30. Memorandum submitted by Professor
Stephen Gorard (DP 48)
COMMENTS ON
THE MEMORANDUM
SUBMITTED BY
PROFESSOR DAVID
JESSON[8]
I agree with much of what David Jesson presents,
and will therefore concentrate on issues in which our conclusions
diverge. I will not repeat in full the criticisms I have already
made (such as the need to remove secondary-modern schools—and
it is interesting that the TCT have no knowledge that ANY secondary-modern
schools are now specialist).
David says:
"Thus, the suggestion that there is no
evidence indicating better performance by pupils in specialist
schools is, quite simply, false". I would not dispute these
figures. Current specialist schools do have higher raw-score benchmarks
than non-spec. There are two crucial provisos that the account
you sent me glosses over.
(1) Does this mean that the schools/pupils
"perform" better? I would say not. It could only mean
that if the pupils were equivalent to start with (matched or value-added).
We know from our own work that specialist schools take proportionately
fewer pupils in poverty than their neighbouring schools do. Therefore,
these local differences need to be taken into account (and it
is not sufficient to say that spec. schools are taking lots of
poor children, or around or even above the national average—spec
schools do not generally appear in areas of low population density
which also have lower levels of poverty, for example, so this
is easily explained, see my 2001 paper).
(2) As he later notes, David's statement
is very different to claiming that the raw-score difference is
due to specialism (which is why I stated that we need four data
points for comparison). For example, no school in special measures
("failing") has become specialist, so there is at least
some selection of the schools that are part of the scheme, so
we would expect them to have higher raw-scores. It means absolutely
nothing by itself. Those schools that are now specialist have
NOT improved their raw-score by more (proportionately) than the
non-spec schools since 1994.
The issue of selection is a difficult one. My
concern is not with the mechanism of selection (if indeed there
is one at all—the work from the LSE is far from convincing
in my mind). My concern lies only in the fact that segregation
studies show specialist schools to be more socially segregated
than their neighbours (according to Form 7 census figures). However,
they share this characteristic with Foundation, Faith-based and
Welsh-medium schools in general. What all of these types have
in common is different admissions routines and over-subscription
criteria (and often travel arrangements) from the "bog-standard"
LEA-controlled school. This is where I seek my explanation for
the differences in pupil composition.
My last comment is a minor one. In the "progress
from KS2 to GCSE" table, there is a discrepancy ("residual")
between the predicted and actual scores for 2001 for both spec
and non-spec schools. This might mean that the groups are achieving
over or below what is predicted (as David concludes) but an equally
plausible argument would be that this simply shows the predictions
to be deficient in some way (omitted variables etc). I would like
to see this point argued at least, rather than assumed. I am keen
that we examine the unexamined.
To conclude, I agree with much of what has been
written and certainly do not dispute the figures presented. Above
all, I heartily agree with paragraph 5, first sentence. Some other
commentators are over-complicating the issues by peddling methodological
avenues that they favour (and from which they make money via software
licences be it noted—caveat emptor). This leads us
away from a consideration of what the figures we do have actually
mean. No amount of statistical jiggery-pokery will overcome the
fact that the only sure method we have for deciding whether specialisation
is effective would be to conduct field experiments (with random
assignment of lots of schools to spec and non-spec for a substantial
period and with equal funding/admissions etc). If policy-makers
genuinely want to make evidence-informed policy then someone needs
to face up to the practical, financial and ethical issues that
this involves.
December 2002
8 Ev 147-8 Back
| 
