More national IQ estimates

Gerhard Meisenberg kindly sent me the following paper, which I was unaware of, in which he and Richard Lynn estimated national IQs using a broader set of international achievement tests:

Meisenberg and Lynn, 2011. Intelligence: A Measure of Human Capital in Nations

To note, I disagree with their method of calibrating the regional African tests (i.e., MLA, PASEC, and SACMEDQ). For these regional tests, scores, which were based on a 500 point scale with a mean set to the regional average, need to be adjusted by finding countries which participated in both regional tests and the regular international tests (e.g., TIMSS). Lynn and Meisenberg individually calibrated the three surveys. But as only a couple of countries participated both in TIMSS/PISA and in MLA and SACMEDQ and none participated in both TIMSS/PISA and PASEC — IQ scores were used to adjust instead — this method was probably somewhat inaccurate. A superior method would have been to simultaneously calibrate scores across surveys as the countries overlapped. (Some of the countries which participated in SACMEDQ participated in MLA and some of those which participated in MLA participated in PASEC.) Doing so would have given 5 to 9 anchoring countries, depending on which set of international reference tests was chosen, and increased the accuracy of the estimates.

Regardless, the estimates are accurate enough for the purposes of interest and should effectively silence the many critics who claimed that Lynn’s (African) IQ estimates were grossly inaccurate.

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

17 Responses to More national IQ estimates

  1. Those “IQ estimates” are nothing but school achievement test scores, so not much more accurate than Lynn’s “estimates” based on fraud and miscalculation. That paper does absolutely nothing to silence his critics.

    Also, when rejecting the different possible explanations for IQ differences between Nigeria and Japan, he completely ignores the best explanation.

    • Chuck says:

      Please don’t comment on my site if you have nothing semi-intelligent to say. I’m not interested in imbecilic comments:

      National IQ estimates

      You: “so not much more accurate than Lynn’s “estimates”

      The hard distinction you make between cognitive ability as measured by IQ tests and as measured by achievement tests is artificial for the simple reason that all cognitive tests measure general cognitive ability to one degree of reliability or another. So one can rightfully make IQ parameter estimates on the basis of performance on student achievement tests. Doing so is common practice with regards to discussion of population differences. See Nisbett et al 2012. This aside, your point is hollow, as “intelligence” as used in the phrase “National intelligence” has a more inclusive meaning than “psychometric intelligence as measured exclusively by IQ tests.” Lynn and Vanhannen are clear about this and explicitly state: “We regard national IQs as measures of general intelligence defined as the totality of cognitive abilities. These include Spearman’s g (the general factor present in all cognitive tasks), and other cognitive abilities that are independent of g. (Intelligence: A Unifying Construct for the Social Sciences, pg. 9). In short, Lynn isn’t claiming that his construct “National IQ” is identical to the construct measured by IQ tests on the individual level. But I guess such technicalities — as properly understanding the issues being discussed — are beyond you.

      Reliability of Lynn’s estimates

      Keeping in mind the point above, if Lynn’s estimates are as unreliable as you claim, again, I ask why do they so highly correlate with other similar estimates? Refer to:

      Hanushek and Woessmann. The Economics of International Differences in Educational Achievement

      Lynn and Vanhanen, 2012. Intelligence: A Unifying Construct for the Social Sciences

      Rindermann, 2007. The g-Factor of International Cognitive Ability Comparisons: The Homogeneity of Results in PISA, TIMSS, PIRLS and IQ-Tests Across Nations

      ALtinok and Murseli, 2006. International Database on Human Capital Quality

      Altinok et al, (work in progress). International Database on Human Capital Quality: An Update

      Meisenberg and Lynn, 2011. Intelligence: A Measure of Human Capital in Nations

      You haven’t been able to answer this one, have you? Instead you resort to vituperation.

      It is interesting that Wicherts et alia, whom you cite so fondly, were so off in their critique, when it came to student assessments. They presented a rather selective review of the data, neglecting the many points mentioned in ALtinok and Murseli (2006), no? But I guess that’s OK because they were “doing the right thing.”

      Parasites

      You: “Also, when rejecting the different”

      Of course, you’re right, L & V completely ignore parasites as a possible explanation and reject all non-genetic explanations.

      “Row 5 shows a high negative correlation of -0.89 between national IQ and the intensity and prevalence of infectious diseases. The authors propose that the widespread presence of infectious diseases impairs the intelligence of populations in low IQ countries. We accept that this is likely the case, but we suggest that the relationship between national IQs and the intensity of infectious diseases is a likely two way causal relationship. The intensity of infectious diseases is a determinant of low IQs, as the authors argue, but low national IQs are also a cause of widespread infectious diseases.”

