A third of a century ago, the late psychologist Richard Lynn started working on an audacious project: assembling a database of national average cognitive test scores, and then seeing whether intelligence positively correlates with GDP per capita.
RE China's score in this dataset - they based on this estimate on the PIRLS performance of a few Chinese provinces from 2001, when China was still relatively underdevelped. Various Chinese provinces were tested in 2009, and they scored a bit above the OECD average.
Thanks for pointing this out. I've lived in Hong Kong for over 30 years, and I find the gap between the out-of-the-mainland Chinese and the PRC averages impossible to take seriously. The mainland's education system is even more competitive and rigorous than HK's, and there's little indication its students do much worse when they go to overseas universities (which is now common).
"Thanks for pointing this out. I've lived in Hong Kong for over 30 years, and I find the gap between the out-of-the-mainland Chinese and the PRC averages impossible to take seriously."
I don't know how it is in HK, but in Beijing and Shanghai they only count the scores of kids whose parents have a local hukou (internal passport).
This skews the scores a lot, because most of the laborers are internal migrants, and their kids simply aren't tested at all.
There are other scores that are kind of dubious: Vietnam's PISA score appears inflated by giving the test to only about 3/5ths of the expected test-takers. But the World Bank only penalizes China.
The biology deniers have set a trap, and I fear that hereditarians are going to walk right into it.
Obviously, Richard Lynn's papers shouldn't be retracted. Retraction should be reserved for cases of fraud, plagiarism, or something equivalent to mathematical error. The article in the screenshot says that Lynn "used biased practices" and "lack[ed]...critical thinking," which is true but irrelevant to the question at issue. 99% of academic papers are biased and fallacious, but no one is calling to retract them all. Activists are targeting Lynn's papers rather than (e.g.) the literature on stereotype threat for purely political reasons. That is what we should say about this case.
But many hereditarians will feel an impulse to defend Lynn's work. This is a big mistake, and exactly what Bird, Sear et al. are hoping for.
Lynn made some contributions. But the things he was best known for—particularly the national IQ database—were often profoundly flawed. In general, you can't trust what he reports unless you check it yourself. We should not be investing our credibility in making excuses for him. Nor should we give the impression that hereditarianism stands or falls on his scholarship. Ditto for Rushton and Eysenck.
The fact that some of Lynn's claims about national IQ were confirmed by subsequent work, or that his national IQs have a high correlation with other outcomes, is neither here nor there. If you gave me a map of the world, I could guess the IQ of each country, and I'm sure that my IQ scores would have a high correlation with GDP per capita, PISA scores, life expectancy, etc. But that would not be science. The point of science isn't to accidentally get something right, but to acquire justified true beliefs by going through the correct procedure.
I don't like criticizing the deceased, but this issue isn't going away, and people need to confront reality about Lynn and some other HBD heroes. We can object to politically motivated attacks on the scientific record without holding up Lynn as our torchbearer.
In old age (at least), Richard Lynn was sloppy and he had some blatant biases (against Irish Catholics, for example). But he hardly "accidentally" first achieved the basic picture of global national cognitive test score patterns that have been replicated.
I'd like it if the first scientist who came up with the huge breakthrough of creating a world database of cognitive test scores didn't make so many typos and had more bulletproof judgment, but Richard Lynn's achievements should be judged from a glass-half-full standpoint.
There is no need to hold up Lynn "as our torch bearer". That's not how 'science' works.Once persuasive evidence for a hypothesis accumulates the switch to a new paradigm tends to be substantial and quick.
Here's an example of the process:
Wegener's 1912 hypothesis of 'Continental drift' was not accepted in the first 1/2 of the 20th century for lack of evidence pointing towards it and the lack of a mechanism for it.
Now, however only 'nutters' (flat earthers?) deny the reality of the hypothesis, reconstituted as plate tectonics. Evidence accumulated through the 1950s and 1960s for it, namely the recognition of an oceanic crustal 'conveyor' inferred from the record of magnetic reversals recognised in the basaltic oceanic crust combined with measured ages of these rocks, and the proposal of a mechanism (expansion at oceanic ridges and subduction at volanic island arcs).
Another that comes to mind, is the acceptance in the 1990s that stomach ulcers are caused by bacterial infection. I imagine a few nutters still believe they are a product of unbalanced stomach acid.
