Gove vs Balls: seconds out!
Statisticians have rushed hither and yon to umpire the row between Michael Gove and Ed Balls in the Commons over whether – as Mr Gove claimed – only 45 of the 80,000 children eligible for free school meals get to Oxford and Cambridge.
Mr Balls said he was wrong, because he hadn’t included those who go on to further education colleges. But the umpires (Kathy Newman’s Fact Check website, and Tim Harford of the Financial Times on Radio 4’s Today programme ) concur that Mr Gove is right.
The parliamentary answer that provided the figures is here. In fact, it looks as if Mr Gove on this occasion delivered a googly, if we can change the sporting metaphor. He’s previously asked why only 189 children on free school meals get three A-levels at A grade – and those figures do exclude FE colleges.
But the university entrance ones include them, so Mr Balls fell into a well-prepared trap. The figures show, incidentally, that decent numbers of these children do make it to university: 100 to Birmingham, 250 to Queen Mary and Westfield, 155 to Salford, 50 to University College London, for example. In all, 10,590 went on to higher education.
But the statistics are pretty basic, rounded to the nearest five, so this calculation shouldn’t be relied upon.The real cause for celebration is that so many people are now seeking to check the facts. Politicians are going to have to be more careful that they usually are in the run-up to the election.

Anonymous (not verified) wrote,
Wed, 10/03/2010 - 23:32
so what is your view on HMT hiding £61.6bn borrowing/lending last year from ONS and does it mean we cant trust the national accounts any more [see Statistics Authority reply to Bill Cash of 18 December]
when are you going to start properly criticisng the many dodgy statistics quoted from academic studies that appear regularly in the news where correlation is often confused or confused with causation, deliberately or otherwise - do we really believe that one study showing drinking wine doesnt increase weight loss makes it ok in the face of a whole body of previous research [perhaps those drinking a little are healthier because they know from experience that they need to exercise to keep their weight down?]...and anybody who knows about random variation understands that if you do enough studies you may get an exceptional result
when will you report on how many studies dont get published by pharmacetical companies ?
keep up the good work
Post new comment