by Terry Wrigley
Attempts
are being made to resuscitate the idea that ability is predetermined by our
genes. Recent research by Robert Plomin claims that 60 percent of achievement
in GCSE Maths or Science is genetic. His work is acclaimed by Michael Gove’s
senior adviser Dominic Cummings.
The
idea that academic ability is fixed and innate has damaged generations of
children. Intelligence Tests were used across Britain from the 1920s to 1970s
to separate children from age 11. On this basis, most manual workers’ children
were consigned to poorly funded schools, and left a few years later without
qualifications. These tests supposedly measured an innate intelligence unaffected by
schooling, yet most children were made to practise day after day to raise their
scores.
This divided education was underpinned
by Cyril Burt’s identical twins studies, subsequently exposed as
fraudulent. Internationally, James Flynn
has demonstrated that IQ tests carried out on entire populations show a dramatic
improvement, due to better health and education, within a single generation (known
as the Flynn Effect). This can hardly be the result of rapidly evolving genes.
Few
accept the theory of innate intelligence nowadays but it has a zombie
afterlife. For example, children are frequently divided into ‘ability groups’
from Year 1, diverting attention away from the different opportunities they
have enjoyed including access to books.
Various
attempts have been made to isolate and quantify this ‘innate ability’ by
studying adopted twins. The supposition is that identical twins, with the same
genes, will have widely different environments when adopted. This is a fallacy:
adoptive parents are carefully selected and likely to be quite well off, well
educated and very caring.
Two
sets of research are commonly cited nowadays, both seriously flawed. The US studies,
led by Bouchard, are based on twins who were chosen precisely because of their similarity. In the
Swedish study most of the twins had not lived separate lives; indeed in half
the cases one stayed with mum while the other lived nearby with granny or an
aunt. No wonder they turned out so similar.
Recent
genetic studies have failed to identify the genes. A massive study was
published in 2013, based on scans of 127,000 people, to find the genes
associated with educational attainment: the genes they located accounted for a
mere 0.02% of the difference.
The
latest study, by Robert Plomin’s team in London, uses GCSE results. Its
calculations are based on the ‘equal environments’ myth: i.e. that all siblings
have identical experiences. This is demonstrably untrue since identical twins are
often persuaded into dressing the same or doing things together. They are
likely to be in the same class, have the same maths teachers, work together on
homework and so on.
Based
on this spurious premise of ‘equal environments’, Plomin’s study calculates how
much GCSE results derive from innate ability, and how little from environment and
experience. It fails to look directly at parents’ qualification or income and
how much that correlates with the GCSE grades.
The
mathematical formulae are misleading. If everybody enjoys an excellent
environment (parental care, nutrition, schools, health service etc.), it will
appear that almost 100 percent of the difference between individuals is due to
genetics. For example, if in some Scandinavian country excellent food and
exercise raised the height of young men to between 6 and 7 feet, the
differences between 6 and 7 footers would be ascribed entirely to their genes,
even though the environment had clearly maximised growth.
Plomin’s
work is acclaimed by Gove’s senior adviser Dominic Cummings, who accepts his
claim of 60 percent ‘heritability’ for maths and science. Cummings even claims
that scores in the phonics tests show 70 percent heritability, and uses this to
justify cutting Sure Start programmes.
There
is a contradiction here for the government: are genes or teachers to blame for
low achievement? Both myths, in fact, serve to distract from political
responsibility for child poverty and spending cuts, in an age where the rich
get richer while preaching Austerity.
Read more from Terry Wrigley at http://www.changingschools.org.uk/
"The latest study, by Robert Plomin’s team in London, uses GCSE results. Its calculations are based on the ‘equal environments’ myth: i.e. that all siblings have identical experiences. This is demonstrably untrue since identical twins are often persuaded into dressing the same or doing things together. They are likely to be in the same class, have the same maths teachers, work together on homework and so on."
ReplyDeleteAs I understand it the studies compared identical and non-identical twins. I can see no real reason why the former would be more likely to be in the same class, have the same teacher or work together than the latter.
So many questions raised, and so few answered in a meaningful manner by this research.
ReplyDeleteHere's a conundrum. What would happen if everyone received the same (let's call it 'ideal') education and social experiences?
Would all have the same IQ, cognitive ability, test scores, or other measure of ability? I think not. Would their be winners and (relative) losers? Yes
Just like the analogy with height and nutrition, and any other COMPLEX trait, there would be a normal(ish) distribution of whatever discriminative ability parameter was being measured with a very major genetic contribution. Putting an accurate number on this is tricky and probably unhelpful.
The ability e.g. of IQ of the whole population would be expected to change with an altered environment (e.g. training for the test, or building specific skills) without a genetic drift, but what about relative ability for the same parameter? Flynn effect an irrelevant red herring. But what about pure cognitive ability?
Would the centile of the child match that of the parents for that parameter? Not necessarily and expect a very loose correlation based on mean of parental centiles (just like height). And accept massive deviations from this mean in both directions. Did I mention that we are dealing with COMPLEX traits?
We have Darwinistic processes going on (no, not natural selection at this point but natural variation) with rather more sinister Galtonesque interpretations of the data for political means.