      Best trust Dienkes on this, as he has rejected even the reasonable possibility of “significant” genetic intelligence differences. A good unbiased source there.

  2. My blog is filled with your imbecilic comments, so don’t complain to me.

    >>> So one can rightfully make IQ parameter estimates on the basis of performance on student achievement tests.

    Not reliable estimates:

    Whereas the correlations indicate that around 50% to 60% of the variance in GCSE examination points score can be statistically explained by the prior g factor, by the same token a large proportion of the variance is not accounted for by g. Some of the remaining variance in GCSE scores will be measurement error, but some will be systematic. Thus, non-g factors have a substantial impact on educational attainment. These may include: school attendance and engagement; pupils’ personality traits, motivation and effort; the extent of parental support; and the provision of appropriate learning experiences, teaching quality, school ethos, and structure among other possible factors (Petrides, Chamorro-Premuzic, Frederickson, & Furnham, 2005; Strand, 2003).”

    http://www.elsevier.com/authored_subject_sections/S05/S05_357/top/intel.pdf

    The present paper criticizes these conclusions and the robustness of the data from which Lynn (2010) derived the IQ scores. In particular, on the basis of recent Italian studies and our databank, we observe that : 1) school measures should be used for deriving IQ indices only in cases where contextual variables are not crucial: there is evidence that partialling out the role of contextual variables may lead to reduction or even elimination of PISA differences.”

    http://www.sciencedirect.com/science/article/pii/S0160289610000541

    >>> Keeping in mind the point above, if Lynn’s estimates are as unreliable as you claim, again, I ask why do they so highly correlate with other similar estimates?

    Because they’re all fake IQs “estimated” from achievement tests. Garbage correlates with garbage. But they’re not in agreement with real, unbiased IQ scores like those provided by Wicherts et al.

    >>> Of course, you’re right, L & V completely ignore parasites as a possible explanation and reject all non-genetic explanations.

    The passage you quoted does not appear in the study we’re talking about. So my point stands: Lynn smugly rejects a bunch of straw-man arguments explaining IQ differences between Africans and Asians, but fails to address the strongest evidence from Eppig et al. (2010).

    Anyway, his argument that “low national IQs are also a cause of widespread infectious diseases” is weak since parasite load is objectively higher in the tropics, and high-IQ Europeans got sick too when they traveled there.

    • Chuck says:

      Where to start:

      (1) “Not reliable estimates”

      “Thus, non-g factors have a substantial impact on educational attainment. These may include”

      Chuck: If the correlation between g and academic tests (AQ) is 0.7 (between and within populations), it is implied that g differences will be between AQ x 0.7 and AQ/0.7. This gives you a good parameter estimate. As for reliability, this is simply the correlation between two variables. Whether or not you want to consider 0.7 to be insufficiently reliable is up to you. Whatever the case, Lynn was vindicated with regards to his AQ scores, no? Rindermann (2012) independently calculated an AQ of 71.

      “1) school measures should be used for deriving IQ indices only in cases where contextual variables are not crucial: there is evidence that partialling out the role of contextual variables may lead to reduction or even elimination of PISA differences.”

      Chuck: So the argument above is that PISA is a poor measure of IQ because it’s influenced by “contextual variables” such as school quality. But then It’s argued by others (e.g., Nisbett et al) that IQ is a poor measure of general mental ability because it’s also influenced by these same variables. So where does this leave us? This is why we distinguish between (a) the cause of a difference and (b) the magnitude of a difference Whether PISA scores can statistically or causally be explained by “contextual variables” is irrelevant to whether they are a reliable measure of IQ either on the individual or group level.

      (2) “But they’re not in agreement with real, unbiased IQ scores like those provided by Wicherts et al.”

      Chuck: I don’t know what your talking about. Heiner Rindermann recently went through and meticulously re-estimated the African IQs:

      “The mean of the two newer unpublished psychometric studies (Nc = 2 countries, Table 1, column 1, S1–2; see online Supplement ‘‘Table 1’’) in 2010 norms is IQ 79. The mean result in student assessment studies (Nc = 14 countries, Table 1, column 2, SAS-k) in international norms (majority of studies 1995–2007) is IQ 71. Lynn’s 1979 norm data have a mean of IQ 70 (Nc = 52), or of IQ 76 (in 2010) after FLynn correction (larger than in UK since 1979). Wicherts’ collection (Nc = 17) results in IQ 77. Rindermann’s collection results in IQ 68 (Nc = 52), or IQ 73 after FLynn correction. HDI predicts a mean IQ of 70 (Nc = 48), skin brightness IQ 68 (Nc = 42), both together IQ 70 (Nc = 50). FLynn-corrected for 2010, the predicted IQ’s are around IQ 75. The total mean range of all studies and different values discussed by the authors is between 68 and 82. If one considers only 2010 estimates and excludes outliers the range is between 71 and 78. Student assessment studies with their larger school-related test content and therefore larger dependence on educational quality seem to boost the difference to more developed countries”

      http://lesacreduprintemps19.files.wordpress.com/2012/09/african-cognitive-ability.pdf

      Average AQ = 71, Average IQ ~75 (Flynn corrected assuming larger African Flynn effect.). In what sense is 71 not in agreement with 75?

