Hacker News Reader

Comment by A_D_E_P_T | Hacker News Reader

The heritability of IQ, from twin studies and via other traditional methods of measuring heritability, is way higher than the author of that article implies. It's as high as 0.8. (80% of variance down to genetics.) See, e.g.: https://pubmed.ncbi.nlm.nih.gov/23919982/

If you measure heritability only via SNP studies, it looks a lot smaller. This is what's called the "Missing Heritability Problem." See: https://en.wikipedia.org/wiki/Missing_heritability_problem

This problem is, as yet, unsolved. There's a lot we don't know about traits that are governed by "many genes of small effect."

The author of the article probably knows better and is being disingenuous, but it's flat-out wrong to say that intelligence is only weakly heritable.

View on HN

Anyone who wants to look at heritability numbers should make sure to watch these lectures first. Heritability is often a misleading stat, because it has an unintuitive definition. Sapolskiy covers common failings of heritability numbers, as well as confounding factors for twin studies.

https://www.youtube.com/watch?v=OareDiaR0hg

https://www.youtube.com/watch?v=FZ6D3-XHmEk

Full Lectures: https://www.youtube.com/watch?v=NNnIGh9g6fA&list=PL848F2368C...

An example of how non-intuitive heritability is: whether a person wears earrings or not would be a heritable trait if we're looking at mid-20th century America, because the variability is entirely explained by whether you're male or female.

Heritability is unable to identify whether a trait is caused by genetics, only whether the variation correlates with genetics.

What do you make of this section that claims recent research has identified an upper bound on the genetic component?

> …prediction accuracy depends on sample size, could the findings drastically change with more samples in the future? In fact, through the magic of statistics, we actually know that this claim will always to be true. We know this because we have estimated a parameter called molecular heritability, which tells us the upper bound on what a genetic predictor could ever achieve. Molecular heritability comes in two forms. “Population heritability” is a non-causal measure of the overall correlation between all genotyped variants and the trait, and it is estimated in large populations of unrelated individuals. Population heritability includes direct genetic influences on the trait that we typically think of as “genetics” but it also includes a lot of other correlated stuff that we typically think of as confounding, like cultural influences on the trait from relatives or prior generations (including the effects of parenting or dynastic advantages) or biases due to population stratification. In contrast, “direct heritability” is a measure of the specific genetic influences that are acting within individuals, and it is estimated in a large number of families. Direct heritability is immune to many sources of environmental confounding and, with some assumptions, can be interpreted as an estimate of causal genetic effects (for theory and derivations see [Veller et al. (2023)])

Of course, IQ only weakly correlates to the intelligence that people actually refer to colloquially. It doesn't test for most behaviors that people do consider intelligent, and it tests a bunch of stuff completely unrelated to intelligence (e.g. cultural trivia).

this isn't accurate at all.

Remarkable that it starts with "Of course"

I mean, who takes IQ seriously? It's a business metric.

How do you figure?

No legitimate IQ test involves cultural trivia. It’s pretty much all Raven’s Progressive Matrices though that has it’s own issues.

> No legitimate IQ test involves cultural trivia.

It's often under the label "verbal comprehension".

That is blatantly false in my experience.

I once was administered an in-person WAIS-IV IQ test, which has been called the "Gold Standard of IQ Tests," from a licensed clinical psychologist for the purposes of diagnosing a medical syndrome.

On the verbal reasoning portion, I distinctively remember being asked two questions that sat wrong with me (paraphrasing):

1. "Who was Sacagawea?"

2. "Who was Catherine the Great?"

There were also vague questions in the word definition/comparison section.

I distinctively remember one of the questions being (paraphrasing):

"Compare and contrast the two words 'Practical' and 'Pragmatic.' What do the two words have in common and what are their differences?"

My issue, especially with the latter question, is what type of answer is acceptable?

I could say something like, "'Practical' and 'Pragmatic' both describe being concerned with reality or feasibility vs. theory."

But what about the differences? What if I said something like, "Well, there is no such things a 'Pragmatic' joke."

Would I have been correct, wrong, a smart ass, or at the mercy of the psychologist's opinion? Would I be marked wrong because it wasn't what the test authors' intended answer despite me being technically correct?

