The system for inspecting teacher-training institutions is so flawed that even very good universities have a 50:50 chance of failing, according to a new statistical analysis, writes David Budge.
Four primary teacher-training courses face loss of Government funding and possible closure after being criticised by Office for Standards in Education inspectors. But Dr Peter Tymms, a leading education researcher from Durham University, believes that the odds were stacked against the four failing institutions the second time round.
Most initial teacher-training institutions passed the first round of primary course inspections last year, but Dr Tymms says that the more stringent framework of the re-inspections makes it highly likely that even the best training colleges will trip up.
Chris Woodhead, the chief inspector, ordered the re-inspections because he felt that the originals painted too rosy a picture. But Dr Tymms's calculations suggest that the second round will be unnecessarily damning. "This is an outrageous state of affairs," he says in a statistical paper he has released via the Internet. "It is against natural justice . . . it brings into question all frameworks constructed by OFSTED."
Under the new framework, colleges are being rated in 14 aspects of their operations. It is possible to score grades of 1-4 in each of the 14 areas - but if colleges collect even one grade 4, the Teacher Training Agency considers withdrawing funding.
Dr Tymms ran a series of computer simulations which involved 100,000 hypothetical permutations of inspectors' findings. It showed that even if the inspectors' judgments were generally very accurate, there was a 73 per cent chance of at least one grade 4 being recorded if the true score for each of the 14 areas was 3.
"My calculations were based on explicit assumptions, and reality will doubtless differ from the idealised model," Dr Tymms said. "Nevertheless it is clear that the framework of inspection has been set up in such a way that any good higher education institution stands a high probability of being identified as not complying with the Secretary of State's criteria."
An OFSTED spokeswoman rejected Dr Tymms's arguments. "The conduct of inspections cannot be predicted mathematically. They are based on reality - the quality of provision, not chance."
Dr Tymms's paper, The security of inspection of initial teacher training, can be accessed via the Education-line site, http:www.leeds.ac.ukeducol