Thursday 31 December 2015

Farewell 2015 and Happy New Year

Milometre events like a new year are of little consequence compared to those moments of history when things really change, but they trigger reflection nonetheless. The progession round our star completed, it is time for a quick roundup of the outgoing year.

For me, with regard to this blog, the major feature of 2015 was that it became usual to read papers on the genetics of human behaviour based on 100,000+ samples showing high heritabilities and low contributions of shared environmental effects. My subjective impression was that a substantial paper was published once a month on either intelligence, personality or psychiatric disorders, all showing that they had substantial genetic components. I sense that the pace of discovery is accelerating, but have no idea how long it will take to be able to predict intellectual levels from genetic data alone.

How does one determine “when things really change”. Wars are the friend of the historian in this regard: they provide key dates around which narratives can be hung. Science is less dramatic, more “drip, drip, drip” than “bang, bang, bang”. However, perhaps 2015/2016 will be seen as the moment when the drips coalesced into a stream.

On a more punctual note, I am reading new papers on the genetics of intelligence, checking back on the links between schizophrenia and violence, trying to review a few books on intelligence, and all this may change when yet another paper gets sent to me by a researcher, which rightly jumps the queue.

Here is the snapshot of blog readership right now, just before going off to celebrate the new year on Bikini Beach. Happy New Year to you all.


Wednesday 30 December 2015

Schizophrenia and violence

I have delved into this issue before, saying that psychiatrists (the missionaries) were torn between trying to normalise mental illness, and (the actuaries) admitting that some psychiatric illnesses have deleterious effects on behaviour, particularly an increase in violence.

Now there are startling new results from a large, representative sample of Norwegians, showing that the rate of violence is about 7 times higher in schizophrenics as compared to controls, and 3 times higher for those with bi-polar disorder.

Genetic and environmental determinants of violence risk in psychotic disorders: a multivariate quantitative genetic study of 1.8 million Swedish twins and siblings

Molecular Psychiatry advance online publication 15 December 2015; doi: 10.1038/mp.2015.184

A Sariaslan, H Larsson and S Fazel

Advance online publication 15 December 2015

Amir Sariaslan, formerly at the Karolinska, and now with colleages at Oxford, is becoming known for delving into the large goldmine of Swedish epidemeolgy and coming up with sparkling nuggets.


These are startling figures, based on excellent and very large samples, showing very high levels of substance abuse for schizophrenia and bipolar disorder, and high rates of violence for both, but particularly schizophrenia. Women seem to get bipolar disorder, men schizophrenia. Immigrants are slightly over-represented in bipolar and over twice as likely to get schizophrenia, a commonly found pattern. Schizophrenia particularly damages employment prospects and marriage prospects.

We observed, for instance, that nearly one in four (23%) schizophrenia patients had ever been convicted of a violent crime, whereas the equivalent prevalence was 11% in patients diagnosed with bipolar disorder and 3% in controls.

Schizophrenia was a stronger predictor of violent crime (r=0.32) than bipolar disorder (r=0.23). Consistent with previous quantitative and molecular genetic studies demonstrating important pleiotropic effects between the examined phenotypes,13, 14, 15 we found that 67% of the correlation between schizophrenia and violent crime, and 51% of the correlation between bipolar disorder and violent crime was attributed to additive genetic influences that were shared between the psychotic disorders, substance misuse, and violent crime. This suggests that the increased risk for violence in these patient groups could largely be attributed to the same genetic factors that simultaneously increased their liabilities to substance misuse and to be diagnosed with the psychotic disorders in the first place. However, we additionally observed a novel finding in that the additive genetic influences that were unrelated to substance misuse explained approximately a fifth (21%) of the correlation with violent criminality in bipolar disorder but none of the same correlation in schizophrenia. In other words, we found support for the existence of disorder-specific genetic effects linking bipolar disorder (but not schizophrenia) to increased violence risk. This implies that aetiological models involving psychotic disorders should not only focus on genetic and environmental factors that are shared between the disorders but also on those factors that are unique to each disorder.




Once again, very little shows up for C, the shared variance of family life we used to think had such an influence.


Here are the two disorders on their own, showing high heritability together with substance abuse, something disorder specific for bi-polar, and unique disorder-specific variance for both.

The authors are circumspect about the implication of their findings. They observe: Clinically, these findings illustrate the importance of risk assessments that consider substance misuse comorbidity, integrated treatments for multiple adverse risks and strong collaborations between criminal justice, substance misuse and mental health services.

If nearly one in four schizophrenics are convicted of violent crime then both the problem and the solution are now much clearer. I had formerly believed what I had been told, which is that violence was so rare in psychosis that it could not be prevented, other than by restricting the freedom of a very large number of schizophrenic patients. These findings suggest that for every schizophrenic patient closely monitored for substance abuse and aggressive behaviour 3 patients will be mildly restricted, a very tolerable ratio. This, in my view, totally changes the Number Needed to Treat calculation in favour of close monitoring. We routinely suggest treatments where the numbers needed to treat are 10 or higher. So long as the coordinated treatment regime were properly managed, few interventions would be show such extraordinary returns on treatment.

I do not want to stigmatise psychiatrists who deal with schizophrenic patients, but they now have a different calculation to make, and new findings to discuss with secure ward hospital administrators and, most of all, legislators. There is a case for preventative medicine in psychosis. This is nothing to do with stigmatisation and a lot to do with evidence-led interventions.

Friday 25 December 2015

Bigger brains are not necessarily more efficient.

