Yes, if my interests are limited to activity-set X, and interest in X is strongly gender-linked, then I should not be surprised if my chances of a random person sharing an interest with me correlate strongly with that person’s gender. And if I have a choice between picking humans at random from either a mixed-gender or a single-gender jar of humans, picking the proper single-gender jar maximizes my chances of finding an interesting human.
But in real life, that’s not the only choice I have. If I’m only interested in X, I can choose social activities that are highly structured around X. Having done so, I’m effectively picking humans at random from an X jar. And, yes, I should expect the gender ratio in the X jar to not be evenly distributed. But also, at that point I should stop using gender as a proxy metric for X, because otherwise I’m in effect double-counting gender.
If instead I continue to select by gender, even on reflection, that seems to indicate that I’m not using gender as a proxy metric for X, but rather interested in gender for some other reason.
Yes, if my interests are limited to activity-set X, and interest in X is strongly gender-linked, then I should not be surprised if my chances of a random person sharing an interest with me correlate strongly with that person’s gender. And if I have a choice between picking humans at random from either a mixed-gender or a single-gender jar of humans, picking the proper single-gender jar maximizes my chances of finding an interesting human.
But in real life, that’s not the only choice I have. If I’m only interested in X, I can choose social activities that are highly structured around X. Having done so, I’m effectively picking humans at random from an X jar. And, yes, I should expect the gender ratio in the X jar to not be evenly distributed. But also, at that point I should stop using gender as a proxy metric for X, because otherwise I’m in effect double-counting gender.
If instead I continue to select by gender, even on reflection, that seems to indicate that I’m not using gender as a proxy metric for X, but rather interested in gender for some other reason.