Closing the Gap

This government, like its predecessor, is obsessed with ‘Closing the Gap’. I am passionately committed to reducing inequality in our education system and in the country as a whole, but the ‘Closing the Gap’ argument based on the supposed key role of social deprivation in constraining educational attainment is completely wrong.

The DfE is consulting on this issue. You can read about their approach here.

You are invited to send comments to

This issue is a major theme of my new book, ‘Learning Matters’. It is a complex issue that requires a number of related misunderstandings commonly shared by those on the left as well as the right of politics to be addressed before it can be fully understood.

Relevant quotations from ‘Learning Matters’ by Roger Titcombe

From Section 1.2

This book argues the case for a developmental approach to education. It is based on the idea that attainment, in all its forms and contexts, is founded on general abilities and that it is the job of schools to recognise and to promote the development of these underlying abilities. At the same time a school should be maximising students’ attainment in their academic studies and nurturing the physical, artistic and social skills that grow out of these talents and abilities. My book draws heavily on the work and ideas of Jean Piaget and Lev Vygotsky.

Much of what follows is based on the concept of general intelligence and the validity of its routine measurement by means of commercially available Cognitive Ability Tests (CATs).

Although the basis for the routine work of Educational Psychologists for more than half a century and the current CATs based admissions systems for hundreds of state funded schools since the inception of the Academies programme, the general intelligence factor ‘g’ is a concept about which much heat has been generated.

Many left inclined educationalists still begin any discussion in this area with an IQ denial statement of some form. I am happy just to accept the fact that cognitive ability, regardless of arguments about its philosophical significance, can be readily measured by relatively simple, albeit increasingly sophisticated, tests and that their results have very high correlations with life outcomes and especially with performance in the education system.

Are there other sorts of intelligence? We certainly don’t all think the same way, which is why standardised cognitive ability tests have three sub-test components: verbal, quantitative and non-verbal. Although most individuals score similarly on each component, some do not, revealing differences in cognitive strategies and abilities in the three areas.

Howard Gardner went much further with his theory of multiple intelligences (1983), which is an attractive, popular and frequently quoted rebuttal of the concept of general intelligence. Gardner proposes seven distinct and independent ‘intelligences’ with two, ‘linguistic intelligence’ and logical-mathematical intelligence’ roughly corresponding with the qualities measured by Cognitive Ability Tests (CATs). The other five, although claimed to be independent by Gardner, in fact correlate to a greater or lesser degree with the first two. To the extent that they correlate highly, they are more clearly understood as components of ‘g’. Those that correlate more weakly seem to be more like ‘talents’; further examples of the rich diversity of human variation to be encouraged and celebrated, but not so strongly predictive of general exam performance and broader life outcomes.

Chapter 12 of ‘Bad Education – Debunking Myths in Education’ (2012) edited by Philip Adey and Justin Dillon, addresses the myths of both ‘intelligence fixed at birth’ and ‘multiple intelligences’.

The arguments in this book are based on the validity of general intelligence as set out by Adey and others but with the insistence that although resilient, such general intelligence is plastic and that its development should be the priority of all good schooling. ‘Plasticity’ is a precise engineering term relating to properties of materials. A ‘plastic’ material is one that can be permanently deformed (shape altered) by the application of an external stress. The opposites are ‘brittle’ (cracks under stress) and ‘tough/resilient’ (does not break under stress or permanently deform – may spring back). In this book ‘plastic intelligence’ means that cognitive ability and level of cognitive sophistication can be permanently changed through perception/experience combined with the right sort of teaching/learning.

‘Plastic’ general intelligence is a significantly different concept to ‘fixed intelligence conferred at birth’. It opens the door to the development of the intellect of all children (and indeed adults) through good quality education. However much education practice commonly believed to be ‘good’ is in fact ‘bad’ and does not result in cognitive growth. That is a theme that runs throughout the book.

Steven Pinker, Professor of Psychology at Harvard University wrote in his book, The Blank Slate (p. 149, Pinker, 2002):

I find it surreal to find academics denying the existence of intelligence. Academics are obsessed with intelligence. They discuss it endlessly in considering student admissions, in hiring faculty and staff, and especially in their gossip about one another. Nor can citizens or policy makers ignore the concept, regardless of their politics. People who say IQ is meaningless will quickly invoke it when the discussion turns to the execution of a murderer with an IQ of 64, removing lead paint that lowers a child’s IQ by 5 points, or the presidential qualifications of George W. Bush.

The [false] academic arguments of the IQ deniers come down to the complex statistics of multi-variable correlations called factor analysis.

