Earlier this year, we ran our second “Using Data in the Early Years” event, bringing together leading-edge data innovators to showcase projects with the potential to transform how we support disadvantaged children during the formative stages of their lives.
At Nesta, we are interested in developing innovations that can put that data into the hands of those making the crucial decisions - whether frontline workers, service level managers or strategic commissioners - in a usable, accessible and practical form that can help inform their assessments and lead to better outcomes for children.
The event built on our previous session, and featured speakers working at the forefront of data linkage, leading the way in the joining up of early years data and using it to create tools that can lead to better interventions. We were joined by attendees from local authorities, charities, academics and businesses involved in the early years space.
Using big data to find the children most at risk of falling behind
Two pieces of work by ADR Wales and Swansea University, and ECHILD at UCL have linked millions of anonymised data records from across health, social care and education in Wales and England respectively, building a unique picture of children’s lives.
Using machine learning and a technique known as multivariate regression (used to understand the relationship between multiple, variable factors on an outcome), researchers were able to gain unique insights into how a child’s development could be affected by the many influencing elements.
ADR Wales/Swansea University’s work was able to provide detail on which groups of children were at risk of not being school ready and other poor outcomes, as well as identify which factors had the most influence. For example, being eligible for free school meals or being a boy doubled the likelihood of not being school ready.
UCL’s ECHILD Database brought together the anonymised health, education and social care information of all children in England. They have been looking to understand how a child’s education affects their health and vice versa, for example researching how the pandemic and lockdown affected children who need extra support.
Building tools to enable data-sharing & decision-making
On the the frontline, Policy in Practice, a social policy software and analytics company, demonstrated the value of data-sharing for child welfare through their Multi Agency Safeguarding Tracker (MAST) tool. By linking datasets across the emergency services, social care, and the NHS, safeguarding decision-making was able to be more informed, presenting a more complete picture of both risk and resilience factors for each case.
At a strategic level, rich, population-level intelligence was also key to the work of both Birmingham City Council and open data project Local Insight. By cleaning and aggregating publicly available data, they both captured the wider structural impacts on children’s development and experiences and turned them into accessible dashboards, putting the data directly into the hands of local decision-makers.
Hearing children’s voices in the data
The Promise Scotland and UCLPartners reminded us of the value of lived experience and the child’s voice, and how in all data-driven work we mustn’t forget the people behind the statistics, bringing their experience into any decisions that are made.
The Promise Scotland is a non-statutory company established by the Scottish government to help deliver its pledge that care-experienced children and young people will grow up loved, safe, and respected.
Creating a composite character, based on the real lives of 12 children, they envisaged two scenarios that brought to life the experiences of children in care. They both tell the story of 10 year old Isla and what life was like for her before the pledge and what it could be like if it is kept.
In contrast, UCLPartners set up the Care City Community Board which helps them to shape the work conducted with their dataset. This can include activities such as informing the direction of the analysis, which means that the people represented in the data could help provide further information about why potential trends or patterns were emerging. This ensured that incorrect assumptions weren’t being attributed to results, that the findings were relevant for the community and crucially, that public perceptions around use of the dataset were understood.
These are encouraging examples of doing data differently, keeping in focus the individual lives that make up our vast datasets.
Personalising data in the future
Finally, the Tony Blair Institute for Global Change looked even further into the future, envisioning a world where a combination of unique identifiers for all children and an improved national infrastructure enables easier and more extensive linking of data across the early years and school system. Ideas such as this could represent the foundation stones on which a more sophisticated set of data tools could be built, meaning that one day, the projects of our other speakers could be the norm rather than the exception.
It was clear from our event that this innovative and brilliant work currently only exists in pockets. While attendees recognised the value of the work and were inspired by the possibilities, there were few straightforward options for translating what they had seen into their own place of work or combining it in a way that would give decision-makers better data capabilities.
There are no easy answers about how we can resolve the dilemma of uneven innovation, but Nesta’s immediate focus is on supporting the development and scaling of data innovations through our partnerships with local authorities and other organisations engaged in early years. We’ll also continue to provide a convening role for the early years data community, holding another “Using Data in the Early Years” event next spring.
In the meantime, you can keep up to date on our work on data and detection in the a fairer start mission as we publish updates from our work with Birmingham and Fairer Start Local, as well as an exploration into the Ages and Stages Questionnaire data.
If you have thoughts about this blog, our work more generally, or would like to be involved in future events, please let us know and drop us an email [email protected].