View Full Version : The Heritability of Intelligence: Not What You Think


sarahsweets
10-31-13, 09:26 AM
http://www.huffingtonpost.com/scott-barry-kaufman/the-heritability-of-intel_b_4174626.html


Interesting article.

Dizfriz
10-31-13, 10:36 AM
It was a interesting article...Thanks for posting it!

Dizfriz

Corina86
10-31-13, 10:52 AM
Interesting, but not at all surprising. If anyone ever worked with people from different cultures and backgrounds, it was pretty obvious.

Amtram
10-31-13, 11:39 AM
Still looking through the figures. . .but it makes sense. You can inherit any number of benefits or deficits and still use them and improve them, or not.

SB_UK
10-31-13, 03:30 PM
What these findings do suggest is that there is a much greater role of culture, education, and experience in the development of intelligence than mainstream theories of intelligence have assumed.

Genes are so yesterday.

Amtram
11-07-13, 05:22 PM
Here's another take on the findings: http://medicalxpress.com/news/2013-10-well-established-views-heritable-intelligence-brought.html#nwlt

Drewbacca
11-07-13, 08:58 PM
Genes are so yesterday.

Who wears short-shorts? :eek:

Drewbacca
11-07-13, 09:08 PM
Here's another take on the findings: http://medicalxpress.com/news/2013-10-well-established-views-heritable-intelligence-brought.html#nwlt

Personally, I hate this kind of study...

In a new study (http://scottbarrykaufman.com/wp-content/uploads/2013/10/Kees-Jan-et-al.-2013.pdf), Kees-Jan Kan (https://sites.google.com/site/keesjankan/) and colleagues analyzed the results of 23 independent twin studies

No independent, reproducible study, in of itself... just data analysis of a bunch of different studies. I have a lot of problems with this approach.

For one, you can cherry pick what studies you want to include or throw out.

Secondly, to really evaluate the author's conclusion and make an educated conclusion, you have to go read the 23 independent studies that we don't have access to.

There is a third major issue, which is that a third party is now analyzing data in order to reach a conclusion that the original studies may not have even been focused on. Variables that the original scientists may have taken for granted, is now becoming a center point for a different group of researchers.

I love this kind of study, in that it generates an interest (and often, controversial conclusions). It yields further research and inspires new papers. Unfortunately, you can never draw any strong conclusions from such a paper... at least, not until you give all of the authors of the original 23 papers on which this was built, a chance to weigh in and possibly even reproduce their own results (with a greater emphasis being placed on the variables and conclusions established in the newer third party paper).

Amtram
11-08-13, 06:57 AM
Yes, you do have to take everything with a grain of salt. But at least this kind of presentation gives you the chance to debate using actual evidence instead of pure speculation. You can attempt to backtrack and see the methodology on some of the papers, and look for editorials, critiques, and the opinions of scientists who blog by doing a search on the papers even if they're not open access yet.

TygerSan
11-08-13, 09:52 AM
I haven't been able to look at the methodology if the original study (though I have to say I'm actually a bit impressed by the detail of the Huffington Post piece).

Without the methods, it is a little difficult to judge bias. Most meta-analytic studies like this one, though, are pretty rigorous in defining and spelling out which search terms they use to grab the original research studies the analysis is based on, what studies met the criterion to be included (if, for example, a study comes up in the search terms, but doesn't give enough methodological detail, doesn't give an IQ test that's properly normed, or if for some reason the authors can't calculate/standardize the results so they can be included with the other data sets, etc., it will be excluded from the analysis).

All of that information should be spelled out in the methods section. Is there bias? Sure, there could be, but I wouldn't throw out a study just because it is a meta-analyisis. In fact, a well done meta-analysis allows for a much larger sample size than one individual study, and can give a fairly comprehensive overview of a particular research topic.

Each individual research study is a very zoomed in view on the issue. Meta-analyses zoom you out so you're not examining just the elephants toe-nail. Of course, you do have to wait until there's substantial data published on a given subject before the meta-analysis has merit. There are always trade-offs.

SB_UK
11-08-13, 10:33 AM
ooo genetics history !

http://galton.org/books/hereditary-genius/

[which we're only now realising is rubbish]

Simply re-inforcing some innate superiority of the invading Imperialists over the savages - to justify barbarism; 'we're doing them a favour'

- it's a pity that the complex disorder geneticist doesn't do us a favour and plant themselves upside down in a plant pot, making muffled 'Bill and Ben' noises from their newfound homes.

At last then, they'll have done some use to mankind.

Of leaving it.

SB_UK
11-08-13, 11:30 AM
Galton believed that if men and women of considerable talents were selected and mated generation after generation, a highly gifted race of people would be the eventual result.And then we discovered that assortative mating of systematizers resulted in enrichment of autism spectrum disorders.Galton - possessed of a remarkably high degree of intelligence (an estimated IQ of 200) and a wealth of novel ideas ... - at least one of which, and no doubt more of which were really DUMB.

http://www.psych.utah.edu/gordon/Classes/Psy4905Docs/PsychHistory/Cards/Galton.html
http://en.wikipedia.org/wiki/Empathizing%E2%80%93systemizing_theory

Amtram
11-08-13, 12:42 PM
Well, there's the abstract. . .

