What makes a good research proposal?
tratt.net
tratt.net
The #1 most important thing in research proposals is making good images. No one cares about the text as long as it fits the proper mold. But the images that concisely convey the novelty of the work are worth $$$.
(Source: advisor and I had over 80% success rate in field that averaged <20% over 16 proposals).
Agreed. Two of these "most fundamental" questions you don't need to concern yourself with as a proposal writer: "What is the problem being tackled?" and "Why is the problem worth being tackled?" are the ones that have been decided for you as part of the call for proposals. At least that's how it works when asking for public money for research funding in most European research programmes.
It's more like:
Researcher: I want to spend the next 20 years continuing the work I have been doing for the past 20 years, please.
Politician: Actually, I want to decide what it is you work on, and I decide based on what I can sell to the public. The public decides what's important to them based on what the mass media write about in a favourable light. The mass media have the attention span of a headless chicken on speed. So expect research priorities to be frequently changing and calls for proposals to not make any sense.
So the system that has evolved around that is to make calls for proposals like "Improve citizen participation in paneuropean policy making through next-generation digital tooling."
"Citizen participation" is the part of the bullshit bingo that the politician needs to argue that the research billions are well spent. "Next-generation digital tooling" is the part that is intentionally super-vague and is the part of the bullshit bingo that the researcher needs, so that he can attach whatever he has been working on for the past 20 years to this newly-minted agenda.
So the job of the proposal writer is to weave the two narratives together: If you've been working on compression algorithms, tell a nice story about how citizens use mobile phones to call their representatives in Brussels, about how mobile phones use compression algorithms, and about how improved compression will therefore lead to improved citizen participation in paneuropean policy making.
In the next cycle, the call for proposals might be "Sustainability 2030: Reducing the carbon footprint of the European digital infrastructure backbone". Go ahead and write about how every bit of a data packet requires electricity to process and about how improved compression will reduce the electricity consumption of digital infrastructures.
Then: Ask yourself who will likely end up reviewing the proposal, and work in as many flattering citations of them as possible.
You really don't need to demonstrate that any of it makes any scientific or economic sense. You just need to pick up on the language game that people in power are playing and demonstrate that you'll participate in it as a willing pawn.
P.S.: When calls have low acceptance rates, the deciding factor is cronyism, not superior substance of a proposal or even superior proposal writing ability. Having the politically-favoured (i.e. quantitatively underrepresented) set of genetalia also helps.
It is sometimes asked to include implications for externalities such as public engagement, climate, equality, but if the call is about compiler design for example, you shouldn't make these central to your research.
> When calls have low acceptance rates, the deciding factor is cronyism
I don't see how it's possible that cronyism feeds into large public European calls that you mention
I haven't directly witnessed it at the Brussels level, but certainly at national level in one European country, I have been a direct witness to how a public call for participation was put out for which it was well known in advance who would get the funding. -- If you want to make the case that Brussels is less corrupt than national governments, be my guest.
All you need is a bit of "procurement process theater". Put N focus areas into your call, corresponding to the N institutions you want to give the money to, and describe them in just the right amount of specificity. Add to that requirements like "to be eligible, an institution must have been in business for X years and have at least Y number of employees" to increase the status-quo bias.
In one instance I happened to be around such an institution in a year where the administration screwed up and didn't manage to put the call out in time. So an institution with something like 50 employees that was totally dependent on the funding would have had to close its doors for a year and reopen a year later after the next funding decision was through. Basically what happened was that the director of the institution went to his crony in politics and said "Hey you screwed up. We need that money, and we need it now." And the politician just kind of went "okay, here, have some money". Apparently he found some taxpayer money in some couch cushion or something to make a bridge loan to the institution against what was basically the certainty that they were going to be awarded the proper amount of money through the proper channels in the upcoming funding decision. I really think that this exposed the whole science funding system for the emperor with no clothes that it truly is.
I wouldn't call the scenario you mention cronyism per se, as it may tie into wider national policy considerations that are legitimate, e.g., national strategies on employment, research, technology etc. It'd be legitimate to fall on the national exchequer for funds if EU funding applications aren't successful.
I'd prefer if groups didn't rely on the likes of EU Horizon. As you said earlier, the success rates are so low for some calls, and the effort for application is so high. It's a poor risk/reward in many cases and leads to a lot of wasted effort.
"Multifunctional specific chemical probes for AMPA receptors, which are implicated in the pathology of:"
- "Alzheimer's Disease" (NIH)
- "Schizophrenia" (NIMH)
- "Addiction" (NIDA)
Your proposal will be judged non-compliant and will never even get to the panel.
Why write obviously false comments like this?
[1] A discussion of risk and mitigations, last paragraph under "Secondary questions".
This is an excellent way to squarely confront possible objections by reviewers. It shows you understand what problems might come up, and shows the mitigation(s). It can be presented as a "risk list" with possible mitigations for each (a two-column table). Risk mitigations can include alternative technical approaches, or schedule adjustments, or descopes if the sponsor allows it. More formally, columns for a qualitative likelihood of the risk being realized (1-5), and a severity if it is (1-5).
Think broadly about risks. "Postdoc cannot be found" is a risk. "Experimental data contaminated by systematic effects" is also a risk.
[2] Advocating for a top-level statement, within the first couple of paragraphs, of the problem to be solved, why it matters, and how it will be done. If you can't do this in two or perhaps 3 sentences, you are confused and should think more carefully about what you're proposing. I'd go even farther and say one should aspire to have the first 3 words of the proposal be "We propose to ...", with the whole thing laid out in one sentence.
Do not have the first couple of paragraphs be problem background. It's not a science paper.
[3] Adapt narrative to the proposal format (OP under "Adapt to the funder"). Those categories in the template are there for a reason -- to simplify comparing proposals -- but you have to take some license to be sure you're not repeating, and that the document still is readable from top-to-bottom.
*
My own bad or poorly-conceived proposals have signaled trouble by repeating themselves -- same or similar paragraphs or long sentences in the abstract, then in the introduction, then in the motivation or conclusion. This either indicates a bad cut-and-paste job, or that thought has been suspended.
*
The summary mentions generating excitement. One way to do that is to think about storytelling techniques.
A classic storytelling technique is the "undiscovered country" plot, e.g. a 2D plot of (say) wavelength vs. telescope observing time that illustrates why having (say) a UV telescope with very long integration times will enable a new class of objects to be seen.
Another "win story" can be developed by answering "why now" -- what combination of data and technical know-how makes this the right time to do this work? Basically, inevitable that someone must do this right away.