Thursday, October 28, 2010

OkCupid Tells it

Prof found me a new blog to love: Dating Research from OKCupid.

Ever since I was forced to take a stats class several years back and, oh yeah, started dating a professor, data analysis has started to intrigue me just a little bit. Particularly when it has to do with people, and their ridiculousness.

And OKCupid has me hooked. In my Internet dating days, I stayed far away from OKCupid. Its a free dating site, which is great, but the few times I visited, seemed to be filled with 300% more creepy people than match.com. Creepy and cheap. Who wants to go on an Internet date with a dude too cheap to pay to meet you? Not me. I wanted them to pay to meet me and pay to take me to dinner.

But, OKCupid has some serious data, and a blog. And it is super interesting. And you should check it out.

They recently compared the most common words in people's profiles by race and gender. Did I ever mention that sometimes stereotypes are true?

Black men most used the words "soul food" and "ESPN". Black women most used the words "soul food" and "the color purple." So, black couples can't watch TV together, but at least they agree on food.

Latino men like "merengue" and "bachata". Latino women most like, guess what (!), "merengue" and "bachata!" Not only do Latino men and women both enjoy dancing, they enjoy the same kind of dancing.

Now, to the reason that it took me 7 years to find a white guy who I could date more than six months...

White guys like Tom Clancy, Van Halen, golfing, Harley Davidson and Ghostbusters. In that order. White women like the Red Sox (what?), Jodi Picoult, Nascar (what?), boating and mascara.

Not only do white men and women like NONE of the same things, the things that white women pretend to like, like nascar and boating, aren't even the things that white guys like! What is wrong with us! What is wrong with white guys? Can't they like just one girly thing, or the Electric Slide?

I've added a link to OKCupid's blog. Be sure to scroll down and check it out if you are kind of nerdy and like dating data. Like me.

1 comment:

Anonymous said...

It's important to note that their dataset is not a representative sample of people (since most people probably have never even heard of this site, much less been a member of it). It's only a snapshot of data of people who self-select to become members of a website of this nature. Which may -- obviously -- have its own biases.

In other words, this is some sort of research, but not really research you could generalize much about. In academia, they'd call this sloppy research, because the data sample is biased. It means it's good for business -- and marketing the site -- but little else.

That's why you won't find these bloggers publishing any of these "data" in peer-reviewed scientific journals.