      (3) “The passage you quoted does not appear in the study we’re talking about. ”

      Chuck: The passage is straight from Lynn’s 2012 book. In the study that we are talking about, above, the only statement I could find concerning cause is:

      “Studies of the determinants of human capital. There is no generally accepted theory to explain why, for example, the level of human capital (a.k.a. intelligence) is so much higher in Japan than in Nigeria. Genetic theories have been proposed by some (e.g., Lynn, 2006). These theories have been attacked by others (e.g., Wicherts et al., 2010), but without convincing alternative explanations. Precise knowledge about the current level of intelligence in different countries is required for these investigations.”

      “without CONVINCING alternative explanations”

      And from this you get: “Also, when rejecting the different possible explanations”

      Is saying that an explanation is not convincing the same as rejecting it?

    • Chuck says:

      “My blog is filled with your imbecilic comments”

      No doubt. But just because you tolerate such comments doesn’t mean that I have to.

  3. 猛虎 says:

    RR, when you cite Eppig, you omit to note this passage of the study, which seems to vindicate the hereditarian position :

    “Peoples living in areas of consistently high prevalence of infectious disease over evolutionary time thus may possess adaptations that favour high obligatory investment in immune function at the expense of other metabolically expensive traits such as intelligence.”

    Besides, if Lynn’s estimates were “garbage” as you say, it would be curious to see that Eppig’s result holds “whether using either of Lynn & Vanhanen’s (2006) two datasets or Wicherts et al.’s (2010b) data.” Given this, how about your claim that “garbage correlates with garbage” ?

    I also think that you should read Murray’s study “The Secular Increase in IQ and Longitudinal Changes in the Magnitude of the Black-White Difference” (specifically pages 14-19).

  4. Meng Hu says:

    As for this claim :

    his argument that “low national IQs are also a cause of widespread infectious diseases” is weak since parasite load is objectively higher in the tropics, and high-IQ Europeans got sick too when they traveled there.

    I think I could recommend this old book Race (by Baker). There were several places unaffected by such diseases, but africans were still poor, regardless of infectious disease. Diseases and the like cannot explain why civilizations didn’t emerged in Africa. Baker is very convincing on this topic (see pages 397-400, 527, 528).

  5. >>> Whether or not you want to consider 0.7 to be insufficiently reliable is up to you.

    I don’t, and neither do most serious researchers.

    PISA tests, however, were developed to measure achievement and not intelligence. In fact, the aim of PISA is to measure “how far students near the end of compulsory education have acquired some of the knowledge and skills that are essential for full participation in society”….

    Nevertheless, Lynn (2010a) uses achievement tests as “proxies for intelligence” (p. 95) adopting the logic that educational attainment and intelligence are highly correlated (from r=0.5 to r=1.0) across nations (Lynn & Meisenberg, 2010; Lynn & Mikk, 2007). However, in his studies it is not clear what kind of IQ tests have been used, and the other factors affecting achievement such as school quality, sociocultural level, and so on, are not controlled.

    http://www.sciencedirect.com/science/article/pii/S1041608011001610

    >>> Whatever the case, Lynn was vindicated with regards to his AQ scores, no? Rindermann (2012) independently calculated an AQ of 71.

    I don’t see what you think that proves. Lynn and Rindermann both “calculated” fake IQs from achievement tests, and both arrived at similar numbers. So what? Like I said, garbage correlates with garbage.

    >>> So the argument above is that PISA is a poor measure of IQ because it’s influenced by “contextual variables” such as school quality. But then It’s argued by others (e.g., Nisbett et al) that IQ is a poor measure of general mental ability because it’s also influenced by these same variables. So where does this leave us?

    IQ tests are much less influenced by the kinds of contextual variables that affect school performance. That’s precisely what they’re designed for. Of course they’re not perfect, but a lot better. Remember, according to school achievement tests, poor black students in the UK are “smarter” than poor white students.

    >>> Average AQ = 71, Average IQ ~75 (Flynn corrected assuming larger African Flynn effect.). In what sense is 71 not in agreement with 75?