IQ is only a narrow specific type of intelligence biased to western education indeed, but it certainly isn't a pop quiz like you make it our to be

> IQ is only a narrow specific type of intelligence biased to western education indeed

I see this claim a lot, but what is even meant by "western education" here? IQ tests show similar results in most countries that have /any/ form of formal compulsory education. It's clear that education or lack of education can affect IQ scores, but not so much that IQ tests are biased by the specification type of education.

The best research I've seen on the matter concludes that education biases the results because in order to receive and participate in an IQ test you need to have some base level of language understanding of the terminology used in the IQ test itself, which implies that you've received some minimal amount of education. The most common IQ test used in the United States, the Stanford-Binet (SB5) generally recommends not administering it to people under the age of 8 years old for this reason, and there are specialized tests for children that show reduced bias for education when used on non-educated populaces.

Or, in summary, it seems standard IQ tests bias for /any/ education, not specifically "western" education. In other words, there's no cultural bias as claimed.

> what is even meant by "western education" here?

IQ tests how good of a computer you are.

It certainly seemed like a pop quiz the last time I took it. I just can't imagine it's useful outside ranking people arbitrarily for the purposes of shrinking a hiring pool or some similar need.

>unrelated to intelligence (e.g. cultural trivia).

When you think those facebook "test your IQ!" tests are actual IQ tests....

> IQ only weakly correlates to the intelligence […] it tests a bunch of stuff completely unrelated to intelligence (e.g. cultural trivia).

These are unfortunate myths. See the following on IQ testing, relationship of various tests to general intelligence, and the absence of cultural bias in testing:

- Standardized exams measure intrinsic ability, not racial or socioeconomic privilege https://milkyeggs.com/society/standardized-exams-measure-int...

- Fair and Square: A Conclusion on IQ Test Bias https://menghu.substack.com/p/fair-and-square-a-conclusion-o...

"Generally, the most commonly used IQ tests aren’t biased or only minimally biased as to be of no practical value."

- Bias is Often Unpredictable https://www.cremieux.xyz/cp/140810182

"Bias does not simply mean that a measure gives different scores to different groups. This is one of the most common misunderstandings about “bias”. What it really means is that members of different groups obtain different scores conditional on the same underlying level of ability."

- The Relationship Between the Scholastic Assessment Test and General Cognitive Ability https://sci-hub.st/https://journals.sagepub.com/doi/abs/10.1...

"Furthermore, empirical evidence suggests a substantial relationship between the SAT and g. In a study of 339 undergraduates, Brodnick and Ree (1995) used covariance structure modeling to examine the relationship between psychometric g, socioeconomic variables, and achievement-test scores. They found substantial general-factor loadings on both the math (.698) and the verbal (.804) SAT subtests."

- Systemwide Review of the Report of the Academic Council’s Standardized Testing Task Force (STTF) (2020) https://senate.universityofcalifornia.edu/_files/underreview...

Intrinsic ability to what, cram for a test? Doesn't seem that useful to measure.

> The heritability of IQ, from twin studies and via other traditional methods of measuring heritability, is way higher than the author of that article implies. It's as high as 0.8.

No one here can seem to be able to prevent an LLM from saying 3+5 is something other than 8, yet you talk about "IQ" as if some test cam boil down human brains to one integer and then rank them.

Also if one person is born in a stable, educated house, and another in a poor unstable house where they are not educated, how is environment separated from heritability? People try to make social science studies to do this, but the problem is they have limited data to draw on.

This ideal of a Calvinistic Brahmin chosen people is rooted in parasitism and idleness. We can have a meritocracy society where everyone starts on the starting line, or we can have one where at birth someone is either on the worker bee or rich kids of Instagram born on third base path. The idle class aristocratic parasitic view favors the very weak idea of heritability - some hidden genetic change in the past 50,000 years after behavioral modernity.

I am not disagree with what you said, but I wanted to point one thing out:

> how is environment separated from heritability?

It isn't. That's why heritable traits are different than genetic traits.

Alcoholism is a heritable disease but is not a genetic disease. One cannot suffer from alcoholism in an environment where alcohol does not exist.

Reply on news.ycombinator.com