Unless I am misreading the survey design, they are randomly sampling the 16-65 age group. There is a significant trend of lower scores with age, even within a b

Unless I am misreading the survey design, they are randomly sampling the 16-65 age group. There is a significant trend of lower scores with age, even within a birth cohort. Under those conditions, it's not clear that the weighting is taking care of that bias. Also their USA results have near the worst non-response bias of any country, and come with cautionary notes about that.