My initial assessment is that there isn’t any correlation between IQ and test scores, ergo the tests are measure of preparation and preference not merit or performance.
This is completely false - I'm not sure how you got that assessment.
Many test scores (e.g. ACT) are heavily IQ/g-predictive.
SAT used to be better at predicting IQ/g but it's still pretty good.
g-factor is the single largest predictor of academic success (which, these days, is mediated mostly by test scores).
Is there evidence of that? Many university recruiters at some of the most prestigious schools believe otherwise.
Is there evidence for that?
As has been pointed out, some national Mensas accept standardised tests as a substitute for an IQ test.
For references, see here eg.: https://www.sciencedirect.com/science/article/abs/pii/S01602...
Wikipedia:
Some measures of educational aptitude correlate highly with IQ tests – for instance, Frey & Detterman (2004) reported a correlation of 0.82 between g (general intelligence factor) and SAT scores; other research found a correlation of 0.81 between g and GCSE scores, with the explained variance ranging "from 58.6% in Mathematics and 48% in English to 18.1% in Art and Design".
https://en.wikipedia.org/wiki/Intelligence_quotient#Social_c...
> The correlation between IQ scores and grades is about .50. This means that the explained variance is 25%. Achieving good grades depends on many factors other than IQ, such as "persistence, interest in school, and willingness to study" (p. 81).
While that paragraph specifically mentions school grades instead a more narrow subset of certain specific tests the point still stands as affirmed by the APA. The numbers you quoted from the third paragraph allow are only about 80% aligned. That is a tiny bit better than half between perfection and a random guess.
The unfortunate reality is that performance on those tests is primarily a direct correlation of time and money spent on preparation specifically for those tests. A major differentiator between those standardized tests and many general IQ tests is centrality versus assessment distribution. You wouldn't for example test creativity with a multiple choice question with only one correct answer the same for everybody.
Look, most people who are from a higher income population want to believe that standardized tests are a fair reflection of intelligence, because with enough practice and preparation they can score higher and thus believe they are more intelligent than the population at large. If the population at large spent less effort preparing for the test then its value as a tool of performance compared to other people is less valid. That is selection bias.
That's a bullshit guess.
More importantly that study did not account for performance disparities versus preparation.
You could also spend 5 seconds Googling the question and find that that this is not exactly an untrod ground in academic literature. The relationship is strong and well-established, far more so than almost any other result from the social sciences.
Also, yes, test results, can, in fact, be improved by studying (or other things!). This does not mean that they are not strongly correlated to IQ. Reality is generally not monocausal. I don't see how this contradicts the original claim.
It does not mean they are related or that they aren’t, which only suggests you are missing the point: selection bias.
> You could also spend 5 seconds Googling the question and find that that this is not exactly an untrod ground in academic literature.
It has been studied exhaustively. The relation between IQ and academic performance is poor according to research from the APA in addition to the personal opinions of people who perform college selection professionally. That correlation is better than a random guess but isn’t anything close to alignment.
People, despite the evidence, need this to be true. After all it is a money making industry. Standardized college selection tests are like tax preparation, an industry that can be easily eliminated by a more simplistic unbiased and more effective tool. With enough effort to the contrary the arguments begin to sound like a self-sustaining fraud built around a fictional necessity and a bunch of bandwagon fallacies.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6963451/
https://www.psychologicalscience.org/pdf/ps/Frey.pdf
I certainly don't "need" it to be true - things are either true or they aren't, and needing or wanting differently doesn't change that.
"That correlation is better than a random guess but isn’t anything close to alignment."
Do you want to make a more specific claim? Again, the original claim you're attempting to rebut is "Many test scores (e.g. ACT) are heavily IQ/g-predictive." How strong does the effect need to be? Since, again, nobody is claiming that it's 100% predictive, merely a significant factor. Do you want to give me an effect size? Keep in mind the typical social psychology effect size comes out to r usually between 0.15 and 0.3:
https://jenni.uchicago.edu/Spencer_Conference/Representative...
What is the practical difference?
> I certainly don't "need" it to be true - things are either true or they aren't, and needing or wanting differently doesn't change that.
You are confusing truth for fact. Truth is where you hang your belief.
To be clear this is the basis of the APA's opinion on the matter: http://psych.colorado.edu/~carey/pdfFiles/IQ_Neisser2.pdf and the APA suggests there is only about a 81% alignment between IQ and academic/test performance. A 50% alignment, when there are only 2 items to compare, is random guess.
All opinions that test results correlate with performance are, either directly or indirectly, rooted in Lewis Terman's famous work on IQ which has been disqualified many times. His research wasn't even good and failed in his own assessment over a longitudinal study. Yet, people still follow this garbage.
EDIT: May this serve as some food for thought:
https://www.psychologytoday.com/us/blog/beautiful-minds/2009...