Sunday, 1 November 2015

Thoughts on testing, more testing, earlier testing, later testing, testing testing...

1. The great struggle over testing is about 'validity' and 'reliability'. The tests move more and more towards being 'valid'. To do this they have to be narrower and narrower i.e. testing things on the basis of right and wrong answers (as they claim). This makes the tests more 'reliable' (as they claim). However, as the tests do this - and they do - so do they become less 'valid'. That is, they discriminate between the tested people on narrower and narrower grounds. This wouldn't matter if the tests were quick guides for teachers and pupils on what to work on, say. But they aren't. They are increasingly a matter of high-stakes: ranking schools, determining the outcome of schools, determining teachers' livelihoods, children's routes through education, and leading to the setting up of schools which are not accountable to local democracy or scrutiny of accounts. It is all linked. It is in the interests of those who advocate more and more testing to deny or ignore these links. To them, these are just 'unintended consequences'. Not so. There is a system and 'highly reliable' and not 'valid' testing are part of a system.

2. Tests for 4 year olds are tests of parents. The first direct consequence will be a rash of 'booklets' for parents to 'improve' scores.

3. All testing exhibits a world view...even of the validity of testing itself! When we test 4/5 year olds we immediately impose a value-system that when you are asked something you mustn't ask for help! Aha, say the testers, this only applies to the test itself! Not so, because the tests set up a tailback into education and parenting. Because that question-answer relationship is set up by the test, so inevitably teachers and parents imitate it in the lead-up or indeed more generally as part of practice: if there's something you don't know, just sit there, sweating; you cannot ask for help. I've heard myself doing it with my own children in order to 'help' them do the test! This is a value-system imposed by testing itself.

4. One test I saw for 4/5 year olds (some years ago) asked children to distinguish between a 'spade' and a 'shovel'. According to the pictures, the spade had no upturned edges. The point here is that there would be cultural reasons why some children would be able to get this 'right' and others wouldn't know. This has nothing to do with 'intelligence' or 'language ability' and a lot to do with whether your parents have a garden, or indeed if your parents were farmers! This is an example of being 'reliable' but not 'valid'. There is only one answer but the answer is ludicrously culturally biased.

5. If we want to analyse and understand what is going on in 'education' today, we have to get to grips with the implied and actual philosophies of the tests and testing principles.

6. A crucial part of testing is to know how to be tested! In the KS2 SPaG test (sample online at you can see that some of the questions are 4-way multiple choice. If you have any doubt, best thing to do is tick any old square. If you're right, the test will have tested your ability to know how to be tested not in any SPaG knowledge. Ultimately, what you're tested here though is whether your teacher has told you this.