huh, I am boggled if there exist people with measurably better working memory which doesn’t translate to IQ. I had believed working memory was a key intelligence bottleneck.
I think the disappointing failure of the WM/DNB training paradigm 2000-2015 or so to show any meaningful transfer to fluid intelligence, while being able to show transfer to plenty of WM tasks, proved that the high correlation of WM/Gf was ultimately not due to WM bottlenecking intelligence.
The nature of human intelligence remains something of a puzzle. The most striking recent paper I’ve seen on what human intelligence is is “Testing the structure of human cognitive ability using evidence obtained from the impact of brain lesions over abilities”, Protzko & Colom 2021. That is, interpreting brain lesions as surgically precise interventions into (non-fatally) disabling specific regions of the human brain, there… is no bottleneck anywhere? Yet, we can totally predict intelligence (up to like >90% variance as the ceiling) from neuroimaging and stuff like brain volume has causal influences on intelligence as indicated by LCV etc. So overall I’ve been moving towards the bifactor model with general body integrity as the other (causal but temporally prior) factor, and trying to reconcile it with DL scaling. Highly speculative, but seems reasonably satisfactory so far...
huh, I am boggled if there exist people with measurably better working memory which doesn’t translate to IQ. I had believed working memory was a key intelligence bottleneck.
I think the disappointing failure of the WM/DNB training paradigm 2000-2015 or so to show any meaningful transfer to fluid intelligence, while being able to show transfer to plenty of WM tasks, proved that the high correlation of WM/Gf was ultimately not due to WM bottlenecking intelligence.
The nature of human intelligence remains something of a puzzle. The most striking recent paper I’ve seen on what human intelligence is is “Testing the structure of human cognitive ability using evidence obtained from the impact of brain lesions over abilities”, Protzko & Colom 2021. That is, interpreting brain lesions as surgically precise interventions into (non-fatally) disabling specific regions of the human brain, there… is no bottleneck anywhere? Yet, we can totally predict intelligence (up to like >90% variance as the ceiling) from neuroimaging and stuff like brain volume has causal influences on intelligence as indicated by LCV etc. So overall I’ve been moving towards the bifactor model with general body integrity as the other (causal but temporally prior) factor, and trying to reconcile it with DL scaling. Highly speculative, but seems reasonably satisfactory so far...
It’s a good point! I failed to notice my confusion there.