This is the first post in a series looking at within-school variation. Other blogposts in the series can be found here.

As Dave Thomson highlighted recently, overall differences in the Progress 8 scores of schools in the ‘middle range’ are relatively small.

Overall school averages, like Progress 8, can, however, mask variation within the school. For example, here are charts showing the distribution of Progress 8 scores for two real schools. Nationally there were 20% of all pupils in each band, or quintile (data used is from 2015/16).

Both schools have the same overall Progress 8 score (-0.02) but they achieve this in very different ways.

In school B there are relatively few pupils in the highest and lowest bands – this means that most pupils are making broadly similar progress.

Meanwhile, in school A the distribution is wide – 50% of pupils have Progress 8 scores in the lowest and highest quintiles nationally.

Why do such differences occur?

There can be many reasons. In this series of blogposts we will focus upon two that can be assessed using data available to us: variations in the performance of pupils with different characteristics, such as ethnicity, gender and level of disadvantage, and variations in performance between different subjects.

Making similar progress

Firstly, let’s consider how much difference it might make if schools were able to eliminate these differences i.e. if most pupils made similar progress. One way to do this is to look at what might happen if the variation between subjects within schools could be reduced.

To do this it’s important that we look at value added scores, for reasons that will be explored later in this series.

What would happen if, in each school, the 50% of subjects with the lowest valued added scores were to improve over time so that they had the same value added scores as the average for the whole school?

Let’s imagine that the original intentions of Progress 8 had been fully implemented i.e. that each year would be compared to a ‘base’ year and the national average Progress 8 score would be allowed to rise, rather than staying at zero every year as at present. Let’s also imagine that the national reference test works and grades are no longer constrained by comparable outcomes.

Our estimates suggest that the impact of this would lead to an increase of +0.2 in the national Progress 8 score. This might not seem much. However, numbers in the Progress 8 world are quite small. In 2016, 1,563 schools had Progress 8 scores below zero. A national increase of +0.2 would mean that 656 of these schools would move from below zero to above zero Progress 8 scores.

By way of comparison, let’s consider what would happen if schools with Progress 8 scores below -0.5 improved to just above this level. If all other schools stayed the same then there would be an increase of +0.02 in the national score.

So reducing within-school variation in value added between subjects would have 10 times more impact on overall attainment when compared to moving schools below the floor standard to just above it.

So, within-school variation is important, and reducing it could have a significant impact on overall national standards of attainment. Parts 2 and 3 of this series will explore within-school variation between subjects and between pupils with different characteristics in more detail.

This is the first post in a series looking at within-school variation. Other blogposts in the series can be found here.

Want to stay up-to-date with the latest research from FFT Education Datalab? Sign up to Datalab’s mailing list to get notifications about new blogposts, or to receive the team’s half-termly newsletter.