How we redesigned our Mental Health Data Prize for scientists in Africa

Last year, we launched our first Mental Health Data Prize, with the goal of funding at the intersection of data science and mental health. We wanted to include datasets from at least one low- and middle-income country and one high-income country. Our initial scoping suggested promising data sources in the UK and South Africa, so we decided to focus on these two countries.

When we designed the prize, along with our partners Social Finance, we engaged with mental health researchers, data holders and young people with lived experience to shape the prize for the UK and South Africa.

This engagement generated excitement and interest before the launch, suggesting we had designed something that works for the community.

But we didn't receive as many applications from South Africa as we'd hoped. In the first phase of the prize, we funded ten teams from the UK and only one from South Africa.  

We had a hunch that this was due to the way we had structured and designed the prize. So, we went back to the community who expressed interest, and others we hadn’t spoken to before, to understand why. 

Why hadn’t people applied? What can we do differently in South Africa?  

To find out, we ran a follow-up evaluation involving a qualitative survey, a workshop with African data scientists and expert consultations with relevant stakeholders.  

We also ran a workshop at the Data Science for Health Discovery and Innovation in Africa (DS-I Africa) Initiative’s annual meeting, to hear from that community what funders need to consider when supporting health data science in Africa. 

The consultation was revealing – we heard lots of constructive feedback about how to tailor the prize more appropriately to research communities in South Africa.  

Some of the areas that came up repeatedly were:  

• Our eligibility criteria should be more flexible, especially when it comes to (co)leadership (for example, mandating the lead to be from the setting in which the data has been collected – which we had done – may have been overly limiting; we may need to ensure the benefits from the work goes to local communities via other means). 
• Longer timelines for application preparation and team formation would be useful. 
• The work required to apply was viewed as too burdensome in resource-limited settings, given the relatively small amount of funding on offer.  
• Forming a team with the necessary skills to both analyse data and build a digital tool was challenging within the timeframe.  
• A lack of clarity on future funding opportunities after the prize was off-putting. 

We also heard that rather than focusing the prize on South Africa, we should consider opening the prize to teams across Africa. We heard from researchers from Kenya, the Gambia and Uganda who would have applied if the scope allowed.

Based on what we’ve heard, we will be running a new Mental Health Data Prize open to teams across the African continent, which we hope will be better tailored to the needs and priorities of data scientists and mental health researchers in Africa. 

Right now, were searching for a partner, preferably based in Africa, that will set up the prize [PDF 412KB]. 

The delivery partner will manage the engagement, training, and team formation stages, involving data scientists and mental health researchers from across Africa. Areas for training and support will be identified by crowd-sourcing ideas from the interested community to best tailor it to their needs.  

After the initial engagement and set up phase, teams will be invited to submit proposals that use existing data from across the continent. Their goal? To improve understanding of the trajectory and resolution of anxiety, depression, and psychosis in different African contexts.