Pop Psychology: Okcupid Edition (Gay vs. Straight Sex)

I'm of the mindset that gay men and women are pretty much just like straight men and women, the real difference being their preference for which gender they find attractive, which is why the following presentation struck me as interesting:

http://blog.okcupid.com/index.php/gay-sex-vs-straight-sex/

Their sample size was 3.2 million, which is pretty fucking impressive. The thought of having access to the data set they do is sexually thrilling for me (perhaps we could tack another letter onto the end of LBGT for people with my proclivities), so I decided to do another edition of Pop Psychology to take a closer look at what our Okcupid blogger decided to present from that data set and raise some questions that will remain unanswered, but serve as practice (for me, anyway) for critical evaluation. After all, having access to something so amazing demands that it not be squandered or misused.

The first question I'd pose is about that impressive sample size itself: what percentage of it represents straight/ gay men and women? You'd think, this being an article about straight vs. gay sex, that would be the first thing they'd mention. If we assume population level frequencies, that would mean that sample is about 1.6 million men and women total, about 80,000 of which will be gay men (5%, published range from 3-10%) and about 40,000 of which will be lesbian women (2.5%, published range from 1 - 3%). That is assuming we can assume population levels; there is no data presented concerning the the percentage of Okcupid users who identify as straight or gay in the first place. I'm sure the Okcupid people have this data available, but it's not shown here for whatever reason. Since it's not shown, there's no way of inferring whether this population is representative or not, which could throw a possible wrench into its interpretation. I'm not saying it does, just that it might.

Let's start by looking at the first point raised by the article:

Gay people are not sexually interested in straights.    

Match Search Returns

• only 0.6% of gay men have ever searched for straight matches.
• only 0.1% of lesbians have ever searched for straight matches.

What that tells us is that most gay and lesbian people do not go looking explicitly for straight people on the website, which is pretty expected, especially when people's sexual orientations are readily visible. What that does not tell us is is that gay people are not sexually interested in straight people; I assume most of them are, and I assume that for the same reason that most heterosexual men are sexually interested in lesbians, even though they'll probably never sleep with them (but more on that in a bit). I have found that knowing a woman's sexual orientation has not made me any more or less attracted to them, though it does affect my judgments of whether or not I'm likely to have sex with them. I imagine other people don't need to know someone's sexual orientation before they feel any sexual attraction, nor do I feel that knowledge would do much to change that attraction. The website just allows for screening based on sexual orientation, skipping what would be an in-person trial and error process, since not everyone has their sexual orientation tattooed on their forehead.

Those absolute percentages are very small, granted, but they also tell us something else: relatively, gay men are six-times as likely to have ever searched for straight men than lesbians are to have searched for straight women; an interesting finding to be sure. Bear in mind those numbers refer to people who have ever searched for a straight match, not how frequently they do it. My guess is that gay men also search for straight matches more frequently, but without the data available I can't say with any certainty. Moving on:

Gay people aren't promiscuous.


Median Reported Sex Partners

• straight men: 6
• gay men: 6
• straight women: 6
• gay women: 6

This stuck out to me above everything else in the article because it stands in stark contrast to everything I've ever read in the published literature. Simply put, men and women (straight ones) don't report identical numbers of sexual partners, generally speaking, and men and women (gay ones) don't report identical numbers of sexual partners. Has Okcupid, with its huge data set just blown all those other results out of the water? I don't think so; consider the following number sets to understand why:

(1) 0, 0, 1, 3, 5, 6, 10, 30, 100
(2) 3, 3, 3, 3, 5, 7, 7, 7, 7

If we assume those numbers represent number of sexual partners, then in both samples, the median (the middle) number of partners is 5; the mean (the average) number of partners is still 5 for sample (2), but in (1) it's 17.22. See the very big difference? Most every study I recall reading reports the mean, and if they report the median at all, it's in conjunction with the mean. That the Okcupid post only reports the median at best doesn't allow them to make the statement they did, and at the worst reeks of spin to attempt to make groups look  more similar than they actually are.

