Sunday, June 17, 2018


The rise and fall of average IQ test scores

I was just cranking up my aged brain to say something about the latest IQ findings when I found that young Oliver Moody of "The Times" has spared me the trouble.  His summary is below.  There are a  few things I would like to add, however.

Perhaps the most interesting fact to emerge is that dumb women having more babies is not a problem.  As long as I have been reading the literature on IQ, people have been worried about that.  Are all these smart ladies who think no man is good enough for them degrading the average intelligence of the human race?  Wonder of wonders, the latest research from Norway was able to rule that out. 

Various people have pointed out that the dumbest females for various reasons tend to have NO babies and a majority of high IQ females do have some babies.  And it was always hoped that those two effects would cancel one another out. And we now have grounds to believe that exactly that has happened.

A lot of interesting IQ research comes out of Norway and Sweden.  The reason is that the Scandinavian countries are very authoritarian, which leads them to keep extensive records about each individual person in their countries.  So if you can get access to government data you can base your research on the whole population, not just a sample, with all its attendant doubts and difficulties.

So what we now know with some confidence is that IQ scores rose during the first three quarters of the 20th century but then flattened out before going into a decline.  And that could clearly not be due to genetic changes.  Evolution doesn't work that fast.

So what WAS going on?  There are two major possible explanations: Computer games and education. Blaming computer games has been going on as long as there have been computer games and it is in my mind just snobbery or some such:  A convenient whipping boy for all sorts of ills.  There is actually a fair bit of evidence that games and internet exposure generally are most likely to be good rather than bad for our brains (e.g. HERE and  HERE  and  HERE)

Additionally, like Piaget, I have tended to find the kids in my care to be instructive.  My son, for instance, could load up and play his favourite computer game when he was two and he plays a lot of games to this day now he is in his 30s and works as an IT professional.  And what I saw was that game playing is normally quite social. There will usually be other kids hanging around and talking even with single-user games and some games are quite educational in themselves.  My son learnt most of his ancient history from "sims" set in that era. He learnt precious little ancient history at school.  So I personally exonerate games from being bad for most people.

So what DID go wrong?  Just one thing can account for both the rise and fall in measured IQs:  Testing.

During my schooldays in the '50s testing was all the rage.  We even did IQ tests at least once a year.  And there were heaps of in-school tests. From about grade 3 on, for instance, we would have weekly spelling tests -- in which a kid got a list of 10 words that he had to learn how to spell. Being a born academic, I always got 10 out of 10 and was regularly praised for it.  Which was a bit unfair because I put zero work into it.  I just had to see a word once to know how to spell it. I still do.

And I think that is one example of a huge difference between then and now.  Education used to be COMPETITIVE and "winners" got all the praise. And nobody apologized for that.

It seems to me that there should be no great  difficulty in arranging prizes for both ability and effort but the Left have simply closed their eyes to ability

By about 1975 or thereabouts, however, the political Left had got a vice-like grip on education worldwide.  Even in chapter  48 of my 1974 book, I noted its encroachment. And Leftists HATE competition because it clashes with their idiotic and counterfactual belief that "All men are equal".  To validate that gospel, therefore, all had to have prizes, not just one kid.  And if you believe that all men are equal, there would be no point in testing.  If the marks come out all the same, what would be the point? But the marks don't come out all the same so to avoid that reality, you just don't do testing if you can avoid it.

The rise in measured IQ scores during the first three quarters of c20 has been the cause of much discussion and the most usual explanation for it is that it was due to the steady expansion of education during that era.  More kids gradually got more education as the century wore on.  And that was highly relevant to performance on IQ tests.  All the testing you did at school made you "test-wise" and that helped you to do well on IQ tests. 

You learnt, for instance that ever useful strategy of: "If you don't know, guess". Some guesses will be right and that will raise your overall score. IQ subtests that were not facilitated by testing -- breadth of vocabulary for instance -- showed very little rise in scores.  You know what an uncommon word means or you don't. So it was environmental rather than genetic factors that explains the rise in average IQ scores -- known generally as the "Flynn" effect.

But the dominance of Leftism wiped all that. Leftists have a horror of competition so avoided testing at all costs.  So an education no longer helped you to do well on IQ tests. And as Leftism gradually tightened its grip, the education effect on IQ scores shrank and shrank.  So IQ scores declined gradually over the years.

It's consoling to note however, that the genetic contribution to IQ test score has not changed.  We are still as bright as we ever were and what we are genetically is increasingly the sole thing reflected in the IQ test scores.



The IQ scores of young people have begun to fall after rising steadily since the Second World War, according to the first authoritative study of the phenomenon.

The decline, which is equivalent to at least seven points per generation, is thought to have started with the cohort born in 1975, who reached adulthood in the early Nineties.

Scientists say that the deterioration could be down to changes in the way maths and languages are taught, or to a shift from reading books to spending time on television and computers.

Yet it is also possible that the nature of intelligence is changing in the digital age and cannot be captured with traditional IQ tests. The turning point marks the end of a well-known but poorly understood trend known as the Flynn effect, in which average IQs have risen by about three points a decade for the past 60 or 70 years.

“This is the most convincing evidence yet of a reversal of the Flynn effect,” Stuart Ritchie, a psychologist at the University of Edinburgh who was not involved in the research, said. “If you assume their model is correct, the results are impressive, and pretty worrying.”