An anonymous commentator on the post “Bigger brains cause Flynn effect” said as follows:

So we are supposed to believe that digit span is plummeting, color perception is getting worse, writing is being dumbed down for everyone, our reaction times will soon be so bad we will be unable to play Doom - and also that brain sizes are increasing and this explains the Flynn effect and the effect is genuine rather than hollow?

Michael Woodley of Menie replies:

Simple answer:

Bigger brains are not necessarily more efficient.

Complex answer:

There is a considerable amount of evidence indicating that brain size and general intelligence (g) are actually quite independent from one another. In humans it has been found that the evolutionary lability of brain volume is low, relative to what would be expected of g (Miller & Penke, 2007). This suggests that different selection pressures operate on each trait. The best evidence for this comes from the observation that there has been a general tendency towards decreasing brain volume in human populations over the last 30,000 years (Beals et al., 1984), despite molecular indications of increases in g over the same time period (Rindermann et al., 2012). This finding generalizes in cross-species comparisons involving primates – brain size appears to be relatively phylogenetically conserved across taxa, whereas g (or G as we are talking about species differences) shows signs of having been subject to accelerating directional selection (Fernandes & Woodley of Menie, 2014).

Selection for g therefore seems to be associated with increases in the efficiency of neural substrates, which does not require an increase in overall brain mass, and can even occur in tandem with reductions in brain mass. Several causes of this have been proposed, to my mind however the most plausible explanation is that it is a simple matter of thermodynamics – the boosting of g via a relatively low investment into streamlining the brain’s neural architecture rather than via a high investment increasing of its bulk is the path of least resistance in terms of energetics.

It is important to note also that brain structure does not exhibit the property of measurement invariance across individual humans – large brains are not simply scaled up versions of smaller ones (e.g. Semendeferi et al., 2011). This finding is consistent with the observation that g-loading is only a weak to modest moderator of the strength of the association between brain volume and measures of cognitive ability (Woodley of Menie et al., in preparation), as it indicates that volume-related changes in the structural organization of the brain have substantial effects on specialized cognitive abilities, not just g. It also helps explain the recent ‘downgrading’ of the magnitude of the brain volume-IQ correlation to .24 (Pietschnig et al., 2015), as much of the mass of the brain is associated with functions that are completely unrelated to cognition, but can nonetheless vary considerably from individual to individual.

The relative independence of brain size from g can therefore plausibly account for the co-occurrent secular trends towards decreasing neurological efficiency (suggesting decreasing g) on the one hand coupled with indications of increasing brain mass. I do not anticipate that the larger brains of modern populations are structured in the same way as the relatively smaller brains of a century ago. I predict that most of the mass increase will have been driven by disproportionate investments of grey matter into features of the brain responsible for memory and learning, such as the right hippocampal formation, the right dorsolateral prefrontal cortex, the cerebellum and the septum pellucidum. The net of this will be modern brains that, whilst less generally efficient, are nonetheless better able to develop the sorts of g-independent specialized abilities latent in the Flynn effect


Beals, K.L., Smith, C.L., & Dodd, S.M. (1984). Brain size, cranial morphology, climate, and time machines. Current Anthropology, 25, 301–330.

Fernandes, H. B. F., & Woodley of Menie, M. A. (2014, June). Differences in

evolutionary patterns and rates of general cognitive ability compared to

neuroanatomical indicators in the primate phylogeny. Oral presentation, Evolution

2014, Raleigh, NC.

Miller, G.F., & Penke, L. (2007). The evolution of human intelligence and the

coefficient of additive genetic variance in human brain size. Intelligence, 35, 97–114.

Pietschnig, J., Penke, L., Wicherts, J.M., Zeller, M., & Voracek, M. (2015). Meta-analysis of associations between human brain volume and intelligence differences: How strong are they and what do they mean? Neuroscience and Biobehavioral Reviews, 57, 411–432.

Rindermann, H., Woodley, M.A., & Stratford, J. (2012). Haplogroups as evolutionary markers of cognitive ability. Intelligence, 40, 362–375.

Semendeferi, K., Teffer, K., Buxhoeveden, D.P., et al. (2011). Spatial organization of

neurons in the frontal pole sets humans apart from great apes. Cerebral Cortex, 21,


Woodley of Menie, M.A., Fernandes, H.B.F., te Nijenhuis, J., & Metzen, D. (In preparation). Modest support for the processing volume theory of general intelligence: A meta-analysis of the Jensen effect on brain volume (under review).

Tuesday 22 December 2015

Bigger brains cause Flynn effect?


more mob

Considering that I spent 7 years studying the cognitive effects of cortical lesions sustained in childhood, I should have been open to the possibility that brains might have increased in size over the last century. If I thought about it at all, I think I assumed that any such increase was too small to notice.

Now the barefoot urchin arrives again, with a manuscript from Woodley of Menie, and makes not the slightest pretence before disappearing into the shrubbery with the undermaid, who put up not even a token show of reluctance. How can even those of us with Private Means carry out our Investigations and Reserches when at any moment we could be dismembered by The Mob? It might be prudent to distribute some nourishing meals, as well as the usual Temperance pamphlets. What does the good Woodley adumbrate?

