IQ and Higher Education
Readers in the UK may have seen recent press reports about a controversial article by Bruce Charlton, an evolutionary psychologist at the University of Newcastle. Charlton points out that average IQ differs in different social or occupational classes (e.g. doctors or lawyers have higher IQ than casual labourers), and that in consequence, if IQ is relevant to higher education, we would expect participation in higher education also to vary according to class. Predictably, ritual curses and denunciations have rained down on Charlton’s head.
I wanted to read Charlton’s article, but found it more difficult to find than I expected. The press reports suggested that it appeared in Times Higher Education (the former Times Higher Educational Supplement), but on tracking down the relevant issue I found a report about Charlton’s article, but not the article itself. The article is however available as a Word document on the THE website. (See the right margin of the webpage here). I make a few comments of my own below the fold.
Probably most readers will agree with the broad thrust of Charlton’s article, but there may be a confusion between the IQ of parents and that of their offspring. Charlton seems (as far as I can see) to assume that the mean IQ of applicants to higher education is the same as that of adults in their parental social class. This is not generally the case. The correlation between the IQ of parents and offspring is only about .5, which implies considerable regression towards the mean. The IQ of the offspring of parents with IQ of, say, 130 will on average be lower than 130, while that of parents with IQ of, say, 85 will be higher than 85. There is a difference of up to 40 IQ points between adults in the highest and lowest occupational classes (depending on the classification used), but only about 15 to 20 points between children from those classes. (For some data see Anastasi, chapter 15.) The difference in average IQ between social classes is kept roughly constant by social mobility, as the dimmer children of the higher classes tend to fall in the social scale and the brighter children of the lower classes tend to rise. (See Mackintosh, pp. 144-8). This does not invalidate the main point of Charlton’s article, but it may affect some of his specific quantitative comparisons.
I think it may also be unfortunate that Charlton describes the present system of entry to higher education as ‘meritocratic’. Entry to publicly funded higher education should not be seen (primarily) as a reward for past achievement, or a badge of ‘merit’. The proper criterion for entry decisions is how far an individual can benefit from the course of study concerned. In general, individuals who have struggled at school are unlikely to benefit from higher education at all. If applications for a particular course of study exceed the number of places available, those applicants should be chosen who will benefit most from the use of scarce resources. It is the scarcity of high quality resources that justifies the selectivity of the ‘elite’ universities. One would not expect an haute couture seamstress to stitch potato sacks, and one should not expect a Fellow of Trinity College Cambridge to teach mediocre students. Such students would not derive the greatest benefit from the teaching, and indeed the teaching would probably not be the best available for such students.
I am assuming in all this that higher education is by definition at a more advanced and demanding level than that of ordinary school education. It is not to be confused with post-school education at a similar level to that of schools, such as is provided by Further Education Colleges in Britain. This may be admirable in its own way, but it is not higher education.
References:
Anne Anastasi: Differential Psychology, 3rd edn., 1958
N. J. Mackintosh: IQ and Human Intelligence, 1998
In practice, however, the British higher education system also serves the same purpose as the French grandes ecoles entrance system in basically picking out those who are destined for high-level managerial and professional roles. The original and rational guiding principle of selection may have been “how far an individual can benefit from the course of study concerned”, but employers assume, probably correctly, that people who were allowed entrance to the most demanding courses at the most demanding universities are the most able. There is also a (probably quite accurate) perception that students at the older and more prestigious universities have to do substantially more and more difficult work to get the same degree level. My uncle, a graduate recruitment manager for a large software firm, refuses to interview anyone who went to a university without a good reputation, and I’d assume he’s pretty typical in that respect. The result is that, from the point of those being selected, getting a place at a good university is an early form of job competition.
The British upper class, who are desperate to get their children into these universities, are a little too proud to go in for blatant American-style donations in return for “legacy” places. Instead they rely on public (in the UK, this means private and prestigious) schools to shoehorn little Arthur, who’s jolly good at rugby but always found Maths and English rather confusing, into Oxford or Cambridge or LSE by any means possible. And – here’s the hole in Mr Charlton’s argument – these schools are very, very good at doing so. Some of the people on my undergraduate course at one of the aforementioned universities were given practice interviews twice a week for the better part of a year by their schools. Most hired private tutors to cram in all the knowledge they were too uninterested or slow to pick up class. Most of the more unpopular and obscure courses are filled with public school alumni, because the schools know the system well enough to predict which course will be undersubscribed and their parents are so desperate to get their children into the right university that they don’t mind what they study. Material Engineering and Theology are particular favourites.
The British higher education system is really pretty far from ideal either as a meritocracy or a public education service.
