Image: Phil Roeder [CC BY 2.0], via Flickr
Well-targeted projects may appeal to non-academic funders
The Joseph Rowntree Charitable Trust, a Quaker organisation that funds work tackling the causes of conflict and injustice in society, awards funding through five main programmes covering the following issues: power and accountability, peace and security, rights and justice, and sustainable futures. There is also a cross-cutting programme. This year’s registration and application deadlines are in August.
Although the JRCT does not specifically fund academic research, it will do so if the research forms an integral part of policy and campaigning work that aligns with its main interests.
Ahmed Izzidien, a post-doctoral research associate at the University of Cambridge’s Judge Business School, who specialises in artificial intelligence research, describes his win with the trust through its rights and justice programme.
What societal problem is your research trying to address?
The title of the project is Using Artificial Intelligence for Policy Outcome Prediction. Say you have a line of legislation that you are about to pass, or a policy document—it could be an education policy, it could be a higher education policy, it could be something to do with justice—and you want to know whether in five years’ time this policy or legislation will produce the desired outcome. The goal of the algorithm is to read those documents and other contextual documents, and then to conclude, based on a probability, that the outcome will be X or Y.
I won’t say it’s a crystal ball, but it’s trying to look at probability and things that have happened in the past, and then draw analogies to predict what is likely to happen in the future.
How did you sell that idea to the JRCT?
I used the Richard Feynman technique. Feynman thought you should try to explain your idea as you would to a six- or seven-year-old child. This is not so the idea is clear to the reader but so it is clear to yourself.
He found that most academics were not able to do this, because they didn’t know the basic premise of their own research. And because of that, when they wrote grant applications, these holes were present in the substance of the application.
Only once you’re satisfied that you can explain your proposal to a six- or seven-year-old child are you ready to write the application. I have found that process incredibly helpful for my research.
How did the technique relate to your own proposal?
People are more tech-savvy than they realise, and it is just a matter of using an analogy to bring that to the fore. When we type in email or on our phones, we get next word prediction. Often it does fit, and that is machine learning in our hands. If you scale that up to paragraphs, documents and then policy papers, the hope is that we can say, OK, given precedent and historic outcome, this is what is probably going to happen.
There are specialists in policy who know more about government and so on, but there are technical methods to make it more efficient and save time, energy and costs. That is the aim of this project.
Was the application itself much different from a conventional research application?
There is an online portal that all applications go through. The questions asked are similar to other portals I have used. My impression was that the JRCT wanted to know more about the whole application than just the technical aspects of my project.
How did you tailor your application to the trust—a non-traditional research funder?
I had some kind feedback from one of the gentlemen at the trust. We had an hour-long interview because my proposal was not exactly in their remit, so he wanted to have a chat. He said that, generally, what they are looking for is not the funding of pure science but an application that is relevant to them.
Why did they decide your bid was one of those?
They are keen to fund research which tackles inequalities. With my application, I was trying to say that with this technology we can nip problems in the bud by foreseeing them and trying to make the changes early. If, for example, we have a particular policy that is going to lead to certain injustices inadvertently in society, if we can foresee those, we can avoid them.
Was the JRCT the only funder for your project?
I had a few. My main funding is from the NGI Trust, which receives funding from Horizon 2020. Their grants go through technical peer review. They funded two-thirds of the costs. The JRCT agreed to contribute the remaining third of funding, at £14,000.
What advice would you give to researchers considering a bid to a non-academic funder?
You should try to align your vision and the funder’s. That became one of the drivers behind my approach to the JRCT, as well as to a few other funders who have similar interests.
This is an extract from an article in Research Professional’s Funding Insight service. To subscribe contact sales@researchresearch.com