Conversation

Scott thinks it might be due to STEM effects...older siblings more likely to go into STEM fields. If so, it might also contribute to twitter case. STEMmies being overrepresented here in general and probably specifically for 2 of 4 of us (Brian and me)
1
3
I do know indirectly via other polls that my followers are strongly STEM biased. So if that’s part of the explanation, it would apply at least in my case.
Image
Image
1
Note: the bigger the sibling-set, the stronger the left-tail clipping effect. If you have many siblings, like Brian, gap between oldest and youngest will be bigger. If you are oldest/middle and joined Twitter at ~22, almost no chance that 5+ years younger sibling will be here.
1
2
And finally, to the extent twitter attracts a “type” independent of age, even in steady state there’s going to be tail clipping. If you’re 20 and joined today, older sibling is probably also on here, but younger one likely will not join. So it’s not a 1-time early adopter effect
1
1
Of course, gotta also note usual caveats: absence of evidence is not evidence of evidence, can't prove a negative, replication crisis in social psych etc., but the unfoundedness of several specific birth-order effect hypotheses does call the general idea into question as well.
2
1
Above my statistics pay grade, but clearly if you do a convolution of a moving unstable distribution like a series of intersecting adoption curves spaced 2-5 years apart, with a symmetric one like birth order, you should get an asymmetric result.
Replying to
Hmm, there is the fact that all 4 of you are friends so some confounding effects will crop up from that. (Eg, I followed you after following Visa, and answered both your polls.) I'm also the right age for twitter with a younger sister on snap, although she is more STEM-inclined.