    The correct number (based on real, unbiased IQ tests) according to Wicherts et al. is 78-82. Lynn’s “estimate” was 67.

    >>> the only statement I could find concerning cause is: […] And from this you get: “Also, when rejecting the different possible explanations”

    It’s at the very beginning of the study — first paragraph — where he compares Nigeria to Japan. He mentions geography, resources, history, culture, economic institutions…everything but infectious disease.

    >>> No doubt.

    I’m glad we agree that your comments are imbecilic.

    • chuck says:

      “I don’t see what you think that proves. Lynn and Rindermann both “calculated” fake IQs from achievement tests, and both arrived at similar numbers. So what? Like I said, garbage correlates with garbage”

      Lynn and Rindermann calculated Achievement Quotas from achievement tests. They then averaged these with their calculated Intelligence Quotas to create “National IQs”, which Lynn defined as discussed above. Their methodologies have been made clear. I’ll assume that you simply misunderstood what was being done and that you haven’t been intentionally trying to mischaracterize the situation.

      So, as I was saying … both arrived at similar AQs — along with others — confirming Lynn’s original AQ estimate and disconfirming Wicherts. Agree or disagree?

      The issue then is the IQ estimates. Correct?

      “The correct number (based on real, unbiased IQ tests) according to Wicherts et al. is 78-82. Lynn’s “estimate” was 67.”

      The most recent paper on this — still in press — said::

      “Lynn’s 1979 norm data have a mean of IQ 70 (Nc = 52), or of IQ 76 (in 2010) after FLynn correction (larger than in UK since 1979). Wicherts’ collection (Nc = 17) results in IQ 77. Rindermann’s collection results in IQ 68 (Nc = 52), or IQ 73 after FLynn correction”

      So there you go…..

  6. >>> RR, when you cite Eppig, you omit to note this passage of the study, which seems to vindicate the hereditarian position :

    Try reading the whole paragraph. They mention that as a possible alternate explanation, and then REJECT it:

    The Flynn effect (Flynn 1987) indicates that conditional developmental causes must be at work at least in part. Large increases in intelligence across a few generations cannot be attributed to genetic differences caused by evolutionary processes. Hence, it does not seem probable that region-specific genetic adaptations are the primary cause of the worldwide variation in intelligence.

    Either way though, it’s still the parasitic environment that’s responsible.

    >>> There were several places unaffected by such diseases, but africans were still poor, regardless of infectious disease.

    You already tried that stupid argument on my blog. Tropical Africa has the highest parasite load and rates of infectious disease in the world. A few “disease-free” pockets wouldn’t be enough to change anything.

    • 猛虎 says:

      That’s because they believe in the Flynn Effect (note the “Hence”). But what if the flynn effect did not occur on G ? You could also read Wicherts (yes, himself) “Are intelligence tests measurement invariant over time?” (p. 529-531) and Kaufman “In What Way Are Apples and Oranges Alike?” (p. 389-392).

      Concerning the fact that “Tropical Africa has the highest parasite load”, this doesn’t help your case. As I already said, several parts of Africa were the inhabitants were healthy. Again, see Baker’s Race (1974). It is out of print, of course, but I scanned the entire book. It is available for free.

  7. >>> to create “National IQs”, which Lynn defined as discussed above.

    Don’t be so gullible. Lynn’s definition of his “National IQs” is just a lame way to cover his ass. Anyone can see right through it. He knows they’re not accurate IQ scores and now everyone’s on to him, so he’s trying to explain that away. But it still doesn’t work. Achievement tests don’t accurately measure “other cognitive abilities” either, because of the many contextual variables.

    >>> both arrived at similar AQs — along with others — confirming Lynn’s original AQ estimate and disconfirming Wicherts. Agree or disagree?

    Disagree of course. Wicherts’ data is actual, unbiased IQ data. Lynn and Rindermann’s “data” is still garbage “estimated” from achievement tests and through fraud. Get it?

  8. >>> But what if the flynn effect did not occur on G ?

    Irrelevant.

    “You cannot dismiss the score gains of one group on another merely because the reduction of the score gap by subtest has a negative correlation with the g loadings of those subtests. […] The assertion that ‘if population group differences are greater on the more g-loaded and more heritable subtests, it implies they have a genetic origin’ is simply false.”

    http://www.iapsych.com/iqmr/fe/LinkedDocuments/flynn2010a.pdf

    >>> As I already said, several parts of Africa were the inhabitants were healthy.

    And as I already said, a few “disease-free” pockets wouldn’t be enough to change anything. Civilizations are built on trade and exchange of ideas. How are isolated “healthy” populations going to do that if everyone around them is too sick to contribute?

