The problem with big data is that it’s not always that easy to wade your way through it. The UK Department for Education, for example, provides a mammoth amount of information about schools – average class sizes, absence levels, details on how many pupils are eligible for free school meals (and how many are actually having them), and a wealth of measures that aim to assess how well children are performing across the school years. But trying to make sense of a sea of numbers in spreadsheets and tables doesn’t always help you to get a sense of what’s actually going on in schools around the country.
This is where Timo Hannay comes in. “About 18 months ago some local schools where I live in East Finchley asked for help in data analysis. Being the slightly obsessive geek I am, I ended up downloading reams of data from the DfE website and building a huge database covering every school in the country”, he says. Hannay has an eye for data. He has a PhD in neuroplasticity, and for a number of years was the director of web publishing for Nature Publishing Group. He then went on to found Digital Science, a startup company focusing on software and services aimed at scientists and research administrators. Now, he can add educational cartographer to the list.
“A few months ago it became apparent to me that other people might be interested some of this work too, and that it would be a good idea to put it online”, he says. “At the same time, though for unrelated reasons, I decided to leave Digital Science. It had grown from one person (me) to over 250, the revenue had gone from zero to a healthy 8-figure level, and it was continuing to grow very rapidly. But I much prefer getting things off the ground than scaling them up, and I prefer startups to established organisations. So as I was thinking about what to do next, schools data seemed like an obvious route.”
And so SchoolDash was born. Taking the mass of data produced by the DfE and the Office of National Statistics, the aim is to present interesting educational data in an easy-to-understand, visually sensible way. For a given measure, it’s possible to see at a regional, local authority, or parliamentary constituency level what is happening in schools in particular areas. It’s worth pointing out that it’s very much in the early stages of development at the moment – “version 0.1”, according to Hannay. But nevertheless, the initial maps highlight some fascinating things. In the inaugural blogpost on the site, one point that Hannay has highlighted in particular concerns the astounding effectiveness that London schools have. On the one hand, they seem to have what could be considered to be a number of disadvantages. For example, London schools tend to be very large:
And, using long term eligibility for free school meals as a proxy measure, children going to these schools show high average levels of social deprivation:
Despite these sorts of factors though, students in London schools seem to be doing academically very well. For example, the proportion of students exceeding national expectations in reading, writing and maths at Key Stage 2 (i.e. from 7-11 years old) is well above the national average:
Why might this be the case? Elsewhere in the data, we can see that London schools get relatively generous grant funding in terms of pounds per pupil, and average teacher salaries are relatively high (although by no means the highest in the country – Hampshire wins in that respect). Obviously, we can’t imply a causal link – these are basic correlations that we can see in the data, and there are a myriad number of likely factors that might impact upon how well pupils are performing academically. However, what the maps do allow for is a quick and comprehensible overview of what’s going on across the country. As Hannay notes in his initial blogpost, “they at least enable us to ask better questions, and they inform the debate about how we should educate our children.”
One area of data highlighted by SchoolDash that’s particularly fascinating concerns Key Stage 4 science performance across the country. The data set that Hannay’s maps use contain data from 2014 about the English Baccalaureate (EBacc). This isn’t to be confused with the French Baccalaureate – it’s not a qualification in and of itself, rather a school performance metric that’s linked to GCSEs. In order to meet the criteria for the EBacc, pupils must have obtained grades A* to C in two sciences and maths, as well as in English, history or geography, and a language. Looking at the maps for the science subject area of the EBacc, in 2014, just under 60% of pupils across the country were entered, and more students were entered in the South than in the North (although the South East was a bit of an outlier in that respect):
And of those pupils who are entered into the EBacc science subject area, about three-quarters of them are achieving A*-C grades.
The DfE data also contains a ‘value added’ measure, which is a figure that attempts to show how well pupils are doing relative to expectations inferred from their previous attainment – in other words, a sort of progress, or improvement, indicator. Breaking the data down into subcomponents looking at pupils who have shown either low or high levels of prior attainment, then it seems as though different areas of the country are having varying levels of success in enabling pupils with previously low levels of attainment in science to achieve better marks. For example, pupils in London and the West Midlands who previously had lower levels of attainment show the most progress in science, with the North East apparently struggling the most (although it’s worth noting that the North East may be partly skewed by a lack of data at the local authority level). For pupils who were previously showing high levels of attainment, there seems to be a much clearer North-South divide, with schools in the South showing more success in driving their high-achieving students towards better science marks:
It’s difficult to say why these differences might be happening, but nevertheless the SchoolDash maps provide a good way to get to grips with a hugely rich and informative dataset. While he eventually hopes to earn a living through SchoolDash, Hannay’s priority over the next few months is to develop the site into something useful for others to be able to explore the data coming out of the DfE. “Everything that I’m building during this phase will remain free for as long as SchoolDash exists”, he emphasises. The first step, though, is to start adding from a long list of extra features, most notable of which will be school dashboards, which will provide a broad range of interactive data about individual schools across the country.
Hannay’s project is ambitious, and could prove to be immensely useful to a wide range of school stakeholders as well as anyone with a general interest in the performance of schools around the country – from estate agents to journalists and policy-makers. And while he’s optimistic about SchoolDash’s potential utility, Hannay is taking a pragmatic approach to its development. “I go into this with a high degree of humility because – though I know about data analysis and visualisation, and I know about building online businesses, and I have school-age kids – I’m not an expert in education and I don’t want to assume that all my ideas are necessarily new or even valid”, he says. “In short, I still have a lot to learn, so I’m taking things one step at a time.”
This article titled “SchoolDash: the school performance data tool for teachers, parents – and even estate agents” was written by Pete Etchells, for theguardian.com on Monday 14th September 2015 10.15 UTC