Some stress-testing on these scores. Converting he PISA scores to the familiar IQ range (given the data provided):
- Singapore IQ: 111,
- Hong Kong IQ: 107,
- South Korea and Japan: 105.5;
Meanwhile, he highest-IQ White nations:
- Canada and Finland, both at 105.
The supposedly lowest-IQ White nations:
- Romania, Bulgaria, Moldova, and Montenegro: down around 90-91.
- Bosnia: 87.5
- Republic of Georgia: 85
I see clues sprinkled throughout that suggest what we have here is really a combination of: (1.) "g" IQ, and (2.) local emphasis put upon schooling per se -- or, more relevantly, how strong test-taking culture is locally. Also possible is (3.) a national desire to get high scores, in the way some countries invest heavily in Olympic teams (for prestige). Asian countries are known to do this. China's "gaming" the system in the way described at TakiMag ("China only administers the test in the brightest parts of the country"), being unsurprising.
We have to expect that some of the above will push things up or down, by a considerable amount in some cases. As with most data, error bars ought to be attached.
The implied 20-point IQ gap between Finland and Georgia. Is it real? I see that Lynn & Vanhanen's 2002 estimate was Finland, 101; Georgia, 87. A fourteen-point gap, still substantial. But does Georgia really have a national IQ about on par with Black Americans? Is there some other explanation?
Lynn flagged Georgia as an outlier in his Figure 3.2: Georgia produced far more R&D than their measured-IQ suggests they would. The other countries above the expected amount of R&D were all very-wealthy educational powerhouses. Georgia's level of R&D suggests it is near the lower-end of the European average for IQ, not down in the mid-80s doldrums.
Let's say that Georgia does get cheated of up to 10 IQ-points below its true level, in this data-set. Why? What did they do 'wrong'?
The countries you listed are pretty fucked-up and dysfunctional. All ex-commie dictatorships with rough post-commie transitions. Environment isn't as important as genetics, but it certainly doesn't have no effect. It'd also depend on when the tests were given, as some of the better off countries like Romania* might be doing better with over a generation between them and the fall of the Eastern Bloc.
* Not the greatest country in the world, but they've handled their post-commie transition better than, say, Yugoslavia or Albania.
Estonia was really Scandinavian in orientation, effectively under German protection and administration during all of of modern history, even during the Tsarist period. (See: the "Baltic Germans").
.
The aberration was really only the period 1944 to 1989 (or a few years more on each side, if you want), and even that relaxed up in the latter few decades. The enthusiasm the plucky Estonians have for NATO and the EU, since the 1990s, is because they still seek German[ic] protection; they also idolize Sweden as their greatest years.
The later World Bank/PISA/etc-derived IQs (Sailer, 2022):
- 1st. Estonia: 106.5 [+6.5] (rose from 2nd to 1st)
- 2nd. Finland: 105 [+4] (dropped to 2nd)
- 3rd. Sweden: 103 [+4]
- 4th. Norway: 102 [+5]
- 5th. Latvia: 100.5 [+4.5] (rose from 6th to 5th)
- 6th. Russia: 100 [+3] (dropped to 6th)
- 7th. Lithuania: 99.5 [+5.5]
.
Some of the absolute and relative movements, here, we cannot escape thinking are "environment," and political conditions and such things, surely? All of which would tend to influence educational policy and the specific rubber-meets-road moment of what school administrators do with these tests.
(Caveat: these twenty-years-apart data-sets were derived in very different ways, making direct comparisons shaky but interesting for a glance, anyway. What is most striking is the _stability_ of the relative rankings, which is the whole point of the 2022 article.)
During one of the PISA tests, the Austrian teacher's union was threatening to strike and called for teachers to give an anti-pep talk to students to encourage to not try hard on the PISA test. Fortunately, the use of multiple datapoints smooths out some of those idiosyncrasies. Still, conducting an international test like the PISA or the regional ones given to African countries is obviously a sizable organizational challenge. So I wouldn't take anomalous datapoints too seriously.
For example, on one datapoint Kenya scored at a northern European level. I found the report for that test and couldn't see anything wrong with it at a glance. I don't really believe it, but on the other hand, I suspect Kenya is better than the average sub-Saharan state.
The IQ vs national prosperity plots need some weighing by country size. It is absurd that equal weight is given to billion plus behemoths and island of few thousands. That is UN not science.