It's getting bigger all the time: Estimating the Flynn effect from secular brain mass increases in Britain and Germany LEARNING AND INDIVIDUAL DIFFERENCES 45:95-100 · JANUARY 2016

Michael A. Woodley of Menie, Mateo A. Peñaherrera, Heitor B.F. Fernandes, David Becker, and James R. Flynn

Secular increases in brain mass over nearly a century have been noted for both males and females in the UK and Germany. It has been argued that such trends may be associated with the Flynn effect. The IQ gain predicted on the basis of these trends is 0.19 and 0.08 points per decade for UK, and 0.2 and 0.15 points per decade for German males and females respectively,indicating a small contribution to the Full scale IQ trends in these countries (2.95% of the German decadal gain and 12.73% of the UK gain). There is also a sex difference in the rates of brain mass gain in both countries, favoring males. Temporal correlations between the secular trend in UK brain mass and European Flynn effects on Fullscale IQ, Crystallized, Fluid and Spatial abilities reveal correlations ranging from 0.751 in the case of Fluid ability to 0.761 in the case of Crystallized ability. The brain mass increase may be an imperfect proxy for changes in specific neuroanatomical structures important for IQ gains. Its small contribution to these gains is also consistent with the influence of other contributing factors. Increasing brain mass is predicted by the life history model of the Flynn effect as it suggests increased somatic effort allocation into bioenergetically expensive cortical real estate facilitating the development of specialized cognitive abilities.

British data

Miller and Corsellis (1977) reported increases in brain mass, utilizing autopsy materials sourced from the London Hospital Pathological Institute amounting to 52 g (from 1372 g to 1424 g) over 80 years (between birth years 1860 and 1940) among their male sample (N=4319), and 23 g (from 1242 g to 1265 g) over the same period for their female sample (N= 3878). Miller and Corsellis admitted into their comparison groups all individuals aged between 20 and 50 years at time of death for whom the brains were not considered pathological (approximately 36% of the brains were excluded on this basis). To determine the secular mass change they simply regressed the mean brain mass of those aged between 20 and 50 at time of death against birth year. The running five-year means for both males and females employed by Miller and Corsellis in their analysis were extracted from their figure.1 (p. 254) and are reproduced graphically here in Fig. 1.


To my eye the pattern is slightly odd, in that it departs from linearity more for women than men (they may have been fed less well) and both sexes do well in the early 1930, a time associated mostly with hardship. It would be particularly good to see the war years in this series, since rationing was said to have improved nutrition.

German studies

Haug (1984) presents the results of 12 studies reporting aggregate brain masses for both males and females (tables 6 and 7, p. 493) collected via autopsy from various pathological and forensic institutes and broken out by age. The studies span the period from 1861 to 1978. Six of the studies involve German-sourced samples. A smaller seventh study (Kretschmann, et al., 1979), not considered by Haug, also reports brainmass means for German subjects. The trend in brain mass across the seven German studies spanning the study years 1880 to 1979 will here be analyzed in order to determine the presence of secular trends within this country. A weighted average of brain masses collected from those with ages ranging from 30 to 49 and 50 to 59 is utilized as the basis for cross-sectional comparison, via regression against study year.The secular trend in height is not controlled in the present analysis, as gains in height and brain mass share variance stemming from a more general secular increase in body mass, which indicates substantial collinearity. Consistent with this Haug (1984, p.492) found a correlation between the two of > 0.8.



Estimating secular gains in IQ from increasing brain mass Jensen (1998, p.326) proposed a method for estimating Fullscale IQ gains stemming from secular increases in brain volume/mass. The method involves converting the gains into a change in standard deviation units by dividing the gains in grams by the reference standard deviations of brain mass for males and females. Based on the assumption that increasing brain mass is boosting IQ this d value must be multiplied by the correlation between brain mass and IQ. The resultant d value can then be multiplied by 15 (the "standard" standard deviation of IQ) yielding IQ points gained throughout the birth and study years covered by Miller and Corsellis (1977) and the German studies respectively.

Correlation between brain volume/mass and IQ

A recent comprehensive meta-analysis has established that the population correlation between brain volumeand IQ is significantly positive at ρ=0.24 (Pietschnig et al. 2015). This value is somewhat smaller than the value presented in a previous meta-analysis (i.e.McDaniel, 2005; ρ= 0.33), however Pietschnig et al. sampled more representatively than did McDaniel. Multiplying the increase in brain mass (in standard deviation units) by this estimate will yield the standardized IQ gain (as per Jensen, 1998). Rushton and Ankney (2009) noted that brain volume is an extremely strong proxy for brain mass, and that, while rarely ever investigated, similar correlations with IQ are obtained when mass is directly estimated instead of volume. Thus the IQ brain volume correlation can be considered synonymous with the IQ brain mass correlation. Pietschnig et al. (2015) noted no significant sex-differences in the strength of the IQ–brain volume correlation; therefore the value of 0.24 will be used in computing IQ gains for both sets of male and female data.

The direct contribution of increasing brain mass to increasing IQ in the UK is 0.19 of a point per decade for males and 0.08 of a point per decade for females. In Germany it contributes 0.2 of a point per decade to male IQ and 0.15 of a point per decade to females.

Males appear to have gained brain mass to a significantly greater degree than females over time, which suggests that there may have existed inequalities in terms of exposure to brain mass enhancing environmental factors that have historically favored males. Alternatively, that females have gained on males on the Raven's Matrix test over the last century (Flynn, 2012), hints at a possible sex difference in terms of how environmental and cultural changes may be assimilated into neuroanatomical and cognitive change.

