Monday 30 November 2015

Does affirmative action cause mental illness in black and Hispanic students?

A reader asks me to comment on a post with the above headline, so this is a quick note with a few thoughts. The link is given below, and that itself links to a New York Times article, which I reproduce in case you want to start with that article.

The true answer is that I have absolutely no idea whether affirmative action causes mental illness in black and Hispanic students. Of course, I have many ideas which have been generated by the question, but I can’t answer it because many of the parts required for a good answer are not available to me, and I don’t want to hunt them down at the moment. For example, I would have to research affirmative action generally, its prevalence in US colleges, its prevalence in the colleges currently experiencing protests versus those not protesting, college SAT entrance score requirements, some base rates of mental illness in college student by race, and general US college dropout rates. The protesting students may have excellent grades, and I cannot determine whether they gained entry through affirmative action just by looking at them, though I can gain some impression of their abilities and character by watching the videos of their protests. Those brief glimpses of behaviour may not be representative, but they can be very informative.

The plausibility of the thesis (affirmative action causes mental illness in its beneficiaries)  is only moderately in its favour, because psychology leads easily to ad hoc explanations, of which the most popular would be: placed in a college where the course content reveals them to be clearly less able than Asian and White students, Black and Hispanic student feel demoralized, depressed and resentful. They attribute their personal failings to institutional shortcomings, and (like many weak students everywhere) complain: “I’m bright but the exam was unfair/I wasn’t taught properly”. Before jumping to that conclusion (which may be right, but is not proved right just by being plausible) it is more important to have a knowledge of the territory. Here are some general findings in the UK, simply to show my thought processes. Dropout rates in British universities are closely related to university quality: better universities have lower dropout rates. That is not surprising: they get much brighter students, far more capable of completing a university course.

You need, on average, 608 UCAS points to get into the University of Cambridge, making it the most competitive university in the UK. Understandably, only 0.35% of students drop out.

Only about 1% drop out of the Russell Group, either because their school exam marks flattered them and they lack ability, or more likely because of psychological problems: loneliness, home-sickness, anxiety and depression.

Lesser institution have higher dropout rates, roughly 15%. They get less able students, and probably offer them less personal attention. Also, the degrees they confer will themselves confer relatively little occupational advantage, so dropping out may be sensible if there are jobs available. Someone who gets three years of job experience may do better than those who hang on attempting to be studious. There are also differences according to the courses taken, with prized medical education places having a low dropout.

College life is great but has drawbacks. Losing parental and school peer support is one aspect, but there is also the unsettling change for many that they move from being top of the class to being close to the bottom of the university. At a real university you will find many persons brighter than yourself, and many, many persons more knowledgeable than yourself. Unsettling. (I still haven’t got over the shock of finding that other students had read Franz Kafka, and some had made notes about his writings). Finding yourself not as precious as you thought you were can be depressing. If you were let in on some sort of racial, religious or sociological quota you may struggle to compete with those judged only on ability and accomplishments. If the gap between the quota entrants and the others is big, they will be seen a queue-jumpers. That will be true on average, however hurtful.

College attendance aside, lower intelligence is related to more psychological disturbance anyway, so weaker colleges will have more students with those sorts of problems.

Does college life cause mental illness? Well, only in the sense that life imposes demands on people, and strain develops if external stress is not met with inner resilience or elasticity. Colleges are intellectually demanding, but also empowering. Learning makes you better able to understand life, and slightly better able to understand yourself. You achieve the beginnings of mastery of some basic techniques, and begin to correct your many mistakes. Compared with struggling to find and keep a job, college life is relatively easy, and much easier than bringing up children.


Psychologists haven’t actually got a Young’s Modulus for human beings because, ahem, Psychology is a young science, and has been for a century, but the general predictors of psychological breakdown are known: being a woman is one of the most important, but being neurotic, less intelligent and less well educated are also important, as are life-events, better known as the slings and arrows of outrageous fortune, bringing strains and grievous losses.

My opinion is that college entrance should be based on academic ability. Standardised testing is the best way of determining this: well-validated general intelligence tests and cognitively demanding scholastic ability tests. g-loaded is best. Anyone who gets in on another basis (including Daddy paying for a University Library) will experience some strain because they cannot keep up with the others, though the stresses will be much less than working as a nurse in a busy hospital ward. Colleges don’t cause mental illness, but vulnerable students may find them taxing. Taking good care of students is both kind and wise, because suicide is a preventable risk.

So, full confession now, when Keele University students had a protest in the sixties I did not participate, even though I was a victim of their “no women in the bedroom after 6 pm rule”. This was not surprising, because virtually nobody took part except the usual suspects, looked at with mild amusement by the majority. The university authorities were benign if paternalistic, and the protestors were seen as far too self-important. They submitted a list of mild suggestions and cleaned up the offices before leaving, to the great satisfaction of their Class Comrades in the Revolution, the university cleaning ladies.

But now, in the spirit of the time, how about my feelings about the current US University protests, as seen in a few videos

Wow, what bad manners!

Sunday 29 November 2015

US academics: Lefty and Liberal because of high IQ?


There is a long academic tradition of regarding opponents as mentally disturbed. Such condemnatory revelation is done more in sorrow than malevolence, the investigators would have us believe. Academia has looked long and hard at Right wingers and pronounced them a pretty odd bunch, suffering from something called The Authoritarian Personality. Fascists is the more usual appellation, and their horrible opinions are calibrated on the F scale. What do Right wing academics say about Lefties? Not much, because there are few Right wing academics in social psychology.

One interpretation popular among Lefties is that the brighter you are, the more likely you are to be a Leftie. At the higher levels of intellect you see things as they really are: Socialist. There is a fair bit of data to support this position. The alternative explanation is that the brightest people want to run things and make money, and regard academia as a boring waste of time. Bright people look at the bitter way academics argue, all the more petulantly and virulently in obscure matters because “the stakes are so low” and escape into the real world. Very bright people tend to the Right, and in their pursuit of power and wealth simply don’t care what academics chatter about, because their mutterings don’t make any difference. The Brightest have left college to compete for the glittering prizes, and the lefties are the social incompetents who remain perpetual students, striking absurd postures about world events over which they have no influence.

At this point you might like to look at a previous post which sets out the general thesis that the Left gather together the very brightest people and the dullest, thus joining together in a U shaped curve the clever idealists who really believe in Leftism and the dull ones who only sign up because they are the beneficiaries. This last group are not bright, but know what is good for them.

Then go through the comments, and note that on 3rd October 2014 Noah Carl writes in to say: My reading of the literature is somewhat different to Solon's. I am currently working on a critical comment piece, which I intend to submit as a response.

Here, after the usual paper-journal delay, is Noah Carl’s interesting paper: Can intelligence explain the overrepresentation of liberals and leftists in American academia? Intelligence Volume 53, November–December 2015, Pages 181–193  doi:10.1016/j.intell.2015.10.008

The author says: It is well known that individuals with so-called liberal or leftist views are overrepresented in American academia. By bringing together data on American academics, the general population and a high-IQ population, the present study investigates how much of this overrepresentation can be explained by intelligence. It finds that intelligence can account for most of the disparity between academics and the general population on the issues of abortion, homosexuality and traditional gender roles. By contrast, it finds that intelligence cannot account for any of the disparity between academics and the general population on the issue of income inequality. But for methodological reasons, this finding is tentative. Furthermore, the paper finds that intelligence may account for less than half of the disparity on liberal versus conservative ideology, and much less than half the disparity on Democrat versus Republican identity. Following the analysis, eight alternative explanations for liberal and leftist overrepresentation are reviewed.

Overrepresentation of liberals and Democrats appears to be largest in the humanities, the social sciences, and the arts (particularly sociology, anthropology and the performing arts), and appears to be smallest in economics, business, computer science, engineering and military science. For example, the ratio of liberal to conservative English literature professors may be as high as 28:1, while the ratio of Democrat to Republican sociology professors may be as high as 44:1 . Overrepresentation in the physical sciences, the biological sciences and mathematics appears to be intermediate, though still considerable.

