[gwern]

Definitely. This is exactly the sort of study you strongly expect the causal conclusion to not obtain and disappear in a randomized study/genetically-sensitive design (anxiety and other mental issues are highly heritable, of course), right up there with such classic sociology non-results as 'children whose fathers leave do worse' or 'children who experience corporal punishment have more behavioral problems' or 'children whose mothers drank during pregnancy are stupider', both because the actual randomized studies of things like early-childhood interventions like Headstart show no effects or fadeout and because there is tons of self-selection in who goes into scouting and who is able to/chooses to stay in it. (Most of the kids who joined with me dropped out at some point, and the more troubled ones dropped out earliest.) Like most of these results, I expect most of the correlation would disappear if you compared siblings, or better yet (but harder to find), discordant twins, or used polygenic scores of genetic risk for anxiety disorders. This study couldn't do it because they took a one-week sample of UK births so no siblings and too few twins, but I'm sure a twin registry or the UK Biobank or someone has asked about extracurriculars like scouting.

[gwern]

And oh yes, since scouting tends to be a family thing where if one son is enrolled in Boy Scouts, the other sons will be too, that means any supposed causal effect of Boy Scouts would fall under 'shared environment' estimates from family & twin designs, which measure the influence of all variables influencing entire families such as parenting choices. And shared environment estimates for anxiety and mental health disorders in general are almost always small. So... if taken at face value, the choice of Boy Scouting or not could explain most or all of a family's influence on anxiety! This seems highly improbable, to say the least.

[objectivistbrit]

A gwern classic - the second sentence of this wall of text contains 112 words and 14 verbs. I think your tendency to include tons of parenthetical remarks, caveats and every example you can think of, confuses the reader and makes it hard to figure out your essential point.

[gwern]

'I made this comment very long only because I have not had the leisure to make it shorter.'