Changing the right behaviour

Tweet Once again last week, Gove emphasised the need for more testing of children in schools in England; and the lack of testing has been blamed by him and those who think like him for the relatively poor performance of the education system in Wales.  His party in the Assembly seem to share his views on testing, but where is the evidence that testing improves performance?  What is the mechanism by which that might happen?

I can think of two possible mechanisms by which testing might help to improve results, one of which works in a generally positive way, and the other rather less so.

The first is that regular testing of pupils’ knowledge and understanding can help teachers identify problems – whether class-wide or restricted to individuals – and adapt their teaching methods as a result, including targeting extra help at those who need it. 

The second is that the data underlying the tests on which the results of national published league tables are based is analysed to identify that small number of pupils who, with targeted help, can make the small improvement needed to move from one band to another and thereby improve the school’s apparent performance for the minimum expenditure of effort.

The first of those is an entirely positive approach to teaching, and it’s something that many teachers do regularly and informally in their classes anyway.  The second is about managing to targets, and changing behaviour in order to secure a particular outcome in terms of the public perception of the performance of a school or LEA.  It doesn’t deliver for the majority of pupils.

The difference between the two isn’t the nature of the testing as such (although tests used for producing league tables need to be more standardised, formal, and rigid than those used more informally), it’s the way in which the data is used.  League tables, ultimately, encourage behaviour by those running our education system which leads to apparent improvements in the performance of schools.  What those behaviours don’t necessarily achieve, however, is improving the educational outcome for the many; the same result can be achieved by simply ignoring the majority of pupils.

League tables based on testing give politicians and managers nice graphs and statistics; they can be used to chart apparent improvement or decline, and even to reward ‘successful’ teachers and schools, and punish the ‘failures’.  It’s much harder in a complex top-down system to ensure that the behaviour and approach of individual teachers in individual classrooms delivers the best results for all pupils – but that’s what we actually need. 

It seems to me that the Welsh Government, with its increasingly complex banding proposals, is closer to the Tory approach than they like to claim; for all the rhetoric about “systems leaders” and collaboration, they are still coming back to national comparisons and tables because they can only think in terms of centralist top-down approaches.