The ethnicity ‘attainment gap’
The Education Policy Institute (EPI) has produced a table showing ‘attainment gaps’ for ethnic groups represented in the English education system. The ‘gap’ is expressed in terms of months behind (-) or ahead (+). I reproduce this together with corresponding GL Assessment Cognitive Ability Test (CATs) data from a 2009/10 report (see p 10).
You will not find this report or any other such data on the GL Assessment website. They don’t like talking about ‘intelligence’ or admitting that their CATs are intelligence tests. The following reproduces the EPI ethnicity table. The second column is their mean 2018 ‘attainment gap’ (+/- months). The third column is the GL Assessment data mean non-verbal CATs score (on the IQ scale with mean = 100, SD =15) for that ethnic group and the fourth column is that score expressed as a percentile (percentage of the population with that score or less). I have omitted EPI data for which there is no CATs score.
Gypsy/Roma -34.1 90 25th
Traveller Irish -28.9 90 25th
Black Caribbean -9.3 95 37th
White & Black Caribbean -6.7 98 45th
Any other Black -3.9 96 39th
Pakistani -0.5 95 37th
White British 0.0 101 53rd
White & Black African 1.2 101 53rd
Any other White 2.0 102 55th
Black African 2.3 94 34th
Any other ethnic group 2.6 101 53rd
Any other mixed 4.7 102 55th
Bangladeshi 5.4 97 42nd
White & Asian 9.1 104 61st
White Irish 9.6 100 50th
Any other Asian 10.6 102 55th
Indian 14 .2 100 50th
Chinese 24.8 112 79th
When you do the statistics by plotting a scatter graph of the attainment gap against the mean non-verbal CATs score for each ethnic group you get a correlation of 0.85. This is very high. It means that the ethnicity attainment gap measured by EPI is highly predicted by the non-verbal ethnicity CATs scores that include pupils from every type of school (selective and comprehensive) and location (affluent and poor) that have CATs data. Note that compared to the Verbal Reasoning and Quantitative CATs scores, the Non-Verbal scores are the ‘purest’ indicator of general reasoning ability (IQ) and therefore the least affected by ‘coaching’, language or cultural issues. Anybody with Excel can do this analysis. It is in the GCSE maths syllabus at foundation level. If you don’t like my conclusions then please process the data yourself and tell me where I went wrong.
Back in the 1990s all Cumbria pupils took CATs in October of Y7. The LEA produced an annual scatter chart relating GCSE mean points score to the corresponding mean Y7 CATs score for each school. The mean intake CATs scores ranged from 85 (16th percentile) to 108 (70th percentile). It was obvious to anyone familiar with the Cumbria demographic that school GCSE results were closely linked to school CATs scores and that the poorest areas had the lowest CATs scores (and GCSE results) and the the most affluent had the highest. Moreover the correlation was always about 0.85.
A correlation of 1.0 indicates perfect predictability and we can see that there are some interesting anomalies in the general ethnicity pattern. It is reasonable to hypothesise about the possible causes of these, but first we have to address why ‘intelligence’ is a dirty word for many on the left. This position is generally held by those that believe the following.
- That intelligence is a flawed concept used by racists to smear non-white ethnic groups as inferior. The ethnic IQ data actually provides the evidence to destroy the arguments of the racists. First, there are many non-white ethnicities that have greater mean IQs than whites. Second, the assertion that ethnic interbreeding ‘weakens the Aryan stock’ (pure Nazi fascism), finds no support in the IQ data of mixed race individuals. On the contrary, there is support for the concept of ‘hybrid vigour’ from the intake CATs scores of inner London schools. My study of Mossbourne Academy in my book reveals a mean intake CATs score for Hackney, of about 97 (42nd percentile), which is much higher than in many northern ‘attainment gap’ towns with similar levels of social deprivation where there is much less inter-racial mixing.
- That by accepting that intelligence is a rationally sound measure of ‘reasoning ability’ this implies that it is fixed at birth and defines the intellectual potential of individuals for life. This is just plain wrong. Intelligence is plastic, but to raise it does need the right kind of co-operative, developmental approach to learning, rather than ‘instruction’ with the obedience of the conscripts enforced by ever harsher sanctions.
- That all humans must by definition have equal intelligence and that differences in attainment between ethnic groups and social classes are caused by social inequalities and discrimination. This is contradicted by mountains of IQ data for different ethnicities, groups and social classes, but rather than engage with what the data really do reveal, many on the left prefer to close their minds and assert that intelligence is a flawed concept and refuse to engage with anyone that mentions the word. For example, a recent twitter post drew attention to a Special Educational Needs programme in Australia that claimed to boost the IQ of children with severe learning difficulties. What surely should have been recognised as potential good news prompted an avalanche of comments about eugenics and cranial callipers from people that had not even read the article.