In our modern society, with its rich literary, scientific and technological culture, proficiency in manipulating complex information and problem solving within this culture correlates strongly with putting food on the table, a roof over the head and maximising any surplus wealth that can be acquired. Hunter-gatherer societies clearly produce different correlations. Value judgments about the qualities needed to prosper in different cultures that so obsess sociologists seem to me to be increasingly pointless as global capitalism spreads, promoting increasingly commonly shared concepts of meritocracy and opportunity founded in the modern commercial and industrial world.

There is no dispute that scores on cognitive ability tests correlate strongly with exam results and future life outcomes in our society and culture. This book is about recognising the plastic nature of intelligence and the opportunity it creates for enriching the lives of individuals and the quality of society through the promotion of cognitive development through national education systems.

From Section 1.10

The passing of the 1988 Education Act brought about the next major change in the assumptions of [GCSE] grading. Schools soon had to compete in league tables based on the proportion of pupils in the school achieving 5+A-C passes at GCSE. Following the election of New Labour in 1997, any school that failed to achieve 25 percent 5+A*-C (the first floor target) was deemed to be failing by definition, regardless of the average cognitive ability of its intake.

This sought to deny any direct link between pupil cognitive ability and exam performance and placed responsibility for obtaining C+ GCSE results squarely with the school. Failure to obtain at least a C grade at GCSE was at first blamed on ‘low expectations’ on the part of teachers, which conveniently fed into the ‘SAD’ thesis (Social Advantage and Disadvantage conferred at birth that traps people within their social class), with schools and teachers accused of fulfilling the role of the educational jailors of pupils locking them into the class defined prisons they were born into.

This vilification of comprehensive schools serving areas of social and economic deprivation and their teachers has persisted ever since with all attempts at a defence being condemned as ‘making excuses for failure’. So-called evidence for this alleged failure was regularly churned out in the form of the persistently poor results of pupils from poor communities compared to their more affluent peers (the ‘attainment gap’). That there might be significant differences in average cognitive ability between school admission cohorts was never considered, investigated or controlled for.

From Section 2.1

Despite the best efforts of the exam boards and the expensive courses the examiners ran for teachers to tip them off on what the exam questions would be (exposed by the Daily Telegraph in 2011), there will always be a proportion of pupils of lower cognitive ability in comprehensive schools that persist in showing little promise of achieving the C grade target by the end of Year 11 no matter how much the C grade is devalued, how hard they try or however rigorously they are coached and crammed. The shape of the intake ability distribution dictates that the lower the mean cognitive ability in the community served by the school, the greater the number of sub-C grade ‘no hopers’ there will be in a school.

Far from being a sign of failure on the part of anybody such a continuously variable exam performance should be the outcome of any sound exam system. The 2010 coalition government, like its Labour predecessor, wrongly persists in regarding this as an ‘achievement gap’ related to social disadvantage that has to be closed, rather than an outcome attainment spectrum consequent upon predicable natural variation in cognitive ability. The ‘achievement gap’ between the bottom and the top of a normal bell curve distribution cannot be ‘closed’ without lowering overall standards and inflicting damage on the education system. This is a major theme running through this book.

CAT scores (from Standardised Cognitive Ability Tests used for banded school admission policies) accurately predict the exam results of cohorts of pupils (the larger the cohort, the better the prediction) in all exams that validly test general reasoning ability. The greater the cognitive challenge of particular subjects, the better the prediction. As cognitive ability is continuously variable according to the bell curve, so should be pupil performance in exams. There is therefore no obvious threshold of attainment that indicates an ‘acceptable’, and still less, an ‘expected’ level of performance at any given age.  Part 1 explains from a historical perspective why any average or ‘expected’ level could not in any event be anywhere near the GCSE ‘C’ grade without doing great violence to the assumption of the maintenance of standards over time (1.10).

As for GCSE in the secondary phase, so for Key Stage 2 (KS2) tests in primary schools. As there can be no ‘expected level’ in a continuous distribution, the government ‘expectation’ of at least Level 4 for all pupils has no validity either.

This statistical truth is usually dismissed and scorned as ‘making excuses for failure’. However if the fact of continuously variable general intelligence is recognised the focus of schooling can and should change to raising the cognitive ability of the national pupil population as a whole. In most of this book I describe this process in terms of ‘cognitive ability’ that can be measured with CATs tests. In 5.5 I refer to a parallel discourse favoured by Guy Claxton that develops the concept of ‘learning capacity’. To avoid confusion, it is important to note that for me, ‘cognitive ability’, ‘plastic intelligence’, and ‘learning capacity’ are all closely related and are all promoted by the developmental approaches to teaching and learning that are described in Part 5 of the book.

