Let’s start at… the end?

Why putting together your proposal back to front can help make a really compelling case for funding

I don’t eat my Christmas pudding before I tuck into my turkey, and I’m betting you don’t either. I never put my shoes on before pulling on my socks – how would that even work? And beginning a conversation with the word ‘goodbye’ would, quite literally, be a non-starter. But Margaret Mitchell, author of the best-selling novel Gone with the Wind, famously wrote the story backwards. In describing her approach, she said: "I had every detail clear in my mind before I sat down to the typewriter*. I believe […] that is the best way to write a book, then your characters can’t get away from you and misbehave, and do things you didn’t intend them to do in the beginning."

When it comes to writing research proposals, I think we can learn a great deal from Ms Mitchell’s approach.

Spoiler alert?

Writing a grant application is, on one level at least, an exercise in form filling. The application form is often long and rather arduous, and if you’re anything like me the temptation is generally to start at the beginning, filling in the easiest bits first. You might then try to tackle the rest of the form more or less in the same order that the sections appear, perhaps skipping some of the toughest and least-appetising sections to be re-visited at a time when you’re feeling more enthusiastic (ha ha ha). But your proposal isn’t a tax return, it’s a story – the story of why your research project matters so much that it must be funded. To succeed, it will need to be a highly compelling story. And arguably, the most important part of that story is the ending. So I’ll try to make the case that, like Margaret Mitchell and some other prominent authors, you really should start at the end and work backwards.

I should make clear at this point that this isn’t really about filling in the application form. You should only ever do that when you’ve done all the hard work of thinking through, designing and planning the research, and you have a complete and well-formed story to tell (like Margaret Mitchell, you should have every detail clear in your mind before you sit down at your typewriter*). So this is about devising that story. Once you’ve identified a specific problem that your project will tackle and given some thought as to how, in broad terms, you’re going to tackle it, here’s a slightly unusual suggestion as to what your next step should be. Draft the post-project press release.

This is not as mad as it may sound. Sure, you haven’t done the research yet, and so of course you don’t have any project results. But what you should have is a very clear idea of what success would look like. What are the implications for the problem you’ve decided to address if the project goes to plan and yields positive results? Without a clear idea of this, you’d be unable to answer the question "what’s the point of doing this project?" If you can’t answer that, then you can be certain that your reviewers won’t answer it for you.

Consider the following (re-assembled and lightly edited) extract from a 2020 press release about the results of a clinical trial undertaken by researchers at the University of Cambridge:

'Pill on a string' test to transform oesophageal cancer diagnosis

A 'pill on a string' test can identify ten times more people with Barrett’s oesophagus than the usual GP route. Barrett’s oesophagus is a condition that can lead to oesophageal cancer. Around 9,200 people are diagnosed with oesophageal cancer in the UK each year and around 7,900 sadly die. Early diagnosis is crucial to patients’ survival.

The researchers studied 13,222 participants. Over the course of a year, the odds of detecting Barrett’s were ten times higher in those offered the test with 140 cases diagnosed compared to 13 in usual care. In addition, the test diagnosed five cases of early cancer (stage 1 and 2), whereas only one case of early cancer was detected in the usual-care group.

Compared with endoscopies performed in hospital, the test causes minimal discomfort and is quick and simple.

Alongside better detection, the test means cancer patients can benefit from less severe treatment options if their cancer is caught at a much earlier stage.

The test could be a game-changer in how we diagnose and ensure more people survive oesophageal cancer. 

Pretty much everything in that press release except the results of the trial could have been written speculatively before the project began, based simply on the goals and aspirations of the researchers. And, given that it had taken them almost ten years of work to get to this stage, it’s highly likely that their preliminary results could have enabled an educated guess as to what the results might have looked like. Most importantly, the researchers would have known from the very outset exactly what they ultimately wanted to achieve. They had identified a problem (late diagnosis leading to severe treatments and high mortality rates in oesophageal cancer; and the additional problem of unpleasant endoscopic diagnosis), and set out to do something about it with a very clear idea of what they wanted to do and how they wanted to do it. They would have understood clearly what project success would look like.

But research isn't all clinical trials

Ah, but! This was a clinical trial, and while it’s great that they were able to describe such concrete and near-term patient benefit, what about earlier-stage research that’s years away from any real-world impact? What about basic discovery research that’s not designed to achieve any specific impact in the real world?

In neither case do I see any difficulty.

When it comes to early-stage research, it’s widely accepted that single projects seldom lead on their own to substantial impact. Impact is a journey, that typically spans a fairly lengthy programme of research. But imagine setting out on a journey without any clear idea of your destination (gap-year Interrailing aside). Imagine how that in-car conversation might go: "Are we nearly there yet Dad?" "I have no idea, I don’t know where we’re going." There would be tears...

