The ‘attainment gap’ still misunderstood
This is the latest example from a BBC report of 3 August.
“The very poorest children in England have fallen even further behind their non-disadvantaged classmates since 2007, research says. The Education Policy Institute study suggests the most disadvantaged pupils are more than two years behind their classmates when they sit their GCSEs. These children were those entitled to free school meals for 80% of their time at secondary school. The Education Secretary has warned of a social mobility emergency. And in a recent speech, Justine Greening highlighted some areas of the country with an “entrenched disadvantage” – where low skills and poor employment were found in a downward spiral alongside underachieving schools.
The EPI report Closing the Gap? acknowledged the disadvantage gap had been entrenched in the education system for generations and that successive governments had tried to tackle it.”
This last paragraph is true and it is sad that the repeated failure of successive governments to, ‘close the gap’ has not led to a more fundamental examination of the issue, which involves asking deeper questions about the nature of ‘the gap’. I have written about this many times.
There is indeed a gap. It is a cognitive ability (general intelligence) gap. It is not one that be closed because of the fact of natural variation, which is precisely described by the statistical normal distribution (bell curve).
For example, measured on the UK IQ scale (UK mean = 100), the mean Singapore IQ is 9 points higher (73rd percentile). This means that only 27 per cent of the UK population have an IQ equal to or greater than that of the average Singaporean. However, the IQ population profile of Singapore has the same bell curve pattern as that of the UK. The difference is that while the mean UK IQ is 100, the mean Singapore IQ on the same scale is 109. The whole Singapore bell curve is shifted upwards on the IQ scale while maintaining the same mathematical form.
Is this important? Of course it is. It means that on average Singaporeans are brighter than Brits and explains why Singapore tops the PISA international school pupil performance ratings. It also means that school pupils with attainment at the low levels that concern the Education Policy Institute and Justine Greening are thinner on the ground in Singapore than they are in England. Does this mean that the Singaporean school system is better than that of those in the UK? Not necessarily. When the international PISA test results are adjusted to take account of mean national IQs, then both Singapore and England come out badly. I explain and discuss the implications of this here and here.
So what is the real reason for the ‘gap’ reported by the EPI? The answer is simple. The ‘gap’ is an attainment gap (GCSE performance). But GCSE attainment is driven by cognitive ability. How do we know that? Firstly, from the common experience of teachers all over the world that some school pupils are brighter than others and that other things being equal (eg quality of schooling), brighter pupils do better in academic attainment tests than less bright pupils.
Secondly, we have the hard statistical evidence. In the 1990s the Cumbria LEA purchased ‘Cognitive Ability Tests’ (CATs) from NfER and all Y7 pupils in its secondary schools took these tests. This included my headship school. The results were used by the LEA to drive extra funding for schools on the basis of the number of pupils in each school whose CATs scores fell below a designated level. This was the ‘non-statutory SEN allowance’ for Cumbria schools. Throughout my headship I sat on the LEA head’s committee that managed the taking of the tests and the Cumbria system of non-statutory SEN funding. Several clear patterns emerged from the Cumbria wide CATs data.
The first was that in general its schools tended to perform at GCSE in accordance with their mean intake CATs scores. Each year the LEA used this information to plot a chart that recorded GCSE performance against mean Y7 CATs score for every school and a regression line was produced. This identified the schools above the regression line (good performers) and those below it (poor performers).
The second pattern was the vast range in the mean intake CATs scores of Cumbria schools. My school had the lowest (score 85, 16th percentile), but despite this we were always well above the regression line. The Cumbria school performance ‘league table’ based on this intake cognitive ability mediated basis was quite different to that of the government’s published School Performance Tables..
The third pattern was that each school’s GCSE performance was inversely related to the school’s Socio-Economic Status (SES) data, of which Free School Meals eligibility was a good proxy. In other words there was indeed an ‘attainment gap’ between FSM pupils and the rest, however in general all pupils, FSM and otherwise performed according to the CATs score regardless of parental affluence.
The fundamental reason for the ‘attainment gap’ is that pupils from poorer backgrounds tend to have lower cognitive abilities. This is why the ‘attainment gap’ persists despite all the government initiatives, ‘zero tolerance of failure’, head sackings, school closures and Academisation and Free School promotions. None of these have worked because they are remedies for an incorrect diagnosis of the problem. They are the educational equivalent of ‘blood letting’, where the response of doctors to failure of the patient to respond was to take more blood, ultimately ending in death. The analogy with the education policies of New Labour and Conservative governments is apposite.
