A new Prime Minister and a new education policy. Not content with merely continuing the long established journey towards the privatisation of ‘state’ schools by stealth (i.e. academisation), Theresa May is keen on the idea of reintroducing grammar schools.

This is not a post about whether selection by ability is a good thing or not, though that may come in a future post. Instead I want to ponder how we decide which children will go to the grammar schools and which ones will not.

Selective schools, both state and private, already exist of course. These schools have to have a means to identify which pupils to take and which pupils not to take. The traditional method of selection is through a one-off 11-plus exam. Proponents of grammar schools argue, publicly at least, that grammar schools often a better chance of academic achievement for bright children from poorer/ disadvantaged backgrounds. However grammar schools always were, and still are, disproportionately filled with middle class children. Alongside the many advantages of coming from a better-off/ more advantaged home, there is a whole industry around tutoring/training to pass the grammar school entry exam. My eldest son is now in year 6 and if we lived in a grammar school area I would now be throwing the proverbial kitchen sink at him to ensure that he gets into a grammar school-- after all, bringing back grammar schools for the ‘brightest’ (say 25%) children also means bringing back secondary modern schools for the other 75%. I went to a comprehensive school but know a lot of people who talk about ‘passing’ or ‘failing’ the 11+. The so-called ‘failure’ seems to have had a lifetime affect on the post-war generation.

This brings us to the crux of the matter. The twittersphere has been abuzz about the need to ensure that the ‘tests’ (I use the word here to cover any sort assessment) are tutor-proof. In, other words is it possible to design an assessment that will reliably discriminate between ‘able’ and ‘less able’ which does not discriminate on the grounds of previous experience, background or performance? Is there a means of assessment to prevent a less-able child getting into grammar school because she has had private tutoring to help her pass the test at the expense of bright of an able child who does not enjoy these advantages? This takes into the even more dangerous territory of ‘innate’ intelligence that can be separated from previous experience and from teaching and learning. Is there a test that can separate the disadvantaged child who may not have performed well in primary school and may not have ‘engaged in education’, but would benefit from a grammar school education from a pupil who does well because she works hard and has a supportive (or pushy) home environment?

To be totally reliable such a test would need to be:

1. Impossible to game through studying or the practice of learning and teaching. (Despite what proponents of IQ type tests say, like any test, the more you practice the better you get at them). The tutor-proof assessment goes against the whole point of assessment which is to evaluate whether learning has taken place. Assessment might be able to predict future learning achievements, but only on the basis of past learning.

2. Be culture/ value free. The tests would need to ensure that children were not disadvantaged by going to the wrong school, growing up in the wrong type of household or with parents from a different culture to the prevailing local culture. For the most part cultural bias can be taken into account, but not eliminated completely. Cultural bias can many forms and assessment can reward knowledge of both so-called ‘high’ and ‘low’ culture. The belief that cultural bias can be eliminated completely is more dangerous than cultural bias itself.

3. Not rely on luck. There is a subtext in the grammar schools debate that grammar schools are/ will be good schools and other schools not so good. If in attempting to eliminate the above questions we end up with an assessment which is only slightly more reliable than a coin toss, a dice throw or a game of snakes and ladders then the whole point of selective schools in undermined.

4. Transparent and fair. If those taking the test do not know how marks are allocated or exactly what is being assessed then the assessment is neither transparent or fair. Once teachers/ tutors/ parents know what kind of questions are asked on a test they can learn how to do better on the exam and this lead to testing to the test.

I’ll write in more detail my thoughts about selection at age 11 in a later post, though you’ve probably guessed I’m not very keen on the idea.