The city-state of Singapore is a traditional urban demographic sink with low fertility. It repopulates itself by selecting migrants, typically from mainland China.
The Argentines complain that they do a more honest job with the PISA, rounding up 80% of the eligible 15-year-olds, some of them no longer in school, to take the test while the Mexicans only round up 60%.
The honest Dutch in one PISA got 101% of expected 15 year olds to take the test. Finland is usually in the upper 90s, America in the mid 90s.
In the US, Texas students score well on the federal NAEP, in part because Texas seems to moneyball the test while other states don't much care about nudging their scores higher. You must have, IIRC, 90% of your eligible students take the NAEP, and Texas routinely comes in around 91%, with rumors that they encourage the dumbest kids to call in sick the day of the NAEP.
That makes sense. El Salvador was such a wreck until recently it is likely that only the most organized of schools were administering PISAs to students each year.
The Chinese score is probably correct considering the majority of their population are semi illiterate. Their economic issues today are due to this problem highlighted in a book by Natalie Hell and Scott Rozelle called Invisible China: How the Urban-Rural Divide Threatens China's Rise.
It's funny how the main thing which has kept China backwards and poor -- their communist government -- is the very same thing that makes it an existential threat to the World. Here's hoping they never figure out how to break out of the cycle without breaking their dictatorship in the process.
Orwell's greatest fear was that the sheer ruthlessness and brutality of the USSR could only be checked by an even equally ruthless communist Anglo dictatorship. Hence 1984's division of the world into three commie military superpowers -- Sinic, Russic, and Anglic -- amidst a scattering of third-world hellholes where all the wars are fought. He ended up being wrong for the First Cold War, and I hope to hell he'll stay wrong about the Second. Just because the anemia of communism killed the USSR before a hot war could break out between us doesn't mean we'll be so lucky this time around.
These ideological lunatics are a danger to our society' future
RE China's score in this dataset - they based on this estimate on the PIRLS performance of a few Chinese provinces from 2001, when China was still relatively underdevelped. Various Chinese provinces were tested in 2009, and they scored a bit above the OECD average.
https://akarlin.com/analysis-of-chinas-pisa-2009-results/
Yeah, they somewhat arbitrarily penalized China for reporting PISA scores just from four prosperous provinces. I'd take that 441 with a grain of salt.
Thanks for pointing this out. I've lived in Hong Kong for over 30 years, and I find the gap between the out-of-the-mainland Chinese and the PRC averages impossible to take seriously. The mainland's education system is even more competitive and rigorous than HK's, and there's little indication its students do much worse when they go to overseas universities (which is now common).
"Thanks for pointing this out. I've lived in Hong Kong for over 30 years, and I find the gap between the out-of-the-mainland Chinese and the PRC averages impossible to take seriously."
I don't know how it is in HK, but in Beijing and Shanghai they only count the scores of kids whose parents have a local hukou (internal passport).
This skews the scores a lot, because most of the laborers are internal migrants, and their kids simply aren't tested at all.
There are other scores that are kind of dubious: Vietnam's PISA score appears inflated by giving the test to only about 3/5ths of the expected test-takers. But the World Bank only penalizes China.
> Niger: 305
Niger, please!
Nathan Cofnas on Facebook:
The biology deniers have set a trap, and I fear that hereditarians are going to walk right into it.
Obviously, Richard Lynn's papers shouldn't be retracted. Retraction should be reserved for cases of fraud, plagiarism, or something equivalent to mathematical error. The article in the screenshot says that Lynn "used biased practices" and "lack[ed]...critical thinking," which is true but irrelevant to the question at issue. 99% of academic papers are biased and fallacious, but no one is calling to retract them all. Activists are targeting Lynn's papers rather than (e.g.) the literature on stereotype threat for purely political reasons. That is what we should say about this case.
But many hereditarians will feel an impulse to defend Lynn's work. This is a big mistake, and exactly what Bird, Sear et al. are hoping for.
Lynn made some contributions. But the things he was best known for—particularly the national IQ database—were often profoundly flawed. In general, you can't trust what he reports unless you check it yourself. We should not be investing our credibility in making excuses for him. Nor should we give the impression that hereditarianism stands or falls on his scholarship. Ditto for Rushton and Eysenck.