    • Chi says:

      I note that Chuck commented on the Flynn paper previously:

      To quote Flynn:

      “(1) g would be of no interest were it not correlated with cognitive complexity. (2) Given hierarchy of tasks, a worse performing group (whatever the cause of its deficit) will tend to hit a “complexity ceiling” — fall further behind a better group the more complex the task. (3) Heritability of relevant traits will increase the more complex the task. (4) Thus, the fact that group performance gaps correlate with heritability gives no clue to the origin of group differences. (5) When a lower performing group gains on a higher performing one, their gains will tend to diminish the more complex the task. Thus, blacks have gained 5.50 IQ points on whites since 1972 but only 5.13 GQ points. (6) Recent achievement test data confirm these IQ gains but the data as a whole pose problems for the external validity of black IQ. (7) The FE is irrelevant to showing that the racial IQ gap is environmental but it was historically valuable in clarifying the debate.”

      Both of Flynn’s counters are flawed. With regards to the closing of the GQ gap, the best explanation to date for this was put forth by Murray (2006) and Chay et al. (2009). Accordingly, the g gap was partially closed by health improvements (i.e . environmental influences that had immediate biological impact). This explains the change in the substantially biological g and leaves environmental explanations wanting to explain current g-loaded differences, at least between mid to upper SES members of the said populations. With regards to the g-loadedness of group differences, Flynn’s cognitive explanation does not hold in wake of the vast social, psychological, and neurophysiological manifold that general intelligence represents.

      We could use a basketball analogy to capture both positions on this matter. Flynn argues that g is analogous to general basketball ability; it’s important because it correlates with the ability to do complex moves, say like making reverse two-handed dunks. Flynn’s point is that to do a reverse two-handed dunk, one needs to learn all the basic moves. Since environmental disadvantages (poor coaches, limited practicing space, etc.) handicap one when it comes to basic moves, they necessarily handicap one more when it comes to complex basketball moves. Rushton and Jensen argue the g is analogous to a highly heritable athletic quotient; it’s important because it correlates with basic physiology, generalized sports ability, and basic eye-motor coordination. Their point is that it’s implausible that disadvantages in basketball training would lead to across the board disadvantages in all athletic endeavors and, moreover, lead to a larger handicap in general athleticism than to a handicap in basic basketball ability. Rather than disadvantages in basketball training leading to disadvantages in general athletic ability, it’s much more plausible that disadvantages in general athletic ability would lead to a reduced effectiveness of basketball training.

      Flynn and other environmentalists can only circumnavigate g by insisting that a web of g affecting environmental circumstances, in effect, constructs g from the outside in. Given that g is psychometrically structurally similar across populations, sexes, ages, and cultures this seems implausible as it would necessitate that either everyone happened to encounter the same patter of g formative environmental circumstances just at different levels of intensity or that environmental circumstances were themselves intercorrelated.

      Given the weakness of environmental accounts of GQ differences, the hereditarian hypothesis is a more plausible explanation than the environmental (0-genetic) hypothesis.

      http://abc102.wordpress.com/2011/02/17/spearmans-hypothesis-and-the-jensen-effect/

  9. 猛虎 says:

    Well, Flynn doesn’t believe that the Flynn Effect is due to improvement in nutrition. I have once read that it is because he has argued that the while FE has stronger effect among the lower half of IQ distribution, the improvement in nutrition does not occur among the lower half of height distribution. See for example “The end of the Flynn effect?: A study of secular trends in mean intelligence test scores of Norwegian conscripts during half a century” (p. 360). He however believes that the FE can be explained by cultural factors. See what he has said here.
    But this shows that Flynn clearly dismissed (in your citation) Rowe’s studies. See for example “Under the Skin: On the Impartial Treatment of Genetic and Environmental Hypotheses of Racial Differences” and “No more than skin deep: Ethnic and racial similarity in developmental process”.

    As for your second paragraph, I always wondered why africans could not even reach islands just off the african coast for thousand years. Eurasians was using boats thousand years ago. The islands off the coast of Africa have been settled by eurasians, not by africans. I don’t think you could explain this by arguing that it is because there were too few healthy africans to reach the islands which lie just a few miles. Yes. A few miles. The difference between africans and eurasians is so huge. And, if africans were intelligent, they would be certainly able to overcome this little obstacle.

    • 猛虎 says:

      “the improvement in nutrition does not occur among the lower half of height distribution”

      Hmm… Should be … “the improvement in height does not occur” (or at least, did not keep pace with the higher IQ gains among the lower half of IQ distribution).

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s