What Scotland could learn from English school performance tables

8 January 2016

Earlier this week, Scotland’s First Minister presented her plan to turn around the Scottish education system. At the International Congress on School Effectiveness and Improvement, Nicola Sturgeon set out four key priorities: raising attainment, closing the gap, improving health and wellbeing, and improving employability. The rhetoric of her speech was admirably outward-looking: she opened by committing to “learn from evidence and analysis from around the world”. In a tweet later, she said: “I’m not close minded to anything that works to help poorest kids get best education”.

New education policies are certainly needed in Scotland. The Scottish Survey of Literacy and Numeracy 2014 shows that the overall proportion of pupils performing well or very well in reading has dropped since 2012. The performance of pupils in the most deprived communities on this measure has dropped by 11 percentage points to 41 per cent in the third year of secondary school (S3).

The educational challenge is clearly great. Yet the policies Sturgeon has chosen are likely to have a strong positive impact. Alongside proposals to improve access to teacher professional development, the Scottish Government plans to reintroduce national standardised assessment in primary and secondary schools (P1, P4, P7 and S3), and to publish this information by school. The availability of this data amounts to the reintroduction of school performance tables, which Scotland abolished in 2003 after devolution. Currently, achievement at these levels is only assessed by teachers – a method that researchers have consistently demonstrated is unreliable and biased.

The potential for school performance tables to raise pupil attainment is enormous, particularly for disadvantaged pupils. Seminal research by the Centre for Market and Public Organisation suggests the average pupil in Wales achieved 1.92 GCSE grades lower each year (2004-08) than their English counterparts after the Welsh Government abolished the publication of school performance tables in 2002. The effect was most pronounced for pupils at schools in the poorest areas, where the average effect was a reduction of 3 GCSE grades per pupil per year. In the US, similar evidence suggests the introduction of state-wide high-stakes accountability has brought about significant improvements in pupils’ maths results.

Yet despite the evidence, both Scotland and Wales have kept their eyes closed to this policy initiative – until now. So what can Scotland learn from the English system, where performance tables are an established feature of our education system?

What is measured matters

The first lesson is that the qualifications included in performance tables non-trivially affect what is taught in schools. Performance tables have taken many forms since their introduction in 1992 (for a history up to 2013-14, see here). In 2010, the Coalition Government introduced a new Ebacc performance measure, based on the percentage of pupils achieving A*-C grades in English, maths, history or geography, a science and a language GCSE. Evidence suggests that this has increased participation in some of these qualifications.

This has led sceptics to argue that performance tables encourage ‘teaching to the test’, meaning that only what is measured is taught. If true, this would make the selection of qualifications counting in the performance tables important further still. However, evidence suggests schools take a more considered approach. For example, data shows that schools have continued to teach some subjects that no longer lead to a recognised qualification in the performance tables. The most likely reason is that schools deem these subjects as important for some pupils.

How it is measured matters

The design of performance data is also likely to impact school behaviour. The criticism of the existing headline measure at secondary school (5+ GCSE A*-C grades) is that it encourages schools to focus on pupils at the C/D borderline at the expense of others. The introduction of a new headline measure of pupil progress from 2016-17 (Progress 8) is intended to remove this perverse incentive. The measure will compare pupils’ attainment in eight subjects with the national average for pupils with the same prior attainment (measured at the end of primary school).

The introduction of this new measure will mark a subtle change in the purpose of performance tables. The Government previously measured progress by using a contextual value-added score, which took account of a range of pupil characteristics known to affect attainment (including ethnicity and month of birth). This theoretically enabled parents to choose a school based on more reasonable expectations of what their pupil (given their background) would achieve at the school. As the research paper recommending Progress 8 found, students from different ethnic backgrounds have different rates of progress (p.35). The discontinuation of this measure and introduction of Progress 8 (which takes account of prior attainment solely) means that factors affecting attainment and largely beyond the schools’ control will be omitted.

Still better than alternatives

Despite the challenges of designing a fair and effective accountability system, the Scottish Government should hold steadfast on performance tables. The alternatives are to continue with teacher assessments (which are clearly not working) or to completely abolish national assessments altogether – which would leave employers and further education institutions with the burden of assessment.

Instead, performance tables have the potential to raise overall attainment and help narrow the gap. The amount of thought, research and development that successive governments have put into designing them is a testament to their strength as a mechanism for changing behaviour. There is a lot that Scotland can and will learn from this policy process.



No comments yet.