Training Funds (T32, F32, F31) Distributions by Institution

Jun 03 2014 Published by under Uncategorized

The recent commentary by Alberts, Kirschner, Tilghman, and Varmus on Rescuing US biomedical research from its systemic flaws (which discussed using training mechanisms in preference to research mechanisms to support graduate students, among other topics) as well as a recent post by DrugMonkey on individual postdoctoral NRSA fellowships (F32s) reactivated my interest in analyzing data on the distribution of training (as opposed to research) funds across institutions.

Below is a plot of of the total T32 institutional training funds awarded to 195 institutions that received at least some T32 funding for FY2013 versus the total NIH funding for FY2013 for these institutions:

T32 Institutional

These two parameters are highly correlated (correlation coefficient 0.94). This correlation implies that institutional training grant funds are relatively concentrated with approximately 50 institutions accounting for about half of the total annual institutional grant investment. Some of these awards support predoctoral trainees, some postdoctoral trainees, and some a mixture. The distribution of the sizes of training grants as a function of the number of years for which they have been funded is shown below:

T32 years size

Many of the largest T32 awards are Medical Scientist Training Program (MSTP, Combined MD/PhD) awards. These are highlighted on the plot for clarity.

Note that there is a trend with average award size increasing as a function of the increasing duration of the award as shown below:

T32 size by year

This trend is driven in part by the large size and long duration of many MSTP awards but the trend is present even excluding the MSTP awards.

In addition to institutional training awards, NIH also funds individual postdoctoral (F32) and predoctoral (F31) awards. The distribution of postdoctoral F32 awards across institutions is shown below:

F32 Institution Total-Names

While there is again a strong correlation with the total institutional funding, it is less strong than that for T32 awards (correlation coefficient = 0.79 versus 0.94). There are outliers from the general trend which have a larger number of F32 awards than one would expect from the general trend.

The distribution of predoctoral F31 individual training awards as a function of total institutional funding is shown below:
Institution Total F31-3

Again, the correlation is less pronounced than it is for T32s (correlation coefficient 0.79) but with outliers below rather than above the trend line. Some of these outliers are two hospitals associated with Harvard Medical School which is an outlier above the trend line. The others may at institutions where the fraction of students applying for F31 support may be low although this is just a conjecture.

Interestingly, the number of NIH Institutes and Centers participating in the parent F31 award announcement in current fiscal year jumped to 23 from 13 at the time of the previous announcement. This may reflect NIH embracing the recommendation that more Ph.D. students being supported by fellowships rather than research grants.

25 responses so far

  • DrugMonkey says:

    So obviously there should be a factor for trainee awards when you assess lab funding, right?

    WRT to your comment about switching to more individual awards to support trainees, the timeline must be shortened. Can't start a grant and then finally fund the F32 postdoc two years later to work on it.

    • datahound says:

      I agree that trainee awards should be linked to an assessment of lab funding but this is easier said than done.

      I see your point about the timing issues about F31 and F32 awards and research grants. I wonder if NIH is considering moving to an NSF-like fellowship program for graduate students where fellowships are awarded generally prior to even selecting a laboratory for dissertation research (or even before starting graduate school).

      The increased IC participation in the F31 program in interesting (and surprising at least to me at this point). As you noted in an earlier post on the F31 program, many ICs did not participate. This is no longer the case. The broad participation has the looks of a corporate NIH decision. It seems to me that a major challenge to a shift in graduate student funding from research grants to fellowships is that I do not believe that NIH has the authority to restrict who is supported on research grants. Legally, graduate students are employees working on a research project. I suppose NIH could not allow tuition costs to be covered on research grants, but the implications of this policy should could be substantial and would need to be examined carefully.

  • mollieb says:

    Mass General and Brigham and Women's don't have Ph.D. programs of their own -- the grad students there are almost exclusively from Harvard Medical School, and are required to apply for F31 funding through the medical school's office of sponsored programs.

    (Source: I was an HMS Ph.D. student working in an MGH lab, and I had to go through HMS to get my F31.)

    • datahound says:

      Makes sense. That is what I concluded although I did not state it explicitly since I did not know for certain.

  • rxnm says:

    What is the problem with just earmarking training awards to research grants? Besides preventing the ~ILAFs and BSDs from getting an *even more* disproportionate share of the pie and *yet another* unearned job market advantage for their trainees, what would the downside be?

    • datahound says:

      rxnm: I am not sure what you mean by earmarking training awards to research grants. Can you clarify?

      • rxnm says:

        PhD students or postdocs become line items in RPGs. Already obtaining an F-award while you are working on an R-funded project makes no sense -- you have to claim it is a different project to avoid a bogus conflict of "funding same thing twice." Tying salaries to actual projects is potentially more humane for trainees as well.

        The stated purpose of F awards to is to ensure that the nation has enough trained scientists. As that is an over-met mandate, despite the wage-suppression propaganda of STEM-employing industries, the awards should be eliminated.

        • datahound says:

          Thanks for the clarification. This gets at the heart of big issues about training. Should training by about developing the individual being trained broadly or be about his/her ability to execute on a specific project?

          Proponents of mechanisms to support trainees through specific training mechanisms (as opposed to research mechanisms) note that training needs can be broader than a particular project. If a trainee is supported purely to work on a particular project, then it is difficult to justify working on side projects or taking courses or attending meetings that are not directly relevant to that project. On the other hand, if the trainee is supported "with their own funds", s/he has more control in broadening training.

          A second driver for training awards (particularly institutional training awards) is that they allow NIH to help promote good training practices (thesis committees, IDPs, etc.).

