Richard Lynn was so kind to send me a signed copy of his latest book. i immediately paused the reading of another book to read this one. some comments and quotes are below. quotes are from the ebook version of the book which i found on the internet.
Some general conclusions about the book. All in all this is a typical Richard Lynn book. It has a very dry style, and is somewhat repetitive. On the other hand, it is not overly long at 400 pages. Many of these are long lists of tables, so are not normally read except if one wants to look up specific countries. It would perhaps have been a good idea to just publish them on the internet for the curious and other researchers. The book contains a wealth of citations revealing a very impressive scholarship. The areas investigated on a global level are many, and the results interesting. The people who think that national IQs are “meaningless” and that human races do not exist or are social constructions (whatever that means, if anything) have the difficult job of explaining why, if these numbers are meaningless, do they fare so well in predicting things on a global level? In other words, why do they have so high validity for a multitude of things? One cannot just regard IQ as “academic intelligence” or some such thing if one can effectively use national IQs to predict things like the lack of proper sanitation. Most often national IQs are found to be better predictors than various non-IQ variables. Although one some occasions I would have liked the authors to use some more variables to see whether they made an impact. I think the authors are sometimes a bit too pessimistic about the possibilities of changing the situation for the low-IQ countries, but I agree with them that one should not expect many of these correlations to change drastically in the near future.
Thoughts and comments to various things
The introduction of the book neatly and shortly explains what the book is about:
The physical sciences are unified by a few common theoretical
constructs, such as mass, energy, pressure, atoms, molecules and
momentum, that are defined and measured in the same ways and
explain a wide range of phenomena in physics, astrophysics,
chemistry and biochemistry. This has been beneficial for the
development of the physical sciences, because it has allowed the
transfer of concepts from one field to others. It has allowed
interface subjects like chemical physics and biochemistry to
develop their own insights and concepts on the basis of those
already developed in their parent fields. Physics is the most basic
of the natural sciences, because the phenomena of the others can
be explained by the laws of physics. For this reason, physics has
been called the queen of the physical sciences.
Hitherto, the social sciences have lacked common unifying
constructs of this kind. The disciplines of the social sciences,
comprising psychology, economics, political science,
demography, sociology, criminology, anthropology and
epidemiology are largely isolated from one another, each with
their own vocabulary and theoretical constructs.
Psychology can be considered the most basic of the social
sciences because it is concerned with differences between
individuals, while the other social sciences are principally
concerned with differences between groups such as socio-
economic classes, ethnic and racial populations, regions within
countries, and nations. These groups are aggregates of
individuals, so the laws that have been established in psychology
should be applicable to the group phenomena that are the concern
of the other social sciences.
Our objective in this book is to develop the case that the
psychological construct of intelligence can be a unifying
explanatory construct for the social sciences. Intelligence is
measured by the intelligence test that was constructed by Alfred
Binet in 1905. During the succeeding century it has been shown
that intelligence, measured as the IQ (the intelligence quotient),
is a determinant of many important social phenomena,
including educational attainment, earnings, socio-economic
status, crime and health. Our theme is that the explanatory value
of intelligence that has been established for individuals can be
extended to the explanation of the differences between groups,
that have been found in the other social sciences, and in
particular to the explanation of the differences between nations.
Thus, we propose that psychology is potentially the queen of
the social sciences, analogous to the position of physics as the
queen of the physical sciences. (p. 1-2)
It is difficult to disagree with this.
one of the things that bother me with the Health chapter is that it doesnt try to compare with and adjoin with the data from The Spirit Level. The authors of SPL contend that many of the things that Lynn&Vanhanan (LV) thinks is due to intelligence, is really due to economic (in)equality. unfortunately, LV does not try to control for this. it wud be interesting to see if the effects of high econ. equality goes away if one controls for intelligence. in other words, that the effects of econ. equality is really just intelligence working thru it.