The fact that some of Lynn's claims about national IQ were confirmed by subsequent work, or that his national IQs have a high correlation with other outcomes, is neither here nor there. If you gave me a map of the world, I could guess the IQ of each country, and I'm sure that my IQ scores would have a high correlation with GDP per capita, PISA scores, life expectancy, etc. But that would not be science. The point of science isn't to accidentally get something right, but to acquire justified true beliefs by going through the correct procedure.
I don't like criticizing the deceased, but this issue isn't going away, and people need to confront reality about Lynn and some other HBD heroes. We can object to politically motivated attacks on the scientific record without holding up Lynn as our torchbearer.
In old age (at least), Richard Lynn was sloppy and he had some blatant biases (against Irish Catholics, for example). But he hardly "accidentally" first achieved the basic picture of global national cognitive test score patterns that have been replicated.
I'd like it if the first scientist who came up with the huge breakthrough of creating a world database of cognitive test scores didn't make so many typos and had more bulletproof judgment, but Richard Lynn's achievements should be judged from a glass-half-full standpoint.
I see now that Cofnas posted the same comment on Substack, so you can respond to him directly.
There is no need to hold up Lynn "as our torch bearer". That's not how 'science' works.Once persuasive evidence for a hypothesis accumulates the switch to a new paradigm tends to be substantial and quick.
Here's an example of the process:
Wegener's 1912 hypothesis of 'Continental drift' was not accepted in the first 1/2 of the 20th century for lack of evidence pointing towards it and the lack of a mechanism for it.
Now, however only 'nutters' (flat earthers?) deny the reality of the hypothesis, reconstituted as plate tectonics. Evidence accumulated through the 1950s and 1960s for it, namely the recognition of an oceanic crustal 'conveyor' inferred from the record of magnetic reversals recognised in the basaltic oceanic crust combined with measured ages of these rocks, and the proposal of a mechanism (expansion at oceanic ridges and subduction at volanic island arcs).
Another that comes to mind, is the acceptance in the 1990s that stomach ulcers are caused by bacterial infection. I imagine a few nutters still believe they are a product of unbalanced stomach acid.
Some stress-testing on these scores. Converting he PISA scores to the familiar IQ range (given the data provided):
- Singapore IQ: 111,
- Hong Kong IQ: 107,
- South Korea and Japan: 105.5;
Meanwhile, he highest-IQ White nations:
- Canada and Finland, both at 105.
The supposedly lowest-IQ White nations:
- Romania, Bulgaria, Moldova, and Montenegro: down around 90-91.
- Bosnia: 87.5
- Republic of Georgia: 85
I see clues sprinkled throughout that suggest what we have here is really a combination of: (1.) "g" IQ, and (2.) local emphasis put upon schooling per se -- or, more relevantly, how strong test-taking culture is locally. Also possible is (3.) a national desire to get high scores, in the way some countries invest heavily in Olympic teams (for prestige). Asian countries are known to do this. China's "gaming" the system in the way described at TakiMag ("China only administers the test in the brightest parts of the country"), being unsurprising.
We have to expect that some of the above will push things up or down, by a considerable amount in some cases. As with most data, error bars ought to be attached.
The implied 20-point IQ gap between Finland and Georgia. Is it real? I see that Lynn & Vanhanen's 2002 estimate was Finland, 101; Georgia, 87. A fourteen-point gap, still substantial. But does Georgia really have a national IQ about on par with Black Americans? Is there some other explanation?
Lynn flagged Georgia as an outlier in his Figure 3.2: Georgia produced far more R&D than their measured-IQ suggests they would. The other countries above the expected amount of R&D were all very-wealthy educational powerhouses. Georgia's level of R&D suggests it is near the lower-end of the European average for IQ, not down in the mid-80s doldrums.
Let's say that Georgia does get cheated of up to 10 IQ-points below its true level, in this data-set. Why? What did they do 'wrong'?
The countries you listed are pretty fucked-up and dysfunctional. All ex-commie dictatorships with rough post-commie transitions. Environment isn't as important as genetics, but it certainly doesn't have no effect. It'd also depend on when the tests were given, as some of the better off countries like Romania* might be doing better with over a generation between them and the fall of the Eastern Bloc.
* Not the greatest country in the world, but they've handled their post-commie transition better than, say, Yugoslavia or Albania.
And yet Estonia was Communist-ruled too.
Estonia was really Scandinavian in orientation, effectively under German protection and administration during all of of modern history, even during the Tsarist period. (See: the "Baltic Germans").
.
