Millennium Villages Project Data Explorer
A ten-year project operating in ten sites in sub-Saharan Africa, the Millennium Villages Project (MVP) was an attempted to address extreme poverty in Africa. Through scientists and development experts working with local governments and communities to deliver an integrated package of interventions, the project aimed to “lift” developing country villages out of the poverty trap. With controversy surrounding the MVP, ITAD, a global organisation whose strategy, monitoring, evaluation and learning services work to make international development more effective, were funded to undertake a five-year evaluation of the DFID-funded Millennium Village Project in Northern Ghana. Employing a robust, mixed-method approach, at the end of the five years, ITAD had the results of their evaluation and a large and varied dataset.
With an extensive dataset, ITAD saw the potential for their data to be useful beyond the scope of the evaluation. Their data was already accessible to others, following an application and approval process, but with low uptake. To further increase accessibility, a second redacted dataset was created in collaboration with the UK Data Service which could be downloaded without approval, but barriers to accessing and engaging with the data still existed. What Itad wanted was a simple and engaging way to explore the datasets. To make this happen they needed facilitation, support and strategy to involve a local partner and deliver a tool that supported exploration of the data produced from their evaluation.
How we helped
We began by working with Itad to gain an understanding of their purpose and the context. They could see there was potential for their data to be of wider use but they what they needed was the knowledge and practical input to allow them to release this potential.
Having gained an understanding of the client’s needs, we worked with the UK Data Service to provide access to the data and to overcome any barriers that developers might face in accessing the data. We then established an understanding of the data available in terms of its format, potential, and limitations. This was necessary to ensure that any tool developed would be an appropriate fit for the dataset available.
Itad were interested in the potential value of their data in informing Ghanaian researchers and policy actors in-country. With experience of local developer support and hackathons in developing countries we were able to network and built relationships with developers in Ghana, setting up a competition to [need a sentence on the competition]. We then worked together with the competition winners, Mobile Web Ghana, to develop and visualise an innovative data explorer to help investigate and contextualise Itad’s research on the Millennium Villages Project (MVP) in Northern Ghana.
To ensure the tool met the client’s needs, we worked with the in-country partner to scope and specify essential elements for the data explorer co-creating ideas about what could be delivered, what would be most useful and what would provide value for money. Throughout this iterative and exploratory process we worked as a bridge between the client and the in-country partner, translating the needs of the client to the partner so they could deliver a tool that met the client’s needs.
Once the tool was developed, we set up user testing and bug tracking to ensure the tool was functionable across different browsers and devices.
Results / benefits
The data explorer tool allows users to interrogate the research data, comparing indicators, and control and Millennium Village Project villages as well as being able to see trends over time.
The tool helps others explore the data allowing freedom to engage with the data, innovate, and to develop their own interpretations
The tool provides visual analytics of the data available. It allows users to quickly check what data is available, visually illustrate trends and to test assumptions in an engaging way depending on their needs and interests.
While there is a lot of talk about sharing data and open access, when it comes to evaluation in particular, this rarely happens in a meaningful way. This project illustrates how opening up data might be achieved and provides an example of good practice for other evaluators and researchers who might want to expose their datasets but don’t know how to make this possible.