Selective schools’ advantage based on heritable factors not school type, says new report

Janet Downs's picture

Selective schools have long been associated with higher achievement in exams at 16.  This achievement plays a major part in marketing for ‘top’ private schools.  And it’s behind the Prime Minister’s support to increase selection. 

But research by King’s College London found ‘exam score differences between selective and non-selective schools are primarily due to the genetically influenced characteristics involved in student admission’.   These traits were behaviour, personality, home environment and health.

When these ‘student and family factors’ were accounted for, the researchers found the type of school attended explained ‘less than one percent of the individual differences in educational achievement’ as measured by GCSE mean grade at age 16.

Academic achievement at GCSE level, the researchers reminded us, is linked with prior achievement, ability and socioeconomic status.  These are all factors involved in selection at age 11.  And if pupils are deliberately selected for the first two and obliquely chosen for the third, then what exactly do selective schools add?

Yet parents are willing to invest money, either by paying huge amounts of money to selective fee-paying schools or buying 11+ tuition in order for their children to attend state grammar schools.

The researchers cited a recent report which said pupils who attended fee-paying schools earned about £200k more than their state-educated peers between age 26 and 42 (£12.5k pa).  But this report didn’t differentiate between selective and non-selective state schools.  More research was required ‘to see whether differences in university attendance, career choice and earnings are still predicted by school type once individual student factors have been accounted for’.  And it would also be interesting to find whether there are differences between school types regarding non-cognitive traits such as confidence.

The researchers recognised there was huge variation within the schools studied.  Some were exceptional and some were under-performing. This variation was most apparent in the non-selective schools because they comprised most schools. 

There were two further limitations to the study.  Firstly, fee-paying schools and state-funded grammars aren’t evenly distributed throughout the country.  Secondly, researchers considered only English, science and maths in their GCSE analysis.  They acknowledged school type may have a greater influence on take up of such subjects as languages, art and social sciences. 

If selective schools don’t have as much advantage as their supporters claimed, then it appears parents are wasting their money.  And that’s without considering the negative affect that selection has on pupils who aren’t selected. 

It also shows how daft is the idea that there should be a ‘private tutoring tax’ taken from parents who use tutors to boost their child’s chance of passing the 11+.  The money raised would be used to pay for tutors for disadvantaged pupils.  Leave aside issues such as privacy, data protection, the enormous cost of administering the scheme and the potential to avoid the tax by paying cash-in-hand, it would appear the money would bring no real advantage.  Far better to scrap selection altogether and have a fully comprehensive system

Share on Twitter Share on Facebook

Be notified by email of each new post.


Roger Titcombe's picture
Tue, 27/03/2018 - 10:26

Research from Durham university featured in today's 'i' comes to the same conclusion. You can download the report here.

This is the Abstract.

"This paper forms part of a larger investigation of indicators of disadvantage and how they may be improved or supplemented in order to track school intakes and results better. Here our evolving dataset based on the National Pupil Database in England over 11 years is used to assess the impact of selective schools. At time of writing, the UK government is planning to increase the number of pupils attending state-funded selective grammar schools via a number of routes. They claim that this will assist overall standards, reduce the poverty attainment gap and so aid social mobility. Using the full 2015 cohort of pupils in England, this paper shows how stratified the pupils attending grammar schools actually are (worse than previous estimates) in terms of poverty, ethnicity, language, special educational needs, and even their age in year. It also shows that the results from grammar schools are no better than expected, once these differences are taken into account. There is no evidence base for a policy of increasing selection; rather the UK government should consider phasing the existing selective schools out."

Janet you are brave to mention 'genetically influenced' effects. The NUT influenced website 'Reclaiming Schools' refuses to post my comments, because I talk about 'general intelligence', even though I strongly support and promote the analyses they publish.

It is vital to recognise that 'genetically influenced' does not mean 'genetically determined'. These traits (in the research ) are behaviour, personality, home environment and health. These are among many heritable factors that combine to produce general intelligence 'g', which is therefore also genetically 'influenced' but not 'determined'.

These 'heritable factors' are also associated with relative poverty, which can and should be addressed through progressive social policies and taxation, but the mistake of the intelligence denying left is to mistake relative poverty as a causative factor in school attainment, when the causative factor is cognitive ability, which crucially is plastic and can be developed by good teachers and comprehensive schools freed from the corrupting pressures of our marketised education system.

Janet Downs's picture
Tue, 27/03/2018 - 12:22

Hi Roger - I've just written about the Durham report here.  

You're right that 'genetically-influenced' doesn't mean genetically determined.  It's important to make that distinction.  However, I'm not sure 'environment' is genetically-influenced to any great extent.  I would have though environment was more nurture than nature.    And it's also affected by economics.

