Just how good are academy schools? A new database makes it easier to tell
By Blog Editor, IOE Digital, on 13 March 2018
There has been huge interest in the performance of schools that have changed from mainstream to academy status in recent years. Since 2010, successive governments have backed the opening of more academies, arguing that they drive up standards by allowing headteachers more freedom to innovate. Critics claim that there is scant evidence to show that becoming an academy is a guarantee of a better school.
So, can education researchers help to resolve this debate?
Until now it’s been quite difficult to track the performance of schools that change to academies. Previously, academic researchers and policymakers could make use of the School Level Database (SLD), an administrative dataset collected annually by the Department for Education.
This fantastic resource has detailed information on every state school in England, with each school given a unique identifier. However, the SLD has a significant amount of missing data and virtually no information on schools before 2006. If certain characteristics of a school change from one year to the next – such as becoming an academy, or merging with a different school – the SLD assigns the changed school a new identifier, different to its previous one. On average, this occurs for around 5-10 per cent of schools in any given year, and is a big problem for any researcher interested in observing a school’s progress or changes in its characteristics over time.
Indeed, researchers have shown that not accounting for the changing identifiers of schools has real impacts on research findings. A different solution was therefore needed.
Funded by CLOSER, we have created a new database to enable researchers to track school performance more accurately and efficiently than ever before, helping to paint a clearer picture of education in England. Our new ‘consistent schools database’ (CSD) minimises the number of schools which can’t be linked and therefore tracked across years due to changing identifiers.
To create the new resource we cleaned publicly available annual school data to make the variables consistent over time. We then matched schools across years using a range of link variables, using information on each school’s numeric identifier, name, type, postcode and local authority.
Currently spanning the years 1999 to 2014, this new database provides information on school mergers, splits, closures, openings, and changes in school institutional type. Schools are also linked both forwards and backwards, allowing links to be traced in either direction, whatever year the researcher starts from.
Such improvements to the scope and quality of research that will come about from this database have the potential to benefit education policy and practice, and ultimately the general public.
And, hopefully, it’ll also help us get to the bottom of that academy-status question.
Guidance on the consistent schools database: Researchers interested in using the CSD can request the dataset by emailing email@example.com, explaining who you are, where you work and what research you intend to carry out using the data.