What is truly astonishing is that the concept of intelligence is so universally accepted in all contexts except education, where it is obviously of most relevance. Here is an extract from a Sunday Times article of March 2017.
People exposed to high levels of leaded petrol as children are still suffering from lower intelligence 30 years later, according to the largest study of its kind. Since the 1970s, lead in petrol has been phased out across the world amid concerns that it affected health. It was not until 1999 that it was finally removed from petrol pumps in the UK. While scientists in the US have estimated that removing the additive has raised average IQ by almost five points, establishing the link between cognitive decline and lead has been difficult, in part because those most exposed to lead are often in lower socioeconomic groups.
The last sentence is interesting because although the main thrust of the article would now be unlikely to raise even an eyebrow, let alone an explosion of cranial calipers or eugenics based apoplexy, the author declines to mention why the fact that the groups most exposed to lead in the air ‘are often in the lowest economic groups’, causes difficulty in establishing the link with leaded petrol. This is because lower socio-economic groups have lower mean IQs regardless of such exposure. This too is revealed in the GL Assessment CATs data, of which more later.
In the US IQ is taken into account in the sentencing of murderers. In some States, but not all, an IQ below 70 (2nd percentile) might get you off death row. In the gentler UK judicial system, learning disability (low IQ) not linked to the commission of the offence, may result in sentencing leniency even for minor crimes like burglary.
In England many Academy chains have ‘fair banding’ admission systems that use CATs to ensure that the admission ability bands reflect the national normal distribution. It is very clear that the Academy (and LA) schools that get into trouble with OfSTED tend to be those located in poor areas, where IQ scores are lower, but they do not have the protection awarded by banded admissions. Without CATs data these schools cannot prove that it is low intake IQ that limits GCSE attainment rather than poor teaching. How can the NUT support these schools if they deny the data source that proves their case and refutes the commonly deployed ‘excuse for failure’ allegation?
While it may be risky to discuss intelligence in the NUT corner of the school staff room, (although you might get away with ‘bright’ or ‘more able’) this is not the case with Educational Psychologists, where it has always been the main currency of the diagnosis of Specific Learning Difficulties, where otherwise high IQ pupils struggle with reading (dyslexia) or maths (dyscalculia). In the current era of extreme under-funding of SEN, reports by Educational Psychologists have become ever more difficult to obtain, but should you ever come across one you will be almost certain to find an IQ score.
Where I am as one with the NUT is in relation to grammar school selection, but even here the NUT case is diminished by failure to recognise higher mean intake IQs as the reason why grammar schools outperform their neighbouring comprehensives (really secondary moderns) that have been robbed of their higher IQ catchment pupils that have passed the 11 plus IQ selection test. In the absence of such a clear and irrefutable explanation, resort is made to variations of poverty, class or ethnically based discrimination to account for the middle and upper class domination of grammar school intakes. This leads to arguments about unfair privilege that ignore the fact that pupils of all abilities would be better off attending a good comprehensive and that the selective system inhibits the development of cognitive ability of both 11 plus successes and failures.
How to account for ethnic differences in mean IQs
The first point is that there is no need to involve any consideration of the genetic inheritance of IQ. The real differences shown in my table may be entirely down to culture: nurture not nature. The argument would be that some cultures are more education-friendly than others. This is often used to explain the low attainment of poor white communities: blame the parents for not supporting the school, encouraging completion of homework, ensuring good attendance etc. It is also the favoured explanation for the astonishing success of Chinese children: ‘tiger mothers’ provide the stereotype. It is also commonly asserted that the rigid, highly competitive, disciplined, Chinese instruction-based education system is a major factor, resulting in calls for English schools to adopt the same approach. There are serious flaws in both explanations that are revealed when the ethnic IQ data is taken into account. In this article, which has been viewed nearly 3000 times all over the world, is supported by academics of international repute, and is yet to be challenged, I refute the argument that the Chinese education system accounts for the high Chinese PISA ratings. The mean IQ of Chinese students is so high that the PISA rating should be much better, and are in fact held back by the comparatively poor Chinese education system. However, the clincher is the fact that Chinese students, often of mixed race, born outside China do well in their host countries. In the US they are massively over-represented in the elite ‘Ivy League’ universities, despite attending the US school system that performs poorly in PISA. All of which points to a genetic contribution to this success.