This paper is interesting, because it shows that the crude measure of brain size accounts for a small but noticeable part of the Flynn effect.  In future more detailed results from brain scans may show more important changes in the constitution of the brain. As the authors say, brain size is a real measure (with a true zero) which is a biologically objective correlate of cognitive ability.

Reading the manuscript in my country hovel I conclude: feed the mob and they will get brighter, in part because their brains get bigger, and probably also because their brains become better wired.

Whether they will burn down the stately homes in rage or queue up patiently to visit them and imbibe their culture is another matter, though the latter is more likely. With these deep thoughts I cast another yuletide log on the cottage fire, and wish you all a Merry Christmas.

Sunday 20 December 2015

Sunday Christmas Carols




To the 12th Century St James’s Church, of the Knights Hospitaller of St John of Jerusalem, as is my wont, walking though the light winds of an oddly balmy December day. The orchestra and choir of former years, all professional musicians, were elsewhere. They had graced us in previous times only because of their friendship with the organist, who in her earlier long career had taught many of them at the Royal College of Music. Now we have been thrown onto our own devices, thought the noble lady herself walked from her house to play the organ for this day only, the rest of the year the organ being played by someone else.

Ritual, of course, depends on repetition, the desire which makes young children beg for entertainments “Again, again” and laugh as if at the first time to rhymes and stories: desire that is of like comfort to the elderly. The love of the old favourite Carols should be easy to indulge, particularly because Christmas Sunday is the one service nominal Christians attend, so any carol will hardly be repeated very often. However, such verities offend true music connoisseurs, who sneer at songs that become popular, and intend to introduce the congregation to new material, more challenging and more relevant to the modern age. Un-named persons chose these new pieces with musicological verve, with mixed results.

Older congregants grew up with the old Carols, and know how to sing them. Few younger worshippers have learnt any, but well constructed songs with good words are easily picked up, and are familiar via televised carol services. About 20 Carols have achieved high recognition, and breach the agnostic barrier, making even apostates break into song. Many are in old or oldish English, which is often Victorian English looking over the shoulder at the Middle Ages, but redeemed by a love of the sparkling phrase, and the grandeur and majesty of the King James Bible, that apotheosis of devout declamation.

The more recent stuff has all the drama of a company report. The three executives wish to let it be known that a child has been born somewhere of low socio-economic status, and not properly registered, so the following actions will be taken, going forward. Form A is for legal births, Form B for births out of wedlock, Form C for parthenogenesis (attributed).

Even the Churchwardens, whom I know to be persons of kindly Christian dispostion, were spitting blood.  Carol 1 “A great and mighty wonder” came in for particular venom.  “Never again” one vowed later, over mulled wine. I myself had lost the will to live as it staggered about at funereal place, and even normally polite young girls nudged their grandmothers at the couplet: “rejoice, ye vales and mountains/ ye oceans, clap your hands”. It was a mumble, not a Carol, and left the Congregation in a confused and dissatisfied mood. 

Worse was to come. The original design of the service included so much that was dire and unsingable that an attempt was made by a kindly deputation to put in some traditional carols. This was turned down, so in a typically British compromise, several random verses were omitted from many carols, by giving verse numbers relating to a Carol book without verse numbers. This was intended to speed things up, but involved singers counting verses in unfamiliar carols whilst trying to pick up the tune, if any. One note said “omit 3,4,5,8, (and then had a blank, under which there a clear but spectral 9)”. The carol in question “This is the Truth” had 9 verses, so it was a finely judged question as to whether it might not have been better to have omitted it entirely. The ninth verse ended on the immortal phrase “and thus I close my Christmas song” so I could understand why persons of taste wanted this verse omitted, but not why their entreaties had been turned down.

Not everyone had read these notes. “Once in Royal David’s City” was once the centrepiece of our little service, with a proper soloist singing the first verse. The cryptic warning “verse 1 solo; omit V3) proved a challenge. What might have been a solo progressed into an uncertain gaggle of women’s voices, since the congregation had a vague folk memory that the women used to do one verse, the men another. Frankly, things never properly recovered. One later carol included the leaden phrase “teach us to resemble thee” to which one could only answer “but if that vague approximation proves too taxing, don’t worry about it too much”.

On a brighter note, none of the lessons were read by children, so though grandparents were denied the indulgent celebration of the uses of literacy, at least the cadences scanned properly, and the words carried meaning.

During this travail we all knew that help was at hand, not of the Divine sort (though there were intimations of immortality when the winter sunlight suddenly illuminated the white interior bathing everyone in redemptive clarity) but in the final, familiar and majestic “Hark the Herald Angels Sing” which allows three verses of yelling, the last at the very highest volume with imagined trumpeters at the four corners. The first two verses put a smile on everyone’s faces and then, just as they were ready to raise high the roofbeams, the organist stopped. Not dead. Just stopped. Like the fabled Macondo in which villagers have lost their memories for words, we have all lost our ability to count. It was a Very British Problem. Should we all sing the last verse unaccompanied, or would that shame the organist? Should one give a nervous cough?  There was a long pause, redolent with pre-orgasmic disappointment. The priest judged it the correct time to give a blessing, the unperturbed organist broke into a cheerful organ voluntary, which was the cue for mulled wine and a muttered post mortem.

Next year the Carols, the Churchwarden promised, will be familiar and within the singing range of the common folk. But the sun shone in on the constant generations, and nothing is lost out of nature, although everything is altered: faces, mannerisms and hairlines transmitted, pew by pew, in sure and certain hope of the Resurrection.

