Thread by @KirkegaardEmil

Back in the good old 1970s, Richard Herrnstein wrote a newspaper article about IQ. The reactions were not much better than nowadays.

From The Intelligence Men: Makers of the I.Q. Controversy, Raymond E. Fancher, 1985

goodreads.com/book/show/1923…

John Stuart Mill was very wrong about the genetic influence on human diversity. This is typical of feminists, and empiricists. This doesn't have to be the case (one can accept that knowledge comes from experience in general but not variation among traits), but practically it is.

It is curious how many of these great men of science somehow fail to see that generalizing from their own upbringing is not a good idea. And even if it was, they fail to take into account their much less eminent siblings.

Also typical of egalitarians, Mill thought one should apply moral reasoning to science, that is, one should grant one hypothesis epistemic preference over another for non-evidentiary reasons. Same as Turkheimer and many others.

Francis Galton and the invention of the adoption study. As far as I know, no one has ever actually done such a historical study of adoptees in ancient Rome. But I presume it could be done. @gtredoux

@gtredoux And likewise, Galton on the invention of the classical twin design using MZ and DZ twins. Galton's evidence was a bit flimsy, but his intuitions were spot on and the conclusions likewise. A good prior goes a long way in science.

Galton's hypothetical eugenics program designed for Victorian England. Not so unlike that eventually tried in Singapore.

In a curious turn of history, Alfred Binet -- inventor of the first practical intelligence test -- spent quite a few years conducting quite poor studies on hypnosis and the like, only to be publicly humiliated when his results were based on obvious study limitations.

Binet experimented on his own daughters and made several discoveries: 1) reaction time decreases with maturation (along with intelligence, 2) and ditto for the individual variation in this, 3) and to some extent this reflects attention control.

Binet's experimentation also lead him to figure out that vocabulary tests and their reliance on abstraction is a great way to measure intelligence.

Moving on to Charles Spearman. He read psychology textbooks in his spare time as an army officer. And he had an emotional revulsion to Mill's associationism. Surely an odd reaction, but it doesn't matter ultimately in science what the original motivation was.

Amazingly, Spearman read the Wissler, which seemingly disproved some of Galton's ideas regarding correlations involving reaction time and basic sensory discrimination with intelligence. Spearman figured out it had in part to do with low reliability.

Spearman then tried to do what was essentially a latent variable analysis to get the disattenuated correlation between composite variables, but did it wrong (these reliabilities are incorrect).

William Stern on the origin of "IQ", the intelligence quotient. He realized that growth of mental powers was a rate, and this differs. IQ was originally a quotient, but hasn't been for many decades (we use deviation scores). The term stuck though.

"mixed success" actually very high correlations with rated intelligence.

Terman was also an admirer of Binet, not just Galton. Typically, the histories of psychology present rather simple hierarchical structures of how ideas spread, whereas cross-fertilization was often the case.

You've heard of the Mensa fallacy by now (I hope), but how about joining Densa?

Didn't grow up together, yet both dropped out of high school, became electricians, and even had the same kind of dog with the same name. Unbelievable.

Jensen's method of correlated vectors before it was cool. In this case not applied to race gaps, but school gaps.

Burt was ahead of his time. In fact, so far ahead that he even was one of the first to adopt the modern mantra regarding heritability: within groups yes, but not between and deemphasize gaps. Maybe Harden etc. can take note of their eminent predecessor.

Unfortunately, Burt was also a narcissist. Take this curious case of removing yourself as a coauthor and rewriting the paper to make yourself look good. On the other hand, Eysenck is not the best source for these kinds of stories.

Given his character and the implausible results of reporting the exact same 3 decimal correlations after increasing the sample sizes, some people had doubts about the data and results.

Many people have said that if new convincing evidence had come out, Arthur Jensen would have changed his mind. There is some evidence of this hypothetical because Jensen did change his mind in a major way prior -- about psychoanalysis.

Somehow I didn't know this, but Leon Kamin -- the IQ critic -- was one of the subjects of the McCarthy senate hearings.

Kamin deserves credit for bringing doubt on the Burt studies, and for good reason.

Burt's results are however about the same as those in other studies published before and since, though MZ apart data are very rare.

Jensen held some surprising naive views on the practical role of intelligence testing. At least, in the 1980s which is when these quotes are from (Bias in Mental Testing, or Straight Talk books).