The aberration was really only the period 1944 to 1989 (or a few years more on each side, if you want), and even that relaxed up in the latter few decades. The enthusiasm the plucky Estonians have for NATO and the EU, since the 1990s, is because they still seek German[ic] protection; they also idolize Sweden as their greatest years.
.
.
-- IQs for Estonia and the vicinity --
.
First, as given in Lynn & Vanhanen (2002):
- 1st. Finland: 101
- 2nd. Estonia: 100
- 3rd. Sweden: 99
- 4th. Norway: 97
- 5th. Russia: 97
- 6th. Latvia: 96 (considerable Soviet-migrant presence)
- 7th. Lithuania: 94
.
compared to...
.
The later World Bank/PISA/etc-derived IQs (Sailer, 2022):
- 1st. Estonia: 106.5 [+6.5] (rose from 2nd to 1st)
- 2nd. Finland: 105 [+4] (dropped to 2nd)
- 3rd. Sweden: 103 [+4]
- 4th. Norway: 102 [+5]
- 5th. Latvia: 100.5 [+4.5] (rose from 6th to 5th)
- 6th. Russia: 100 [+3] (dropped to 6th)
- 7th. Lithuania: 99.5 [+5.5]
.
Some of the absolute and relative movements, here, we cannot escape thinking are "environment," and political conditions and such things, surely? All of which would tend to influence educational policy and the specific rubber-meets-road moment of what school administrators do with these tests.
(Caveat: these twenty-years-apart data-sets were derived in very different ways, making direct comparisons shaky but interesting for a glance, anyway. What is most striking is the _stability_ of the relative rankings, which is the whole point of the 2022 article.)
During one of the PISA tests, the Austrian teacher's union was threatening to strike and called for teachers to give an anti-pep talk to students to encourage to not try hard on the PISA test. Fortunately, the use of multiple datapoints smooths out some of those idiosyncrasies. Still, conducting an international test like the PISA or the regional ones given to African countries is obviously a sizable organizational challenge. So I wouldn't take anomalous datapoints too seriously.
For example, on one datapoint Kenya scored at a northern European level. I found the report for that test and couldn't see anything wrong with it at a glance. I don't really believe it, but on the other hand, I suspect Kenya is better than the average sub-Saharan state.
The IQ vs national prosperity plots need some weighing by country size. It is absurd that equal weight is given to billion plus behemoths and island of few thousands. That is UN not science.
That's why I'm not that impressed by Singapore
The city-state of Singapore is a traditional urban demographic sink with low fertility. It repopulates itself by selecting migrants, typically from mainland China.
The relative positions of Argentina and El Salvador are surprising.
The Argentines complain that they do a more honest job with the PISA, rounding up 80% of the eligible 15-year-olds, some of them no longer in school, to take the test while the Mexicans only round up 60%.
The honest Dutch in one PISA got 101% of expected 15 year olds to take the test. Finland is usually in the upper 90s, America in the mid 90s.
In the US, Texas students score well on the federal NAEP, in part because Texas seems to moneyball the test while other states don't much care about nudging their scores higher. You must have, IIRC, 90% of your eligible students take the NAEP, and Texas routinely comes in around 91%, with rumors that they encourage the dumbest kids to call in sick the day of the NAEP.
That makes sense. El Salvador was such a wreck until recently it is likely that only the most organized of schools were administering PISAs to students each year.
It would be good to add the number of data points next to each country's name.
The Chinese score is probably correct considering the majority of their population are semi illiterate. Their economic issues today are due to this problem highlighted in a book by Natalie Hell and Scott Rozelle called Invisible China: How the Urban-Rural Divide Threatens China's Rise.
It's funny how the main thing which has kept China backwards and poor -- their communist government -- is the very same thing that makes it an existential threat to the World. Here's hoping they never figure out how to break out of the cycle without breaking their dictatorship in the process.
Orwell's greatest fear was that the sheer ruthlessness and brutality of the USSR could only be checked by an even equally ruthless communist Anglo dictatorship. Hence 1984's division of the world into three commie military superpowers -- Sinic, Russic, and Anglic -- amidst a scattering of third-world hellholes where all the wars are fought. He ended up being wrong for the First Cold War, and I hope to hell he'll stay wrong about the Second. Just because the anemia of communism killed the USSR before a hot war could break out between us doesn't mean we'll be so lucky this time around.