My understanding is that IQ and g are distinct. g is the quality that IQ attempts to (imperfectly) measure. Are there tests that present themselves as 'g' tests rather than IQ tests?
Yeah, 'g' is what IQ wishes it was but couldn't quite be.
Turns out it still can't be easily pinned down as a single objective number, best we get is more of a coefficient. After taking a bunch of different cognitive tests, weighting the results and integrating things down a level or two you can try to figure out how much those results were down to generalized cognitive aptitude and how much was specialised/learned talent for each task.
CHC cognitive testing (edit- changed from CAS, I mixed up my three letter cognitive test acronyms, -0.03g for me. You can also use the CAS2 battery of tests to calculate 'g' too, but they're against the idea that 'g' is a single constant so it seems kind of rude to do so 😂) is one battery of tests that has been around for a while that doesn't purport to be an "IQ test" but is comprehensive enough to be used as a base to try and calculate 'g' from. But that's the thing, the tests themselves stop at giving you the results of multiple independent cognitive performance tests, the 'g' calculating is something else afterwards.
The 'g' is only semi-dependant on the test results, you can get the different 'g' results from the same answers just by changing how you weight each test result, and they're all debatably plausible. 'g' is supposed to be the essence of general intelligence, it is not learned or specialized, it is something structural and rock steady across all tasks. The hallmark of a good formula for 'g' interpretation is that it is even more tightly heritable than traditional IQ values. But it's also kind of an academic distinction, good IQ values still track very closely with 'g' and the only diagnostic advantage of 'g' is potentially differentiating between edge cases of someone with high 'g' and low learned ability across almost all cognitive tasks and someone with moderate 'g' and high learned ability across almost all cognitive tasks, but the real world problem solving ability of both individuals is going to be comparable except on one or two types of task.
My understanding is that IQ and g are distinct. g is the quality that IQ attempts to (imperfectly) measure. Are there tests that present themselves as 'g' tests rather than IQ tests?
Yeah, 'g' is what IQ wishes it was but couldn't quite be.
Turns out it still can't be easily pinned down as a single objective number, best we get is more of a coefficient. After taking a bunch of different cognitive tests, weighting the results and integrating things down a level or two you can try to figure out how much those results were down to generalized cognitive aptitude and how much was specialised/learned talent for each task.
CHC cognitive testing (edit- changed from CAS, I mixed up my three letter cognitive test acronyms, -0.03g for me. You can also use the CAS2 battery of tests to calculate 'g' too, but they're against the idea that 'g' is a single constant so it seems kind of rude to do so 😂) is one battery of tests that has been around for a while that doesn't purport to be an "IQ test" but is comprehensive enough to be used as a base to try and calculate 'g' from. But that's the thing, the tests themselves stop at giving you the results of multiple independent cognitive performance tests, the 'g' calculating is something else afterwards.
The 'g' is only semi-dependant on the test results, you can get the different 'g' results from the same answers just by changing how you weight each test result, and they're all debatably plausible. 'g' is supposed to be the essence of general intelligence, it is not learned or specialized, it is something structural and rock steady across all tasks. The hallmark of a good formula for 'g' interpretation is that it is even more tightly heritable than traditional IQ values. But it's also kind of an academic distinction, good IQ values still track very closely with 'g' and the only diagnostic advantage of 'g' is potentially differentiating between edge cases of someone with high 'g' and low learned ability across almost all cognitive tasks and someone with moderate 'g' and high learned ability across almost all cognitive tasks, but the real world problem solving ability of both individuals is going to be comparable except on one or two types of task.