Progress 8
from Twitter
Thread from Andy Byers, Headteacher Framwellgate School Durham:
P8 is a bit dodgy, isn’t it? The regional figures illuminate a systemic problem. Is anyone suggesting that schools in the NE are significantly worse than schools in London?... As a profession, we need to stop playing the P8 game. We need to make these successes scalable; we can control how we teach, but the profession must also speak out about the many other disparities which lead to differences in P8; long-term disadvantage, funding, recruitment.
As a value-added measure P8 is definitely a fairer measure than raw exam performance, which largely measures intake... all VA models like P8 are basically based on calculating a mean expected performance and then looking at the difference between that expected performance and what pupils in the school actually attain. What’s the problem here? Well, in any such model 50% of schools will be on or above, and 50% will be on or below the expected mean. So however we’ll the education system and indeed individual schools do, the statistical model means there will always be schools that ‘underperform’. So even in the ‘perfect’ education system up to half of schools would still be labelled as underperforming. That is the fundamental problem with using VA measures for accountability.
Muijs thinks that P8 is a "zero-sum game" but has value in learning from schools that do "really well" with their cohort. He points out that:
- schools with large proportion of EAL students do disproportionately well
- students from disadvantaged backgrounds require more support and are more likely to be set back by life events