Contra Scott on Lynn’s National IQ Estimates
I'm Still Worried and I Haven't Learnt to Love Lynn's National IQ Estimates.
It is very odd of Scott to discuss Richard Lynn's IQ Estimates without talking about, you know, Richard Lynn's IQ estimates and where they come from. This is why he misses a simple but major factor as to why people disregard Lynn's estimates: They’re bad.
Richard Lynn’s Estimates
Let’s talk about the most famous Lynn estimate: the one cited in The Bell Curve.
The estimate of African IQ given in The Bell Curve is 75. Where did Lynn get this number? Well, he averaged 11 random studies from 6 African countries, most of which were not meant to be IQ tests.
Here’s a taste of what is contained in these studies (this section is blatantly plaigarized from Shaun):
A 1969 Nigerian study of 86 Male factory workers, which doesn’t actually measure IQ or give any estimate. However, Lynn cites an IQ number of “86” from that study. This is presumably a rookie mistake in which the number of participants was mistaken for an IQ score.
A 1974 Zambian study of 152 copper miners which neither measures nor reports an IQ score. Lynn doesn’t explain how he calculated an IQ score of 75 from it.
A 1989 South African study of 1093 16-year-olds designed to test the ‘fairness’ of a Junior Aptitude Test that was in use throughout the country at the time despite being standardized only on white students (apartheid tingz). The designers of the study had to disregard some of the data from black schools because the kids did not understand English (the language the test was administered in) among other kinds of sociocultural biases.
The second Lynn estimate, the one referred to in the linked Aporia Magazine article, gives an estimated score of 68. I agree with Witcherts et all that “[Lynn’s] estimate of average IQ among Africans is primarily based on convenience samples, and not on samples carefully selected to be representative of a given, targeted, population. Unfortunately, in many developing countries, such representative samples are lacking”. Their analysis leads them to conclude that Africans have an average IQ of 80.
As the Aporia article notes, much of the debate concerns the representativeness of samples and if we go down this line of argument, we’ll be stuck in a never-ending cycle of claiming that the one I disagree with is “debunked and well-refuted”. So I will talk about the one thing that both sides agree on: we do not have good studies/ representative samples for African IQ as compared to literally every other continent, which leads me to my next question -
Why??
Scott states that:
Nobody does high-quality IQ tests on the entire population of Malawi; to get his numbers, Lynn would often find some IQ-ish test given to some unrepresentative sample of some group related to Malawians and try his best to extrapolate from there.
and then doesn't interrogate that line of thought further. But I'm here to ask: Why?? Why is my IQ (or at least the average IQ of my country) not taken seriously enough to warrant legitimate study? Why is a measure of "African" IQ so haphazardly estimated being compared to high-quality studies of Western countries? You can't have it both ways; either take the question of IQ differences seriously enough to do good studies or leave it altogether. This is the worst of both worlds.
Looking for insights while admitting that we don’t have the relevant data is counter-productive. We are the proverbial drunk man searching for the keys under the streetlight. Do the results of 150 (likely illiterate) copper miners in 1974 tell us anything about the IQ of Zambians, let alone Sub-Saharan Africans or the “Negroid” race? Let’s take a step back and rethink this. Let’s be serious people.
Does The Environment Matter or Not?
I completely agree with Scott’s central thesis that since IQ is partially environmental, people with a subpar environment will have lower IQ scores. However, as he’s defending Lynn’s estimate he states that -
(don’t worry too much about West African vs. East African ethnic differences here - Lynn’s IQ estimates for both regions are similar)
- a statement that ignores the fact that environmental factors matter just as much as ethnic differences. Sub-Saharan Africa is very environmentally diverse, with relevant factors like literacy rates ranging from 27% to 95% and GDP per capita ranging from $824 to $15,519. Some countries have been mired in civil war since their independence six decades ago, while others are thriving democracies.
An “average” IQ of “Africans” derived from studies that weren’t designed to measure IQ, spanning 8 decades (that include periods of colonization and apartheid i.e. institutionalized discrimination), with several unrepresentative groups of people in seemingly random countries simply does not make sense to me.
Conclusion
IQ tests are not created in a vacuum, they are standardized so that the average test taker (in the West) has a score of 100. Tests are updated regularly to keep this result constant. Meanwhile, “African” IQ is forever defined by highly unscientific estimates cobbled together from whatever studies can be retroactively reworked to function as proxies for IQ. This is important. IQ does not “naturally” appear in a normal distribution. Tests are designed to return a normal distribution on a given population. This is even done in the book “The Bell Curve” where Hernnstein and Murray have to re-standardize the data they have from Army tests in the early 20th century to fit a bell curve. This set of assumptions is then carried on to tests done on Africans not meant to test IQ. A series of random studies, re-analysed using unstated assumptions to suit purposes they weren’t designed to suit does not a good dataset make.