Some years ago I wrote an article for the Guardian newspaper that addressed this issue, but the then education editor would not publish it unless I removed any suggestion that poorer pupils were on average less cognitively able than more affluent ones, despite this being a fact readily verifiable by GL Assessment who now market the CATs, and every school and Multi Academy Trust that uses them, usually to drive ‘fair banding’ admissions policies. The London Borough of Hackney, in which all the pupils have been taking CATs in Y6 for many years, also has the data that show this relationship. Part 4 of my book, ‘Learning Matters‘, contains a case study of the implications of Mossbourne Community Academy’s CATs driven banded admissions policy for its curriculum and GCSE results. Mossbourne, The (Hackney) Learning Trust and GL Assessment all assisted in this work, which has been shared with them. None have ever challenged its conclusions.
The relationship between poverty and cognitive ability has been known to heads and governors of secondary schools for decades, which is why after the 1988 Education Reform Act compelled all schools to compete on the basis of the raw aggregated attainment of their pupils, those schools that had powers over their admissions policies (Academies and Religious Schools) were increasingly inclined to find ways of cutting down the numbers of FSM children they had to admit. The most common approach has been to vastly increase the cost of the compulsory school uniform, but many other strategies are in use. LEAs were also well aware of this relationship, which is why most tried to achieve cognitively balanced intakes by imposing catchment areas that contained a mix of affluent and poor postcodes. Academisation destroyed the ability of Local Authorities to impose catchment areas and introduced an admissions free for all. This resulted in those schools located in affluent post codes (therefore admitting fewer less bright children on the basis of proximity) got better exam results, climbed the local league table, had more Y7 applications and were therefore able to deny entry to ever greater numbers of poorer pupils that lived further from the school, so increasing the mean intake cognitive ability, therefore getting better exam results etc. etc. The exact converse happened to schools located in the heart of poor communities. This admission inequality issue was solved by the introduction of CATs driven fair banding, which was not available to LA schools, except in Hackney.
The fact that poorer postcodes tend to produce lower cognitive ability pupils is not controversial. In every society for which we have data, educational achievement is positively correlated with their parents’ level of education or with other indicators of socio-economic status. This topic is central in social science. What remains controversial is the range of speculation as to the reason for this pattern, ranging from genetic inheritance, through qualitatively different parenting, to class and ethnicity based discrimination by teachers and schools.
It is not necessary to get into any of this, because regardless of the mechanisms, the result is that better educated parents, that tend to raise more cognitively able children, are usually more successful in their careers and so can afford to live in more affluent areas. This tendency towards social stratification has grown in recent decades. I write here from my personal experience of the cognitive decline of a South Birmingham council estate.
If the foregoing is not shocking enough, it is not the worst consequence of the misunderstanding of the real nature of the ‘attainment gap’.
This arises from the general failure of the government and the increasingly ‘business culture’ background of leaders of schools that are immersed in the ‘marketisation’ paradigm to recognise the fact of plastic intelligence and the opportunity that this provides for schools to teach in a way that increases the cognitive ability (general intelligence) of all their pupils including the most cognitively disadvantaged.
I write about the profound potential benefits of Labour’s proposed National Education Service here.
Unfortunately, in the marketised English education system, the response of the government to low attainment is to blame the schools, resulting in the sacking of heads and mass Academisation. This being the case it is no surprise that heads with lots of FSM children may be panicked into ‘quick fix’, behaviourist approaches to improving their school’s exam results. This has a theoretical underpinning from the neo-con economic movement of the US that inspired Margaret Thatcher and her ideological disciple, Tony Blair. I write about this here.
With the privatisation and marketisation of the English GCSE Examination Boards these methods can appear to work, but with the consequence of abusively severe pupil disciplinary systems and shallow ‘training based’ learning that suppresses the development of cognitive ability and fails to equip pupils for transition to cognitively demanding higher education. The consequence is an ‘Anti-Flynn’ effect of declining national IQ.
When it comes to variations in attainment between pupils it is necessary to understand not only that this cannot be reduced, but any attempt so to do makes the ‘gap’ wider (as EPI reports) and degrades the education system as a whole by robbing all pupils of all abilities of their entitlement to a cognitively development education throughout life.
The solution to unacceptably low standards is not to try to ‘close the gap’, or mathematically impossible attempts to bring the poorer achievers up to the average, but to develop the cognitive ability of all pupils at all levels in the bell curve distribution, while simultaneously addressing specific gaps in knowledge and understanding. In other words we must stop trying to deny the bell curve, or attempting to squash the ends into the middle, but lift it as a whole to higher levels for all pupils of all abilities. Slower learners may justify greater investment in aspects of their learning development to help them overcome specific cognitive hurdles, but not for the purpose of reducing the ‘gap’ between the highest and lowest achievers. This is not only doomed to fail, but its pursuit, by raising apparent attainment through the behaviourist approaches of didactic instruction backed by incentives and harsh discipline, risks damaging the cognitive development and deep learning of all pupils.