“Well, suppose that factor analysis was a perfect model. Would that mean that we’re all born with some single number g that determines how good we are at thinking?”
″Determines” is a causal word. Factor analysis will not determine causality for you.
I agree with your conclusion, though, g is not a real thing that exists.
(As far as I can tell, there are only three problems with the study I linked: 1. due to population structure, the true causal effects of the genes in question will be misestimated (this can be fixed with within-family studies, as was done with a similar study on Externalizing tendencies, 2. the study might lack the power to detect subtle differences between the genes in their specific degrees of influences on abilities, which if detected might ‘break apart’ g into multiple distinct factors, 3. the population variance in g may be overestimated when fit based on phenotypic rather than causally identified models. Of these, I think issue 2 is unlikely to be of practical importance even if it is real, while issue 1 is probably real but will gradually get fixed, and issue 3 is concerning and lacks a clear solution. But your “g is not a real thing that exists” sounds like you are more pessimistic about this than I am.)
My response is we have fancy computers and lots of storage—there’s no need to do psychometric models of the brain with one parameter anymore, we can leave that to the poor folks in the early 1900s.
How many parameters does a good model of the game of Go have, again? The human brain is a lot more complicated, still.
There are lots of ways to show single parameter models are silly, for example discussions of whether Trump is “stupid” or not that keep going around in circles.
This seems to be an argument for including more variables than just g (which most psychometric models IME already do btw), but it doesn’t seem to support your original claim that g doesn’t exist at all.
(You seem to have put your comments in the quote-block as well as the thing actually being quoted.)
Since immediately after the bit you quote OP said:
No. A perfect fit would only mean that, across a population, a single number would describe how people do on tests (except for the “noise”). It does not mean that number causes test performance to be correlated.
it doesn’t seem to me necessary to inform them that “determines” implies causation or that factor analysis doesn’t identify what causes what.
(Entirely unfairly, I’m amused by the fact that you write ‘”Determines” is a causal word’ and then in the very next sentence use the word “determine” in a non-causal way. Unfairly because all that’s happening is that “determine” means multiple things, and OP’s usage does indeed seem to have been causal. But it may be worth noting that if the model were perfect, then indeed g would “determine how good we are at thinking” in the same sense as that in which factor analysis doesn’t “determine causality for you” but one might have imagined it doing so.)
What would your response be to my defense of g here?
(As far as I can tell, there are only three problems with the study I linked: 1. due to population structure, the true causal effects of the genes in question will be misestimated (this can be fixed with within-family studies, as was done with a similar study on Externalizing tendencies, 2. the study might lack the power to detect subtle differences between the genes in their specific degrees of influences on abilities, which if detected might ‘break apart’ g into multiple distinct factors, 3. the population variance in g may be overestimated when fit based on phenotypic rather than causally identified models. Of these, I think issue 2 is unlikely to be of practical importance even if it is real, while issue 1 is probably real but will gradually get fixed, and issue 3 is concerning and lacks a clear solution. But your “g is not a real thing that exists” sounds like you are more pessimistic about this than I am.)
My response is we have fancy computers and lots of storage—there’s no need to do psychometric models of the brain with one parameter anymore, we can leave that to the poor folks in the early 1900s.
How many parameters does a good model of the game of Go have, again? The human brain is a lot more complicated, still.
There are lots of ways to show single parameter models are silly, for example discussions of whether Trump is “stupid” or not that keep going around in circles.
This seems to be an argument for including more variables than just g (which most psychometric models IME already do btw), but it doesn’t seem to support your original claim that g doesn’t exist at all.
(Also, g isn’t a model of the brain.)
(You seem to have put your comments in the quote-block as well as the thing actually being quoted.)
Since immediately after the bit you quote OP said:
it doesn’t seem to me necessary to inform them that “determines” implies causation or that factor analysis doesn’t identify what causes what.
(Entirely unfairly, I’m amused by the fact that you write ‘”Determines” is a causal word’ and then in the very next sentence use the word “determine” in a non-causal way. Unfairly because all that’s happening is that “determine” means multiple things, and OP’s usage does indeed seem to have been causal. But it may be worth noting that if the model were perfect, then indeed g would “determine how good we are at thinking” in the same sense as that in which factor analysis doesn’t “determine causality for you” but one might have imagined it doing so.)