All I’m asking for is that studies of our intelligence are afforded the same level of scientific rigour that those in Western and Asian nations are. The score derived from higher quality representative studies might be the same as Lynn’s, it might be higher, it might be lower, I don’t know. I’m certain that it will be lower than the average Western IQ for the very good environmental reasons outlined in Scott’s post. But can we stop defending bad methodology because we agree with its conclusions?
P/S, with apologies to Mike Godwin: I don’t want to get into the ‘Richard Lynn was racist, funded by eugenicists and that coloured the quality of his research into Negroid IQ’ argument here. All I will say is that this article is highly informed by Shaun’s amazing in-depth analysis of The Bell Curve which in turn is informed by even more in-depth books critical of The Bell Curve. I highly recommend you listen to it.
I suggest reading Cremieux's response to all of this:
https://www.cremieux.xyz/p/national-iqs-are-valid
> I don’t want to get into the ‘Richard Lynn was racist, funded by eugenicists and that coloured the quality of his research into Negroid IQ’ argument here.
For those who don't know what Lizzard is referring to, let me quote my comments on Scott's post.
>He was the editor-in-chief of Mankind Quarterly, a white supremacist journal that was founded by people like:
[...] Otmar Freiherr von Verschuer who was director of the Kaiser Wilhelm Institute of anthropology human heredity and eugenics in Nazi Germany. He was a member of the Nazi Party and the mentor of Josef Mengele, the physician at the Auschwitz concentration camp infamous for performing human experimentation on the prisoners during World War 2. [...]
It's funded by the pioneer fund, an organization he was a board member of and that has been classified as a white supremacist hate group, with one of its first projects being to fund the distribution in US churches and schools of "Erbkrank", a Nazi propaganda film about eugenics. <
Did this 'color' (pardon the pun) the quality of his research? Well, to expand on Lizzard's example of the 1989 South African study, let me quote my comment again:
>
He would use the IQ of 69 for South Africans, cited by Owen (1989), which is referred to by Richard Lynn as:
> “the single best study of Negroid intelligence.”
The study in question examines test and item bias in the Junior Aptitude Test (JAT), a standardized test originally designed for white pupils in South Africa. Owen’s study aimed to assess whether this test was suitable for students from different racial backgrounds—white, Indian, and black pupils.
The abstract of the study states:
> “This study was undertaken to shed light on problems concerning the construction and use of a common test battery for various South African population groups.”
Essentially, the study evaluated whether the JAT, developed for white students, was appropriate for non-white groups. It selected schools across South Africa but faced challenges, particularly with black schools. The study notes that the majority of the selected black schools could not participate “owing to the unrest situation.” As a result, the testing was limited to black schools in the KwaZulu region, one of the segregated areas designated for black residents under apartheid.
The study analyzed the performance of students on various JAT subtests. A significant issue identified was that the test was administered in English to black students in KwaZulu. Several JAT subtests heavily relied on language ability. Owen assumed that language would not significantly impact the results because the black students in KwaZulu had ostensibly been learning English in their schools. However, this assumption ignored or disregarded critical issues, such as the lack of teaching equipment, the low number of certified teachers, and pupil-teacher ratios that were more than double those of white schools. As the study itself states:
> “Language was not expected to play a significant role in test performance in this investigation.”
However, the results showed this assumption to be entirely incorrect. Language played a critical role, and the black students’ poor knowledge of English rendered certain sections of the JAT, such as the synonym test and the memory paragraph test, “virtually unusable.” Yes, that's an actual quote, the study explicitly states:
> “Certain tests proved to be virtually unusable.”
Owen further noted that language bias was not the only issue. In a section titled “Item Bias in the Tests of the JAT,” he identified other forms of cultural and economic bias. For example, several test items presupposed familiarity with objects like electrical appliances, microscopes, and Western-style ladies’ accessories. Owen writes:
> “In the case of both the Indian and black testees, it seems that the single largest cause of bias lay in the fact that the pupils were not familiar with the objects represented by the pictures. Cultural and socioeconomic status factors probably also played a role in this regard.”
Despite the study concluding that the JAT was biased and that some results were “virtually unusable,” Richard Lynn still considered this “the single best study of Negroid intelligence.” Lynn used this study, which, again, tested children in segregated schools using a non-native language, as a key foundation for his estimate of the average IQ of black Africans. In his own words:
> “The mean IQ of the sample in comparison with Caucasoid South African norms is 69. [...] It is proposed, therefore, to round this figure up to 70 and take this as the approximate mean for pure Negroids.”
<
[ original comments here: https://www.astralcodexten.com/p/how-to-stop-worrying-and-learn-to/comment/86319730 ]