... my question is how he picked the "sample size of 90". I suspect he knows more than 900 women, which in turn suggests that he knows more than 90 lesbians by simple probability. I suspect that the lesbians he notices are the bull dyke type, so it becomes a self-selecting and unrepresentative sample.
Involvement in various "gay community" activities at college, some friends that were gay, and then of course meeting friends of those friends. Yes, selection bias rears it's ugly head. There is a possibility that the only people who show up to gay support organizations or who help with pride week are raging bull dykes, but I tend to think there's more to it then that, especially sense I saw the same distribution through various sources, and I saw a correlation, that is, sort of pretty girls were almost always bi, girls who looked like adolescent boys were almost always totally lesbian, and hot girls lasted a month and then decided they were straight.
Yes, there are exceptions, but let's not let confirmation bias spin us the other way. Just because there's a 1 out of a 100 exception does not discount there being some kind of trend. I think there's something to be said of a bias towards seeing equality where it doesn't exist. Seems to me with what we know about homosexuality and biology, one would assume there is a distinct possibility of some trends showing physical differences, but everyone jumps at the idea that samples showing inequality are just biased, even if they have no good reason to believe that, other then that it's simply possible, while ignoring the fact that their own experiences could be biased.
If homosexuality is biologically driven, we're talking about some very radical changes to brain chemistry and structure. It tells us to either like breasts or wide shoulders, it tells us to like curves or triangular bodies, it attracts us to the particular musk given off by men or women, drives us to or away from vagina. Any one of those things by itself is complex, differentiating between a breast and a circle drawn in the dirt, not being attracted to utters, placing value judgments on "nice" breasts, liking larger breasts, disliking saggy breasts. We can induce homosexuality in animals by giving the proper blast of hormones at just the right moment; we can make female spider monkeys run around trying to mount other female spider monkeys, but these same hormones at slightly different times will make radical changes in other areas of the brain and to the body. I think it's weird to assume that a natural process creates homosexuality, but that it's impossible for the same natural process to make any changes to physiology.
Even if you dispute that homosexuality is biologically driven, that's even more to my point. One would expect social pressures toward heterosexuality to be much stronger for "hot" women.
I would also be more apt to believe it is just my imagination if I got some false positives now and again, all I get is false negatives. That is, you can point to some hot lesbians, but can you point to any girls who look like adolescent boys who are totally straight? I use to point to Rossie O'Donald and Ellen as examples of false positives, Rossie looked like a dyke but had a huge crush on Tom Cruise, Ellen was kinda boyish but made jokes about sneeking into hot guys appartments and laying in wait for them, but yeah, you're right, obviously physical appearence predicts nothing.