#At this point of there isn't evidence of reproducibility in not trusting the research at all

[lamia-lmiae]

I mean, if the peer review was reliably real and the research was reproducible.

[max1461]

I take it that you're gesturing at two sets distinct of critiques here, one scientific (sensu lato) and one political. I'd like to avoid the political discussion as much as possible, aside from making one comment: indeed I suspect that no one is at fault for you not being von Neumann, but fault is a useless concept anyway. If you were suffering poverty or some other social ill as a downstream consequence of not being a von Neumann, or not being regarded as a von Neumann, or any other such variation on the theme, I might suggest that we reorganize society somewhat to mitigate this fact! Material conditions and all that. But this is probably reflective of deeper axiomatic differences in our ethics, so I won't push it.

On the scientific side, I'll copy-paste what I wrote in another reply:

There are, as far as I can tell, three basic claims about IQ testing that its supporters seem to make. The first claim is that IQ tests measure something consistent and reproducible at all. The second claim is that this thing which they measure corresponds with what we might reasonably dub “intelligence”. The third claim is that this thing which they measure is innate, not amenable to practice and roughly invariant for each person. If I’m to buy the general picture of human intelligence that, you know, people who are really into IQ testing would like me to buy, I think I’d need to be convinced of each of these points.

Most of the debate around IQ seems to derive from people contesting point (3), but my skepticism is deriving primarily from points (1) and (2).

Point (1) concerns me mostly because psychology has proven to struggle with this in the past, even with regard to what appear to be relatively well accepted results, and I'm not sure what makes IQ different. There's an object-level question here: what is the specific evidence that sets IQ research apart from e.g. those studies claiming willpower gets depleted when you use it (which every pop-sci outlet touted to the public for like 20 years, before going "oops lol none of those are replicable"). There's also a meta-level question, the central debate of social epistemology: how does a non-expert identify actual experts and know when to trust them? Certainly, the events of the last two years have drawn into sharp focus the epistemic precarity that follows from unexamined trust of experts—and epidemiology is, as a far as I can tell, on much better epistemic footing than psychology in general.

It's possible that the only way for me to resolve this issue would be to read all the research myself, but I hope there's an easier answer than that, because I don't really have that kind of time.

You seem to have a lot of confidence on point (2), but I'm not totally sure it's warranted. I'm not sure of a really good, objective criterion for defining "general intelligence" as distinct from just "being good at specific things". Some people are good at math, some people are good at painting, etc., and these are not necessarily or even usually the same people. The point here is not to say "everybody is special in their own way" or something, it's to say: the concept of g as I understand it is as a cognitive trait (or collection thereof) that makes you better at all or most other cognitive tasks the more of it you have. I'm very willing to believe in such traits; memory, as you mention, seems like a straightforward example. On the other hand, there might be many such traits, and they might not correlate with one-another, and they might each have different effect sizes, and so forth. What IQ advocates have to defend is that (a) these various factors can all be meaningfully summarized by a single number, and (b) that IQ tests in particular actually do this. Specifically, (b) is crucial and is where most of my skepticism is coming from.

Points (1) and (2b) seem to be what you're dismissing when you say "[t]he actual point of contention isn't IQ itself". Maybe that's true for others in this debate, but those are my biggest points of contention!

Measurement is to science as definition is to math. In math we may start with some intuition, and that intuition may guide us in various directions of research, but ultimately we are only able to actually do math once we pick a specific and rigorous definition to capture that intuition. The results we then prove are ultimately dependent on what definition we picked, regardless of what our initial intuition was. Similarly, a quantity is only scientifically meaningful once we've specified a way to measure it. We can talk about g as an intuition, that's fine, but without specifying a metric we can't actually do science about it. Once a metric is picked, the validity of all the science thereafter will be staked very specifically on the features of that metric. The specific validity of the test or set of tests by which we judge IQ is the crux of this entire debate.

I know I'm probably inviting some terrible discourse (especially in this part of tumblr) by saying this, but... given the track record of psychology as a field in general, I as a layman am yet to be provided with any compelling reason to take IQ testing seriously. I've heard IQ scores tend to be persistent (as in, take the test again get the same result), but IIRC so do personality test scores like Myers-Briggs and five-factor, and at least significant numbers of people seem to regard those as bunk.