“little Arthur, who’s jolly good at rugby but always found Maths and English rather confusing” is unlikely to get into any Oxford or Cambridge college, and even less likely to get into LSE (since I doubt that rugby is valued highly there). Your comment seems to reflect prejudice. Not that I am claiming the selection processes are perfect. Selectors could (and sometimes do) make some allowances for the disadvantaged education of some applicants when judging their potential.
Perhaps I was being a little harsh: most of the students pushed into Oxbridge by public schools are probably above well average by national standards. Without public education they’d probably still manage to get into a half-decent university. But Oxford/Cambridge/LSE? I doubt it. I realise that tutors often try to see potential in students from disadvantaged schools, but, being realistic, it’s not at all easy to judge potential without judgement being swayed by the prospective student’s prior knowledge and presentation skills. Add to that the intimidation factor of an Oxbridge interview, which is much greater for someone outside the “system” and who hasn’t had practice interviews, and it becomes really quite unfair to expect tutors to be able to judge fundamental ability in a twenty-minute conversation.
There is a very useful application relevant to this discussion on Razib’s SciBlogs.
This is written to calculate predicted heights of offspring given parental height, but it is flexible so you can change the parameters and instead of getting height you get the trait that interests you. Substitute any metric you like for height, use reasonable parameters for average and SD on that metric, and calculate what your kids, or somebody else’s, will be.
Relevant to this post, this can serve as an IQ predictor (e.g. make avg for both parents 100, SD=15, heritability 0.5). For two parents at 145 (+3 SD above population average) the expect IQ is 122.5 = +1.5 SD >population average. Only about 2.25% of children will be above 137.5 (+2.5 SD above pop average).
I did admissions interviews for a Cambridge College once. My fellow interviewer had two prejudices that he described to me, though he didn’t call them that. One was that every applicant was either a musician or a sportsman – and that one must never admit the sportsman.
the only caution i would offer re mike’s point is that there was a reason i used 2 standard +/- interval, the heritability can get weird out to the tails (remember that the tails are fat for most quantitative tarits).
Thanks for covering this article of mine – I agree it was not made at all easy to find by the Times Higher Education, so I posted a copy on a blog:
hing.blogspot.com/2008/05/social-class-iq-differences-and.html
ardian.co.uk/specialreports/lecturerspay/story/0,,2283132,00.html
http://charltonteac
I did not bother mentioning the regression to the mean point for three reasons: 1. The target audience would not have understood it; 2. I wanted to emphasize the neglected point that the varied precision of *definition* of socio-economic status has a much bigger effect on the size of measured IQ differences; and 3. there is an increasingly important element of assortative mating for IQ which also needs to be taken into account.
This was one reason why I gave two estimates for the effect of social class differences – both of which reached the same general conclusion.
But despite my attempts to ‘keep it simple’ my main point, the crux of the article, was still missed – and has been missed here.
The observation of social class differences in IQ is at least 80 years old, and the predictive vaue of IQ is perhaps even older. What was slightly new and a bit surprising in my article, although obvious enough to mathematical types, was the way on which the social class differentials favouring the higher social classes get exponentially bigger as the universities become more selective.
Fat tails or not, the effect is very big at the level of selectivity of Oxford and Cambridge – which may be the most academically selective undergraduate universities in the world because they are world famous, beautiful, have excellent teaching and facilities as well as very good research, are heavily subsidized by the state in various ways – and the fees at ‘Oxbridge’ are exactly the same as the fees at the other 100 plus UK universities. So there is no financial reason *not* to go there, Oxbridge is just as cheap as the lowest-ranked local university.
For the past 8 years the UK government has been obsessed with trying to mirror the proportion of social classes in ‘Oxbridge’ and the national population – this is government policy, there is a government agency called OFFA tasked with monitoring and implementing this goal, there are sanctions and shaming used against Oxbridge when each year shows no ‘progress’ on this policy.
This policy is motivated by a class war agenda obvious in the governments untruthful and misrepresented response to my article. The universities minister Bill Rammell called it ‘drivel’. Since the article was published, denial of class differences in IQ has been suggested by Rammell as a belief around which the government and teachers union could form an alliance.
http://education.gu
The UK class war agenda is fuelled by the belief that it is snobbery, or private education, which explains the relative fewness of low social class groups at Oxbridge. But IQ considerations make clear that no such explanation is necessary, and indeed there is no evidence that snobbery and private education are responsible.
In fact the best evidence from longitudinal IQ studies (eg. by Daniel Nettle, and by Ian Deary’s group) suggests that the UK has been an IQ meritocracy for many decades. In other words, social class is mainly determined by IQ (which is mostly inherited but varies between siblings), and not by inheritance of class from parents.