To further knowledge concerning the nature and nurture of intelligence, we scrutinized how heritability coefficients vary across specific cognitive abilities both theoretically and empirically. Data from 23 twin studies (combined N = 7,852) showed that (a) in adult samples, culture-loaded subtests tend to demonstrate greater heritability coefficients than do culture-reduced subtests; and (b) in samples of both adults and children, a subtest’s proportion of variance shared with general intelligence is a function of its cultural load. These findings require an explanation because they do not follow from mainstream theories of intelligence. The findings are consistent with our hypothesis that heritability coefficients differ across cognitive abilities as a result of differences in the contribution of genotype-environment covariance. The counterintuitive finding that the most heritable abilities are the most culture-dependent abilities sheds a new light on the long-standing nature-nurture debate of intelligence.

I'd like to know more about the culture-loaded vs. culture-reduced tests. Those two types can yield a wide range of difference. Unfortunately, the references weren't provided at PubMed or for free on Sage. Soooo. . .

Here's some more info on the methodology from Scientific American. (http://blogs.scientificamerican.com/beautiful-minds/2013/10/17/the-heritability-of-intelligence-not-what-you-think/?print=true) Mike the Mad Biologist offers his opinion (http://mikethemadbiologist.com/2013/10/21/iq-covariance-and-heritability/), along with a link to a pdf (http://scottbarrykaufman.com/wp-content/uploads/2013/10/Kees-Jan-et-al.-2013.pdf) (Drew shared this, but I didn't scroll down far enough) to the full article that includes the criteria and references. So now we should be able to have a discussion based on real evidence and its viability, and understand the strengths and flaws of the actual research, eh?

Drewbacca
11-08-13, 12:46 PM
All of that information should be spelled out in the methods section. Is there bias? Sure, there could be, but I wouldn't throw out a study just because it is a meta-analyisis. In fact, a well done meta-analysis allows for a much larger sample size than one individual study, and can give a fairly comprehensive overview of a particular research topic.

I wasn't really throwing it out, just stating why I don't like this sort of paper. My critique is directed at a class, not the individual research. It's important to look at all studies with an open mind and judge it on its own merit, which I have not done. Unfortunately, the most controversial studies always seem to fall into this category and often get misrepresented by the media. I think that a big part of the problem is lack of access to the general public, I for one would be thrilled if I could view journal articles via my local public library but journals are insanely expensive to subscribe to... I really need a college job, just to have that access.

Amtram
11-08-13, 01:39 PM
I know - meta-analyses can be fraught with bias, so you have to really look at them with a critical eye. OTOH, many of them provide links that you might not otherwise have seen. And once you've looked at the linked research, you have a much better idea of how much bias might be involved.

Because of the number of flawed meta-analyses I'd read that featured references to Joel Nigg, I initially thought he was a crank. When I finally read things he had actually authored, I realized that his data was being cherry-picked and misused for a number of these analyses.

Not too many scientists who blog or write accessible articles are interested in dissecting meta-analyses unless they get a lot of public hype and are either incredibly wonderful or irredeemably awful.

So, for example, Scientific American jumped right onto the cultural bias issue that concerned me in the abstract. To wit:

For instance, here is the cultural load of the Wechsler Intelligence Test subtests:
http://blogs.scientificamerican.com/beautiful-minds/files/2013/10/Figure-1.png (http://blogs.scientificamerican.com/beautiful-minds/files/2013/10/Figure-1.png)
They discovered two main findings. First, in samples of both adults and children, they found that the greater the cultural load, the greater the test was associated with IQ:*
http://blogs.scientificamerican.com/beautiful-minds/files/2013/10/Figure-2.png (http://blogs.scientificamerican.com/beautiful-minds/files/2013/10/Figure-2.png)
This finding is actually quite striking, and suggests that the extent to which a test of cognitive ability correlates with IQ is the extent to which it reflects societal demands, not cognitive demands.
Second, in adults, the researchers found that the higher the heritability of the cognitive test, the more the test depended on culture. The effects were medium-to-large, and statistically significant:
http://blogs.scientificamerican.com/beautiful-minds/files/2013/10/Figure-3-1024x567.png (http://blogs.scientificamerican.com/beautiful-minds/files/2013/10/Figure-3.png)
As you can see above, highly culturally loaded tests such as Vocabulary, Spelling, and Information had relatively high heritability coefficients, and were also highly related to IQ. As the researchers note, this finding “demands explanation”, since it’s inconsistent with the traditional investment story. What’s going on?

It goes on to emphasize the genotype-environment covariance (http://img2.timg.co.il/forums/1_169055133.pdf) - that some intelligences depend more on genes, others are more capable of changing over time depending on environment and motivating factors, meaning that this study doesn't provide as many answers as it does questions!