT in my point 7!