There are two other problems: the first is that those numbers refer to number of partners, not number of same/opposite sex partners. It's not uncommon for lesbians to have several male partners (in some cases, more male partners than female ones), and that can matter a lot (especially because the same isn't true of gay men).
The second problem, again, returns to the sample itself: the age of these people are not reported. Older people tend to have more sexual partners than younger ones, simply because they've had more time to rack up the numbers. Gay men tend to report establishing a clear sense of their orientation before lesbians, which could mean our lesbian sample is substantially older. Perhaps they controlled for this, but if they did, they make no mention of it (they do mention controlling for things in other blog posts, so I can only assume they did not here).

• 45% of gay people have had 5 or fewer partners (vs. 44% for straights)
• 98% of gay people have had 20 or fewer partners (vs. 99% for straights)

What is curious about this is that the previous stats had broken down the data by gender and sexual orientation; here, we see it recombined to just sexual orientation. I can't think of any good reason to do that from a strictly informative point of view, which implies to me there's probably something else going on (namely, controlling the light in which these findings are presented).
While those differences in percentages appear small absolutely, as was the case previously, they can also be read to say something else: gay people are twice as likely to have more than 20 partners, relative to straight people. I also assume that effect is largely being carried by the male portion of the gay side, which could potentially mean that gay men are up to four-times as likely to have over 20 partners, depending on what percentage of that gay 2% and straight 1% are men. So gay men could be somewhere between 2-4 times as likely to have over 20 partners, relative to their straight counterparts, which strikes me as more promiscuous.On that note:

we found that just 2% of gay people have had 23% of the total reported gay sex, which is pretty crazy.

 An interesting finding to sure, but the same probably holds for straight people. Male numbers of sexual partners (straight) are distributed differently from female numbers; there are a small number of men with a large number of sexual partners along with many men that have low numbers to no number at all. On the other hand, women are less likely to have the peaks that men do in the high end, but they're also less likely to be shut out of the game entirely (fewer women with no numbers). Accordingly, if this was broken down by gender, I'd predict that most of that effect was being carried by gay men, not lesbians.

However, since we are dealing with a single gender in the case of gay men or women, the distribution curves would have to look different from their heterosexual counterparts: either the peaks of the gay community would have to be smaller than those of the straight community and the distribution more even, or larger numbers of gay men and women would not have (m)any sexual partners. Since we already know from the previously reported data that, at the upper end of 20+ partners, gays outnumber straights 2-to-1, that would require more gays not having, relative to straight people. That would mean, cutting off the top end of the distribution, the gay community should be even less promiscuous than the straight community, on average.

Then again, perhaps gay men are more promiscuous all around (which would simply shift the graph up), while lesbians are less so (shifting the graph down), but that article seems to imply that's an unfair stereotype.

So there are good reasons to doubt the statement that promiscuity is equal across the four combinations of straight/gay men/women; it's very expected, and established, the that groups are not the same sexually (men are not women, and women are not men). However, there are two more possibilities that the article doesn't touch on that I think could underestimate the extent of gay promiscuity.

The first is the social risks openly gay people may face, which could result in them being less willing to try to hit on people they otherwise would like to. In that sense, it's not that homosexuals are less promiscuous then they'd prefer by choice.

The second is the size of the dating/mating pool that gay men and women have to deal with; were only 2.5% of the population potentially sex-able to me (50% aren't men, and of that 50% that are men, only about 5% are gay, and of that 5%, not all will be attractive), I'd either have to know many, many more people than average to reach the average number of sexual partners, or be particularly more motivated to make it happen, provided that 2.5% was no more promiscuous than the roughly 50% I have potentially available now. Since I don't think gay people know substantially more people than straight ones, that raises the possibility of motivation and/or promiscuity.

Of course, none of this says anything about judging the worth of a person by their number or choice of sexual partners. I can't help but wonder if the people at Okcupid were trying to present the data selectively to make the case that gay people are really just like straight people, which means they're OK. If they were, they went wrong on two major fronts. The first is that the gay community does not need to be the same as straight people to be OK. They're already OK; always have been. The second big issue is that for gay men and women to be similar to straight men and women, gay men are going to be different from lesbians along the same lines that straight men differ from straight women. The real differences in behavior emerge because when a man is pairing with other men, their sexual interests tend to converge more than the interests of men with women; that same goes for women pairing with other women. There's no need for everyone to be identical for everyone to be equal, and trying to present select bits of the data to make them look more the same isn't helping.