The following argument is based on sexual selection as a mechanism by which culture (memes) can get into the genome in only a few generations. It is based on how the sexual preferences of females may favour intelligent ‘geeks’ (wen) over dim ‘alpha males’ (wu). The highest status and sexually ‘fittest’ females select and mate with the most intelligent males, resulting in gains in mean ethnic Chinese IQ in just a few generations. The biology of the argument is sound.
In my headship school a parallel pattern emerged in the election of the joint chairs (a boy and a girl) of our school council. The social structure of the KS4 population in secondary schools is characterised by boys’ and girls’ popularity hierarchies based on sexual attractiveness, with peer group status depending on who can gain the regard of the highest status individuals of both sexes. We instituted student elections for the joint chairs of our school council. In other schools these would be the ‘head boy’ and ‘head girl’ roles appointed by the headteacher. Our school council had real responsibilities with students trained in the running and disciplines of formal meetings. See this article for details.
This resulted in the development of both confidence and rhetorical ability. Candidates for the joint chair positions were required to issue statements and campaign for support. It was a secret ballot in which school staff did not interfere. In the early years there were some unfortunate winners with qualities more related to former peer group hierarchy status than ability. This was duly noted by the student electorate such that in subsequent years it was ‘geeks’ that were nearly always elected. This not only brought about impressive qualities in the general functionality of the school council, but all students of all abilities were positively affected. We were sure that this resulted in cognitive gains with life changing consequences, not just for individual students, but for the entire culture of the school in which ‘geekiness’ became respected such that peer-peer bullying was judged by OfSTED to have been completely eliminated.
So where is the UK in terms of the health of our national social esteem hierarchy? I fear that we have for some time had a celebrity admiring culture that is unlikely to promote any cognitive improvement in the genome through either nurture or nature. This culture gave rise to appalling popular ‘reality’ TV shows like ‘Love Island’ and ‘The Jeremy Kyle Show’ together with depressingly shallow social media celebrities able to monetise their millions of ‘likes’.
Then there are other escalating social epidemics like knife crime and its association with drug culture and the exploitation of ‘in care’ children in the out of control criminal ‘county lines’ network. This latter appears to be based on exploiting juvenile drug runners from children’s homes by getting them addicted to high strength cannabis and then controlling them through cannabis ‘payments’ to satisfy and feed their addiction. My headship experience convinced me that the benign ‘happy hippy’ view of cannabis is completely and dangerously wrong. In the 1990s my school had only a handful of cannabis dependent pupils, but the consequences were always disastrous. I recall devoted and distressed single mothers in my office, in tears recounting how cannabis had damaged their children changing their personality, destroying all structure in their lives and exposing their caring relatives to abuse and assault.
While drug gang leaders murder each other for control of their ‘businesses’, the women gang members compete for the attention and protection of the most violent and feared gang leaders. There is no likelihood of any emerging ‘wen’ culture here. It is hard to see how such social conditions can be mitigated. However I do suggest an approach that could work.
The attainment gap
This flawed concept dominates the education debate. It is based on the assertion that northern working class schools are failing their ‘socially disadvantaged’ pupils. The argument is that such pupils judged to be ‘high attaining’ on the basis of their SATs scores underachieve at GCSE. While this is true, where CATs data are available it is clear that SATs scores for ‘disadvantaged’ pupils are inflated and that the pattern of GCSE attainment is actually as predicted by the lower CATs scores. So there is no ‘attainment gap’. The ‘killer fact’ revealed by EEF research is that a similar size ‘attainment gap’ exists in all types of schools with all OfSTED judgements from ‘outstanding’ to ‘failing’, proving that the real issue lies in the depressed mean IQs associated with ‘social disadvantage’. This truth is so unacceptable to the establishment view that it is dismissed out of hand despite irrefutable evidence that it is true.
John Mountford and I have researched a number of schools where SATs, CATs, social disadvantage and SEN data are available and it confirms our argument and refutes the claims that northern schools are discriminating against disadvantaged pupils. This is something that you might think that left inclined educationalists would welcome, but not apparently if it requires the acceptance of IQ-based evidence.
I have written many articles on this subject, which I commend for study. The evidence and the conclusions that flow from them can of course be ignored. What cannot, is the fact that it is only through examining IQ/CATs data that the truth emerges and the dangerous fallacies that stand in the way of really improving the life chances of disadvantaged children can be exposed.
I welcome comments/corrections/criticisms in response to my articles. In the interests of informed debate I always publish comments unless they are abusive.