How we listened to our users to improve our education statistics
The Department for Education (DfE) is one of the largest producers of official statistics in government, covering a wide breadth of policy areas. These statistics are valuable to the public and allow us to deliver evidence-based policy, inform public debate and provide transparency around the education system.
Our current model for making these statistics available is very resource intensive, and does not meet the needs of our users. To improve this, we are building a new ‘Explore Education Statistics’ web service to replace our use of gov.uk and overhauling how we write our content and publish open data so we can better meet our user’s needs.
The vision for the project is to:
- build a reliable service which makes it easier for all users to find, access, navigate and understand education statistics
- empower producers to work more innovatively, provide more coherent and engaging information and save time and effort through automation
Our new service
We have been working with an organisation called HiveIT to build our new ‘Explore Education Statistics’ dissemination platform. Providing an interactive point-and-click user experience, as well as downloadable data and a clear upfront summary of the story our statistics are telling.
We will be moving away from a “static” pdf summary and Excel tables – which is our current way of publishing official statistics.
The two main problems we have tried to solve are:
- making it easier for users to access the breakdowns of statistics they are interested in – alongside clear, punchy summaries of what the headline statistics are showing
- streamline our production processes: the way teams produce and publish their statistics is not consistent and could be made more efficient
We have focused on understanding what users need and how we can meet that need – as well as ensuring that what we do is in line with the Code of Practice for statistics.
Since beginning this project we have completed the discovery and alpha phases – which allowed us to identify the main problems, iteratively test new designs with users and explore different options for technical functionality – and we are now in our beta phase.
Why do this now?
Our current approach to producing and sharing statistical outputs is unsustainable – the amount of published data our users want to see is growing and as a result
- we’re either having to extend the amount of burdensome excel tables included in our releases
- or release bespoke requests via Freedom of Information and adhoc submissions
Both take up valuable time and analytical resource meaning we cannot service current demands. We also recognise the limitations of gov.uk as a platform for enabling us to tell a better story with our data that will engage more users, make the experience more inclusive, make it easier for people to find what they are looking for and to understand the data once they do.
The introduction of the new dissemination platform will make our statistics easier to find, access, navigate and understand. Our discovery provided evidence that we need to improve on this to better meet the wide range of our user’s needs and recommended we provide more interactive content and allow easier self-service.
The changes we are making to our data and drafting standards will support our move to the new platform but are driven primarily by the need to improve consistency across our publications and to act in line with GSS best practice guidance and Code of Practice publication principles.
We also hope that by expanding and improving the quality of our open data we will stay aligned with the aims of the GSS data project making it easier for education data to be included in the future.
What we’re most proud of…
Throughout our project we’ve focused specifically on how we can build a service that is designed for inclusivity with users who have access needs such as blindness, dyslexia, colour blindness and other physical limitations, as well as those with low levels of data literacy. We test with these users regularly and are confident we are creating a service which allows them to access and understand our data where they were previously struggling to do so.
We’ve focused on how our data feeds the platform first – this platform couples very closely with the work we’ve been doing using Reproducible Analytical Pipelines, ensuring we can produce consistent and extensive machine readable data, with appropriate quality assurance and validation, in an efficient and reproducible manner. This means our statisticians are free to spend more time focusing on the quality of our statistics and providing insight. We’ve built a full administrative application alongside the platform to make the uploading of data and production of releases as simple as possible for our statisticians.
We’re also proud of how quickly we’ve been able to turn this around. We are on track to have built prototypes through to a full end-to-end service and admin portal within just 10 months!
So what next?
Later this summer we will be opening the service to a select group of users for private beta testing and then, assessment permitting, will be moving to public beta in autumn 2019.
From January 2020 we expect to be using the new service as our primary route for all of our National Statistics releases going forward.
Please help us make sure our service meets users’ needs…
Having users be involved in our user testing is vital to the success of our project and in ensuring that the service we build is fit for purpose.
If you want to take part in any of our testing activities please complete our beta recruitment screener survey.