Holy sampling issues, Batman!

Women’s preferences (for the subset of women who use OK Cupid) for the incidence of sex goes up as they age.

They showed a graph previously that showed that as age increases, their population on OK Cupid decreases. This makes sense for several reasons - older women may be unfamiliar with or have negative feelings towards online dating, and there might simply be fewer single older women who are actively looking for a partner.

I speculate that the subset of women who use OK Cupid from ages 20-25 correlate pretty well with the general singles population, as internet dating is well-known and well-accepted among women in these ages. As we look at the older women using OK Cupid, I think they will trend more metropolitan and more generally interested in dating and in sex than the general population. Essentially, we're looking at a self-selected sample.

Cool stuff. Can anyone identify distinct regions here? The Gulf coast seems intuitive, but others less so. The Ozarks/Smokey Mountains appear to be showing up pretty starkly. Any ideas about the Wisc/Minn/Iowa blob? It would also be cool to see how this correlates with larger population growth/employment trends.

They should want more sex. Their Testosteronlevel is rising. And Testosteron is making humans and animals horny.
The Biologial reason most likely is, that this is Natures Way to tell them "Not much eggs left. Get what you can get."

Are users being asked this question in anonymous surveys, or are the replies visible to other users?

If the latter is true, it seems like a significant number of responses would be influenced more by strategy and societal pressures than actual sexual desires.

Also, when it comes to data collected exlcusively online, the older the survey sample is (or the less statistically likely to be using the website), the less representative of the entire demographic it becomes. I'd be a bit suspicious of the value of OKCupid graphs as data, especially ones like these, which are visually skewed by the huge variance in population density from county-to-county. Malheur County, Oregon and Fremont County, Wyoming look like huge hives of horny young women, but there's only about 35,000 people in each, with a population of twentysomething female OKCupid users you can almost count on one hand.

Yeah, don't women reach their sexual peak as they hit their 30s or older?

I know this is the wrong post but you turned off comments and I just have to psst to you that... the song "America" is not a back and forth between white men and Puerto Rican immigrant women. Those dudes are the Sharks! They are also Puerto Rican in the film. The white people are on the other side of the camera.

the english major in me can't help noticing the typo on the graphs.

It's nice to see that America's wild cougar population is rebounding.

Oh, man...I live in Minnesota...I'm screwed.

Apparently I am one of the horniest twenty-somethings in Wisconsin.

I'd like to see a corresponding graph for men. That is, if we could get them to tell the truth. I believe that there is a cultural tendency to exaggerate the male sex drive.

Once a day seems to show up in lots of rural areas. Which makes me think its bad question writing: Once a day is an imagination of 'lots of sex', which I suspect people on online dating sites in rural areas are not getting, though lack of available partners. The situation seems a little more dire in a community of 500.

What happened to the over 40 crowd?

Is there something in the water in eastern Oregon?

I cannot trust this or any anonymous survey. The marketing office I used to work in once did an anonymous survey on sexual habits: since it was anonymous, the boss decided to save the money we would normally have paid the respondents for their time. The entire survey was completed by our employees: three or four teenagers and two or three retired ladies in their sixties or seventies.

While people are right to point out that this data is not representative of the population as a whole, the OK Cupid blog doesn't represent it as such, and the post is addressed to male OKC users (with big presumptions of heterosexuality all around). If I'm a single straight man on OKC, I don't really care how women of a given age in the general population feel about sex. I care about how women of a given age who are OKC users feel about sex, since those are the women I could pursue for a date. So in that sense the various charts from OKC are potentially useful as a way of illustrating the need to match data to the purpose you're using it for.

I'm frustrated by the legend on these maps. We have no idea what "higher concentration of 'once a day'" means. Is that 30% want it once a day? 5%? 100%? For number-crunchers, the OKC bloggers seem awfully averse to showing us any numbers to contextualize their findings on relative strengths of phenomena.

40 is the upper age limit? Oh, come on!

"But, combing through the data, I intuitively felt like this graph didn't tell the whole story. So I dug deeper, and found something interesting. If you separate out the absolute best-looking women, almost all of whom are very young, and also remove the people you won't realistically want to date (the worst-looking women), you find that everyone else's attractiveness doesn't change much with age"

Perhaps I missed something, but how did she arrive at a statistical measure of who the most attractive women are?

I'm familiar with OK Cupid, but didn't realize they were using those 0-5 star ratings to gather data on user behaviors. When I was a user, I never took those star ratings literally, but only used them if I wanted to raise the chances of a quickmatch. I suspect other members use this feature in the same way, and I wonder about its utility in statistical analysis.

I'd be more interested to see any results from the 0-5 star personality ratings.

Interesting theory, all 20 somethings on the internets are liars lol

I am a member of OK cupid, I took that test. I'm 23 and male. I didn't think about it but I immediately answered once a week. Although, I know it would be everyday. Why? To not seem like a creeper or inadequate male or something stupid that would be read too deeply into. Which is the perception you get if you even think of mentioning online dating in the first place. Yet, I seen my dad using an online dating service and try to hide it like a 13 year old hides porn. He's not computer literate.

Why would women lie? Easy to not seem like whores. A hard stereotype to escape. God forbid they ever admit to liking sex while not in prision... I mean marrage. As much as I would like to believe the world runs on cougar-power. I just think cougars are honest. Only because they aren't used to properly witholding info when asked about themselves at 40. Go figure.

Back to the computer literacy. I believe that's the difference. We never, ever, ever! Be honest, on purpose, online. Especally, on a dating site that compares you to thousands of others based on cunning wordplay. We all begin with the end result in mind and alter the data accordingly. That map would run all blue. Maybe, we have all had too much standarized testing.

Also notice, that the ones who are 20 and answered honestly are in the middle of nowhere. I barely know the two states that scored in the blue... Oregon and Washington, right? Maybe they are excellent communities, maybe they are computer illiterate idk.

He're an example if your not convinced.

OKcupid: Do you want kids?
Me: nnnnnooooooooooooooooooooo waaayyyy... wait....... check maybe, I don't want limit options (it didn't ask in how many decades) its not a lie, its a bad question

This is not as on point as your usual post. Watch out data is a tricky mess. I don't think I can forgive it after it helped Karl Rove get Bush elected.