Busting gay stereotypes with data

As a gay guy, you sometimes have to put up with some pretty offensive stereotypes that get thrown your way by extremists in the community and the media. These stereotypes are usually deployed in the form of anecdotes about how gay people are "promiscuous" or "corrupting". These misrepresentative anecdotes have serious consequences, not just in the continued denial of fundamental civil rights to gay people in places like Iran and Nigeria and the United States, but also at the individual level. Recently here in the US, there's recently been a rash of suicides by gay teens. (Some of the stories were recounted in this heartwrenching statement (video) by Forth Worth (TX) Councilman Joel Burns, and the It Gets Better Project is a wonderful new resource for gay teens subjected to bullying.)   

But as every good data analyst knows, the plural of "anecdote" is not "data". But without data published outside of the academic literature, it's been hard to combat the anecdotes with reality.

Until now: Enter the clever folks at dating site Ok Cupid, who have used their database of 3.1 million user profiles to put these stereotypes to the test. Because each profile includes a self-reported sexual orientation along with a wealth of demographic, profile and essay information, data analysis reveals that:

• Gay men (and women) are no more promiscuous than their straight counterparts, based on self-reported statistics on number of sexual partners, and
• Gay people are not sexually interested in straight people, based on searches for profiles according to reported sexuality. In fact, there was not a single gay user in the OK Cupid database who primarily searched for straight people.

To be sure, this isn't a peer-reviewed academic analysis. You can argue that the population of Ok Cupid members isn't representative of the population as a whole: it certainly skews young and single. (For example, I'm nearing 40 and have been married to my wonderful husband for 6 years: I don't have a profile on Ok Cupid.) But, it's data, and as always the data say what the data say: amongst the population of people with profiles on Ok Cupid, these offensive stereotypes just don't hold water.

That said, some stereotypes are based in truth, looking at this analysis of personality traits (based on key phrases appearing in profiles):

Straight men are more interested in sports? Knock me down with a feather. For the complete details of this analysis and much more besides, check out the link below.

oktrends: Gay Sex vs. Straight Sex

Comments

[Ajay]

Until recently gay relationships were officially illegal in India- which led to whimsical harassment by police as well as the overwhelmingly conservative Indian society. Techie and geek circles are majority male, but surprisingly the above average intelligence of the techies has not made them any more tolerant. As for teens committing suicide, it reminds me of Oscar Wilde's and Lord Byron's creativity being hounded, though Alexander the Great had better luck as gay guy. The OK Cupid study is both fascinating in it's use of data and not anecdotes , as well the bold stand to take up a social issue for a business. To tolerate evil is to participate in it- just like Anti Semitism and Racism on color (or nationality)- hatred has no place in civilized society. If you want to hate gays, or jews,or asians- go to Iran

[Gary Strader]

Mr. Smith,

I am very sympathetic with your political cause. However, I believe you are engaging in what you are accusing others of doing - stereotyping. All stereotypes have some basis in truth, which is why they are stereotypes, but we shouldn't promote those that make our interests look good, and hide those that make us look bad. A stereotype is a stereotype.

"the data say what the data say" - is it good statistics to take a single corporate funded "study" that you admittedly say is not peer-reviewed academic analysis, and is self-reported, and turn around and say that this is "truth"? Because you agree with it? Is that confirmation bias?

I don't mean any disrespect. I think your post is morally right, but I don't think it's good statistics.

[David Smith]

Gary, thanks for your comments, but I'm not sure I understand your point about stereotyping.

I'm going to stand by what I said. The data *do* say what the data say, here and always. Amongst those who have profiles at Ok Cupid, the conclusions as stated in the OkCupid post are valid. Now, you could take issue that the data analysis was done incorrectly, or (more likely), that the sample isn't representative of the population as a whole.

In the end though, this is all just data summarization, not statistical inference. Nonetheless, I find it to be useful and interesting data.

Thanks for contributing to the discussion, and no disrespect intended here either.

[Dan]

Fascinating and intriguing. Thanks for the heads up on the info and findings.

What is especially interesting to me is that the personality traits graph shown here nails me to a tee. And while this, I suppose, in some ways is stereotyping me as a gay man, it is not being trotted out to nail me to some negative demographic, it is merely "reporting" results. Could someone use this in a negative way to scar? Yup, but the odd thing is that (from my gay prism-of-vision) it is the straight guy who might be in the cross hairs of negative implications.

Either way, good stuff!