You’d think an Education Secretary who had twice been censured
by the UK Statistics Watchdog for dodgy use of statistics would take care when spouting data.
But it was not to be. In her speech
to the annual conference of the Association of School and College Leaders she said three quarters of teenagers in Trafford, in Greater Manchester, reached the benchmark of 5 GCSEs (or equivalent) A*-C including Maths and English. But “just a 30-minute drive away” in Knowsley “the number is less than half of that at 35 per cent
But she's failed to understand the significance of sample size.
There are just six mainstream state secondary schools in Knowsley. A sample of six is rather too small to make a safe comparison with another local authority especially when the six schools are in a pocket of deprivation.
Morgan makes the mistake of not comparing like with like. Trafford is an affluent area which maintains selection - this draws high achieving pupils away from neighbouring LAs. Knowsley is not an affluent area - pupils eligible for free school meals any time in the last six years (FSM6) formed 54% of the GCSE cohort in 2014. In England, 26.9% of the last GCSE cohort was FSM6. Knowsley’s proportion was more than double that. In Trafford, 20.1% of the GCSE cohort were FSM6 - less than half of Knowsley's and lower than the national average.
The proportion of children of children with Statements or on School Action Plus (SEN) in each of the two areas is also different. In Trafford, the proportion in the cohort was 9.3%, the same as the national average. In Knowsley, the proportion was 15.6%.
Trafford has 18 mainstream state secondary schools. Seven are grammars. Only one, Stretford Grammar, has a proportion of FSM6 more than 9% in the cohort (it had 18%). The other grammars had between 1% and 9%. As would be expected, the proportion of SEN children in the grammars was low – between 0% and 4%.
FSM6 children in grammars would, of course, have been chosen for their high ability. It’s hardly surprising, then, that 100% of disadvantaged pupils at selective St Ambrose College reached the benchmark. But there were only six of them in a cohort of 156. In three of the grammars, the numbers of FSM6 pupils in the cohort was so small, the results were suppressed. This rather blows a hole in the argument that grammars help bright but disadvantaged pupils out of poverty.
Morgan, then, has compared a local authority with a tiny number of non-selective secondary schools, which had more than twice the national average of FSM6 children and an above-average proportion of SEN children in the 2014 GCSE cohort with an affluent, selective local authority with three times the number of secondary schools, where the GCSE cohort had a lower proportion of FSM6 children than the national average and where the proportion of SEN children equalled the national average.
This is hardly a fair comparison.
EXTRA: Morgan also told the Conference the curriculum should be decided by elected ministers. So much for autonomy.
NOTE: The above article has been slightly changed to tidy up tenses and also to make it clear the figures refer to the 2014 GCSE cohort and not the entire intake. All figures from the School Performance Tables.