After reviewing the political affiliations of American law professors, Lindgren (2015) concluded that, “By some measures, in 1997 the most underrepresented racially defined groups were Non-Hispanic white Republicans and non-Hispanic white Protestants”. ” Similarly, when the psychologist Jonathan Haidt asked attendees at the 2011 meeting of the Society for Personality and Social Psychology to indicate their political affiliations via a show of hands, he counted only 3 conservatives and only 12 libertarians, but approximately 800 liberals (Duarte et al., 2014). The 23rd annual Commencement Speakers Survey carried out by Young America's Foundation documented a ratio of six liberal speakers for every one conservative speaker among the top 100 universities. And notably in 2014, invitations to at least six prominent commencement speakers' were met with protests on campus from liberal or leftist student groups, leading to the cancellation of four.(Strauss, 2014; Chotiner, 2014). It is important to keep in mind that American academia has probably not always been so skewed toward liberalism and leftism. Duarte et al., (2014) compiled historical figures on academic psychologists' partisan affiliations, which indicate that the ratio of Democrats to Republicans may have been as low as 2:1 or even parity in the 1920s.

High intelligence may explain Leftism for the following reasons: Academic advancement requires very high intelligence, and since few individuals with conservative or rightist views possess very high intelligence, such individuals are comparatively scarce within the academy. At present, there is a certain amount of circumstantial evidence for this hypothesis. Numerous studies have found that individuals with higher intelligence to be more socially liberal on issues such as gay marriage, abortion, working women, free speech and marijuana legalisation.

Americans with higher intelligence are apparently more likely to identify as liberal on a liberal/conservatism scale. And compared to Americans with only high intelligence, those with the highest intelligence are more likely to identify as Democrat, more likely to support welfare for the poor, and more likely to favour affirmative action for minorities. In addition, scholarly elites such as Nobel laureates, Pulitzer Prize winners and Putnam fellows have donated to the Democratic Party far more often than they have donated to the Republican Party. However, there is also some circumstantial evidence against the hypothesis. In particular, several studies have found that individuals with higher intelligence tend to be more economically rightist in areas such as redistribution of income and government control of the economy.

Method: Fig. 1 illustrates the method used for assessing how much of the overrepresentation of liberals and leftists in American academia can be explained by intelligence. I first bring together data on the political beliefs of three separate populations: academics, the general population, and a high-IQ population. I then calculate the proportion of each population that identifies with various political positions (e.g., thinking of oneself as a liberal, supporting the Democratic Party). The extent of overrepresentation for any particular position is simply the percentage point difference between academics and the general population (i.e., the total length of the right-hand bar in Fig. 1). And the fraction of this overrepresentation that can be explained by intelligence is simply the percentage-point difference between the high-IQ population and the general population divided by the percentage-point difference between academics and the general population (i.e., the grey portion of the right-hand bar divided by the total length of the bar). In the hypothetical case of Fig. 1, there is a 10 percentage-point gap between academics and the general population of which 50% (i.e., 5 percentage points) can be explained by intelligence.


This is a crude variance estimate method, but at least is clear in how it is being calculated, and seems reasonable.

I define the high-IQ population as the roughly 4% of GSS respondents who scored 10 out of 10 in the vocabulary test. Note that a score of 10 equates to a mean IQ of ~128, which is just under two standard deviations above the population mean. This is in line with estimates for the average IQ of academics that have been reported in the literature, though it may understate the intelligence of academics in the physical sciences, whilst possibly overstating the intelligence of academics in the social sciences and humanities (see Dutton & Lynn, 2014). Gibson and Light (1967) tested 148 male academics at Cambridge University, and reported a mean IQ of ~128 among physicists, and of ~122 among social scientists.

So, given that method, here are some of the political views showing what proportion of those viewpoints might be accounted for by intelligence.


Here, in tabular form, are the striking differences between academics, the general public, and high IQ people.

Proportions backing various views


I have summarised Noah’s possible explanations very briefly, but the paper is particularly interesting on the more detailed arguments:

1 Self selection: Academics may be more Open to Experience in personality.

2 Self selection: Academics less interested in making money and raising children.

3 Self selection: scholars can cope with ambiguity, conservatives not.

4 After self selection, social pressure strengthens group think.

5 Self selection: Academics have weedy, feeble physiques, were probably fearful and bullied, hence want restorative “justice”.

6 Academics become more liberal through groupthink.

7 Academics have low pay relative to their educational level compared with independent professionals, resent them, want equality of influence.

8 Academics discriminate against any candidate with right wing inclinations.

Personally, I see all these as boiling down to two factors: self selection and social pressure.

Noah concludes: Intelligence can account for most of the disparity between academics and the general population on the issues of abortion, homosexuality and traditional gender roles. By contrast, intelligence cannot account for any of the disparity between academics and the general population on the issue of income inequality. Furthermore, intelligence may account for less than half of the disparity on liberal versus conservative ideology, and much less than half the disparity on Democrat versus Republican identity. Possible explanations for the remaining overrepresentation comprise: self-selection on personality, interests, cognitive style or preferences; social homophily and political typing; self-selection on strength and stature; individual conformity; status inconsistency; and discrimination.

In sum, this paper shows that intelligence is a large part of the answer, but not all of it. The paper also shows the very clear difference in political and social attitudes between academia and society as a whole. It does a great service in showing this yawning chasm of attitudes.

If you evaluate arguments by questioning the motives of the proponents (does the Left do this more frequently than the Right?) then academia is so biased that its findings should be set aside. Nothing these apparatchiks “find” can be trusted. They are Lefties, espousing Leftie causes, and they are a burden on the public purse. On the other hand, if all these disciplines have strong methods and a respect for facts, then the standard procedures of empirical science should prove sufficient to reveal errors in research, so science should self correct. On this, possibly too optimistic reading, there is no need to pack academia with Right wing scholars (and such affirmative action should be anathema to them anyway) because the truth will out in the end. In conclusion, we have to examine arguments regardless of motives, and seek to improve the accuracy and soundness of research. A topic for a blog, I think.



Friday 27 November 2015

Does social class affect intelligence only in America?


Paper publications impose delays on thought. In compensation, these laggard luminaries of lackadaisical journals claim that the final result is of a much higher standard, polished as the texts are by the sparkling minds of anonymous reviewers. Perhaps so.

On 13 December 2014 I reported from the ISIR conference: In a very big meta-analysis Tim Bates showed that social class interacts with intelligence to some extent in US samples, but not in other parts of the world. It suggests that the much quoted Turkheimer (2003) is something of an outlier in the US funnel plot, but there is a US/rest of world difference, though hard to be sure why, possibly less supportive welfare environment for poor Americans.

Only now in late November 2015 has that paper made its way through the review process, with the the result that I can report to you that in the published paper Elliot Tucker-Drob and Timothy Bates say :

A core hypothesis in developmental theory predicts that genetic influences on intelligence and academic achievement are suppressed under conditions of socioeconomic privation and more fully realized under conditions of socioeconomic advantage: a Gene × Childhood Socioeconomic Status (SES) interaction. Tests of this hypothesis have produced apparently inconsistent results. We performed a meta-analysis of tests of Gene × SES interaction on intelligence and academic-achievement test scores, allowing for stratification by nation (United States vs. non-United States), and we conducted rigorous tests for publication bias and between-studies heterogeneity. In U.S. studies, we found clear support for moderately sized Gene × SES effects. In studies from Western Europe and Australia, where social policies ensure more uniform access to high-quality education and health care, Gene × SES effects were zero or reversed.

Large Cross-National Differences  in Gene × Socioeconomic Status  Interaction on Intelligence. Psychological Science (2015)  1 –12  DOI: 10.1177/0956797615612727

This is an elegant paper, and worth the long wait I have been complaining about. It takes up the Scarr-Salapatek (1971) hypothesis: “IQ scores within advantaged groups will show larger proportions of genetic variance and smaller proportions of environmental variance than IQ scores for disadvantaged groups. Environmental disadvantage is predicated [sic] to reduce the genotype-phenotype correlation in lower-class groups” (p. 1286).

The author’s meta-analysis data set consisted of 43 effect sizes from a total of 24,926 pairs of twins and siblings (approximately 50,000 individuals) participating in 14 independent studies. Data were relatively evenly split across U.S. (18 effect sizes, 8 studies, 10,831 twin or sibling pairs) and non-U.S. (Western Europe and Australia; 25 effect sizes, 6 studies, 14,095 twin or sibling pairs) samples.