For a video introduction to the SPL, see this:
one annoying thing about this book, is that it is full of data tables, and the data from these cannot easily be copied into something useful. at least, i have failed to do it in any easy way. it requires a lot of fiddling to get the formatting right in calc/excel. hopefully, LV will make data tables available on their websites where they can easily be downloaded so that others can test out other hypotheses.
many of the tables span two pages but are not that big and cud easily fit into a single table on one page. unfortunately, having to use the image now requires that one either zooms out a lot to fit it all into one screen before taking a screenshot and hence makes the text small, or take two screenshots and edit them together in an image editor. it wud be very nice if they were made available on the website for free use.
a recurrent thing about the book is that the editor did quite a poor job. there are a lot of easily visible typografical mistakes that are a bit annoying. they dont distract too much from the reading of the book, except in the rare cases where a missing word makes interpretation necessary. for instance, on p. 83-84 table 4.5, the 10th line is missing the prefix “in” which makes it appear as if the data presented varies wildly from a positive 0.61 correlation to three other strong negative correlations between -.52 and -0.60.
there was also another place where a “not” was missing and this left me confused for a few seconds.
as for formatting, look at table 7.1, line 1, the word “All” is strangely located in a line below the other information. look also to lines 10-11 and notice how the two “F” are floating to the left.
these mistakes shud be fixed and a new online edition released. this cant be too difficult to do.
notice how low the dysgenic effects are. i was under the impression that they were stronger. also keep in mind that the lines 14-17 are those with the best data. the reason for that is that:
Rows 2, 3 and 4 give negative correlations between
intelligence and fertility based on a nationally representative
American sample showing that the negative correlation is higher
for white women than for white men, and higher for white
women than for black women. This study is not wholly
satisfactory because the age of the sample was 25 to 34 years and
many of them would not have completed their fertility.
To overcome this problem, Vining (1995) published data on
the fertility of his female sample of the ages between 35 and 44,
which can be regarded as close to completed fertility. The results
are given in rows 4 and 5 for white and black women and show
that the correlations between intelligence and fertility are still
significantly negative and are higher for black women (-0.226)
than for white women (-0.062). These correlations are probably
underestimates because the samples excluded high-school
dropouts, who were about 14 per cent of whites and 26 per cent
of blacks at this time, and who likely had low IQs and high
average fertility. (p. 201-2)
which is to say that if one gathers the data before women are done having children, one will miss out some older women who get children late. since such women are especially likely to be well-educated (and hence, smart), this is an important bias.
still given that there are some consistent negative correlations, then there is a dysgenic effect – its just smaller than i had imagined. at least on a within population basis.
It would be interesting to explore to what extent differences
in geographical circumstances and water resources affect the
access to clean water, but unfortunately it is difficult to find
appropriate indicators of geographical factors. However, there is
one indicator for this purpose.WDI-09 (Table 3.5) includes data
on renewable internal freshwater resources per capita in cubic
metres in 2007 (Freshwater). It measures internal renewable
resources (internal river flows and groundwater from rainfall) in
the country. It is noted that these “estimates are based on different
sources and refer to different years, so cross-country
comparisons should be made with caution” (WDI-09, p. 153). It
could be assumed that freshwater resources per capita are
negatively correlated with Water-08, but in fact there is no
correlation between these variables (0.050, N=139). The
correlation between national IQ and Freshwater is also in zero
(0.014, N=147). Access to clean water seems to be completely
independent from freshwater resources, whereas it is
significantly dependent on national IQ (39%) and several
environmental variables. Therefore, it is interesting to see how
well national IQ explains the variation in Water-08 at the level of
single countries and what kinds of countries deviate most from
the regression line. Figure 8.1 summarizes the results of the
regression analysis of Water-08 on national IQ in the group of
166 countries. Detailed results for single countries are reported in
Table 8.3. (p. 246)
Very interesting! Is this a direct disproof of Jared Diamond (1997)‘s environment theory regarding access to water?