The solution to unacceptably low standards towards the bottom of the bell curve is not to try to ‘close the gap’, or mathematically impossible attempts ‘to bring the poorer achievers up to the average’, but to raise educational standards for all pupils of all abilities, while 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 some specific hurdles that are a feature of modern life, but not for the purpose of reducing the ‘gap’ between the highest and lowest achievers.

From Section 4.15

It is appropriate to end this [Mossbourne Academy] case study where we started, with the article co-authored by the then Guardian Education Editor, Jeevan Vasagar on 18 August 2011. This appeared to make a number of assumptions:

That Mossbourne has a similar intake to its [allegedly failed] predecessor, Hackney Downs school.

It hasn’t. Mossbourne’s intake typically has had an average cognitive ability of 100, the national average, whereas Hackney Downs’ was likely to have been around 84, similar to the 100 pupils refused admission to Mossbourne and whose appeals were turned down. 84 represents the 14th  percentile, meaning that only the lowest 14 percent of the national population had a cognitive ability lower than the average cognitive ability of the Hackney Downs pupils. Not only is Mossbourne’s intake massively more able than that of its predecessor, it is significantly more able than the Hackney average and that of many of the other secondary schools that participate in the same system of banded admissions.

The school’s success [Mossbourne Academy] does not depend on its selective intake.

 It most certainly does. Mossbourne selects on cognitive ability regardless of social deprivation. Mossbourne attempted to admit 60 percent of its pupils from within 1000m of the school, an area of severe social deprivation, but it selected only a small proportion of the pupils that live there. Large numbers of those with lower CAT scores are rejected. In other words the socially deprived surroundings of Mossbourne Academy produce children with a much lower than average cognitive ability (but still covering the full range) from which Mossbourne selects an intake whose cognitive ability (but not social or economic profile) matches the national profile.

That Sir Michael Wilshaw’s assertions in terms of low-income homes, FSM entitlement and bilingualism are more relevant than the CAT score profile he declines to mention.

They are not. It is cognitive ability that counts. The Guardian articles fail to make any reference to admissions issues like CAT scores and banding that lie at the core of Mossbourne’s success. But then all the print and broadcast media are guilty of the same omission. Less defensibly, this is also true of OfSTED reports.

My Post on Local Schools Network (24 Sep 2014) – The ‘gap’ can’t be closed

Read it here.

Two important articles appeared in the Independent of 23 September. The first is entitled, ‘Even excellent schools ‘don’t help poor kids to catch rich’.

“The gap in performance between poor and better-off pupils is just as high in schools ranked ‘outstanding’ as those labelled ‘inadequate’, researchers have found.

“Even if we improved all ‘inadequate’ schools to the level of those judged ‘outstanding’ we would still have a free-school-meal gap of much the same size as we do today,” the study concludes.

Professor Steve Strand, the report’s author, concludes that successive governments have been wrong to blame “failing schools” for the stark differences in performance between the two sets of pupils, Instead, the reason is likely to be “factors outside the school gates (in the home, wider community or peer groups)”.

This leads to another headline in the same newspaper.

‘Half of all five year olds are not ready for school, research shows’.

” Nearly half of all five-year-olds in England have not reached a high enough level of intellectual, emotional and physical development to prepare them for school, new figures reveal today.

“Just 52 per cent of children have achieved a ‘good level of development’ by the end of reception, according to updated figures compiled by Professor Sir Michael Marmot, one of the country’s leading authorities on public health and social inequalities.

“The figures, which Professor Marmot said were a worrying sign for the future health and wellbeing of the country, mean that only little more than half of England’s five-year-olds could be called “ready” for school on all measures of development.”

The conclusions from the first article are stark. Schools that are judged by Ofsted as good or outstanding are no better at ‘closing the gap’ than the hundreds of allegedly failing schools closed by Labour and Conservative governments and replaced by Academies since the early years of this century.

This represents a colossal waste of public money and assets and a gross historical injustice to thousands of teachers. Henry Stewart and Janet Downs, [who post on Local Schools Network] have produced post after post showing that the costly free market experiment of Academies and Free Schools is not ‘closing the gap’ either.

However, this is the point where the conventional analysis of the left goes wrong, with resort to the arguments that emerge in the second article.

If the children of poorer families are failing to ‘keep up’ or make ‘expected progress’ at school then the blame must surely lie with ‘poor parenting’ for which the solution is state intervention, for example through ‘Sure Start’.

Now helping poor parents to provide well for their children while they try to earn a decent living is fine, worthy and certainly not a waste of public money. However can it result in any closing of the ‘education gap’?

