>>12400450>>12400535looking at pic related,
IQ for children is incredibly impressionable, how smart you are as a 4 year old, and how smart your parents are as a 4 year old is only 0.4. the same relationship, but at the age of an adult is then 0.8, so simply iq is more environmentally influenced as a child, and as you mature into an adult, your genetics play a larger roll, and people's iq regress up/down closer to what their innate genotyping iq would be.
Secondly when children get "IQ gains", they haven't actually gotten any smarter, its a flaw in the IQ test.
The iq batteries they use to measure intelligence, consists of multiple subtest, the problem the different subtests individually measure, not just intelligence (g), but other mental abilities that the subtest involuntarily measures simultaneously, and they all have various g-loading's (how much the test measures general intelligence, the thing you want to measure, as apposed to other abilities.) The point of using composite scores is you are finding the overlap from all the test which should be g, and the bordering abilities are meant to be left out. When these kids get IQ gains, literally all of the gains are in the 'other abilities' and not in general intelligence at all. The study below finds that the amount of increases in IQ, in each individual subtest, and the g loading of the subtest is literally a perfect negative correlation of -1.
so simply the IQ gains are increases in particular abilities (not g) that the IQ test accidentally and involuntarily measures.
https://www.researchgate.net/publication/266976086_Are_adoption_gains_on_the_g_factor_A_meta-analysis 
