It is A level results day again and we brace ourselves for another set of spurious articles based on press releases from the Independent sector. The Guardian is in first year with possible the most spurious use of data yet seen:
"There are some indications that private schools may have widened the gap with state schools, my colleagues Jeevan Vasagar and Jessica Shepherd write. The Girls' School Association, which represents heads of independent girls' schools, said that of the first schools to announce their results, 32.9% of grades awarded were at A* level and 70% at A and A*. In last year's results, for all private schools, 17.9% of entries were awarded an A* and just over half secured the two top grades."
How much can one paragraph get wrong? A prediction of a widening gap based on 18 out of 2,500 schools? A prediction with no information at all from the state sector. A prediction based on one set of 18 schools and comparing them, not with those school's results last year but with the whole private sector? Did it not occur to the reporters that these 18 schools might be generally above average?
And why does, year after year, the Guardian choose to base its results coverage on press releases from the independent sector. Do they not realise that the people issuing those releases could, just possibly, have their own interests and agenda?
If you would like to comment on the Guardian's approach, please post thoughts on the Guardian's education blog
To show how ridiculous the Guardian's article is, I did a quick analysis of the first 17 comprehensive schools to report results on the Guardian's own web site:
Indications that comprehensives have overtaken private schools in A level results
For the first 17 comprehensives submitting their results to the Guardian web site an average 28% of A level results have been A*. For all private schools last year, just 18% got an A*. The first indication, therefore, is that state schools have not only closed the gap but sped ahead of private schools.
Of course this has no more basis than the Guardian's own article. It is just possible the first 17 schools reporting are not representative of comprehensives as a whole. But to acknowledge that would be to actually think about what data means, and that might be asking a lot of some journalists.