There had been signs that IQ scores might have fallen since the turn of the millennium. Two British studies suggested that the decline was between 2.5 and 4.3 points per decade. This has not been widely accepted owing to the limited research to date. A study has now shown, however, that Norwegian men’s IQs are measurably lower today than the scores of their fathers at the same age.

Ole Rogeberg and Bernt Bratsberg, of the Ragnar Frisch Centre for Economic Research in Oslo, analysed the scores from a standardised IQ test of more than 730,000 men who reported for national service between 1970 and 2009. The research appears in the journal PNAS.

The vast majority of young Norwegian men are required to perform national service and take a standardised IQ test when they join up.

The results, published in the journal PNAS, show that those born in 1991 scored about five points lower than those born in 1975, and three points lower than those born in 1962.

The reasons behind the Flynn effect and its apparent reversal are disputed. Scientists have put the rise in IQ down to better teaching, nutrition, healthcare and even artificial lighting.

Some academics suggest the recent fall could be down to genetics. Their argument is, crudely, that less intelligent people have more babies, and so over time the gains are cancelled out by the spread of genes linked to low-intelligence.

Yet this theory has been scotched by the Norwegian paper. Because the decline can be observed within the same families, it is unlikely to be the result of a demographic shift.

Dr Rogeberg said it was more plausible that the changes in the way children are educated or brought up – such as less time drilling pupils in reading and mathematics – were at play.

He stressed that the findings did not necessarily mean that today’s young people were any more stupid than their parents. Instead, it may be that definitions of intelligence have yet to catch up with the skillset needed to navigate the digital era.

“Intelligence researchers make a distinction between fluid and crystallised intelligence,” he said. “Crystallised intelligence is stuff you have been taught and trained in, and fluid intelligence is your ability to see new patterns and use logic to solve novel problems.”

Classic IQ tests, with their emphasis on arithmetic and verbal reasoning, tend to favour the kind of crystallised intelligence that is fostered by a more traditional education. “If this is the underlying cause of the decline, this need not be overly worrying,” Dr Rogeberg said.

Robin Morris, professor of neuropsychology at King’s College London, said IQ scores probably had hit a ceiling in the west, but there was not yet any reason to be unduly concerned.

“I think the reverse Flynn effect is real but would urge caution about generalising based on one sample,” he said. “Probably the tailing off is a general effect in high income countries in which the contributor factors generally stabilise.”

SOURCE

Friday, June 1, 2018



Genetic Intelligence Tests Are Next to Worthless (?)

I take the title above from the title of a long and very readable article by Carl Zimmer, a NYT journalist.  You could almost predict the title from the fact that he is a NYT journalist.

But the article is extremely well done.  It is something of a triumph of the journalist's art. He explains the facts and issues of his subject beautifully.  I wish I could write as well.  We both try for lucidity and simplicity but he does by far the better job. That he is a journalist and that I am an academic shows. I never could get teddy bears into my writing.

The article is a long one so I am not even going to excerpt it.  Instead I am going  to offer what I hope is a fuller perspective on the matters he raises.  Put simply, he mistakes the major issue involved.

When we look at his title, we have to ask: "Worthless for What? His answer is that currently available genetic information is useless as a substitute for a normal IQ test.  He is absolutely right about that and his warning is well-taken.  People who claim to assess your IQ from your genetic profile are little better than quacks and their results are of no everyday use.

And the reason for that is that IQ appears to be just one aspect of your body's general good functioning.  The brain is just one organ of the body and if the body overall is functioning well the brain should be working pretty well too. And from the research with IQ tests we find an amazing range of good things that high IQ correlates with -- better health, longer life more harmonious marriages etc.  You name it, more or less. 

So that lies behind the fact that there are a LOT of influences on your IQ. They may be scattered anywhere in your body.  What IQ researchers have said for a long time is that IQ is "polygenetic".  It is the sum of a whole lot of little genetic influences.  Almost anything that influences your overall health could also influences your IQ.

At this stage I have to stress that I am talking about the "in general" case.  As elsewhere in life, there are exceptions to the general rule. There are healthy specimens who are as dumb as a brick (some Hollywood actors?) and there are other unfortunates such as Stephen Hawking and Carl Steinmetz where a brilliant brain inhabits a broken body.  Sometimes you need just one faulty gene to have a big influence on your bodily health.

And it is the general case that interests scientists.  They are not actually much interested in YOU.  They are only interested in what emerges from a study of people in general. So when they find some of the many influences on IQ in people's genes they see themselves as being on the right track in seeing IQ as mainly genetic.  And the advances are already exciting.  As each new study comes out, more and more of the genetic influences on IQ are being found.  Genetics generally is in its infancy and the genetics of IQ are no exception. 

And so far there has not been a single study looking at epigenetic influences on IQ.  Epigenetics are bits of your genetic profile which influence how other genes work.  They can even turn a gene "on" or "off".  So to expect that current studies could give us the whole genetics of IQ is very naive.  The fact that we have at this early stage already been able to detect some of the genetics involved is the interesting and exciting part. 

In other words the issue is not whether or not we can measure your IQ from your genes right now but rather whether we are on the right track towards that.  And from a scientific point of view we are doing amazingly well considering the huge difficulty of such research -- JR.