Here is the funnel plot of results, with black dots being the US and the red dots the rest:


It is pretty clear that US studies find effects which others do not. So, in the US only, here is the picture:


As you realise, the analysis of variance depends on the circumstances being measured. The authors are probably right to surmise that the US versus The Rest is due to more generous welfare in the rather rich selection of The Rest countries. For example, I doubt we have good data for Brazil or Mexico, but those should show a greater SES effect on g, which would strengthen the interpretation being advanced in this paper.

The authors say:  First, studies from the United States supported a moderately sized Gene × SES interaction on intelligence and academic achievement (a′ = .074; Fig. 1). Second, in studies conducted outside the United States (in Western Europe and Australia), the best estimate for Gene × SES magnitude was very slightly negative and not significantly different from zero. Third, the difference in the estimated magnitude of the Gene × SES effect between the U.S. and the non-U.S. studies was itself significant.

There were no other moderating variables, and no publication bias.

We also replicated the well-established phenomenon that genetic influences on intelligence increase and shared environmental influences on intelligence decrease with childhood age. [] Genes account for considerably more variation in intelligence at both higher ages and in higher U.S. socioeconomic contexts. Indeed, both phenomena may reflect a process of increased and accumulated effects of gene-environment transactions with the increased opportunity that comes with both social class and age.

The results indicate that Gene × SES effects are not uniform but can rather take positive, zero, and even negative values depending on factors that differ at the national level. The finding that low SES was associated with attenuated genetic influence on intelligence in the United States resolves an important debate. The finding that this interaction is observed only in the United States, together with the novel discovery here that the effect may even reverse in sign (The Netherlands), suggests that further research on between-nations variability in the effects of family SES on cognitive development is particularly important. Candidate mechanisms that might underlie such variability include national differences in how concepts of letter and number that underpin literacy and numeracy are imparted, educational quality more broadly, medical and educational access, and macrosocietal characteristics, such as upward social mobility and income support.

This is a very neat result, from a very detailed and carefully argued paper.

As an amusing corollary, as the political Left celebrates the spread of welfare states and seeks to improve social provision in the US, the political Right can take comfort from the proof that, once people have a level playing field of generous benefits, then breeding counts more than ever before.

(I know that the historical picture since 1870 probably favours social spending; that removing benefits would probably have deleterious effects and some good ones; that we may be already detecting dysgenic effects caused by welfare; and that we haven’t identified the candidate mechanisms, but there is amusement in noting how social policies have unintended consequences).

Monday 23 November 2015

Third Blog Birthday




As is traditional on this blog, the cake shows only one candle, the ironic rite of passage preferred by patrons of a local restaurant who note birthdays with minimal fuss. Orbit completed, precise ages not seemly to disclose.

Having already described, in a previous post, why I blog, this anniversary is about you, the reader. 

Here are the places where readers live:

Top Ten countries

United States              317,722

United Kingdom          63,992

Germany                      27,603

Ukraine                         18,792

France                           18,731

Canada                          17,130

Russia                            13,954

Australia                        13,775

Finland                           4,310

Spain                               2,912


Top Ten Posts

Are girls too normal? Sex differences in intelligence 8 Sep 2013, 24 comments 8284

Gone with the Wind 20 May 2015, 38 comments                                                6109

The 7 tribes of intellect 2 Dec 2013, 59 comments                                                4762

Income, brain, race, and a big gap 31 Mar 2015, 32 comments                          4550

Give me a child until he is seven, 20 May 2013, 6 comments                             4317

Do women find bright men sexy? 18 Sep 2015, 27 comments                             4137

Flynn effect as a retesting, rule-based gain 2 Nov 2013, 14 comments               3481

Chanda Chisala: An African Hereditarian? 7 Jul 2015, 19 comments                3479

“It’s the people, stupid”: review of Wade 14 May 2014, 30 comments               2970

Intelligence in 2000 words 9 Dec 2013, 14 comments                                        2963


The most read post “Are girls too normal” is about sex differences in the variance of intelligence, a matter which was well known in the 1960s but which is news today, probably due to poor psychology teaching. My sole contribution was to depict the differences in a simpler way.

Hot on its heels is the more recent “Gone with the Wind”, which is a case of a snappy title bringing attention to the soberly entitled “Meta-analysis of the heritability of human traits based on fifty years of twin studies”, though that paper was already being widely read from conventional sources.

“The 7 tribes of intellect” keeps going strong.

“Income, brain, race and a big gap” is in fourth place, and is notable for going on to generate a good debate with the author of the paper I was criticising, leading many new readers to go back to have a look at it, before going on to read the author’s reply and my rejoinder. The sudden spike in interest, leading to 4,729 readers in one day was due to a mention by Steven Pinker, for which many thanks. All authors are invited to reply to posts, and it is great when they do so.

Referring URLs  and sites


Google                      15,200

Unz + iSteve              3,950

HBD Chick                 2,808

Marginal Revolution 1,350


Twitter                       42,647

Google                       34,566

Unz                              6,722         

Reddit                          4,897

HBD Chick                  3,602

iSteve                           2,557

feedly                           2,254

facebook                      2,232


As you would expect on this blog, here is a cautionary note about metrics. I have used “pageviews” throughout, as the common currency of blogs. I am aware that some of those page views are very brief: a new reader sees some numbers and a graph on the page and quickly decides to read something else. Average time on a page is 4 minutes. Remember, this is much better data than we have for bookshelves. Right now I can look at some books on my shelves and note that I have read only a few pages, and some almost none.

My readers, or unique page viewers, are 50% new visitors 50% returning visitors. Mostly men, there is a peak for young adults.



Their interests are news and politics and education. What I would like to know is how many of you are teaching or studying psychology.


Twitter is taking on a life of its own. Formerly I used it simply to announce each post, and to propagate some aphorisms to tempt people to read the blog. I have a mere 1457 followers (I came to the party late), tweet sparingly (4 tweets per day), but still get roughly 323,000 impressions every 28 days, which is 11,200 impressions a day. I get 110 re-tweets per 100 tweets. I have delusional expectations about any tweet composed in the early dawn, but nonetheless get attention with 30 tweets a month garnering more than a thousand impressions (one with 10,751) almost enough to make me feel I should drop the blog and tweet all day.

Top at 58 re-tweets is a graph showing effect sizes for early intervention studies. Next at 32 re-tweets are two aphorisms, as are most of the other top nine depicted.

Top 9 tweets


79% of my Twitter followers are male.

Blog supporters

My blog readers vary from experts in the field (though they almost never post up comments) to recent voyagers into these cognitive waters, (who ask questions and welcome references). I receive a steady stream of papers and books, and welcome those, even if it often takes me time to post something about them. Keep them coming.

Thank you to all of you who have loyally re-tweeted my tweets about each blog post, which is specially kind when done by celebrated bloggers like HBDChick, Jayman, iSteve and others, all of whom have their own blogs to tend to. Commendations, mentions and re-tweets by figures like Steve Sailer,  Charles Murray and Steven Pinker greatly assist me. 

Now we are 3

On the first blog birthday I said : Finally, I can claim that in one year 71,701 readers have given my words a look, as opposed to the modal 6 if I had published a paper. (At that stage I had 199 Twitter followers.)

On the second blog birthday I said: At the end of two years I have written 418 posts, which is 4 a week, come rain or shine. Page views all time history at the end of two years: 313,753 (At that stage I had 597 Twitter followers).

On this, the third blog birthday, there are a total of  627 posts, I have an all time total of 657,875 readers.


Thank you, all of you.

If you have any ideas to help me reach more researchers and students, please let me know. (In particular, I would like to be read by university teachers writing psychology text books). As a matter of general preference, I seek readers who understand the basic rules of evidence based arguments, and prefer focussed discussion, with references. They are doubtful, cautious, helpful, open-minded but easily startled. Approach them carefully, with a gentle recommendation that they might like to take a brief look at these pages. Perhaps.

Sunday 22 November 2015

Why I blog: Small particles of knowledge


I started this blog three years ago because I wanted to justify to myself having paid to go to a conference in San Antonio in 2012, and thought I should let other psychologists know something about the papers presented there. Usually, I stay at home in my study.

I continued blogging because I often found myself muttering during news items about the psychological variables which had been left out, often regarding intelligence differences, or failures to follow correct methods, or errors in argument.

The largest reason for blogging is even more personal. My self-evaluation is that I am primarily a translator, bridging a gap between researchers and readers.