          With regard to the issue of whether too many (or too few) scientists are being trained, the data support that notion that research funding dictates the number of Ph.D. trainees much more that specific training funding. See the third figure in . As I understand it, this is one of the ideas behind the Alberts et al. push for more Ph.D. trainees supported on training mechanisms (with the assumption that the number supported on research mechanisms would decline), namely that there could be more direct control on the number of trainees.

          • rxnm says:

            I agree with all that....part of earmarking training funds to research grants would be the elimination of spending research funds on trainee salaries (but perhaps allow for technician and staff scientist salaries? not sure here, but these positions should be expanded).

            The great majority of postdocs in the US (including almost all foreign postdocs) are supported by non-training funds, and are explicitly employed to perform the work on the project. I am not sure that broadening training activities or "alternative career preparation" outside the scope of the grant they are paid from would be technically permissible for such postdocs. Any training reforms that are limited to F recipients would not make much difference to the postdoc population as a whole or be fair to the majority of postdocs.

            On the other hand, tying training funds to research grants as a separate and mutually exclusive category of expenditure allows 1) the strictest control over trainee number, 2) redefining "training" in a way that encompasses necessary reforms.

  • "This may reflect NIH embracing the recommendation that more Ph.D. students being supported by fellowships rather than research grants."

    This is a way to shift a lot of the costs of training off of the NIH and onto institutions, as the indirect cost rate for NIH fellowships (and training grants) is capped at 9% (I think). So a graduate student supported on an F31 or T32 costs NIH a fuckeload less than one supported on an R01, with the balance in costs being eaten by the degree-granting institutions. I am certain that the NIH powers realize this and are embracing it.

    • Oh, and obviously, it's not just the costs of "training" per se, but the costs of actually getting the science done.

    • datahound says:

      I think you are giving NIH too much credit for plotting here. Just because the number of Institutes and Centers signing on to the announcement has increased, the number of awards need not increase. The number of training slots (termed FTTP for full-time training positions, both pre- and post-doctoral on individual and institutional awards) fell from 17,150 in FY2010 to 16,888 in FY2011, to 16,305 in FY2012, to 15,539 in FY2013. My prediction is that this number will continue to fall in FY2014. NIH may be preparing to increase the number if there is a push to use fellowships more (although this likely requires other changes). Alternatively, this could be driven by frustration from graduate student or training groups over the fact that some ICs did not accept F31 applications, limiting student options or for some other reason.

      You are correct that having students supported on training grants is cheaper in terms of indirect costs. Of course, other sources must cover the costs of the research outside of the stipend.

  • drugmonkey says:

    I'd be interested to know how ICs would feel about it if a trainee funded by their F award was working on the project funded by some other IC in a given lab.

    • datahound says:

      In my experience with non-NIGMS ICs, it is okay as long as it is in the area of science overlaps enough. Certainly, many students supported by NIGMS T32 grants do their research in labs supported by many other ICs ) but NIGMS has a broader mission and general training mission).

      • drugmonkey says:

        But you are advocating a massive change which will take this scenario from one-offs to "fairly frequent".

        • datahound says:

          Agreed. If NIH is really going to move toward supporting all or most graduate students on fellowships, there will have to be a substantial restructuring of the program in this and several other ways.

          • DrugMonkey says:

            How could things be structured to be fair across ICs? From the perspective of their interests in their funded RPG awards being productive.

  • NIGMS awarded an F32 to a fellow in my lab when I had no NIGMS RPG support.

    • datahound says:

      Yes, but NIGMS has a broad mission and the fellow was likely working on something that could fit.

  • Ola says:

    Regarding T32s, it will be difficult to assess this as "available" to any given lab when assessing the funding of that lab, since not all T32s at a given institution will be accessible to all labs.

    In my limited experience, the people in charge of the T32 get to decide who goes on it, and those decisions are based far more on personal and local politics than anything such as student merit or PI needs. If'n you're a faculty young'un, one of the first things you need to do at your new institution is cozy-up to the folks in charge of the T32s.

  • datahound says:

    DM: One possibility is that NIH could move to a system where graduate students compete for fellowships prior to or very early in their Ph.D. programs. Funds would have to be set aside and administered centrally through one or more ICs. Thus, the graduate fellow would have his/her own support independent of a thesis project. This is like the NSF Fellowship model although this could be even done to an even greater extent than it is at NSF. Indeed, some such as Nobel Laureate Roald Hoffmann has been advocating this change for some time in a different context (cuts to education at Universities (such as Cornell where he works) at the expense of research). See .

    I am not advocating this (there are still many issues to be resolved), but it is one model.

    • Joe says:

      This is the way many T32 slots are handled at my university, i.e., they are offered to students being recruited to the university and grad program as an enticement: " Come here and you will be a free agent able to work in virtually any lab, and every lab will want you." It works OK, but what it really selects for is students with superior GRE scores and grades plus some consideration for great letters and prior research experience. These students look great on paper, and some are superstars. However, some are really not. Plus, what is missed by this method is the students that catch fire in grad school. She might have gotten B's in Organic when she was an undergrad sophomore, but now that she is doing real experiments, she can't stay away from the bench and is churning out beautiful data. We need a system that can reward (and fund) both students that look good on paper and those who develop into great scientists during their grad school years.

  • DrugMonkey says:

    The more disease focused ICs will silo-up. Hard. Preference to apps from environments in which there is no question the trainee will work on the topics of interest to that IC.

    Hampers collaboration, cross-pollination and flexibility.

  • […] a follow-up to my previous post on the institutional distribution of NIH training funds, I have analyzed 1 year's worth (FY2013) of […]

  • jcolearyiii says:

    You don't have to come from a big lab in a big school to get an F31. I'm proof of that. You just have to have very specific things that they are looking for. Also, the applicant needs to be smart and persistent in their application.

Leave a Reply