>I find it somewhat incredible that the results would be going essentially in opposite directions so close in age range. I wonder if the kids studied were more

>I find it somewhat incredible that the results would be going essentially in opposite directions so close in age range. I wonder if the kids studied were more from higher socio-economic strata than the median. The methodology section of the appendix to the study linked[0] by the Atlantic article is vague on socio-economic status. Since many of those who are on the high end have had significantly less disruption than those on the lower end. If the sample was skewed toward the higher end of the spectrum, perhaps that could account for the results. [0] https://ifstudies.org/ifs-admin/resources/final-teenquarantine2020.pdf https://ifstudies.org/ifs-admin/resources/final-teenquaranti... Edit: Included the missing link.