The issue was best described by Samuel Johnson:

"The greater part of students are not born with abilities to construct systems, or advance knowledge; nor can have any hope beyond that of becoming intelligent hearers in the schools of art, of being able to comprehend what others discover, and to remember what others teach. Even those to whom Providence hath allotted greater strength of understanding can expect only to improve a single science. In every other part of learning they must be content to follow opinions which they are not able to examine; and, even in that which they claim as peculiarly their own, can seldom add more than some small particle of knowledge to the hereditary stock devolved to them from ancient times, the collective labour of a thousand intellects."
Johnson: Rambler #121 (May 14, 1751)

The great man did not need to spell out that his intellect was superior to many of the best and brightest with whom he conversed (or gored and tossed, as Boswell lamented). He knew he had a great faculty of mind, and enjoyed it, while retaining his humility.

My aim is to be an intelligent hearer, able to comprehend what others discover; able to describe their findings clearly, mostly in commendation though sometimes with detailed reservations; and thus to add a small particle of knowledge to the collective labour of a thousand intellects.

Are your feelings easily hurt?


If so, you are probably also a worrier, moody, irritable, nervous, fed-up, tense, lonely and guilty. In a word: Neurotic.

I have a distant and mature understanding of such propensities. Not that I am a worrier, of course, but simply that, very occasionally, I find myself worrying about things which may never happen, and becoming gloomy and anxious as a result. Only every now and then. I would explain it further and give more lurid examples, but why tempt Fate? Neurosis is bad enough without Nemesis.

So, how do we explain what makes people like us neurotic? Many people (myself included) are tempted to look back at their childhoods, identifying events which were painful and which would make just about everyone worry if life was worth living. Losing one’s keys, for example.

Perhaps the cluster of anxious or “vigilant” attitudes to life have a genetic component.

Genome-wide analysis of over 106,000 individuals identifies 9 neuroticism-associated loci

The authors say:

We report a genome-wide association study of neuroticism in 91,370 participants
of the UK Biobank cohort and a combined meta-analysis which includes a further 7,197 participants  from the Generation Scotland Scottish Family Health Study (GS:SFHS) and 8,687 participants from a Queensland Institute of Medical Research (QIMR) cohort.  All participants were assessed using the  same neuroticism instrument, the Eysenck Personality Questionnaire-Revised (EPQ-R-S) Short Form’s Neuroticism scale.  We found a SNP-based heritability estimate for neuroticism of approximately 15% (SE = 0.7%).  Meta-analysis identified 9 novel loci associated with neuroticism.  The strongest evidence for association was at a locus on chromosome 8 (p = 1.28x10-15) spanning 4 Mb and containing at least 36 genes.  Other associated loci included genes of interest on chromosome 1
(GRIK3, glutamate receptor ionotropic kainate 3), chromosome 4 (KLHL2, Kelch-like protein 2), chromosome 17 (CRHR1, corticotropin-releasing hormone receptor 1 and MAPT, microtubule associated protein Tau), and on chromosome 18 (CELF4, CUGBP elav-like family member 4).  We found no evidence for genetic differences in the common allelic architecture of neuroticism by sex. 
By comparing our findings with those of the Psychiatric Genetics Consortia, we identified a large genetic correlation between neuroticism and MDD (0.64) and a smaller genetic correlation with schizophrenia (0.22) but not with bipolar disorder.  Polygenic scores derived from the primary UK Biobank sample captured about 1% of the variance in trait liability to neuroticism. Overall, our findings confirm a polygenic basis for neuroticism and substantial shared genetic architecture
between neuroticism and MDD (major depressive disorder).  The identification of 9 new neuroticism-associated loci will drive forward future work on the neurobiology of neuroticism and related phenotypes.


As you all know, individual differences in neuroticism are highly stable across the life course and being neurotic is associated with considerable public health and economic costs, premature mortality, and a range of negative emotional states and psychiatric disorders, including major depressive disorder, anxiety disorders, substance misuse, personality disorders and schizophrenia, so it is an important aspect of personality and may also explain the some of the causes of psychiatric disorders.

Mean neuroticism scores were lower for men than for women (men mean EPQ-R-S = 3.58, SD = 3.19; women mean EPQ-R-S = 4.58, SD = 3.26; p = 0.001).  Principal component analysis of the 12 EPQ-R-S items showed that all items loaded highly on a single component, and the internal consistency (Cronbach alpha) coefficient was


Women worry much more than men. Crudely speaking, 28% more. Although only 1 man in 10 is totally phlegmatic and stable, they are twice as common as totally phlegmatic and stable women. Remember this when things go bump in the night. The authors found no evidence for genetic differences in the common allelic architecture of neuroticism by sex, suggesting to me that they have more of the same rather than something different.


A very good paper with a large sample and a clear result, which identifies 9 loci of interest where only 1 had been shown in previous research.

So, we have a first step, and future work may well push up the variance in neuroticism accounted for by the genome. There is still plenty scope for much of neuroticism to be caused by unfeeling parents, boarding schools, war zones other than boarding schools, sudden noises, and the smell of steak in passageways.

Thursday 19 November 2015


If you ever needed an illustration of the vast yawning gap between the mainstream narrative about disaffected Muslims travelling from Europe to join ISIS and discussions conducted in the better informed parts of the blogosphere, here is an illustrative example. We start with a post by Steve Sailer.

Steve takes up a discussion on Marginal Revolution, where a commentator known as does an interesting calculation: PISA scores for first and second generation immigrants are compared to the PISA scores for Europeans,  and the gap with the second generation is plotted against the number of people who have left Europe to join ISIS.

First, while mainstream media revolves round named journalists, who are indeed trying to make a name for themselves by establishing bylines, the blogosphere is a mixture of named and anonymous contributors, with a preponderance of the latter. Presumably they think  their comments will draw hostility and even sanctions against them, possibly losing their jobs. It happens.

Second, while mainstream accounts are historical, political, and cultural in their primary focus, and likely to discuss Muslim enclaves in terms of poverty, unemployment and cultural disaffection; blogosphere accounts cover those, but also include cognitive and scholastic ability and other behavioural measures like levels of violence in their countries of origin.

Third, mainstream accounts are usually mostly news reportage, with some general political and economic content and opinion. Blogosphere discussions tend to dig up data sources and more detailed publications. The big difference between mainstream reportage and blogs is that the latter tend to give links to data sets. Perhaps we should always distinguish between linked and unlinked reportage, or just set aside any discursive account which does not give references.

In that spirit, here is some background reading, with references:

The IQ estimates in this post are drawn from PISA and the (rough) estimates of numbers of jihadis from:

However, this is a very rough estimate because a) who knows for sure how many would-be jihadis have made this journey? and b) the numbers are being compared with total populations in each country, rather than Muslim populations in each country. Digging up those numbers would give us a better estimate of the conversion rate from moderate to militant Muslim beliefs and actions. With more time we could compare PISA intelligence estimates with actual national IQ measures, but at least PISA gives a common yardstick, and is more readily accepted.



You will see that the light blue line of Jihadi numbers per million corresponds quite well with the darker blue discontinuous line for the Gap between natives and 2nd generation immigrants. At the top of the graph you can see the purple line which gives the IQ for natives. Remember, this is not a time graph, but a country graph, so the only time element is the difference between 1st and 2nd generation immigrants, which will partly reflect the individual immigration patterns of each nation.

The number of nations for which we have fairly reliable jihadi estimates (11) is too small for detailed statistical analysis, but the implication is that ISIS membership is a second generation effect. Crude correlation coefficients are: ISIS membership and 1st generation immigrant IQ r= –0.24 (n.s.), ISIS and 2nd generation immigrant IQ r= –0.78 (p<.01)  ISIS and gap between 2nd generation immigrant and native IQs r=-0.84 (p<.01). Just as a check, correlation between ISIS and native IQ is r= 0.12 (n.s.).

In some ways the most interesting data are the simplest: second generation ability has dropped from first generation by 2.7 points in Belgium, 4.54 points in England, and 2.5 points in Portugal. The overall results for the 11 countries are:

First generation immigrants    89.04

Second generation immigrants 90.91

Natives                                       97.85

On average there is a 7 IQ point deficit between second generation immigrants and natives. This is highly significant. If jobs are given on ability alone, then there will be a big reduction in the number of immigrants obtaining cognitively demanding occupations which, because of the rarity of high ability, tend to be high status and well paid. “Small differences in means are great at the extremes”.

