But whats in for League tables and accountability?
The use of Contextual Value added in league tables has been dropped by the government with the official reason given that it is too complex for parents to really understand. Possibly true, but there is also an unresolved debate about how best to measure value added. It has long been used by the SSAT to measure schools performance but there are grave doubts about the accuracy of value added measures and academics still cant agree on the best methodology . The debate is at its most acute in the United States where much effort is focused on measuring the amount of value teachers add in schools as part of the education reforms there. But it remains hard to identify consensus about how fair and effective such measurements are, and how one can accurately take in to account and factor in other variables affecting pupil performance.
But if contextual value added has been rejected as too complex, here in the UK what might be a better measure? Mike Baker in the Guardian recently reminds us of the work of the CMPO Bristol. Rebecca Allen and Simon Burgess published a paper for the IFS last year entitled Can School League Tables Help Parents Choose Schools?. Allen and Burgess devised a measure that gives parents the “expected GCSE performance for a child of similar ability to theirs” for all schools in the local area. Because it is reported in terms of average GCSE grades, rather than points, they argue that it is “relatively simple” for parents to interpret.
An important advantage of their proposed measure is that it militates against schools focusing their efforts on pupils at the C/D threshold. It is based on the best eight GCSE subject grades for pupils at three different points in the ability distribution: those who scored at the 25th, 50th and 75th percentiles in their key stage 2 tests at the end of primary school.
A further advantage is that it should be more useful in helping parents to choose schools because it works better as a predictor of their own child’s likely exam grades if they attend a particular school. This would help to dispense with the misleading idea that there is somehow a “best school” when it is much more a question of which is the “right school”.
Bakers view is ‘No doubt some will still regard this proposed measure as more complex than the current five A*-C measure. Certainly, it does not lend itself to a simple football-style league table. But maybe it is all the better for that, since the current measure is misleading if parents believe it will tell them what their own child is likely to achieve at a particular school.’