1. Exquisite Tweets from @toadmeister, @ElToroSolo

    PreoccupationsCollected by Preoccupations

    Interesting piece and well-written. But you over-estimate impact of differences in childhood SES on difference in life outcomes and underestimate impact of genetic differences — and you set too much store by epigenetics. Lot of good recent research in this area, as I’m sure you

    Reply Retweet Like

    toadmeister

    Toby Young

  2. You know what? I'm bored of marking, and too many people misunderstand this. But not many of them try to set educational policy on that basis, unlike the chancer below or his equal in unwarranted self-regard Dom Cummings (@OdysseanProject)

    Thread follows.

    Toby Young @
    Interesting piece and well-written. But you over-estimate impact of differences in childhood SES on difference in life outcomes and underestimate impact of genetic differences — and you set too much store by epigenetics. Lot of good recent research in this area, as I’m sure you

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    The first thing you need to understand about heritability is that it has nothing to do with outcomes in a particular child.

    *Nothing whatsoever*.

    It is a property of groups and the differences within them, not individuals.

    2/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    Put another way, knowing IQ is highly heritable tells you nothing at all about what is responsible for the IQ of any particular child, whether it is due to "nature" or "nurture".

    In fact, the question is ill-posed. Everything is 100% "nature" and everything is 100% "nurture.

    3/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    That is, in development, everything is caused by genes, and everything is caused by the environment. Environment does nothing without genes, and genes do nothing without environment.

    Donald Hebb phrased this nicely; he said

    4/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    Asking whether nature or nurture contributes to any phenotypic trait more is like asking whether height or breadth contributes more to a rectangle's area.

    The outcome is a product of the two. Both are necessary, both are sine qua non.

    5/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    So what does it mean to say something is heritable? It's a property not of individuals, but of groups and the differences between the individuals that compose them.

    Take the rectangles again.

    Say you've one group of rectangles with similar breadth, but v. different height.

    6/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    In that group, most of the difference in area between the rectangles can be explained by differences in height.

    Take a second group, with v. different breadth, but similar height; then most of the difference in area is explained by differences in breadth.

    7/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    We calculate the degree to which differences in each variable correlate with differences in the whole by a mathematical technique called ANalysis Of VAriance; ANOVA.

    8/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    ANOVA tells us, for any population, how much differences in IQ correlate with genetic differences, and how much with environmental differences.

    Heritability (H2) measures the relative influence of the two; variation due to genes divided by total variation (in simple cases).

    9/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    Now, to see what's so completely, devastatingly wrong about @toadmeister and @OdysseanProject's claims, ask yourself this:

    In a population of genetic clones, what will H2 be?

    That is, what proportion of the population-wide differences in IQ are down to genetic differences?

    10/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    The answer is <...drumroll...> 0.

    Zero. Zilch. Nada.

    There's a very simple reason for this, which you may have guessed; in a population of genetic clones, there are no genetic differences at all, so no differences in IQ can be explained by genetic differences.

    11/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    Now take a population raised in precisely identical environments. What is the heritability of IQ there? That is, what proportion of differences in IQ are down to genetic, as opposed to environmental differences?

    Everybody who said 100%, give yourselves a pat on the back.

    12/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    Because there are no environmental differences for this population, enivronmental differences cannot explain any of the differences in IQ.

    Now, the clincher. What does it mean, then, to say that IQ is highly heritable in most developed countries?

    13/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    IT MEANS THAT SUCH COUNTRIES HAVE COMPARATIVELY EGALITARIAN EDUCATION SYSTEMS, SO THAT THE COMPARATIVE EFFECT OF ANY ENVIRONMENTAL DIFFERENCES THAT DO EXIST IS SMALL.

    14/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    That IQ is highly heritable doesn't mean that egalitarian approaches to education policy are doomed to failure.

    Quite the opposite.

    It means that they have been hugely successful.

    15/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    You would think that education policy-makers would understand this, even if the general public don't.

    You would think, at a minimum, that the education policy-makers like @toadmeister and @OdysseanProject, who bang on about behaviour genetics, would understand this.

    16/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    But you would be wrong.

    You would think that they would demonstrate a modicum of shame or humility when people who know what they are talking about point all of this out; it is a *basic* concept of behaviour genetics, after all.

    17/

    Reply Retweet Like

    ElToroSolo

    Shane N. Glackin

    ElToroSolo

    Shane N. Glackin