For example, using Emil’s visualiser, we can see the real life implications of the natives being IQ98 and the second generation immigrants being IQ91

The natives, (painted blue like the ancient Saxons) will have 3.5 times more bright citizens (IQ130) than the immigrants (coded red). Immigrants will be rarely found in the top universities and most prestigious professional occupations. If Europe as a whole demands a Greenwich Mean IQ of 93 for any proper, paid occupation (probably a reasonable estimate of what is required in a non-subsidised occupation), then 63% of the natives will be in employment, but only 45% of the immigrants. Finally, if we look at very low level jobs, which result in compensatory payments from the taxpayer, say those reserved for those below IQ85 then 19% of the natives will be in this unfavoured category, but a substantial 34% of the immigrant group.

If European societies feel that, through embarrassment, they cannot talk about ability, then the only interpretation of different outcomes (all people being judged equal in ability as a matter of principle) is that the immigrant group have been subject to a massive injustice due to native prejudice. In point of fact they are imposing a significant cost on the natives, but if IQ cannot be discussed, and the possibility of a substantial genetic component in intelligence can never even be contemplated, on pain of banishment, then Europe has created a fertile soil for resentment, envy and hatred, while remaining mute about its contribution.

In the spirit of the blogosphere, I should point out that all these figures should be looked at again, and replaced with better estimates wherever possible. We certainly need better estimates of Muslim numbers in each European country, particularly for young men. The causal hypothesis that IQ is a major factor should be compared with other testable hypotheses. So, can we please get some blogosphere data checkers to work on this material, and come back with corrections and improvements?

If you post as Anonymous, try to add a number or imaginary name so I can distinguish one Anon from another Anon. I don’t know which particular unknown person you are.

Fun and contentment (plus death and IQ)

Just to give you an indication as to how the citizens of Edinburgh entertain themselves when not hosting the Festival, here is the St Andrew’s day lecture, which will be given by Pat Rabbitt, and will be worth hearing.

The 2015 special invited St Andrews Day seminar will be given by Professor Patrick Rabbitt of Oxford University. The talk will be held at 5pm on 24th November in room F21 of the Department of Psychology, 7 George Square. A drinks reception will follow the talk.
Seminar title:
"Death, intelligence, fun and contentment in old age"
"Many large and reliable longitudinal studies now allow us to explore the relationships between how the extent to which we can keep our wits about us in old age, our past and present health, our nearness to death, the pleasure we get from hobbies, interests and social life and our general level of contentment interact and determine each other. The talk will discuss how the results of analyses completed this year illustrate these relationships."
You do not need to register for this talk. All are welcome to attend.
Tel: 0131 650 4639

Tuesday 17 November 2015

Genetic story jumps ahead

Does the Edinburgh Deary gang never sleep? Just when you expect them to put down their pneumatic drills and have a cup of tea, they come up with two very interesting papers which link the genome to intelligence, personality and health.

We need a name for these sorts of findings: Pleotropic Pandora-ism? The System Integrity Nexus? The Infinite and Ubiquitous Network of Causes? (OK, I stole the last one from Borges, “la infinita/Y ubicua red de causas”  but why not? It was taken from “Elvira de Alvear” a beautiful poem written about a generous and courteous lady who suffered a long mental collapse, so it is close to this team’s work on mind and health and ageing).

Anyway, what have these industrious Scots been up to?

Shared genetic aetiology between cognitive functions and physical and mental health in UK Biobank (N = 112 151) and 24 GWAS consortia.

They say: The causes of the known associations between poorer cognitive function and many adverse neuropsychiatric outcomes, poorer physical health, and earlier death remain unknown. We used linkage disequilibrium regression and polygenic profile scoring to test for shared genetic aetiology between cognitive functions and neuropsychiatric disorders and physical health. Using information provided by many published genome-wide association study consortia, we created polygenic profile scores for 24 vascular-metabolic, neuropsychiatric, physiological-anthropometric, and cognitive traits in the participants of UK Biobank, a very large population-based sample (N = 112 151). Pleiotropy between cognitive and health traits was quantified by deriving genetic correlations using summary genome-wide association study statistics applied to the method of linkage disequilibrium regression. Substantial and significant genetic correlations were observed between cognitive test scores in the UK Biobank sample and many of the mental and physical health-related traits and disorders assessed here. In addition, highly significant associations were observed between the cognitive test scores in the UK Biobank sample and many polygenic profile scores, including coronary artery disease, stroke, Alzheimer's disease, schizophrenia, autism, major depressive disorder, BMI, intracranial volume, infant head circumference, and childhood cognitive ability. Where disease diagnosis was available for UK Biobank participants we were able to show that these results were not confounded by those who had the relevant disease. These findings indicate that a substantial level of pleiotropy exists between cognitive abilities and many human mental and physical health disorders and traits and that it can be used to predict phenotypic variance across samples.

Well, in decades gone by, few would have predicted that. Intelligence was something which applied to school work (if it applied to anything at all) and perhaps to the evaluation of brain-damaged patients and child development, but it had nothing to do with serious matters like physical and severe mental health disorders. Even though the general drift of modern genetic research has been going in this direction, the results are still astounding.

Three cognitive tests were used. The Reaction Time test was a computerized ‘Snap’ game, in which participants were to press a button as quickly as possible when two ‘cards’ on screen were matching. There were eight experimental trials, with a Cronbach α reliability of 0.85. In the Memory test, participants were shown a set of twelve cards (six pairs) on a computer screen for five seconds, and had to recall which were matching after the cards had been obscured. We used the number of errors in this task as the (inverse) measure of Memory ability. The Verbal-numerical Reasoning task involved a series of thirteen items assessing verbal and arithmetical deduction (Cronbach α reliability = 0.62).

Linkage Disequilibrium score regression was used to derive genetic correlations to determine the degree to which the polygenic architecture of a trait overlaps with that of another. Next, the polygenic risk score method was used to test the extent to which these genetic correlations are predictive of phenotypic variance across samples. Both LD score regression and polygenic risk scores are dependent on the traits analysed being highly polygenic in nature, i.e. where a large number of variants of small effect contribute toward phenotypic variation.

Here is a table which one day will find its way into psychology textbooks. The correlations above the diagonal shouldn’t really exist, and yet there they are. They show a link between the chemical code of life and the skills of life itself.


In Table 2 they show links between the cognitive measures and: coronary heart disease, ischaemic stroke, Alzheimer’s Disease; Autism, Bipolar disease; Major depressive disorder, Schizophrenia, Intercranial volume, Infant Head Circumference, BMI, height, childhood cognitive ability, college education and years of education. The last three load on verbal-numerical ability, but not memory or reaction time. Perhaps all this is better depicted on a heat map:



The picture is becoming clearer. There is a link between cognitive ability and health because the causal code of life for a good body and a good mind operates on shared pathways.

How about personality?

Pleiotropy between neuroticism and physical and mental health: findings from 108,038 men and women in UK Biobank.

There is considerable evidence that people with higher levels of the personality trait of neuroticism have an increased risk of several types of mental disorder. Higher neuroticism has also been associated, less consistently, with increased risk of various physical health outcomes. We hypothesised that these associations may, in part, be due to shared genetic influences. We tested for pleiotropy between neuroticism and 12 mental and physical diseases or health traits using linkage disequilibrium regression and polygenic profile scoring. Genetic correlations were derived between neuroticism scores in 108,038 people in UK Biobank and health-related measures from 12 large genome-wide association studies(GWAS). Summary information for the 12 GWAS was used to create polygenic risk scores for the health-related measures in the UK Biobank participants. Associations between the health-related polygenic scores and neuroticism were examined using regression, adjusting for age, sex, genotyping batch, genotyping array, assessment centre, and population stratification. Genetic correlations were identified between neuroticism and anorexia nervosa(rg = 0.17), major depressive disorder (rg = 0.66) and schizophrenia (rg = 0.21). Polygenic risk for several health-related measures were associated with neuroticism, in a positive direction in the case of bipolar disorder (β = 0.017), major depressive disorder (β = 0.036), schizophrenia (β = 0.036), and coronary artery disease (β = 0.011), and in a negative direction in the case of BMI (β = -0.0095). These findings indicate that a high level of pleiotropy exists between neuroticism and some measures of mental and physical health, particularly major depressive disorder and schizophrenia.

