Secondary Performance Measures – better but not there yet.

The government has now published its decisions on secondary accountability. Unusually, it would seem that they have paid some attention to comments received during the consultation. As a result they’ve dropped the Grade C threshold measure in English and maths from the core measures. Instead they have taken up the idea of counting English and maths double in the main attainment and progress measures.

The two key measures, then, are both based on the concept of the best 8 grades. It’s not quite that simple as the 8 have to include English and maths and three other EBacc subjects. There may be a temptation to over-react to the mere mention of EBacc in this context. That should, I think, be resisted. Most students do two science GCSEs so, in the interests of a broad and balanced curriculum asking them to take one out of history, geography, modern languages and computer science is not unreasonable and will stop some of the abuse of “equivalences”. Vocational courses can count in the remaining three slots but we still await any real detail of what these will look like in future.

It is proposed that the key measure will be the progress measure. It will be calculated by comparing a student’s GCSE grade with the average performance of students with similar prior attainment (in KS2 tests). There are two genuine positives in this. First it should reduce the hysterical C/D borderline fixation and encourage schools to pay attention to all their students. Second, by giving priority to progress over raw attainment, it gives schools with more disadvantaged intakes a better chance and will challenge the comfortable coasters who ignore less able kids.

But relief at getting something half-way sensible should not lead us to assume that there are no problems left. It is proposed that the progress measure will be presented as a single grade for a school. In an effort to make things simple for parents, we will actually be ignoring all the complexities that lie behind the figures. They are not exact measures and shouldn’t be given such a level of spurious precision. KS2 results are notoriously unreliable and often don’t really measure what children really know and can do. Rather they measure how good the primary school is at test preparation. Nor of course are GCSE grades all that reliable these days. So any progress measure based on them is likely to have a substantial level of statistical unreliability.

More importantly still, we have to question the use to which these measures will be put. We are still in the world of using them to judge and compare schools. There will still be a very simplistic floor target – defined as half a grade below the average performance for pupils of the same ability. This won’t take out of the system the fear that no one’s job is secure beyond the next set of results. Rough measures like these are in the true sense indicators – and no more. They may give us a clue as to what is happening but they are never conclusive evidence – especially if we are looking at one year’s results that may be affected by any number of extraneous factors. You can only really judge a school by studying in depth what goes on from day to day. There will be random variation from year to year and real improvement takes time, commitment and continuity.

The key thing is to get away from knee jerk judgements. A low progress measure is a reason for keeping a school under review and for considering support and intervention. So are many other things – low attendance rates, high exclusions, high staff turnover and so on. These are all things that a good local authority would continually be scrutinising and that a good inspectorate would be picking up. But they won’t be picked up by civil servants trying to monitor schools from Whitehall. Hitting a school over the head in response to one piece of data is not a particularly well-crafted form of intervention.

So two requests. One is to recognise the real complexity of school performance data and find ways of better reflecting that in what is published rather than simplifying it down to a single sound bite. And two, get rid of the simplistic floor target notion. The data is not good enough to justify it. Treat it as one piece of evidence amongst many that enable us to judge how a school is doing and how it needs to be supported.