Many people state categorically that private schools get people into selective universities more effectively than state schools – and they may indeed have such an effect – but the fact is that nobody knows. There are no IQ-controlled studies of the effect of private education in the UK (at least, I haven’t been able to find any) – and clearly the selectivity of some elite private schools will have a huge effect on admissions to the most selective universities.
Although I would be surprised if private schools did not have some beneficial effect on education outcomes, strictly speaking whether or not the prinate schools generate added value to IQ – and if so how much – are still open questions.
By contrast, the effect of class differences in IQ on university admissions is about as definite as these things can be.
Bruce, in your reading, have you ever come across any evidence on the educational and social backgrounds of Oxbridge dons? I ask because I have often wondered why the generation now approaching retirement, many of them from state Grammar schools, and the children of schoolteachers, minor civil servants and the like, should be assumed to be interested in being snobbish about the youngsters they admit. (My allegation about their schooling and parentage is wild guesswork of course – I just wonder whether anyone knows.)
Razib is correct that the tails are raised for IQ. I don’t know about other traits. My decidedly PC Psych 1 prof taught me this and attributed it to assortative mating, unintentionally admitting heritability (possibly leaving environment in play).
In any case, using the Razib Heritability/Regression to the Mean Application is useful, as long as we recognize that it may, because of raised tails, UNDER represent the degree to which parents at the extremes of the scale may be represented at that level by their children (i.e. their kids will be above (top end) or below (bottom end) the statistically predicted level). The error favors Dr. Charlton’s argument.
Razib is correct that the tails are raised for IQ. I don’t know about other traits.
i think it applies to most quant. traits. the issue is that the guassian approx. starts to be wack on the margins because of prevalence of gene. architectures which will deviate from the regression line a lot (think gene-gene interaction, dominance, etc.). obviously a portion of the below 70 IQ set are mendelian mutant tards.
The article in question was a big diservice to the IQ community because:
1. it grossly misrepresents the facts. Any study with big N I care to remember presents the correlation IQ – parental SES as no larger than .35 (see a paper in press in Learning and Individual Differences with a huge German sample; also another one by Deary from around 2005). The German paper r = .35, the Scottish paper r = .33
This means that 87% of the variance in IQ is not explained by parental SES. Arthur Jensen’s main effort in his books has been to insist on this point and to show that IQ tests – quote from memory – “help seeing through the veneer of social class”.
The presentation of IQ as closely related to parental SES is not only false but also can easily be explained in Marxist terms, by saying that IQ is nothing but a reflection of socially constructed class privilege. The readers of GENE EXPRESSION would not like that…
Although I am a true believer in IQ research, I admit I was angry when I read the article in question, for the reasons above.
I have a question about regression to mean here. I remember reading (I think in A.R. Jensen’s g Factor) that children of black and white parents regressed towards their respective population means; hence the expected IQs of the children of a black couple with a midparent IQ of, say 130, would be significantly lower than the expected IQs of the children of a white couple with a midparent IQ of 130. Would this also apply to the various English social classes, with the children of high and low SES parents regressing towards different means?
For UK, Rindermann, 2007 indicates the population mean at IQ 102. I am not aware of big racial differences in the UK – the population is overwhelmingly white, so most folks in the lower classes would be white as well. The kids of both rich and poor would gravitate towards the 102.
Please note that although the parental IQ – child IQ correlation is .42 (Mingroni, 2007), the correlation between one’s IQ and one’s destination SES is only about .34-.5 (Stremte, 2007). In other words, plenty of those in the upper classes are not that smart, and plenty of those in the lower classes are not that stupid.
This lack of perfect correlation between one’s IQ and one’s destination SES explains the difference between the .42 (parent IQ-child IQ correlation) and the .33 – .35 (parent SES – child IQ correlation).
Actually, the fact that it is as big as .33- .35 seems to suggest that parental class inflates children IQ scores independently of genes (but later on the effect might vanish – factor in the declining contribution of shared environments with age)
To clarify my question above, I wasn’t referring to the racial make-up of the UK. I was under the assumption that regression to the mean was a statistical/probabilistic phenomenon. What I meant was that if you define groups (black or white, high SES or low SES, etc.), effects within the group will tend to regress towards the group mean. I was asking whether breaking the UK population into class groups would make a difference? Say the UK average IQ is 100 and with and SD of 15 and the correleation of parental and child’s IQ was 0.5. If you picked a random couple in the UK with a midparent IQ 2 SD above the mean (130), you’d expect their children to be 1 SD above the mean (115). However, here we’re not talking about random couples with mid-parent IQs of 130, but couples from different social strata, each with mid-parent IQs of 130. If, say, high SES families average IQ 115 with and SD of 10 and low SES families average 90 with an SD of 10, assuming the correlation between parental and child’s IQ in both groups is 0.5 as above, would the children of the high and low status groups with mid-parent IQ’s of 130 regress to expected IQs of 122.5 and 110, respectively? If so, this would presumably make it likelier that the children of the higher SES parents would attend a good university?