Participants completed the Neuroticism scale of the Eysenck Personality QuestionnaireRevised Short Form (EPQ-R Short Form). This scale has been concurrently validated in older people against two of the most widely-used measures of neuroticism, taken from the International Personality Item Pool (IPIP) and the NEO-Five Factor Inventory (NEO-FFI); it correlated -0.84 with the IPIP-Emotional Stability scale and 0.85 with the NEO-FFI Neuroticism scale. A previous study found a high genetic correlation (0.91) between the EPQ-R Short Form Neuroticism scale and psychological distress assessed in a nonpsychiatric population using the 30-item General Health Questionnaire.

Neuroticism is not a good trait to have (say I with a worried expression on my face). People who are higher in neuroticism have an increased risk of developing common mental disorders such as mood, anxiety, somatoform and substance use disorders, and also schizophrenia, bipolar disorder and attention deficit hyperactivity disorder. Higher neuroticism is associated with personality disorders, major depression, generalised anxiety, panic disorders and phobias, and alcohol and drug dependence, antisocial personality and conduct disorders. It is even linked with risk of developing Alzheimer’s disease. Perhaps, if you are of similarly “vigilant” disposition, I should have warned you to skip the above paragraph, but I knew that your eyes would be anxiously drawn to this gloomy list anyway, so why waste time in pointless reassurance? We are doomed, utterly doomed.


Here is a picture of the results:


In very quick summary, three domains of enquiry: cognitive ability, personality and health, have been brought together and shown to rely on common genetic pathways, a causal overlap of significant proportions. These two papers follow a steady drumbeat of high quality research, showing genetic associations with a broad range of important human behaviours. However, although they are in that tradition, they also reveal a considerable speeding up of the discovery rate, and in the power of the findings. These results will cause excitement in informed circles (and has already done so among the first readers) and have a very good chance of being seen as landmark papers. 

Saturday 14 November 2015

Getting murder into proportion


A mourner pays his respect outside the Le Carillon restaurant the morning after a series of deadly attacks in Paris that killed at least 127


After the atrocities in Paris last night, it may seem the wrong time to talk about statistics, yet much of the subsequent commentary has involved statistical considerations, albeit implicit. A common concern is that because the terrorists chanted Islamic slogans it might be interpreted as proving that all Muslims are terrorists. Hence the frequent explanations that not all Muslims are terrorists. If anyone draws that conclusion, they would of course be mistaken.

Most Muslims are not suicide bombers.

However, that does not end the statistical discussion, because there is another statement which is equally true.

Most suicide bombers are Muslims.

The following commentary on those two observations is hardly snappy, but it covers the observed facts: Most Muslims are not suicide bombers. Suicide bombers are a statistical rarity among Muslims. However, if you look at the last 20 years of suicide bombing (people with bombs strapped to their bodies, who may also carry other weapons with which they kill members of the general public) then the majority are Muslim. Yes, those deaths are few when compared to other deaths, including violent deaths, but these murders have a particular extra sting of intentionality: they are conducted at random against people for for simply being what they are: civilians going about their lives.

People are not fools. They can understand the normal processes of disease, and of the accidents that sometimes happen. They understand that national murder figures include criminals fighting each other over territory. They find it much harder to accept being attacked by someone bent on the pitiless destruction of their society.  Hence the reason why members of the public are often alarmed by militant Muslims, fear them, and want to avoid them. They do so fully knowing that most Muslims are not terrorists and also knowing that most political, random violence attacks against them are carried out by Muslim terrorists. Being Muslim is a distinguishing characteristic of contemporary terrorism directed against the West. It is very, very weakly predictive, but it is not zero, and people tend to think in general categories, not in statistical gradations.

During what were euphemistically referred to as The Troubles, the mainland bombings  and gun attacks on the United Kingdom were carried out by Nationalist, Northern Ireland Roman Catholics with Irish Republican Army membership. Their being Christians, at least nominally, was not a distinguishing feature, though Roman Catholicism in Northern Ireland was arguably a relevant aspect of their upbringing. (Schooling was largely religiously segregated). British citizens were variously worried about Catholics, nationalists, and the Irish in general. They knew the bombers were a minority. They suspected that they had covert support from many Catholics in Northern Ireland, and big turnouts at IRA funerals seemed to confirm that. Irish accents were not a prized characteristic.

Political murders aside, the proportions of murderers vary considerably from one nation to another. Murder rates per million range from 3 in Japan, to 10 in the United Kingdom and France, roughly 20 in North Africa and Middle East (Syria before the war 22), 30 in Taiwan, 40 in Fiji, 50 in Mauritania, 61 in Ghana, 71 in Eritrea, 80 in Eastern Africa, 90 in Russia, 100 in Middle Africa, 111 in Madagascar, 120 in Ethiopia, 133 in Grenada, 147 in the Cayman Islands, 152 in Myanmar, 170 in Guyana, 184 in Botswana, 193 in Equatorial Guinea, 200 in Nigeria, and then upwards through the Caribbean, Africa and Central America, till we get to 308 in Columbia, 412 in El Salvador, 526 in US Virgin Islands, 537 in Venezuela, and 904 in Honduras.

Not everyone in Honduras is a murderer (merely 1 in 1106 Hondurans), but there is an appreciable 300 fold increase in the risk of being murdered when compared with Japan. You are perfectly entitled to avoid Honduras and holiday elsewhere. Indeed, it is up to you at precisely what murder rate you choose not to visit a country. As an adult you can set your own risk preferences, and act accordingly.

To get things into proportion, this new attack on Paris is a very significant event. Some reports say that the perpetrators clearly had French accents. That is also very significant.  The wider issues of the slaughter in Paris were prefigured in January, with the attack on journalists and a Jewish supermarket.

I said then: Islam has a prominent militant wing, a death meme which finds resonance in many a self-important young idiot.

I do not know how one achieves a Reformation in fundamentalist Islamic beliefs, or even if that is possible, but I hope that someone does, soon.

Thursday 12 November 2015

London slaves case comes to Court

You may remember the London slaves case in November 2015, which suggested to horrified newspaper readers that people had been brought into the country and kept as domestic slaves. The case was shocking because the women concerned were held for 30 years.

Helped by the revulsion to this example of “slavery” the Modern Slavery Act 2015 was accepted by Parliament shortly afterwards, and following agreement by both Houses on the text of the Bill it received Royal Assent on 26 March 2015.

In fact, within a few weeks of the women walking out of the house the whole narrative had collapsed. This turned out to be a case of a household of browbeaten former Maoist women groupies taken over by a domineering paranoid male. He is now on trial, and will very probably be convicted. I imagine that after arguing that the women consented to everything he did, he will be sent for psychiatric examination and be found to have either paranoid schizophrenia or a severe personality disorder, or both.

Odd, isn’t it, when psychiatric diagnoses seem better founded than Parliamentary legislation?

Wednesday 11 November 2015





By chance I was in Aylesbury today, for the first time in 50 years since that day long ago when I went with a college friend to see his father, who gave us lunch in his garden. Aylesbury was built on an 4th century BC Iron Age hill fort and was an important market town from Anglo-Saxon times onwards. The Grammar School was founded in 1598.  The town was of Puritan sentiments, and backed the Parliamentarians in the Civil War.

I searched for the town centre, of which some traces remained in the quiet streets by the church of St Mary, with fine houses, alms houses provided by a benefactor in the 18th century, pubs and the idiosyncratic architecture of independent minds. In the church the school kids gave a spirited rendition of “It’s a long way to Tipperary” and  past Prebendal House (where the radical MP John Wilkes lived) I met a woman who appeared to be a local. I mention that fact because in the market square the crowd were roughly 20% not English (by 2011 10% of the population were Islamic), with many women with head scarves and some with fully veiled faces, many Middle Easteners and Africans, a few of the latter in wheelchairs being pushed by carers. The lady and I discussed the demolition of so many historical buildings (done in the 1960 to make way for new shops), the beauty of the remaining streets, and she gave me advice on the least-bad coffee shop.

Near the market square the poppy wearing legions waited and, remembering the date, I joined the crowd. Town Mayor with chains of office, British legions with regimental caps, flag bearers, respectful crowd. In the background by the market stalls stood a respectful, fully be-gowned and be-wigged judge who, being spotted, was invited to join the other dignitaries. Then the traditional ceremony, the flags lowered, “at the going down of the sun and in the morning we will remember them.”, two minutes silence,  the local clock striking 11 (eleventh hour of the eleventh day of the eleventh month) and then a mistimed single gun salute which startled all of us. “Got the timing wrong” muttered a knowledgeable lady afterwards, but she agreed with my suggestion that their intentions were better than their watches.