A test of Open Live Writer

Altruists turn down advantage

Thursday 17 December 2015

Fairness for me and then (perhaps) for thee

Here’s a simple game to assess children's sense of fairness, with important implications. Children seem to learn from an early age to reject an offer which is unfair them (offered less than the other child), even if the rejection means that neither they nor the other child gets anything. However, at what age, if ever, do they reject an unfair distribution which favours them (offered more than the other child)?
Blake et al. (2015) The ontogeny of fairness in seven societies. Nature. 10 December 2015

It turns out that 3 societies have high minded altruists, and 4 don’t, though we can live in hope that they develop them in due course. Here are the groups who were studied. Glance at them, and make your predictions.

In fact the angels come from Canada, US and Uganda.

Upward pointing scores on the right show rejection of unfair advantage. So, kids in Mexico, Peru, India and Senegal rake in an unfair bounty of sweeties (though the Mexicans seem to be learning as they mature, while the Peruvians are getting less equitable). Ugandan, US and Canadian kids are paragons of virtue, turning down allocations of goodies which are unfairly in their favour.

Like most results, the interpretation is a projective test. The US and Canada are white, wealth and Christian, and just the sort of societies which encourage altruism. Uganda? Um….African, Christian, but not wealthy. Other nominally Christian countries do not show Christian charity. Protestantism is a possible factor, but that does not account for Canada.  Candidly, at first glance no grand hypothesis suggests itself. Funny things, facts. This is the delight of research: presumptions are confounded, and must be re-examined.

I think it best to strip off the national labels, because these are not representative national samples, but a smattering of villages. Sample sizes and the basis for selection need to be examined.

My rule of thumb is that samples should be 30 and over, simply because at roughly that level gains in significance start levelling off. On that basis only the US sample meets standards. We should forget the 13-15 year comparisons. The Indian and Mexican 10-12 year figures are also far too low to draw firm conclusions. The proportion of familiar/unfamiliar pairs varies wildly, but doesn’t seem to match the key results, so is probably not relevant, though I don’t see it analysed in the statistical tables.

The game does not tax the intellect, so might not seem the sort of thing which would be linked to intelligence. At the same time, there is an intelligent argument that if you reject unfairness even when it is to your advantage, it will propagate the ideals of fairness, from which all will benefit in the long run. Slow life history, deferred gratification, wisdom: you know the score by now. However, why should Uganda have a slow life trajectory?

The supplementary data gives very good descriptions of the societies from which the children were drawn, but I do not know why these were picked, other than convenience. They are certainly very different, which is probably the main thing.

The paper is well written and the findings cautiously presented. I would have preferred that the sample sizes were given in Table 1, but that’s experimentalists for you. Psychometricians always prefer larger samples, and that samples should be as representative of the population as possible.

What else should be measured in these children? Perhaps Piagetian conservation of mass? Piagetian perspective-taking? Personality? Some additional other test of altruism? Perhaps a little school test with a chance for them to cheat?  If the children were granted anonymity, how would that affect the results?

I think this is an intriguing finding, which leads to new hypotheses, and should be replicated on larger samples before we can draw any strong conclusions. 

Tuesday 15 December 2015

Chanda Chisala and African heredity

Chanda Chisala has published “Closing the Black-White IQ Gap Debate, Part 3 : Thomas Sowell vs. Richard Lynn”. He has kindly made mention of my comments on his Parts 1 and 2, so I will respond to his two points.For background, here is what I have posted so far:

Three introductory observations:

First: Chisala refers to me as an hereditarian, and I previously referred to him quizzically as an African hereditarian, championing African cognitive elites, namely Ibo and Yoruba. In my case, and I assume in his, I am in fact “50% hereditarian/50% environmentalist, plus error”. I think we are both considering data supportive of an inherited component in human abilities.

Secondly, I think that evolution applies to all living things and that certainly includes homo sapiens. Any group, given time and relative isolation, can evolve in different ways, often according to the characteristics which are prized or required in that group. Cognitive elites can occur on any continent. (That does not mean they will be equally bright across continents). They will usually be prominent in commerce and scholarship, and be bred for their wits, in the sense that marriage choices will be very influenced by intelligence indicators such as educational and occupational attainments.

Third, when I object to “Powerpoint publishing” I am in fact indifferent as to how the summary results are displayed, so long as there is a way of getting to the full research document with methods, results and technical appendices. It is when those are absent, or very difficult to track down, that proper analysis becomes impossible.

Now to the specific points Chisala makes in response to my comments:

“currently the UK is a magnet for immigrant groups, particularly elites.”

Yes, this is so. The UK had 330,000 immigrants in the year to March 2015. They are unselected, and that is a great problem. Capable immigrants contribute significantly, but those who do not find a job are not a benefit.  In my last post I said: Not shown in the Chisala paper is that the social profile of African immigrants is probably bimodal. They have almost as many parents in the professional ranks as the UK average, but also a very large number of unemployed persons. It is an odd distribution, suggestive of at least two different sources of immigrants as regards social status.

So, the UK is attracting African elites, and also Africans unable to get work. Given the wide disparity in occupational level of Africans in the UK, you can simultaneously get elites at particular schools, and lower than average African performance in the country as a whole. In order to see whether African immigrants are representative of their countries of origin the ideal would be to have intelligence and scholastic attainment results from their countries of birth. Absent those results (though there are other data suggesting that African immigrants are more educated than average Africans) the occupational data at the very least implies heterogeneity of socio-economic status.

