Nick Seaton of the Campaign for Real Education says CVA measures for schools offer a cover for failure

Everyone involved with education at local level will have come up against ‘contextual value added’ (CVA) measures of school performance. But how many question such manipulated statistics as they should?

In January this year, Bristol University’s left-leaning Centre for Market and Public Organisation published a paper, The Limitations of Using School League Tables to Inform School Choice by George Leckie and Harvey Goldstein.

It needs a maths degree to understand it fully, but no matter. The general thrust is that because the system now uses pupils’ test results when they were 11-years-old and the same pupils’ exam results
when they are 16 or 18, performance tables are a poor measure of school performance. In the years between, it is argued, a bad school may have become a good school and vice versa.

The catch, of course, is that it is not tables of raw test and exam
results that are at fault. It’s the way raw results are manipulated by
the CVA process, then presented to suggest an under-performing school
is better than it is.

CVA  is  a handy tool for officials, headteachers and politicians to
justify their  ineffectiveness. It also allows ideologically driven
members of the establishment  to attack good schools (especially
grammar schools) on the grounds that their success is due to the
‘privileged’ nature of their pupil intake.

Concerns about CVA revolve round the unreliability of the complicated
mathematical formulae it uses to estimate the results youngsters should
achieve when they take their GCSEs or A-levels.  Instead of
emphasising real results, it manipulates them and measures schools on whether or not they achieve their concocted targets.

This manipulation takes several variables into account: gender, age
within cohort, eligibility for free school meals (FSM), special
educational needs, ethnicity, English as a second language and income

Implicit here, of course, is the idea that a child who suffers from any
form of social deprivation (such as being a boy!) should not be
expected to do well at school.

The FSM measure has always been questionable, because self-reliant
families rarely claim such benefits even when they could. The
Department for Children, Schools and Families (DCSF) has gone some way
to acknowledging this. Its Guidance for Local Authorities on Revising
and Resubmitting Expressions of Interest in Building Schools for the
Future, DCSF, September 2008, says: ‘For social need, we will now use
the Tax Credit Indicator (TCI) rating for the school. This responds to
several comments that the previous proxy, eligibility for Free School
Meals, often does not, for local and cultural reasons, fully reflect
social deprivation.’  Do they know each family’s income anyway?

Reporting recently on some of these developments, the Times Educational
Supplement (27 February 2009) quoted  Professor Goldstein, who
described CVA measures as ‘at best misleading, at worst dishonest.’
Professor Stephen Gorard of Birmingham University, who writes sensibly on the subject, advised that CVA measures provide ‘no basis for making policy, rewarding heads, condemning teachers or
closing schools.’

Despite all this, CVA is regularly used by Ofsted, and local authority inspectors and officials, when advising elected members.

But aren’t schools funded and run for the benefit of their pupils, not
those who work in the system?  What matters to youngsters and their
families are the raw results they personally achieve, not whether their
school has reached some politically-driven target.

Each year, millions of taxpayers’ pounds are wasted compiling CVA
tables. Isn’t it time to save some money? Have simple, honest
statistics? And, incidentally, consider the true value of different
exam subjects?