So, nothing. Just a little English country town. I admired the crowd remembering their history. I thought of a cousin who died in the Pathfinder squadron. I regretted much of the recent architecture. I also felt it was the Last Post for a passing age, and perhaps a dying people. I wondered whether the newcomers would understand the past, and respect it. It was not primarily their war, and none were at the ceremony. The cross on the war memorial, which had always seemed normal, now seemed questionable.  There were no trumpeters, just a recording played on a sound system. A century has almost passed since the Great War, and there may be a case for letting these memories fade, but these are one of the habits of our tribe, and it seems churlish to ever forget that for this, our living tomorrow, they gave their today.

Monday 9 November 2015

Chisala and Powerpoint Publishing


Just because research hasn’t been published doesn’t mean it’s right.

Chandra Chisala has continued a line of argument based on results from some schools-based research in the UK and USA. I commented on his first post here:

In brief, it included unpublished work on particular schools chosen because African students were doing particularly well, with hard to track down references and incomplete methods sections.

Chisala has now posted a Part 2 which gives unpublished results from Seattle schools. Compared to Part 1 (which included national UK exam results) there is even less detailed material available. Everything said may be true, but the exhibits are not available for inspection, so it repeats the style of his first article, in which it is very difficult to trace references back to the published source. Call it “Powerpoint Publishing”.

Chisala summarises his argument thus: Remember our goal. We only need to show that blacks in Africa would have a higher average IQ than native black Americans if they were moved from Africa to America since the African environment clearly depresses IQ, as both environmentalists and hereditarians agree in principle. This result would mean that whatever “problem” the black Americans have that result in such a large and intractable IQ gap with whites and other groups, has nothing whatsoever to do with the genetic evolution of races, especially since they even have more white genes than Africans. It is not their sub-Saharan African (black) genes that are responsible for their chronic academic under-achievement; it has to be a factor that is endemic to African American history.

There are many assumptions in this line of reasoning. The main one is the assumption that recent African migrants to the USA are a representative sample of Africans in Africa. This is not unreasonable, but one would need a better understanding of the family backgrounds to be confident about it. The second assumption is that the African environment is universally bad. In fact, Africa is propelling more of its people into better standards of life, by the open-market business methods which work everywhere. Health is improving, as is school attendance. Many African countries are showing a Flynn effect. Kenya has gained 8 IQ points per decade, and the global picture is positive, though convergence in maths will be a long time coming (Meisenberg and Woodley. Intelligence 41, 2013 808-816). The third assumption is that we can sort out the relative contributions of nature and nurture by looking only at Africans. Asians also suffered poverty in the post colonial period, so that recently enriched nations like Vietnam are also relevant. Poor malnourished Vietnamese refugees fled Communism decades ago (unclear whether they were from cognitive elites). The fourth assumption is that “factors endemic to African American history” are unique. As I keep explaining, most African slaves ended up in Brazil, so the achievements of Africans in Brazil are also relevant, and a particularly interesting test case, since there was a much more relaxed attitude to intermingling.

There central element in this article is a Powerpoint presentation of Seattle schools results, which begins with a disclaimer that using “language spoken at home” to infer race does not necessarily map onto genetic groups. The education authority says Please note: this is an important and critical limitation of this study. The slides make it clear that:  The “Admission Form” also collects specific primary race data for Asians and Native‐Americans Americans and ethnicity for Hispanics Hispanics per state regulations. This data is not collected in specific detail for Whites or Blacks / African‐ Americans.  So, we don’t have specific ethnic details on the key groups being compared.  A supplementary table in a published paper might provide further data to assist in estimating this potential error term. It would have been helpful to have included this disclaimer in the article.

Chisala says: The fact that these are only group pass rates (on mathematics and reading) does not matter for purposes of ethnic IQ comparison since the pass rate positions correlate perfectly with expected mean IQ score ranks of the groups before disaggregation (that’s the same logic we were using for GCSE pass rate comparisons, especially when mathematics is included).

The problem with using pass rates rather than actual scores is that if pass rates are raised by making tests easier, then it will appear that group differences are reduced. If any test gives generous marks to students who simply attempt answers to questions, and indicate knowledge of very basic terms, then real differences in competence are obscured. The Seattle data are based on overall district pass rates of up to 70%, which is understandable for an education authority, but loses a lot of detail. Nonetheless, most findings impart some information, and taken at face value the results are interesting. In the Bayesian spirit of “hunt the submarine” one should try to work out the truth of an important matter using the best available data, however slight. Chisala has potentially found something interesting, something which opens the door to new hypotheses. Looking at language differences gives additional information. However, there is also a potential distortion (apart from not knowing how language maps on to race) which is that brighter children pick up new languages more quickly, such that those in the “not needing to go to English classes” are very probably brighter, or conversely, have had much longer to learn the language (another detail it would be good to have).

For example, first including and then excluding those with poor English, pass rates for Maths are:

Asians and Pacific Islanders: Chinese 87-91%, Japanese 88-89%, Korean 88-87%, Vietnamese 75-82%, Filipino 67-72%,  Indian 66-74%, Samoan 39-44%

Hispanic and Latinos: Hispanic English Speakers 58-58%   Hispanic Spanish Speakers 39-54%

Black and African Americans: Amharic 51-62%,  Tigrigna 46-58%, Oromo 39-53%, English 36-36%, Somali 28-47%

Comment: Overall, those who need English language teaching are less able to do maths. African American (Black English-speakers in this classification) do badly on Maths, but in the same sort of range as Samoans and Somalis. Sample sizes are generally reasonable, but rather low for Amharic (143), Tigrigna (106) and Oromo (94)  and their representativeness is unknown.  Chinese and Vietnamese do very well, despite coming from previously very poor countries. Personally, I do not see a clear pattern of environmentally deleterious effects here.

The presentation also gives a measure which includes attendance and discipline, where 3 is the district average and 10 means not likely to complete high school. By this measure the following are at risk: Somali 5.5, Samoan 5.5, African Americans 5.4, Spanish Speaking Hispanics 5.3, Oromo 5.1 which suggests that a mixture of genetic and cultural elements are involved. African descent is still a partial but plausible contributing factor to poor school progress.

Chisala agrees that proper representative samples would be the most informative, but in the absence of those is using some available results on individual school districts, and extrapolating from those. Although this is a weaker method than proper sampling, it can sometimes achieve informative results. For example, if high achievement were found in disproportionate numbers among African students, relative to the number of Africans in the world then one could estimate, from that extreme high-performing group, the likely average intelligence from the population from which they are drawn. This is a particularly useful method if there is no other ability data, or you  want to trust only data based on open competition, like chess tournaments where players get Elo rankings based on win/lose scores.

If the number of very bright Africans is higher than would be expected from those, say, 2 standard deviation above a mean IQ of 80, then the mean of 80 is called into question. The real mean is likely to be higher, thus accounting for the larger than expected number of those achieving +2 sigma performance. Therefore, get a good estimate of peak African achievement, divide by African population, consult normal curve statistics, work out the implied mean.

Chisala has found some high performing Africans in the UK and the US. Good. What does this tell us about the population from which they are drawn? At the moment, not enough to conclude that the mean of 80 is wrong.

Perhaps I need to spell out my concerns more clearly. I don’t think one can draw firm conclusions from this sort of reporting of results. It is simply not good enough to say the research was not intended for publication. It cannot be evaluated until we have been able to read it properly. (Just seeing the full report would be enough: it does not need to be in a journal. I have emailed the Seattle School dept asking if they have any reports they can send me).  It is premature to rush to conclusions about what this means for various hypotheses when we haven’t got sufficient detail on the basic results.

Tuesday 3 November 2015

Deutschland über alles, dann unter allen?


The Flynn Effect is a funny mixture: part IQ inflation, part civilizational advantage. It seems that ever since intelligence testing began, people have been getting smarter. That is not too surprising. The first world war partly destroyed Europe, but the slaughter caused by the Gatling gun was matched by the ability to bring food across the world, preserved in tin cans and refrigerated ships. Nutrition improved. The second world war brought destruction world wide, but also spurred innovation. Since 1945 the world has got richer. More of the world was exposed to better nutrition, health and education. Since the 1990s the world has got healthier and better educated and even richer. Complain as we might, most people on the planet are living better lives than ever before. All this should boost the human condition, and as we become more civilised, both mind and body improve. Test scores record this improvement. This is the rosy view of rising intelligence.