To get a better understanding of the scholastic achievements of immigrant groups we need to get beyond the percentage pass rates and look at the actual figures, which as far as I know are not currently released. For example Deary et al. (2007) has a scoring system based on the grades, which is far better than just pass rates, but he did not have access to the raw scores, nor was able to publish anything on racial differences. This is great pity, because it would give us much of the data we need, which could then be tracked year by year.

“A marshy island called Singapore ends up rich, and resource-rich Nigeria remains poor (and very populous). Could this be because of any differences between Chinese and African peoples?”

Yes, this is about GDP per capita, not country totals regardless of population numbers. Countries get rich in two main ways: either they innovate or they provide raw materials and holidays to countries that innovate. Germany and Japan are good examples of the first; Greece, Arabian Gulf, Nigeria, Venezuela, Cuba and Egypt are examples of the second category. The whole point is that native wit seems to contribute more than raw materials.

General observations

I think that the more detailed approach of looking at genetic sub-groups is promising, and that it is worth looking for cognitive elites wherever they can be found. I would be interested to look at good cognitive data on the Tutsis and the Hutu. The presumption is that the former are brighter than the latter. If so, this would be in concordance with the general pattern in genocide (brighter attacked by duller).

For the avoidance of doubt, if African samples representative of Africans in Africa turn out to do very well simply by coming to the West, in my view that would strongly suggest that African environments were the major cause of low African ability in Africa. Richard Lynn has always argued that poor nutrition, bad health and adverse circumstances were highly relevant. Rinderman, Woodley, Meisenberg and others have looked at ability increases in African countries, and I have discussed their work in various posts.


Deary, Stand, Smith and Fernandes (2007) Intelligence and educational attainment. Intelligence, 35, 13-21.

Monday 14 December 2015

Holiday books please

I have three books to review over the Christmas break, but invite you to send me others on intelligence or personality to add to the list. Digital form probably best at this late stage. 

Meanwhile, Windows Live Writer is playing up, and preventing me from posting up two pressing matters.

Wednesday 9 December 2015

Multiple emotional intelligence


Although it is fashionable to denigrate conventional intelligence, or at least to doubt that it is capable of being measured accurately, there is a concomitant willingness to believe that multiple intelligence or emotional intelligence are the very real answers to all questions about job performance.

Multiple intelligence is a marketing triumph: it is the sketch of an idea, vague, hard to operationalize, difficult to test, and largely without any empirical support, and thus very popular, particularly among educationalists. It always gets a mention in psychology text books, on the basis that any comforting notion deserves favour. Lack real intelligence? Compensate with multiple intelligence! Since we do not have reliable measurements there is little more we can say about it.

Emotional intelligence is another marketing triumph: it conflates personality with the perception of emotion, the latter being difficult but possible to test, and has some empirical support. The proponents have worked hard to create a psychometric test and to collect data. They point out that they are not just testing personality again, but working on the specific issue of assessing whether there is a specific mental skill involved in understanding the emotions of others. There are such skills.

The big question is: what do these fashionable assessments add to the tried, tested, and widely validated measures of general mental ability? To answer this question we need to consult the Oracle: Jack Hunter, Frank Schmidt and co-workers.

The Validity and Utility of Selection Methods in Personnel Psychology: Practical and Theoretical Implications of 95 Years of Research Findings. (see below)

We also need to look at some methodological issues, and one source is: Oh, I-S., Postlethwaite, B. E., & Schmidt, F. L. (2013). Re-thinking the validity of unstructured interviews: Implications of recent developments in meta-analysis. In D. J. Svyantek, & K. Mahoney (Eds.), Received wisdom, kernels of truth, and boundary conditions in organizational studies. Charlotte, NC: Information Age Publishing. Pp 297 – 329.

This group have established a reputation for careful and very detailed work, such that their procedures have set the standards of best practice. In that spirit, here are some problems with using job selection tests to predict ability to do the job. Those who have already got the job in question are brighter than the applicants, and the incumbents not only have a higher mean but a small standard deviation. They are The Right Stuff, selected within a narrow band of capability (small standard deviation), whereas applicants have a broader range of capability. The predictive power of the test is weakened by this biasing effect of range restriction, and needs to be corrected by disattenuation, and the technique must cover direct and indirect effects.

For example, those who are offered the job probably tick all the boxes: bright, personable, and with experience. Using only a cutoff score on a mental ability test gets one of the criteria, but misses the other two. The correction for direct effects misses some indirect effects.  Furthermore, the brightest and best candidates may get several offers. They may turn down the job in question for a better one, which complicates a simple analysis because a very bright candidate will be registered in the records as “did not get appointed to job” simply because they took up a better option elsewhere.

A technically advanced IRR solution, tested against Monte Carlo simulations provides the best estimate (Hunter et al. 2006) and shows that general mental ability is a better predictor than formerly stated by about 25% in that correlations rise from 0.51 to 0.68.

Hunter, J.E., Schmidt, F.L., & Le, H. (2006). Implications of direct and indirect range restriction for meta-analysis methods and findings. Journal of Applied Psychology, 91, 594 – 612.

Two little asides here: having been properly selected for general mental ability those who have a particular job will not be very easily distinguishable one from another on the basis of a general intelligence test. Intelligence will seem to have “disappeared” because it was the basis of selection, and all that will be left visible will be personality and experience differences. Second observation: job applicants self-select to some degree, so the standard deviation and mean of applicants will probably be higher than that of the general population, and not as low and wide as that general population, so another source of restriction of range, and possible cause of under-estimation of the predictive power of intelligence will have been missed.

