I wish the ads for subscription went away if you've actually subscribed.
(i.e. The two bits in the post that say "Knowingless is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber." that then show an icon saying that I've already subscribed.)
"total sample consists of 24,875 bio males, 9,143 bio females, 9,851 cis bio males, 6,692 cis bio females, 1,130 enby bio males, 1,715 enby bio females, 914 transwomen, and 736 transmen."
The number of bio males doesn't add up. Did most of them not answer the question about whether they were cis?
"I slightly regret that I combined the ages 19-25 in one cluster when providing people options to choose from, as that makes what is probably an artificial bump in this graph."
i thought about that, but wasn't sure it was kosher? i feel like theres some god of virtue in the sky who would strike me down if i did that. but if i said i did that maybe it would be fine.
I would think it's the other way around. Making histograms with unequal-width bins but giving each a height based on the total in each bin is the common mistake that is not kosher.
Using the total divided by the bin width (so that *area* is proportional to total) makes it correct. You want to plot "frequency density".
I wish the ads for subscription went away if you've actually subscribed.
(i.e. The two bits in the post that say "Knowingless is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber." that then show an icon saying that I've already subscribed.)
"total sample consists of 24,875 bio males, 9,143 bio females, 9,851 cis bio males, 6,692 cis bio females, 1,130 enby bio males, 1,715 enby bio females, 914 transwomen, and 736 transmen."
The number of bio males doesn't add up. Did most of them not answer the question about whether they were cis?
"I slightly regret that I combined the ages 19-25 in one cluster when providing people options to choose from, as that makes what is probably an artificial bump in this graph."
Divide by the width of the bin?
i thought about that, but wasn't sure it was kosher? i feel like theres some god of virtue in the sky who would strike me down if i did that. but if i said i did that maybe it would be fine.
I would think it's the other way around. Making histograms with unequal-width bins but giving each a height based on the total in each bin is the common mistake that is not kosher.
Using the total divided by the bin width (so that *area* is proportional to total) makes it correct. You want to plot "frequency density".
I would have liked to collect exact ages so we could check hypotheses about the effect of going off to college.
Are there lots of sexy animals in college?
There's a meme on the right wing that going to college causes a sudden spike in deviancy. That's empirically testable.