State schools outperforming independents – really?

The Telegraph claims A-level analysis offers good news for the Government’s educational programme – but something dodgy is going on.

A report in the Daily Telegraph today gives the Government something to seize on – “England's best 500 state schools are outperforming the top 500 private schools”. This is according to an analysis of A-level results. The Government might indeed like to make something of it, as it would seem to justify their educational policy. But it is interesting that the Telegraph could not find a member of the Government willing to be named, only a “senior source” who claimed that this shows that the independent sector should be learning from the maintained sector rather than the other way around.

I suspect that the reason the Telegraph failed to pin down anyone who would put their name to such a conclusion, is that the conclusion is nonsense.

First, it assumes that people send their children to independent schools in order to get good A-levels, and therefore if state schools are ahead in A-levels, then they are better schools. However, parents who choose the independent sector do so for so many more reasons than the exam certificates at then end. They may choose this sector for the greater range of extra-curricular activities and access to sport; or because their children lack confidence and the parents believe that the pastoral care is better; or because the parents work away and they need their children to board; parents may consider that these schools will develop resilience or other virtues in the child. Or, simply, the local comprehensive has a reputation for failing to address behavioural issues and the parents are not prepared to tolerate that. Whether it's right or wrong to choose an independent school is a different discussion, but certainly the choice of school is made on a more subtle basis than academic achievement, and education is more than A-levels.

Secondly – and this is the killer – the article is statistically disingenuous. They have compared the top 500 out of about 3500 state schools, with the entirety of the UK independent schools (approximately 500) that offer A-level courses.

This would indeed be a versatile statistical tool, to compare the top seventh of one category with the totality of another. Here's my idea. Take 70 women at random and ten men at random. Tell the shortest 60 women to go away, then measure the remaining ten women and all the ten men and declare that women are on average taller than men. Or that Greeks are richer than Germans, Austrians are plumper than Americans, or Skoda drivers drive faster than Saab drivers. Or whatever you want to prove, really.

I hope that pupils in or out of the top seventh of schools would be alert to the dubious statistical practices here.

Follow @7000hours

Tweet