Earlier in the week, the Times carried an article about a report on the research findings of Professor David James of the University of the West of England.
Middle-class parents obsessed with getting their children into the best schools may be wasting their time and money, academics say today.
They found that children from privileged backgrounds excelled when they were deliberately sent to inner-city comprehensives by parents opposed to private schooling.
Most of the children “performed brilliantly” at GCSE and A level and 15 per cent of those who went on to university took places at Oxford or Cambridge.
My alarm bells were set ringing by the claim that 15 percent of those who went on to university took places at Oxbridge. Why was this good performance limited to Oxbridge? Were the results for other universities similarly impressive? A classic way of lying with statistics is to subdivide your sample population until you get the answer you're looking for.
Later in the article we read that the sample population was 124 families. This would suggest no more than a couple of hundred children were assessed, so concerns about the statistical significance of the results appear to be fully justified.
The article on which the Times piece is based hasn't been published yet, but in the style so typical of modern "academics" the UWE has chosen to issue a press release and a short report on the projects findings before official publication. It's here.
From this we discover that the families and children assessed covered a range of ages. This significance of this is that only a fraction of those assessed will have actually reached university entrance age. Let's say that this was forty children. That would mean that six went on to Oxbridge. If it had have been five then it would only have been 12%.
The idea that one could make any claims based on results of this kind is a joke. That the Professor is issuing such a misleading report is really rather reprehensible. It looks more like a piece of political propaganda than real research.