4 ms·

I think it's important to note that the first graphic you posted doesn't include longitudinal data (i.e., doesn't correct for the Flynn effect). Compare to: htt

by r-zip 6y ago

I think it's important to note that the first graphic you posted doesn't include longitudinal data (i.e., doesn't correct for the Flynn effect). Compare to: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2683339/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2683339/.

sinenomine 6y ago
Thank you for posting this paper, it is more statistically/psychometrically sound than the one I posted. By testing each person across a significant path of their lifespan, and correcting for test-retest drift, it produces a more defendable result. Still, it does not make decline disappear into statistical insignificance, just lowers its magnitude to 0.8~1.0 s.d, which is still large, thus sparing the qualitative reasoning in my interpretation. It should also be noted, that the Flynn effect has ended somewhere in the 1990s-2000s, at least in Norway[2], so it is likely that long-term it ceases to be a disturbing factor for between-cohorts comparisons. Test-retest gains will still be an issue though (and there is a viewpoint that a large part of Flynn effect could be explained as a large-scale environmentally-driven test-retest effect acting on the population). I also like the review[1] of paper you posted, suggesting several measurable histological/neurological/physiological correlates of cognitive aging, aptly summarized on this graph: https://i.imgur.com/CpQ6Upc.png https://i.imgur.com/CpQ6Upc.png If you are interested in discussing these papers, or joining a community of like-minded people, feel free to ping me via the email in my profile. 1. https://pubmed.ncbi.nlm.nih.gov/19231030/ https://pubmed.ncbi.nlm.nih.gov/19231030/ (sci-hubbable) 2. http://www.iapsych.com/iqmr/fe/LinkedDocuments/Sundet2004.pdf http://www.iapsych.com/iqmr/fe/LinkedDocuments/Sundet2004.pd...

4 ms·

I think it's important to note that the first graphic you posted doesn't include longitudinal data (i.e., doesn't correct for the Flynn effect). Compare to: htt

by r-zip 6y ago

sinenomine 6y ago
Thank you for posting this paper, it is more statistically/psychometrically sound than the one I posted. By testing each person across a significant path of their lifespan, and correcting for test-retest drift, it produces a more defendable result. Still, it does not make decline disappear into statistical insignificance, just lowers its magnitude to 0.8~1.0 s.d, which is still large, thus sparing the qualitative reasoning in my interpretation. It should also be noted, that the Flynn effect has ended somewhere in the 1990s-2000s, at least in Norway[2], so it is likely that long-term it ceases to be a disturbing factor for between-cohorts comparisons. Test-retest gains will still be an issue though (and there is a viewpoint that a large part of Flynn effect could be explained as a large-scale environmentally-driven test-retest effect acting on the population). I also like the review[1] of paper you posted, suggesting several measurable histological/neurological/physiological correlates of cognitive aging, aptly summarized on this graph: https://i.imgur.com/CpQ6Upc.png https://i.imgur.com/CpQ6Upc.png If you are interested in discussing these papers, or joining a community of like-minded people, feel free to ping me via the email in my profile. 1. https://pubmed.ncbi.nlm.nih.gov/19231030/ https://pubmed.ncbi.nlm.nih.gov/19231030/ (sci-hubbable) 2. http://www.iapsych.com/iqmr/fe/LinkedDocuments/Sundet2004.pdf http://www.iapsych.com/iqmr/fe/LinkedDocuments/Sundet2004.pd...