Roger Titcombe's picture
Tue, 27/03/2018 - 13:38

Maybe the sense is of a supportive home environment inherited by the children from their intelligent, healthy parents. In my Mossbourne study I was impressed that the minority of FSM, Pembury Estate children, with the same high CATs scores as those from posher areas further away from the the school, did just as well. The whole point of a comprehensive school is that it should make comprehensive (ie full) provision to support the learning of all children equally well, regardless of parental affluence. When I first started my study Fiona Millar put me in touch with Henry Stewart, as someone with a lot of 'local knowledge'. Henry told me that when he visited Mossbourne he was impressed by the way that the students were supported in their learning. I was expecting something like other 'Spectacular School Improvement' Academy examples of the time that relied on the 'vocational scam' supported by cramming for maths and English. This turned out not to be the case. However, as the first head, Michael Wilshaw emphasised, the foundation strategy was to achieve a genuinely all-ability comprehensive intake that was not 'swamped' by  huge numbers of lower cognitive ability students from the neighbourhood estate. This was achieved by the Hackney CATs driven Fair Banding admission policy. Which is a long way of saying that I agree with you that the 'advantage' of inheriting a very supportive home background is overplayed in a genuinely comprehensive, well funded system (that we so obviously no longer have).

Roger Titcombe's picture
Tue, 27/03/2018 - 11:01

All this is connected with the 'Attainment Gap' fallacy that arises from ignoring the causative role of cognitive ability in school attainment and confusing this with FSM eligibility and base-line KS2 SATs, inflated in poor areas by their high stakes nature driven by flawed accountability measures and OfSTED reports based on them. The latest incarnation of  the attainment gap fallacy is the intervention of the Schools Commissioner for England, once again attacking the comprehensive schools in northern England.
This is my letter to her. I am not optimistic of receiving an acknowledgement or reply.
FAO Anne Longfield
Your report publicised in today's media (26 March) reproduces the ongoing fallacy that I explain in this article, that I hope you will find time to read.
The basis of your argument is that children from economically deprived areas fail to maintain or build upon the gains made through successful government policies in primary schools when they move on to secondaries. The comparison is made with the London Borough of Hackney where levels of deprivation are similar, but secondary school attainment is much higher.
The comparison lacks validity because of the common, but false assumption accepted by both the political right and the left that social deprivation drives low school attainment .
Part 4 of my book, 'Learning Matters' describes my research into the success of Mossbourne Academy and the Hackney Borough secondary schools in general.
The truth is that Hackney school students, like those elsewhere, perform on average in accordance with their cognitive ability as measured in Hackney, and a large number of other Academy Schools throughout London, by standardised IQ type Cognitive Ability Tests (CATs) taken in Y6, marketed by GL Assessment and used mainly to drive 'Fair Banding' based Admissions Systems.
The attainment gap illusion arises because these tests are not widely used in the poorly performing districts of the north of England on which your report is focused. My unique advantage is my experience as a Cumbria headteacher in the 1990s sitting on an LEA advisory group when the pupils in all the Cumbria schools also took the CATs, which were then provided by NfER-Nelson.
The Cumbria CATs data showed a clear link with social deprivation measures. The Borough of Hackney data, to which I had full (anonymised) access shows a parallel link, but crucially, the overall cognitive ability levels are much higher, as they are throughout the London area.
The truth is that former industrial white working class areas of northern England are characterised by astonishingly low IQ levels that are reflected in CATs scores in the few places where the data exist. For example, my headship school in the urban centre of Barrow-in-Furness had a mean intake CATs score of below 85 (-1 SD, 16th perecentile), and in all probability still has, explaining the the ongoing problems of Furness Academy, which opened in 2009.
My school had an HMI Inspection in 1990, followed by OfSTEDs in 1995, 1998 and 2004 (this last a year after my retirement). The school failed to meet government floor targets in every OfSTED inspection, but this did not prevent a succession of good and ever improving OfSTED reports all of which were informed by the CATs data that the inspectors took  into account at every inspection.
OfSTED inspectors are no longer allowed to use such CATs data, which makes little difference in northern schools where it doesn't exist anyway and in London schools where the CATs scores are much higher.
Your problem is that the base-line for school secondary school performance and progress measures is KS2 SATs. These high stakes tests produce inflated results because of their high stakes nature. There are also other flaws in the validity of the new measures that are discussed here.
You may find this deeply unattractive to believe, so I urgently propose that you test my assertions by introducing CATs tests in a sample the northern schools serving severely socially deprived schools that are your concern. You could do this in current Y7 year groups. The results data that GL Assessment will provide will include predictions of the Y11 GCSE profile for these pupils.
Given that while some schools certainly are more effective than others, I am confident that such CATs-based predictions will still on average be much closer to the levels of attainment that you are judging to be unacceptable in northern schools.
So if I am right what is to be done? The first essential is to abandon the high stakes SATs base-line testing regime and replace it with one based on the CAT, which is a low stakes cognitive ability measure.
This will free primary schools to concentrate on the development of cognitive ability (plastic intelligence) in ways increasingly set out by EEF, that the DfE and many Academy MATs increasingly ignore.
The current high stakes SATs regime actually inhibits the development of cognitive ability. The result is that the more effort that you and DfE apply to the problem on your false model, the worse it will get, which is exactly what has been happening for the last 20 years.
Best wishes
Roger Titcombe

Add new comment

Already a member? Click here to log in before you comment. Or register with us.