Gary N.

Sociology, Education, Student Achievement, Social Stratification, Social Theory. Author, Education, Social Background and Cognitive Ability.

[Gary N. @GaryNMarks]

@LilyGrace_09 Eve

[Gary N. @GaryNMarks]

@DamienMorris Plenty of whacko ideas out there that are acceptable to great swathes of the public. The problem is that they are immune from falsification.

[Gary N. @GaryNMarks]

@Steve_Sailer @flytotheday His thesis of legitimation crisis on the inevitable fall of western capitalism wasn't right. It was more a product of hopeful marxism than anything else.

[Gary N. @GaryNMarks]

@gtredoux According to them, there isn't much occupational mobility (there actually is, mostly because of ability) because privileged families game the system just enough so that the odds ratios remain the same. They are so powerful, they collectively are able to do that.

[Gary N. @GaryNMarks]

The point of all that is that the intergenerational correlations can be largely predicted from Fisher-type models the heritability of a trait and the amount of assortative mating (spouse correlation). For education, the heritability is about 0.4, assortative mating is about 0.5, so the predicted intergenerational correlation is 0.24. If the heritability is 0.5 (which some studies estimate) the predicted correlation is 0.31.

[Gary N. @GaryNMarks]

Intergenerational correlations between parents and children have been consistently around 0.4.

Bouchard and McGue {, 1981 #2471@1058} report a weighted average parent–offspring correlation for IQ of 0.42, the same as that for height and fingerprint ridge count—both traits presumed to be minimally influenced by the social environment. According to Duncan {, 2005 #844@34, 43} intergenerational correlations for math and reading test scores measured at much the same age also cluster around 0.4.

For educational attainment, Hertz et al. {, 2008 #117} analyzing data spanning 50 years across 42 countries, reported a global average parent–child education correlation of 0.4. This figure varied slightly by region: 0.39 in Western Europe and the United States, and 0.41 in former Eastern Bloc countries.

Svalastoga {, 1966 #7103@176} concluded from his study of occupational mobility in Western Europe that a father-son correlation of 0.4 is a reasonable approximation for industrialized European nations. Beller and Hout {, 2006 #2465@19} estimated father-son correlations for occupational attainment ranging from 0.3 to 0.4 across a range of contexts. Across many countries, Marks {, 2014 #3974@208} calculated an average intergenerational father-son correlation of 0.37 for occupational status.

Income correlations follow a similar pattern. Esping-Andersen {, 2004 #7245@291} contends that the core intergenerational income elasticity was 0.4. The OECD {, 2018 #5368@36} reported average intergenerational earnings elasticity of 0.37 across 24 OECD countries. Intergenerational elasticities are not the same as the intergenerational correlation but are often close. For the US, Hertz {, 2005 #3345@165} calculated an intergenerational correlation in long-run average income in the U.S of 0.4.

Why do parent–offspring correlations congregate around 0.4 for social stratification variables?

[Gary N. @GaryNMarks]

@krichard121212 @RegimeEnforcer In non longnitudinal studies, rge is typically small. Are rge estimates as high as 0.8 0.9 credible? What comprises E?

[Gary N. @GaryNMarks]

@krichard121212 There no doubt that IQ is important for educational, occupational and economic outcomes and is highly heritable. There's heaps of evidence. So don't lie to yourself and others.

[Gary N. @GaryNMarks]

@QuentinDempster @sussanley This is an example of the stunning intellect of ABC employees. Can all they be that naive? Would any of them disagree with Laura or Quentin?

[Gary N. @GaryNMarks]

@jayjoseph22 How many times have you made that fatuous claim. You ignore hundreds perhaps thousands of studies where heritability is found to be much greater than zero.

[Gary N. @GaryNMarks]

There are many recent examples of that. Just use paramaters that mimic genetic relationships and call it environmental. For example, the high 0.7 MZA iq correlations are argued as purely environment but there are no environmental variables or set of environmental variables that have correlations with iq close to 0.7. It is usually 0.3 or 0.4. And the 0.7 correlation is an R square value.

[Gary N. @GaryNMarks]

The 0.3 father-son and 0.4 brother correlations are predicted by Fisher's 1918 equations. For a heritability of 0.5 for achievement and a spousal correlation of 0.3, the predicted correlation is 0.29 for both pair types. Pretty close despite measurement error. The unhinged politically motivated antiheritarians cannot explain these or any other observed empirical correlation among genetically related kin. They don't take data, theory or the literature seriously if it contradicts their strongly held poltical convictions.

[Gary N. @GaryNMarks]

Analyzing University of Oxford class lists between 1861 and 1892 on the level awarded for B.A. degrees (six categories from first class honors to no degree), Schuster and Elderton {, 1907 #6361@13} reported a father-son correlation of 0.31. This correlation is much the same as father-son correlations for achievement. This correaltion cannot be accounted for within the SES paradigm because there was little variation in the social backgrounds of Oxford students at that time.

[Gary N. @GaryNMarks]

Schuster and Elderton {, 1907 #6361@13} also calculated brother correlations from Oxford class lists of around 0.4.
During the second half of the 19thcentury at Harrow and Charterhouse, elite public schools in London, students were streamed to classes (e.g., the upper sixth) based on their prior academic performance. The brother correlations were again around 0.4 {Schuster, 1907 #6361@23-24}.

[Gary N. @GaryNMarks]

This is getting ridiculous. Yes the sample sizes are way too small but that is all we have. The MZA studies are consistent with each other and with many other studies estimating the heritability of iq in adults. The MZA studies can't be dismissed. And as others have said, this debate was done and dusted decades ago.

[Gary N. @GaryNMarks]

@hecubian_devil @Dogfacedman1 IQ has strong correlations with educational (about 0.55) and occupational attainment (0.45). Its correlation with income varies a lot but it is often between 0.3 and 0.4 with accurate measures of income. So no, IQ isn't simply a measure of test taking.

[Gary N. @GaryNMarks]

@jayjoseph22 No it's not. It's about MZA twin studies in general. That is clear.

[Gary N. @GaryNMarks]

@jayjoseph22 Besides the MZA studies there are hundreds studies published in reputable journals that find substantial heritabilities for iq, achievement, educational and occupational attainment and income. All that evidence cannot be rejected.

[Gary N. @GaryNMarks]

@jayjoseph22 @davidbessis @Biorealism @halocosm @paulg @cremieuxrecueil They are valid. The equal environments assumption was supported by the study of misclassified twins and other studies. The genetic-environment correlation is small, often not statistically significant, and doesn't substantially reduce heritability estimates. I could go on.