The cautious view is that modern life sets us more IQ type questions, and problems which seemed out of the ordinary in the early part of the 20th Century soon became commonplace, because of better and more widespread education. IQ tests have become less puzzling, less of a test, because they look like familiar items in regular school tests. Therefore, the apparent gains are hollow, and there is no real life proof that we are getting as much brighter as the tests suggest. We are suffering IQ inflation, not a real increase in intellectual wealth.

The contrary view is that if you use proper tests: reaction times, digit span, colour perception, then there is slight but unmistakeable evidence of mental decay, and falling real intelligence.

And now to Germany.

A reversal of the Flynn effect for spatial perception in German-speaking countries: Evidence from a cross-temporal IRT-based meta-analysis (1977–2014) Jakob Pietschnig and Georg Gittler. Intelligence 53 (2015) 145–153.

In their abstract Pietschnig and Gittler say: Generational IQ changes (the Flynn effect) have been shown to be predominantly positive but differentiated according to IQ domains and countries. However, evidence from recent studies points towards a decrease of the Flynn effect globally or even a reversal in some countries. In the present meta-analysis, we show an inverse u-shaped trajectory of IQ test performance changes in a large number of samples (k = 96; N = 13,172) on a well known test for spatial perception (the three-dimensional cubes test, 3DC) in German-speaking countries over 38 years (1977–2014). Assessment of both item response theory-based measures as well as more standard measures of classical test theory showed initial increases and a subsequent decrease of performance when controlling for age, sample type (general population vs. mixed samples vs. university students) and sex. Our results suggest saturation and diminishing returns of IQ increasing factors (e.g., life history speed) whilst negative associations of IQ changes with psychometric g may have led to the observed IQ score decrease in more recent years.

Here are the u-shaped historical trends, shown either as ability parameters or percentage of solved items, both giving the same overall picture.

German nv IQ

So, has Germany abolished itself?

The authors give some background: Worldwide IQ gains have been shown to have started to slow down in the past decades across all domains. This has been interpreted as a potential predecessor for a stagnation or even reversal of gains in the future (Pietschnig & Voracek, 2015). In fact, stagnation of gains has been reported in data from Norway and Sweden (Emanuelsson, Reuterberg, & Svensson, 1993; Sundet, Barlaug, & Torjussen, 2004). More recent findings even indicate a reversal of IQ gains in Denmark, Finland, and France in past decades (Dutton & Lynn, 2013, 2015; Teasdale & Owen, 2005). Interestingly, declines in reaction times over more than one century indicate that contrary to the frequently observed IQ gains, generational changes in psychometric g may have been in fact negative in Western countries (Woodley, te Nijenhuis, & Murphy, 2013, 2014; Woodley of Menie, te Nijenhuis, & Murphy, 2015). Therefore, it seems likely that IQ changes in further countries may eventually show a similar trajectory of an initial stagnation and subsequent decrease of IQ test performance

They go on to make a crucial point: IRT-based estimates [are] more accurate than conventional raw scores within the framework of classical test theory, where each correct item solution is equivalent to one raw score point, regardless of the item's difficulty.

We investigate changes in spatial IQ performance of the general population on the three-dimensional cubes test (3DC) over a time-span of 38 years in German speaking countries. In our main analysis, we focus on changes of IRT based mean person parameters of samples and support our findings by supplementary analyses of estimates derived from classical test theory. Moreover, we investigate influences of sample age, sample type (general population vs. mixed samples vs. university students), and sex on spatial task performance.

In contrast to many cognitive ability measures, item difficulty of the 3DC is not progressively increasing throughout the test, but rather easier and more difficult items alternate

In all, 96 independent samples (N = 13,172) from 76 published and unpublished studies were included in the present meta-analysis. Participants in primary studies were predominantly from healthy convenience samples in schools, Universities, and from the general population.

Here is the key result: Results from our linear regression analyses suggest decreases of about 4.8 IQ points per decade when controlling for age, sample type, and sex, thus indicating a substantial negative Flynn effect that is even stronger compared to previously observed positive trends (e.g., Flynn, 1984, 1987; Pietschnig & Voracek, 2015). This trend was observed in linear regression analyses, but our results showed that the present changes over time may be even better described as a curvilinear function, thus indicating initial increases, followed by stagnation (with performance peaking around the mid-1990s), and subsequent decreases of task performance. This curvilinear relationship emerged for both IRT- as well as non-IRT-based measures and remained robust when controlling for age, sample type, and sex, thus corroborating stability of our results and further corroborating validity of the Rasch model in our samples.

The fall in intelligence is quite severe. The authors soberly go through a list of possible explanations for IQ having been boosted by cultural changes but then subject to diminishing returns, finally tentatively landing on: One possible reason may be that the above discussed IQ-boosting factors have masked the g-based ability decrease until a point of saturation was reached. If this is so, the zenith of our curvilinear regression would approximately give the point where beneficial factors were eventually outperformed by the negative trend in g.

If life is dandy, particularly in wealthy Germany, workshop of the world, what could be causing a drop in intelligence?  The authors comment: Other potential reasons that have been cited for decreasing IQ test scores such as changing population ability levels due to non-Western immigration (e.g., Rindermann & Thompson, 2014) or dysgenic fertility patterns (e.g., Lynn, 2011) do not seem suitable to contribute substantially to our present findings. On the one hand, effects of immigration on national IQ levels were observed not to be long-lasting with performance gaps diminishing over time (te Nijenhuis, de Jong, Evers, & van der Flier, 2004). On the other hand, both immigration and dysgenic fertility effects have been shown to be too small to provide substantial contributions for our present findings (Meisenberg & Kaul, 2010; Rindermann & Thompson, 2014).

Comment: Rindermann & Thompson (2014) calculate that, on average, the mean natives' and immigrants' competence gap is equivalent to 4.71 IQ points, which by coincidence is precisely the drop experienced in Germany. How much of an effect do immigrants contribute?

(By way of a brief aside, as of 2012 Germany was 80% German. Another 3.9% are European, but that includes a substantial 1.9% Polish (IQ92) 1% Italian (IQ97) 0.5% Romanian (IQ91), 0.5% Greek (IQ92) and 2.9% others Spanish(97), Croatian (99), Dutch, Portuguese and Austrians. Many but not all are significantly below the German mean of IQ99.   3.7% of the population are Turkish (IQ88) and this should have a bigger effect.

Here is the entry for Germany in our 2014 paper:

Natives constitute 81% of the population, their competence is (IQ) 100.99

Migrants constitute 19% of the population, their competence is (IQ) 92.75

Competence difference –8.26 IQ points

This certainly looks like it should push Germany back a bit, but at only one fifth of the population the migrant effect is still relatively slight and takes the national mean down only to 99.4. However, if we plot migration numbers by year it might match some of the effect, so it could be a partial contributor, though far from the whole story. Dysgenic trends might be a further contributor, so we can follow that argument by seeing what Woodley makes of a similar drop in ability in France, which I hope to post about shortly.

The authors’ final words are: This may most likely be due to saturation and diminishing returns of IQ boosting factors (e.g., life history speed) and a manifestation of declining psychometric g.

Comment: Saturation and diminishing returns, as the authors make clear, would result in improvement being sustained at a plateau. In fact, in these results abilities have fallen as fast and as far as they had previously risen, which is an alarming finding. All that could explain that (if immigration is taken out of the picture) is declining psychometric g, which presents Germany with a substantial German problem. This is particularly sobering because the sample is younger than the general population (84% are below 30 years of age) so the young generation is duller than the older, not the usual pattern. (Incidentally, it makes me wonder whether immigrants, with greater family size are contributing more to this younger population).

Here is a testable prediction: if Germany takes in the projected numbers of Middle East immigrants, then that will boost the immigrant origins population by at least 1 million. Once their families join them the number could be 3 or 4 million. In all probability this will have a negative effect on German national intelligence levels. (It makes no difference whether they are from Syria or Pakistan, since roughly IQ 83 is expected from either). However, we do not need to wait very long to test all this. If a German researcher is quick about it, they could pop round the reception centres and give the three dimensional cube test to as many new immigrants as possible, and we could have the answer in a few months. Indeed, giving a few more intelligence tests would be predictive for training and occupational placement purposes. A thesis in it for someone?