Tuesday, 27 November 2012

Lies, damn lies and statistics. Why I'm not taking OFSTED's league table seriously


The UK schools inspectorate, OFSTED, today published a league table of local authorities’ comparative performance in inspections.

I’ve always considered OFSTED’s methodology in assessing the effectiveness of schools to be deeply flawed. Schools get, or at least used to get, an obscene amount of notice that OFSTED were going to turn up. I vividly recall my secondary school spending what seemed like months preparing for an impending OFSTED visit – tidying classrooms, selecting pupils to be interviewed by the inspectors, prepping them on what they were and weren’t allowed to say, painting the walls, and even installing an electronic information board in the main corridor which, needless to say, was removed once the inspectors bade them farewell for another few years. But the inspectors’ findings have been used for years to compile school league tables forcing schools to compete against each other in vying for for the top spot and, in turn, parents’ business.

My secondary school cherished its OFSTED “Outstanding” rating more than any other prize, which made it a popular choice for parents in Bradford – the school was over four times oversubscribed for its 2012 intake – marginally better, from a probability-of-getting-in perspective, than the University of Cambridge (figures taken for 2010 as last year available). This fact was used to exert control over its student population – frequently we were reminded of the enormous privilege it was to attend and the scores of applicants we had deprived of a place, in order to maintain discipline and attendance – which wouldn’t perhaps have been so bad if the school employed the sibling rule and/or aligned its holiday calendar with the city’s other schools.

This element of control served two purposes. The first of these was that behavioural standards were high, as was the attendance rate – which is some mean feat when unfortunate parents with children in two different schools can’t plan holidays when all their children are out of term-time. Whether these ends justify the means is another question entirely – but it doesn’t take a rocket scientist to work out that being continually reminded of the fantastic privilege with which one has been honoured rather takes the shine off. And for that you can blame my year 8 English teacher for never teaching me the rule about ending sentences with prepositions.

 The second of these is that the OFSTED rating was maintained, which allowed the school to continue the cycle of control into the future. Thinking about it, another consequence was probably an increase in the differential in comparative OFSTED ratings between the school and neighbouring schools – holidays in term-time were forbidden, so parents with children in two schools probably took the other child out. From the school’s perspective, a virtuous circle.

It’s only natural that, once they know they will be tested, schools will pull out all the stops to ensure they attain the highest rating possible. Even schools that don’t seek to lord vague conceptions of “privilege” over their pupil populations will seek to attract applicants – and, of course, there’s the headteacher’s reputation at stake. No-one wants to be seen to be at the helm of a sinking ship. And school league tables arguably serve a valuable purpose in allowing parents to make an informed decision as to where to send their children – even if the underlying methodology could use some work.

What I cannot understand is what useful inferences we can draw from the local authority league table that OFSTED published today. Sure, it’s interesting to be able to ask why the worst-performing local authority, Coventry, only sends 42% of its children to “Good” or “Outstanding” schools when neighbouring Solihull – fourth in the league table – can send 77%. But it’s not as though the league table can actually promote competition between authorities in the same way that a schools league table can. People generally live in reasonably close proximity to their families – it’s not as though a family in Derby, unhappy with the 43% chance that their child has of attending a “Good” or “Outstanding” school can just up-sticks the hundred miles to Buckinghamshire where their chance will rise to 78%, leaving their entire support network behind.

What I think is more dangerous is the potential for misuse, or misinterpretation, of the information. Local authorities are huge areas – and there’s the potential that the combining of data in this way does more harm than good. In Wakefield – a local authority of some 131 square miles, OFSTED calculate that a child has a 52% chance of attending a “Good” or “Outstanding” school. Hop over the boundary into Kirklees, a district of a mere 158 square miles, and that figure jumps to 71%. This may tempt a concerned parent in Wakefield to apply for a place for their child out-of-district in Kirklees.

But these figures don’t give any meaningful insight into the performance of individual schools within an authority. Within each authority there will be enormous variation in terms of schools that underperform, schools that perform adequately and schools that excel. The OFSTED table shows only those schools in the “Good” or “Outstanding” category – more granular information is required to draw an informed conclusion. There are absolutely no grounds to infer that any one school in Kirklees is better than any other school in Wakefield – it’s perfectly possible that a Wakefield school just over the border from another one in Kirklees is the better one. Furthermore, no mention is made of what proportion of other grades there are within the authority. Which is better – an authority with 80% “Outstanding” and 20% “Special Measures” or one with 60% “Outstanding”, 20% “Good” and 20% “Satisfactory”?

The table could, in limited circumstances, be beneficial as a benchmark of the comparative performance of local education authorities. Or, at least, it could if the league table included only those schools within local authority control, but that isn’t what they have done. This league table includes all state schools within each local authority, including Free Schools, Academies and Technology Colleges, over which the local authority has no control. I’m no council apologist – I’ve had too many parking fines for that – but to rank the performance of local authorities against each other in areas where they have no remit to take action is the height of unfairness.

I’m all for presenting complex situations simply, and I can see what OFSTED have tried to do, but this over-simplification of the facts does nothing but obfuscate them. In any kind of ranking there will be a winner at the top and a loser at the bottom, but in this case the ranking is so meaningless, due to the massive variation within the reported units and the lack of control of the local authorities over the schools within their boundaries. Miriam Rosen, HM Inspector of Schools, knows this, and might as well not have bothered.

No comments:

Post a Comment