Glib assertions

However I need to correct a misunderstanding: it is academies overall (not specifically ‘converter’ academies) that fall behind maintained state schools in the number of students achieving five-plus GCSEs or equivalent at A*–C including Maths and English (50.1 per cent and 57.8 per cent respectively) in the DfE data.

Converter academies considered alone actually average better (77.1 per cent) than maintained secondaries but, as only high-performing schools have been eligible to participate in the ‘converter academy’ programme, this outcome is hardly surprising.

Since my letter was submitted, however, a more discriminating statistical analysis of academy performance has appeared (see http://bit.ly/zAigU8). In this study, Leonard James divided English academies and maintained secondaries in the DfE GCSE database into narrow bands according to free-school-meals uptake (DfE’s measure of deprivation).

He compared sponsored academy and community school GCSE performance within each band, thereby ensuring that sponsor-led academies in deprived areas were compared only with maintained comprehensives in similarly challenging circumstances.

James found that sponsored academies performed on average less well in the key GCSE measure (as above) than maintained comprehensives in the same band, to a statistically significant extent, except for bands with FSM uptakes above 36 per cent (representing deprived areas) where maintained school and sponsored academy populations were statistically indistinguishable in their GCSE outcomes.

Apart from a few conspicuously successful academies, therefore, the evidence does not suggest that the Academies Programme is delivering any improvement in GCSE outcomes where the need is greatest.

Councillor Tilley should beware basing the profound changes proposed for the county’s education provision on Michael Gove’s glib assertions, without taking professional advice on the statistical evidence for their effectiveness.

Robin Gill, Oxford