I don’t think that a press release for one of the earlier-stage projects in the pill-on-a-string research programme need have looked much different to the one above. I’m sure the research team had a clear vision, direction and goals from the very beginning. So the press release might simply have used some additional phrases like "This brings us a step closer to…"

Similarly, when it comes to discovery research, it’s vital to be able to articulate a clear understanding of why the work needs to be done. Here, the problem you aim to address could be a gap in scientific knowledge; but you should still have a clear vision for the downstream importance of filling that gap. What could other scientists, including those working in more impact-focused areas, do with that new knowledge? You’re not expected to anticipate every possible outcome, but I think it’s vital to at least be able to envisage some reasonably-plausible outcomes and indicate that these motivate your research. (I accept that theoretical mathematicians and physicists may feel this pressure somewhat less keenly; but I maintain that those working in the life sciences should be acutely aware of it - at least if they want to secure funding for their research.) Our hypothetical press release might allude to eventual impact potential by once again using phrases like "These new insights bring us a step closer to…".

Leave out the hype

You will, I hope, note that nowhere do I advocate resorting to the kind of unjustified hyperbole that some over-eager press officers are guilty of. It’s not uncommon for academics to complain that their university’s press office has over-simplified and twisted their results to make outlandish and overblown claims regarding importance and impact. This sort of thing has no more place in a serious press release than it does in a research proposal where, as I’ve noted before, reviewers quickly tire of over-used overclaims such as 'urgent', 'critically important' and 'transformative'. Important and impressive things, like the numbers in the Cambridge press release above, tend to speak for themselves (okay, I’ll let them have 'game-changer').

All in order

What I’m proposing then is a sequence of inter-related steps in planning and drafting a research proposal which, if followed, should result in a coherent and convincing story throughout which impact is firmly embedded (remember, UKRI research councils and most other funders, particularly medical-research charities, are strongly focused on impact). By starting with a problem, you’ll effectively be starting with impact, and working back from that to construct a strongly impact-driven project (or at least, if it’s basic science, a credibly impact-informed project):

1. Identify the specific (and hopefully important) problem that your project will address or build towards addressing. Your would-be research project now has a point to it.
2. At a high level, hatch a plan for using your team’s knowledge, skills and resources to tackle that problem. You now have the makings of a research project.
3. Identify the specific way/s in which you intend that your project will address (or at least build towards addressing) the problem. This will give your project a specific aim (the main advance that you want your project to achieve) and, if different, an overarching impact aim (the ultimate real-world goal of your wider research programme). It would also inform a hypothetical written-in-advance press release.
4. Identify the handful of concrete, measurable steps that you’ll need to complete during the course of the project to give you the best possible chance of achieving your aim. These are your research objectives.
5. In conjunction with Steps 3 and 4, formulate any specific research questions and hypotheses linked to your aim and objectives. This puts more flesh on the bones of your project, and adds measurable specificity to your project goals.
6. Figure out what work needs to be done to complete each of the objectives, how it will be done, what resources will be needed, who will do the work, and when. Your project now has some work packages.
7. Think about what might not go exactly to plan, and how you’ll mitigate this – now there’s a risk table and a credible Plan B.
8. Describe all of this, and any other details the funder asks for, in the application form. You now have a (hopefully winning) research proposal!

There’s a bit more needed of course. What makes you think your proposed approach is novel, credible and feasible? That’s where pilot work, preliminary data and an appreciation of the relevant literature comes in. What will the next steps towards impact be after this project? Describing these shows you have a good understanding of the impact journey and how the current project is positioned on that journey. How can we be sure that yours is the best team to do the project? There’s always an opportunity to explain and evidence this. All of these are the key components of a good 'why fund me?' story, and they coincide with the main points that a reviewer will be looking out for.  

The above steps are often recursive and iterative as you gradually refine your proposal. You might for example need to revisit your research objectives, with knock-on implications for the work programme and possibly even the scope of the project aim. But by starting with a problem, and making sure the whole project is built around addressing that problem in a specific way, you’ll have ensured that impact is embedded throughout rather than clumsily bolted on – and that no one will be in any doubt as to the point of doing the research you’re proposing.

(* Google it, Gen-Z'ers!)

The views and opinions expressed in this blog are mine alone and are in no way endorsed by my employer. Factual information and guidance are provided on a 'best-endeavour' basis and may become out of date over time. Web-links were correct at time of writing but commonly go out of date. No responsibility can be taken for any action or inaction taken or not in respect of the content of this blog.