We know the answer to this through extensive studies in America with the ‘Headstart’ programme and here in the UK. The English experience of expensive and essentially social programmes like ‘Sure Start’ has been hugely disappointing in terms of measurable educational outcomes.

Dr Christine Merrell of Durham University Curriculum, Evaluation and Management Centre, responsible for a long-term study into the effectiveness of Sure Start was reported in the Daily Mail of 19 April 2012 as follows. She said: ‘Given the resources put into early years’ initiatives, we expected to see a rise in literacy and numeracy scores in schools. So it’s disappointing that there’s been no improvement’.

You can read a summary of the findings here.

Note the scale of the failure. It is not that there has been only a low level of educational impact, but there has been NO IMPACT AT ALL.

So what is going on? The explanation is that the basic suppositions are wrong. What if there is no ‘gap’? Children are all different in all sorts of ways including their physical and mental development. There is a large variation in the age at which girls start their periods. No-one talks about a ‘puberty gap’.

In the cognitive realm Piaget carried out hundreds of experiments on the cognitive development of babies. While the Piaget doubters love to nit pick about the details, no one denies that there are qualitative stages in the development of infant cognition. Some of this is time-dependent maturation. Should we expect all infants to develop at the same rate? Of course not.

Many on the left hate the fact of ‘Bell Curve’ variation that occurs between all individuals and which is part of the driver of evolution that is therefore deeply rooted in the Earth’s living systems.

But we are talking about developmental correlations with poverty and social deprivation. No one should be surprised about these. See my post here and here.

I strongly agree with Philip Adey’s summary at the end of his chapter in his book.

Adey P & Dillon J (Edited 2012), Bad Education, Open University Press

“The persistent correlation between different types of ability shows that a hierarchical model consisting of special abilities underpinned by a general intellectual processor offers by far the most plausible structure of human intelligence.

There is substantial reason to believe that students’ general intelligence can be advanced by appropriate curriculum intervention. Far from general intelligence being a millstone around educator’s necks, once one accepts that it is modifiable it becomes the great educational opportunity.

The main function of the education process from nursery school – maybe as far as first degree level, should be to develop students’ general intelligence.”

For this to take place investment has to be targeted onto the right kind of educational, not social intervention.

Educational underperformance is not rooted in social inequality but in the quality of schooling. In this regard Wilshaw and Gove are right. Where they are disastrously wrong is in thinking that marketising the education system promotes the right kind of educational intervention, when it does the exact opposite.

Marketisation is a vigorous, continuous generator of perverse incentives and flawed ‘common sense’ distractions, that results in the wrong kind of behaviourist rote learning and cramming directed at meeting ‘floor targets’ and other invalid and wholly arbitrary performance indicators. If anything, it is making our children dimmer rather than brighter.

So if we are concerned about the development of cognition in our children then we should be researching and developing pedagogy that has been proven to promote general cognitive development.

Will this ‘narrow the gap’? No, because children of all abilities and developmental stages will benefit from a developmental pedagogy. Will it benefit the children of poorer parents? Certainly, who knows how individuals may respond when stimulated in this way? There is no fixed ‘educational potential’ for any child.

But we also need to be much more relaxed about the facts of human variation. We must vigorously assert the entitlement of all children for their individual development to have equal priority regardless of the stage they are at, in a comprehensive school system that is not corrupted by marketisation.

So how has this misunderstanding that even afflicts to the top statisticians at the DfE come about?

The first reason is that statisticians are not educationalists. If you ask them the wrong questions it is not their fault if they give you the wrong answers.

The second reason is that there is a strong correlation between socially deprived communities and low mean cognitive ability. There is also a even stronger correlation between cognitive ability and school attainment.

The latter is a causal link. Put simply, brighter kids do better at school.

The former is not causally related to school attainment. It is just that socially deprived communities contain high proportions of children with lower cognitive abilities.

However it is low mean cognitive ability that is the main factor that depresses the attainment of children from socially deprived communities, not necessarily the associated social deprivation.

This fact points to the only effective way to improve the life chances of all pupils including those that suffer poverty and social disadvantage. We need a genuinely comprehensive school system that prioritises inspiring and developing all pupils of all abilities from all social class backgrounds and so making them all cleverer, as well as knowing and understanding more stuff.

That is not happening under our marketised education system and it is a very serious problem.

Read my book!

This entry was posted in Blogs. Bookmark the permalink.

2 Responses to Closing the Gap

  1. Pingback: School improvement is reducing social mobility | Roger Titcombe's Learning Matters

  2. Pingback: Is effective teaching taking place in Academies? | Roger Titcombe's Learning Matters

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s