Anyway, let us look at a talk given by Frank Schmidt on selection methods for job performance.


Start at the top to find out what works, or at the bottom to see what doesn’t. General Mental Ability is king of the castle, and all else lies in its shadow. Tests of integrity make useful contributions, as do employment interviews, measures of conscientiousness, checks of references, background data and job experience. In the 1% gain category are Years of Education, Interests, Emotional Intelligence, Grade Point Average and Organisation Fit.

If you want to make valid choices without wasting too much time and effort, avoid overlapping predictors and go for those that make useful independent contributions, as shown below:


Remember, a weak predictor can still make a contribution if it provides something not covered by the generally more powerful ones. However, emotional intelligence is not on the list.

Jobs are given to those who can show ability, honesty, conscientiousness, proper references, education and job experience, the respect of their peers, and emotional stability.  According to your tastes this is very dull, or highly reassuring.

Now let me tell you about the multiplicity of my gastro-intestinal intelligence.

Tuesday 8 December 2015

Search this blog

In a small but crucial change, I found that the “Search this blog” function was not working, so I have picked up some good free advice from Melissa Hero and it is now working again.

It is right at the top of the page, which makes it more visible and causes far less confusion with the “follow by email” field, which is left further down the page. Better still, it is a more functional solution, which makes all the relevant posts available so that they can be read in one go.

Monday 7 December 2015

Christians sag, Muslims rise, Jews ever constant

Yes, I know that Google N-gram is not a perfect window on the world, but in the English literature over the last two centuries reference to Christian has fallen, to Muslim has risen, and to Jew is ever constant. Interesting to know which of those mentions was positive or negative.

ChristianJewMuslim 15

Something happened around 1920 to boost Muslim, very probably the Balfour letter of 2 November 1917. Pretty easy to fit dates to the subsequent rise of the words, but odd that creation of state of Israel has no influence on the mentions of Jews. At this point you can see the battle of civilisations, with a Muslim rise and a Christian fall (and perhaps a more recent Christian renewing of vows).

Let’s look at it a slightly different way:


As mentions of Israel rise, so do mentions of Muslim and Islamic.

Now we need a pointless variable which goes in a different direction, which then gives us sufficient material for a book. Terrorism turns out to be a lagging variable for Islamic. Atheism barely registers. Coexistence little better. Murder has the same frequency as Muslim, and Islamic rises to that level till all three are indistinguishable by 2008. A good story to be made out of that, no doubt. Weapons behave similarly.  After trying a few words which contributed nothing more, I recalled a conversation I had in King Henry VIII’s wine cellar with a colleague who had done long service in researching military psychology. Call him Simon. The venue is in the deep bowels of the Ministry of Defence building, and is one of the last remnants of Whitehall Palace, and conducive to reflection.  Talking to Simon about trauma reactions in the military he interrupted me, saying of something I had just said: “Ah, that’s a word that doesn’t get used any more”.

Have a look at the final picture:

Character of Christianity

   Judging people by the content of their character. Quaint.

Saturday 5 December 2015

The puzzle of falling French intelligence

In June I reported on the work of Edward Dutton and Richard Lynn in “A negative Flynn Effect in France, 1999 to 2008–9” Intelligence 51 (2015) 67–70.

I wrote that they had looked at a small but probably representative sample used by the Wechsler team in their 2011 French standardisation of the adult form of their general intelligence test. They say: The results of the French WAIS III (1999) and the French WAIS IV (2008–9) are compared based on a sample of 79 subjects aged between 30 years and 63 years who took both tests in 2008–2009. It is shown that between 1999 and 2008–9 the French Full Scale IQ declined by 3.8 points.

If true, this is a substantial drop in ability. The authors did not furnish an explanation, though they considered a few possibilities. In my comments I suggested a couple of reasons (apart from small sample size) that the results might not be as solid as imagined, but still felt that there had very probably been a drop in ability.

Now the paper has been looked at by Michael Woodley and Curtis Dunkel who examine causal explanations in more detail.

Michael A. Woodley of Menie and  Curtis S. Dunkel. In France, are secular IQ losses biologically caused? A comment on Dutton and Lynn (2015) Intelligence, Volume 53, November–December 2015, Pages 81–85 doi:10.1016/j.intell.2015.08.009

The authors say: Dutton and Lynn report secular declines in Fullscale IQ evaluated using WAIS of four points a decade in France between the years 1999 and 2008–9. It is posited that the trend may have a partially biological cause, stemming from dysgenic fertility and, to a lesser extent, replacement migration. Given that these, and other biological phenomena are associated with the Jensen effect, it is expected that if they are the principal causes of the IQ decline then the secular change should also be associated with the Jensen effect. Furthermore if it can be demonstrated that the vectors of secular IQ decline, g loadings and the vectors of other biological indicators share variance, then the case for biological causation will be strengthened. Using the method of correlated vectors and error disattenuation, the secular IQ declines are shown to be associated with a high-magnitude Jensen effect (ρ = .833). A multi-vector common factor comprised of the vector of g loadings along with the vectors of three biological variables (subtest heritabilities, dysgenic fertilities and simple visual reaction times) was found to load substantially on the secular IQ decline vector (λ = .723). These findings indicate that the French secular IQ loss likely has a primarily biological cause.

How do the authors come to this conclusion?