PG, i think it would make a difference and you should update the priors a bit. but the extent to which depends on the extent of poplation structure within the society…ie, what’s the extent of assortative mating by class? assorative mating tends to increase heritabilities if you take the assortation into account from what i remember in falconer.
Celia Green expresses it thus: “10 billion pounds has been spent, effectively to lower the average IQ of the undergraduate population.”
“The presentation of IQ as closely related to parental SES is not only false but also can easily be explained in Marxist terms, by saying that IQ is nothing but a reflection of socially constructed class privilege. The readers of GENE EXPRESSION would not like that…”
Or we may point out that Marxism merely makes sweeping statements, rather than proving anything, and considers it’s statements final and unfalsifable. Marxism is impervious to any counter-argument too, so the total corpus of academic economic argument – which is dismissed as entirely bourgeois – is ignored. It is like Intelligent Design argument was merely to dismiss Darwinism as atheist. As dumb as they are ID advocates produce better arguments than that.
In any case the statement IQ is nothing but a reflection of socially constructed class privilege is falsifable and falsified if IQ does not correlate exactly with social class, which it doesn’t.
It could well be false even if it correlate exactly, as the cause of higher IQ in higher social classes could be entirely genetic. We would leave that to scientists to discover and discuss, and let the ever declining cult of Marx play in it’s sandpen.
Glad to have prompted this discussion.
On regression towards the mean, someone asked whether people would regress towards their ‘class’ mean. I don’t know of any data, but on theoretical grounds one would expect there to be some genetic influence from grandparents, etc, not fully reflected in the correlation with parents.
I should have said in commenting on Whiskydrinker that he is probably right that some courses at Oxbridge are softer than others. Rightly or wrongly, Geography always had that reputation. I don’t know about Theology. I note from the prospectus here
http://www.theology.ox.ac.uk/prospective_students/undergraduates/prospectus/course_outline.htm
that students studying Theology as a single subject have to take an option in Greek, Hebrew, or Classical Arabic, which doesn’t sound all that soft.
“Greek, Hebrew, or Classical Arabic”: tougher than Social Science, then?
Bruce Charlton: While I’m not disputing that the “lower” class generally has a lower IQ than their economic superiors, I’m far from convinced that the upper class of financiers, landowners, and managers has a higher IQ than certain sectors of the middle class, especially doctors, academics, and engineers. Do you know of any concrete data on average IQ by occupation?
bioIgnoramus: “I ask because I have often wondered why the generation now approaching retirement, many of them from state Grammar schools, and the children of schoolteachers, minor civil servants and the like, should be assumed to be interested in being snobbish about the youngsters they admit.”
I wouldn’t say it’s down to snobbery at all – it’s down to the fact that public schooling really does produce students with better grades and better presentational skills relative to their innate ability.
DavidB: Theology is not necessarily an easy course to succeed in, merely an easy course to get into due to its unpopularity and in particular the lack of interest from state-schooled pupils. As regards Geography – it does have that reputation, hence the Porterhouse Blues line “We haven’t had a first here for ten years! Even counting Geography!”
whiskydrinker said: “Do you know of any concrete data on average IQ by occupation?”
You could start with the references I gave in my article, perhaps? ;=).
It looks, however, like they’re data on class IQ, not occupational IQ specifically – what I’d really like to see is whether I’m right about certain sectors of the middle economic class having an IQ more comparable to the upper class. I think, and I could be wrong, that the average academic or engineer probably has a higher IQ than the average building site foreman or real estate agent – despite them earning comparable amounts and belonging to the same economic class.
“One would not expect an haute couture seamstress to stitch potato sacks, and one should not expect a Fellow of Trinity College Cambridge to teach mediocre students”
Watch the interview with Steve Binder, the acclaimed director and producer here:
http://www.youtube.com/watch?v=5DMEOSrzGgs
Take note of his story about how he got his start in the mail room of a TV studio.
One should not expect anyone to do anything. However, one should expect everyone to be willing to invest what ever it takes to launch their life’s work, even if that means stitching potato sacks or working in a mail room.