11 Comments
Jun 26, 2023Liked by Windsor Swan

I appreciate you response, but such is not the case. But of course, if one reads certain authors, one gets such an impression. My take is that IQ as measured by the tried and true tests is a lengthy one and composed of a number of subtests—albeit, any number of IQ studies use proxies for IQ out of necessity since the formal tests are of such length and cost. IQ “tests” on the Internet are of course mostly ego driven short tests and of a clickbait nature.

I know of no reputable researchers who have successfully shown IQ is more environmental than hereditary. Indeed the opposite from all studies. “g” is important because of its relation to life outcomes. Those with a high IQ do better in many aspects of life than those with a low IQ. The army has been interested in IQ for 100 years and uses a form of such testing as a placement tool for different specialties training and recruit selection. Such is the use of “g” or IQ (I am using these interchangeably).

The shortest and best book I can recommend to read on the current state of the research is, “In the Know” by Russell T Warne (Author). This book will get you started as a survey.

Expand full comment
author

I think you have misunderstood my comment. My fault - I wasn't clear. I am aware that IQ is mostly hereditary. But I don't want to take up anymore of your time so I will leave it there. Thank you for the contributions.

Expand full comment
Jun 25, 2023Liked by Windsor Swan

Complete overlap would make other measures meaningless as they would be identical. The correlation between individual measurements attempts to get at an approximation of “g”, but there is error involved which makes the measurements imperfect, hence an (acknowledged) approximation of “g”. We have multiple measurements (subtests) in order to cancel out as much error as possible in our estimate (of “g”), but “g” does not entail/require complete overlap or perfect correlation. There may be some “aspects of intelligence” as you put it, but that is not related to the concept of “g” per se. The aspect/concept of different intelligences was indeed that which lead to “g” in the first place. All the attempts at measuring “different intelligences” always wound up with collections of correlated sub-test scores.

I myself have no problem with “g” *and* different intellectual strengths and weaknesses as indicated by subtests. To me, “g” is like a measure of a highly efficient brain. The higher the “g”, the more “horsepower” so to speak. Whether the car powered by this engine has superior ability in handling curves, or off-road traction, or acceleration is not a show stopper for me.

Expand full comment
author

My understanding is that most intelligence researchers believe that there is no need for different subtests as they can give one test that is highly g loaded and this measures intelligence adequately. And that intellectual strengths and weaknesses are not genetic but environmental. Thats my understanding, and maybe its wrong.

However, if it is wrong, and researchers acknowledge that g doesnt cover everything, then what exactly do we need g for at all? Thats what I dont get. I think you and I actually have very similar understandings of intelligence. Thanks for the comment.

Expand full comment
Jun 25, 2023Liked by Windsor Swan

"Meaning there is some overlap between types if intelligence."

Precisely, and that is what is termed "g". You've contradicted your own argument.

Expand full comment
author

I dont think so, g is saying there is complete overlap, not partial. Only partial overlap means there are some aspects of intelligence that having nothing to do with other parts. Or maybe i should say aspects of aspects of intelligence.

Expand full comment

Somehow author hasn't understood how group factors work before deciding to blog on it. A sex difference on a given test depends on its loading on g and the group factors, as well as the sex differences on each group factor. Men have a big advantage on the spatial factor and women a smaller advantage on the short term memory factor. The big debate is about the g factor difference, which Lynn, me etc think is about 4 IQ.

Expand full comment
author

Well im certainly not an expert, I dont even have a degree - which explains the "somehow". But I dont think a sex difference has anything to do with a tests g loading, there are some tests with high loading and some with low - where there is no sex difference in either. So if the answer is not about spatial intelligence being a distinct ability then it must be about this "group factor" which I havnt come across before. I will check it out.

But also, if there arent multiple types of intelligence, how can some people have high variance? Doesn't really make sense to me but happy to be proven wrong. Anyway thanks for the comment.

Expand full comment

High variation in the sex gap across tests is expected when their g loadings aren't 1.00. Some additional variation is caused by the group factors. You can model it with bifactor for the most interpretable model.

PS. I also have no degree in this field.

Expand full comment
author

Fair enough. When you say "high variation in the sex gap across tests"... my understanding is that there is little to no variation in the sex gap across tests. Males consistenly outpeform females spatially, and females consistently outperform males in memory tasks, albeit to a smaller extent. Other categorical differences are inconsistent and generally not large. Thats what I thought.

Anyway I wasnt referring to a variance in the sex gap, but the variance that higher IQ people display when tested. If high IQ people consistently display more variance, and from my understanding they do, doesn't this at least disprove the universality of g?

Expand full comment

Lynn a few years back devoted an entire issue of Mankind Quarterly to the debate of male/female “g” differences (IIRC). He invited opposing viewpoints as well in this issue. It’s a good read to get a feeling for the debate. I remember the 4 point difference postulated, however I’ve never quite had a feel for the effect of such (smallish) differences in IQ. When talking male vs female, I usually can’t get past the inherit sex linked behavioral differences between male and female.

Expand full comment