It has been observed, with a few of exceptions, that variables which associate with IQ subtests to a greater extent, when the measure is more g loaded, tend to be biological (i.e. pertaining to genetic and physiological factors) and include factors such as population, race and species-level differences in cognitive ability ( Fernandes et al., 2014, Rushton, 1999 and Rushton and Jensen, 2010), subtest heritabilities ( Woodley of Menie, Fernandes, & Hopkins, 2015) and inbreeding depression effects ( Rushton, 1999 and Rushton and Jensen, 2010). Collectively these have been termed ‘Jensen effects’ ( Rushton, 1998).

Factors that exhibit the opposite tendency, i.e. associate with less g loaded abilities to a greater extent (and are in consequence ‘anti-Jensen effects’), are typically environmental or cultural in origin, such as IQ gains accrued via retesting ( te Nijenhuis, van Vianen, & van der Flier, 2007), gains accruing from the adoption of low IQ children into high IQ homes ( te Nijenhuis, Jongeneel-Grimen, & Armstrong, 2015), those accruing from intensive educational interventions ( te Nijenhuis, Jongeneel-Grimen, & Kirkegaard, 2014) and the Flynn effect ( te Nijenhuis & van der Flier, 2013).

First, they correlate the vector of French secular IQ decline magnitudes with the vector of subtest g loadings from the French standardization of the WAIS-III. This is the method of correlated vectors. They also correct for measurement errors.

Second, they construct a multi-vector common factor using the vector of the secular IQ declines and subtest g loadings along with the vectors of various biological indicators, including subtest heritabilities, Subtest-fertility (dysgenic) correlations and simple visual reaction time-subtest correlations.


The raw vector correlation is 0.38 and the corrected estimate 0.833. The corrections are reasonable, but the authors are clear that there is no way to get round the smallness of the sample, so I see this as just an indication of the underlying trend.


This adds in the dysgenics estimates and simple reaction time results, and the reliabilities, to give a much fuller and more intriguing picture.

Finally, they show the correlations of the vectors and bring the whole thing together:


Crucially, the loadings for g, heritability and secular IQ declines are all positive.

The authors say: The French secular IQ declines are associated with modest and large-magnitude (based on Cohen, 1988; small = .1 to .29, modest = .3 to .49, large = .5 to 1)1 Jensen effects when both the French WAIS-III normalization g loadings and the meta-analytic manual g loadings from Kan (2011) are used (r = .38 and .658 respectively; the two sets of g loadings exhibit a large-magnitude correlation: r = .626, N = 9 subtests). When corrected for four sources of error, the vector correlation between the French WAIS-III g loadings and secular IQ declines increases to .833 indicating a high-magnitude Jensen effect.

They concede that there is high measurement error, and they and the original authors agree on this point.

The results of the multi-vector analysis are consistent with Rushton's (1999) prediction that the vectors of biological variables and also g loadings should cluster with one another, forming a biological nexus. All loadings of the multi-vector common factor on its components are of large magnitude (λ = .505 to .854). The loading of the vector of secular IQ decline (λ = .723) indicates an association between this variable and other biological variables, which supports the speculation advanced in Dutton and Lynn (2015)concerning possible biological causes of the French IQ decline.

Replacement migration in France involving populations exhibiting lower means of IQ and higher rates of total fertility, such as Algerians, Moroccans, Tunisians and Roma (Čvorić, 2014 and Lynn and Vanhanen, 2012) may be increasing the rate of secular losses at the level of g, consistent with speculations advanced in Dutton and Lynn (2015), however the additional loss in g due to this process is anticipated to be very small. Based on a simulation, Nyborg (2012) estimates that in Denmark, replacement migration may be reducing heritable g by .28 points per decade, which would increase the overall loss in gto 1.51 points per decade ( Woodley of Menie, 2015), this still being only 37.75% of the loss observed in the French cohort.

Another possible contributing factor is the so-called “brain drain” (De Rugy, 2012), which involves high-g French individuals emigrating to countries that offer more competitive salaries and better working conditions. Thus human capital flight between 1999 and 2008–9, especially among the younger demographic, may have contributed to the French secular IQ decline, however the precise impact of the effect is difficult to determine.

Brain drain is certainly a possibility: the “South Kensington effect” of French emigres clustering round the Lycée français Charles de Gaulle is a result of London hosting roughly 270,000 French. The number is less significant than the fact of them being highly qualified and entrepreneurial, a real smart fraction.

Now the authors’ approach should be used to look at the more recent paper by Jakob Pietschnig and Georg Gittler on an equivalent fall in intelligence in Austria and Germany. Will they find the same result?

It is a minor point, but authors reporting falls in intelligence sometimes describe it as “a negative Flynn Effect”. Whilst technically true, this somewhat underplays what is going on. If you strip out from the apparently rising scores that portion which is due to measurement error (changes in material, in scoring, in standardisation and possibly in student guessing) then the rest is assumed to be due to better environments: education, health and prosperity. The original assumption was that once everyone was on a level playing field of externally provided good circumstances, then intellects would reach a plateau of high functioning, but a plateau nonetheless. Jim Flynn said he was always surprised at how long the effect was continuing. He felt that once the deficits in living circumstances had been corrected, it would stop. However, as far as I know, no-one apart from those looking at dysgenics had predicted that intelligence would fall. I think we should refer to falling intelligence as Falling Intelligence. Either that, or follow Charles Murray in calling it the Woodley Effect

The work by Woodley and Dunkel serves as a general method of estimating the biological component in intellectual decline, a key explanatory component of the Woodley Effect. It should be tested widely.