Thread by @ent3c

From time to time you hear it said that executive function, a cognitive ability involving mental control, is 100% heritable. It seems like a compelling basis for hereditarianism- under all that complexity there is a solid gold nugget of pure genetic variation! [wonky thread]

The @kph3k book, for example, says "First,it is nearly 100 percent heritable. That is, within a group of children who are all in school, nearly all of the differences in general EF between them are estimated to be due to the genetic differences between them." /2

This statement isn't wrong, exactly, but it is misleading if you don't understand what it means. The important thing is that the 100% h2 is estimated for a *latent* variable, an abstract statistical concept that "explains" why multiple measures of EF are positively correlated. /3

The key idea is that different measures of EF aren't very highly correlated with each other, so the latent variable doesn't capture much of the variation in the domain. The first paper to make the claim of 100% h2 for EF was Friedman et al (2008). ncbi.nlm.nih.gov/pmc/articles/P… /4

Let's take a look. They fit a hierarchical factor model to nine indicators of EF and estimated an ACE model for the common factor, arriving at the amazing 100% heritability for the top-level latent common factor. This is their Figure 4. /5

Notice that the factor loadings aren't all that high. Multiplying through the loadings on the 1st and 2nd level factors, the median loading is about .45, meaning that only about .45^2 or 20% of the variance in the observed variables is "explained" by the latent factor /6

The important thing to understand is that there is no actual 100% heritable "EF score" that can be assigned to each child; if there were the variable wouldn't be latent. When you look at the heritabilities of the actual tests, none are particularly close to 100%. /7

The mean of the observed scores isn't 100% heritable. Estimated "factor scores" aren't 100% heritable. The latent variable is a modeled abstraction, isolating the part of EF common among the subtests. The proper conclusion is: 20% of executive function is 100% heritable. /end

/ps Latent variables are like microscopes- they allow you to magnify small things, but they don't *really* make things bigger. Getting excited about 100% h2 of EF is like looking at an insect under a microscope and yelling, "Look out a giant bug!"

Extra thought: in addition to everything else, "due to" in this sentence means "associated with" not "caused by". If I had a time machine I would strangle that wording (which is from long before @kph3k or me) in the cradle.