Figure 8.1 shows that the relationship between national IQ
and Water-08 is linear as hypothesized, but many highly
deviating countries weaken the relationship. In the countries
above the regression line, the percentage of people without
access to improved water services is higher than expected on the
basis of the regression equation, and in the countries below the
regression line it is lower than expected. In all countries above
the national IQ level of 90, the percentage of the population
without access to clean water is zero or near zero, except in
Cambodia, China and Mongolia, whereas this percentage varies
greatly in the countries below the national IQ level of 85.
National IQ is not able to explain the great variation in Water-08
in the group of countries with low national IQs. Most of that
variation seems to be due to some environmental and local
factors, perhaps also to measurement errors. ( p. 247-8)
in the case of China it seems very unhelpful to category it as one country. it is a HUGE place. it wud be better to split it up into provinces, and calculate these instead. en.wikipedia.org/wiki/Provinces_of_the_People%27s_Republic_of_China altho this will result in many of them having no data. i doubt that there is IQ data for all the regions of China. perhaps those in the regions away from the ocean are not quite as clever as those near the ocean, and near Japan. but surely there is data about Hong Kong, Macau, and some other city or city-like states.
one thing that bothers me a bit is that when LV discuss outliers to their correlation, they use some seemingly arbitrarily picked number. heres a random example (p. 258):
Table 8.3 shows the countries which deviate most from the
regression line and for which positive or negative residuals are
large. An interesting question is whether some systematic
differences between large positive and negative outliers could
help to explain their deviations from the regression line. Let us
regard as large outliers countries whose residuals are ±15 or
higher (one standard deviation is 13).
they note that the sd is 13, but instead opt to use 15 without an explanation. this is the same every time they adopt such an analysis, which do they every chapter. normally, they choose some number slightly larger than 1sd. in p. 155 sd = 1.7, and they use 2. in p. 146 they use 11 while sd = 10.1. in p. 103 they use 12 while the sd is 12.017. the general rule seems to be: choose an arbitrary but nicely looking number just a bit larger than the sd. i dont think this skews the analysis much, but i wud have prefered just if they used 1sd as the border for counting as an outlier.
one odd thing is that when LV finds that a relationship between national IQs and some other variable is curvilinear, they still go on to use the linear model in their explanation. they do this time and time again. it results in some bad points of analysis, for instance:
It is remarkable that this group does not include any
economically highly developed countries, Caribbean tourist
countries, Latin American countries, or oil exporting countries.
Most of them are poor sub-Saharan African countries (17). China
is not really a large positive outlier for the reason that its
predicted value of Water-08 is negative -6. The other eight
positive outliers are poor Asian and Oceanian countries. Most of
them (especially Afghanistan, Cambodia, Myanmar and Timor-
Leste) have suffered from serious civil wars, which have
hampered socio-economic development. (p.259)
if they had made a proper model, one where negative values are impossible, then they wud have avoided such details. its not that LV doesnt know this, as they discuss on page. 79:
Rows 13 through 18 give six correlations between national
IQs and various measures of per capita income reported. The
author analyzed further the relationship by fitting linear, quadratic
and exponential curves to the data for 81 and 185 nations and
found that fitting exponential curves gave the best results. His
interpretation was that “a given increment in IQ, anywhere along
the IQ scale, results in a given percentage in GDP, rather than a
given dollar increase as linear fitting would predict” (Dickerson,
2006, p. 291). He suggests that
exponential fitting of GDP to IQ is logically
meaningful as well as mathematically valid. It is
inherently reasonable that a given increment of IQ
should improve GDP by the same proportional ratio,
not the same number of dollars. An increase of GDP
from $500 to $600 is a much more significant change
than is a linear increase from $20,000 to $20,100. The
same proportional change would increase $20,000 to
$24,000. These data tell us that the influence of
increasing IQ is a proportional effect, not an absolute
one (p. 294).
heres as example of a plot where LV acknowledges that it is curvilinear:
i wud replicate this plot myself and fit an exponential function to it, and then look for outliers, but i wud need the raw data for that in a useable form. see the previous point about how it is difficult to extract the data from the PDF and the need to publish it in some other format, preferably excel/calc.
Some systematic differences in the characteristics of large
positive and negative outliers provide partial explanations for
their large residuals. Most countries with large negative residuals
have benefitted from investments, technologies, and
management from countries of higher national IQs, whereas
most countries with large positive residuals have received much
less such foreign help. (p.260)
tourism is not the only way to receive money from the rich countries. it wud be interesting to look at the effects of foreign aid to poor countries. is there any discernible effect of it? perhaps it has had effects on water supply, for instance.
Table 8.4 shows that the indicators of sanitation are a little
more strongly correlated with national IQ than the indicators of
water (cf. Table 8.2). The explained part of variation varies from
41 to 60 percent. Differences between the three groups of
countries are relatively small, although the correlations are
strongest in the group of countries with more than one million
inhabitants. It should be noted that the correlations between
national IQ and Sanitation-08 are negative because Sanitation-08
concerns the percentage of the population without access to
improved sanitation services (see section 2). (p. 261)
i understand their wish to stay true to the sources numbers, but i wud have prefered if they had multiplied the numbers by -1 to make them fit with the direction of the other numbers.
Row 7 gives a low but statistically significant positive
correlation of 0.18 between national IQ and son preference. This
may be a surprising result, because it might be expected that
liberal and more modern populations would not have such a
strong preference for sons as more traditional peoples. (p. 273)
Consistent with Frazer’s analysis, it has been found in a
number of studies of individuals within nations that there is a
negative relationship between intelligence and religious belief.
This negative relationship was first reported in the United States
in the 1920s by Howells (1928) and Sinclair (1928), who both
reported studies showing negative correlations between
intelligence and religious belief among college students of -0.27
to -0.36 (using different measures of religious belief). A number
of subsequent studies confirmed these early results, and a review
of 43 of these studies by Bell (2002) found that all but four found
a negative correlation. To these can be added a study in the
Netherlands of a nationally representative sample (total N=1,538)
that reported that agnostics scored 4 IQs higher than believers
(Verhage, 1964). In a more recent study Kanazawa (2010) has
analyzed the data of the American National Longitudinal Study of
Adolescent Health, a national sample initially tested for
intelligence with the PPVT (Peabody Picture Vocabulary Test) as
adolescents and interviewed as young adults in 2001-2
(N=14,277). At this interview they were asked: “To what extent
are you a religious person?” The responses were coded “not
religious at all”, “slightly religious”, “moderately religious”, and
“very religious”. The results showed that the “not religious at all”
group had the highest IQ (103.09), followed in descending order
by the other three groups (IQs = 99.34, 98.28, 97.14). The
negative relationship between IQ and religious belief is highly
statistically significant. (p. 278)
the Bell article sounds interesting, but after spending some time trying to locate it, i failed. it seems that im not the only one having such problems.
one of the interesting datasets that id love to see a nonlinear function fitted to. i want to know how much we need to boost intelligence to almost remove religiousness. perhaps one can discover this from using high-IQ samples. at which IQ are there <5% religious people?
another of those tables that have problems with the direction. Legatum and Newsweek shud be positive with each other, right? since they are measuring in the same direction, that is, the one opposite of HDI and IHC (which correlate positively).
LV mention the 2008 study by Kanazawa: Temperature and evolutionary novelty as forces behind the evolution of general intelligence. The interesting thing about this study is that it sort of tests my idea that i wrote about earlier. Kanazawa goes on with his novelty hypothesis using distance from Africa to predict national IQs. However, compared with Ashraf and Galor (2012) paper, he just uses bird distance instead of actual travel distance (humans are not birds, after all!, nor did they just sail straight from Africa to populate America). So im not really sure what his computed r’s are useful for. It wud be interesting to add together the data from the Ashraf and Galor (2012) paper about distances, and genetic diversity to the climate model. LV does mention at one point that lack of genetic diversity make evolution slower:
anomaly is that the Australian Aborigines inhabit a relatively
warm region but have small brain sizes and low IQs. The
explanation for this anomaly is that these were a small isolated
population numbering only around 300,000 at the time of
European colonization, so the mutant alleles for higher IQs did
not appear in them. (p. 381)
consider also the criticism of Kanazawa’s paper in Why national IQs do not support evolutionary theories of intelligence, Wicherts et al (2009):
5. Migration and geographic distance
Kanazawa (2008) was concerned with the relation between lev-
els of general intelligence, as they were distributed geographically
thousands of years ago, and the degree of ‘‘evolutionary novelty” of
the relevant geographic locations. Lacking data regarding evolu-
tionary novelty, Kanazawa proposed, as a measure of evolutionary
novelty, the geographic distance to the EEA, i.e., a large region of
sub-Saharan Africa. The idea is that the greater the distance from
the EEA, the more evolutionarily novel the corresponding environ-
ment. There are several problems with this operationalization.
First, Kanazawa operationalized geographic distance using
Pythagoras’ ﬁrst theorem (a2+ b2= c2). However, Pythagoras’ theo-
rem applies to Euclidian space, not to the surface of a sphere. Sec-
ond, even if these calculations were accurate, distances as traveled
on foot do not in general correspond to distances ‘‘as the crow ﬂies”
(Kanazawa 2008, p. 102). According to most theories, ancestors of
the indigenous people in Australia (i.e., the Aborigines) moved out
of Africa on foot. They probably crossed the Red Sea from Africa to
present day Saudi Arabia, went on to India, and then through Indo-
nesia to Australia. Thus the distance covered on foot must have
been much larger than the distances computed by Kanazawa. This
suggests that the real distances covered by humans to reach a gi-
ven location, i.e., data of central interest to Kanazawa, are likely
to differ appreciably from the distances as the crow ﬂies. One
can avoid this problem by using maps that exist of the probable
routes that humans followed in their exodus from Africa, and esti-
mating the distances between the cradle of humankind and various
other locations accordingly (Relethford, 2004).
Third, it is not obvious that locations farther removed from the
African Savannah are geographically and ecologically more dissim-
ilar than locations closer to the African Savannah. For instance, the
rainforests of central Africa or the mountain ranges of Morocco are
relatively close to the Savannah, but arguably are more dissimilar
to it than the great plains of North America or the steppes of Mon-
golia. In addition, some parts of the world were quite similar to the
African savannas during the relevant period of evolution (e.g., Ray
& Adams, 2001). Clearly, there is no strict correspondence between
evolutionary novelty and geographic distance. This leaves the use
of distances in need of theoretical justiﬁcation. It is also notewor-
thy that given the time span of evolutionary theories, it is hardly
useful to speak of environmental effects as if these were ﬁxed at
a certain geographical location.
People migrate, and have done so extensively in the time since
the evolutionarily period relevant to the evolutionary theories by
Kanazawa and others. A simple, yet imperfect, solution to this
problem is to use data solely from countries that have predomi-
nantly indigenous inhabitants (Templer, 2008; Templer & Arika-
wa, 2006). However, Kanazawa used national IQs of all
countries in Lynn and Vanhanen’s survey, including Australia
and the United States. This casts further doubt on the relevance
of Kanazawa’s data vis-à-vis the evolutionary theories that he
set out to test. Given persistent migration, it is likely that many
of the people, whose test scores Lynn and Vanhanen used to cal-
culate national IQs, are genetically unrelated to the original
inhabitants of their respective countries. In at least 50 of the
192 countries in Kanazawa’s (2008) study, the indigenous people
represent the ethnic minority.