F**k Statistics

August 31, 2011 5:05 AM

Statistical analysis of OKCupid profiles exposes some sexually fascinating revelations:

- Herbivores like giving oral more than omnivores
- Twitter users are more likely to masturbate today
- Christians and Atheists are just as likely to claim they have never masturbated
- The correlation between men who prefer gentle sex & use of the word 'boating'

I f**king love statistics

[previous OKCupid magic]

posted by 0bvious (73 comments total)

Dear God! He leads with Napoleon's March!
posted by Nanukthedog at 5:09 AM on August 31, 2011

What the hell is the deal with the image behind the "Herbivores" link? Chew with your mouth closed please. ::shudder::
posted by jwhite1979 at 5:11 AM on August 31, 2011 [2 favorites]

use of the word 'boating'

This is one of those times when someone fucks with my childhood and a piece of literature. Before anyone tells me, I'm off with my copy of Wind In The Willows. Go to hell the lot of you.
posted by Fiasco da Gama at 5:13 AM on August 31, 2011 [4 favorites]

I am interested in these Folk Statistics.
posted by flapjax at midnite at 5:16 AM on August 31, 2011 [5 favorites]

self-reporting from a skewed sample is bizarro world science
posted by DU at 5:19 AM on August 31, 2011 [1 favorite]

use of the word 'boating'

He's only done it twice; the first time he got seasick, and the second time he lost his hat!
posted by GenjiandProust at 5:19 AM on August 31, 2011 [8 favorites]

Treehugger needs to stick to smokeless stoves and solar lighting

and what is boating?
posted by infini at 5:25 AM on August 31, 2011 [1 favorite]

If there's one thing that might turn Republican men into public-infrastructure-supporting socialists it's that GDP vs casual sex graph.
posted by DU at 5:28 AM on August 31, 2011 [5 favorites]

twitter users? masturbating? hm.
posted by sexyrobot at 5:34 AM on August 31, 2011 [6 favorites]

Well, there's a difference between what someone professes and what someone actually believe. It could be that people who self-identify a herbivores are less inhibited about revealing that than the larger population of omnivores.
posted by inturnaround at 5:40 AM on August 31, 2011 [3 favorites]

My favorite part was the comments.
posted by box at 5:41 AM on August 31, 2011

self-reporting from a skewed sample is bizarro world science

If you figure it's not meant to inform people about the world at large but rather about their pool of potential mates on OKCupid then it's not quite as horrible.
posted by jedicus at 5:50 AM on August 31, 2011

The correlation between men who prefer gentle sex & use of the word 'boating'

Did someone miss off the 'motor' prefix here?
posted by PeterMcDermott at 5:58 AM on August 31, 2011 [4 favorites]

If there's one thing that might turn Republican men into public-infrastructure-supporting socialists it's that GDP vs casual sex graph.

Real conservatives meet their soulmate via her Dad.

For casual sex.
posted by straight at 6:00 AM on August 31, 2011

Wasn't this posted when it was new?
posted by empath at 6:00 AM on August 31, 2011 [2 favorites]

So that's what Toad meant when he said: "There is nothing--absolutely nothing - half so much worth doing as simply messing about in boats."
posted by Flashman at 6:04 AM on August 31, 2011 [3 favorites]

Re Chart #2: Just lie back and think of "england."
posted by Naberius at 6:06 AM on August 31, 2011

Real conservatives meet their soulmate via her Dad.

For casual sex.

Wait, they have casual sex with women's fathers, and then meet the women for matrimonial purposes? That is a really uncomfortable family dynamic. It does, I suppose exhibit a sort of "family values," but...
posted by GenjiandProust at 6:11 AM on August 31, 2011 [3 favorites]

Vegetarian-Friendly Sex Slang
----------------------------------------
Peeling the banana.
Tossing the salad.
Squeezing the melons.
Zeroing in on a grown man's nuts and nutsack.
Putting Monsanto in yoursanto.
Ordering the split pea soup.
Sorry, that's got ham.

Cornholing others.
posted by Obscure Reference at 6:20 AM on August 31, 2011 [5 favorites]

I don't get it.
posted by seanmpuckett at 6:25 AM on August 31, 2011

Reducing the "map that made a nation cry" to a pie chart. It's like they're trying to make the Baby Tufte cry.
posted by en forme de poire at 6:29 AM on August 31, 2011 [6 favorites]

I f**king love statistics

I do not care for the FTFY meme, but this should be: "I love fucking statistics." Both because these are statistics about fucking, and because Metafilter is not the minutes of The Ladies' Temperance League, so you can actually use all the letters here.
posted by ricochet biscuit at 6:40 AM on August 31, 2011 [27 favorites]

Metafilter is not the minutes of The Ladies' Temperance League, so you can actually use all the letters here.

9:40 PM: Member-in-standing ricochet biscuit has blown our cover. Ladies, to the battlements. We must accelerate our plans tenfold and lay the siege tonight.
posted by griphus at 6:42 AM on August 31, 2011 [13 favorites]

...but this should be: "I love fucking statistics."

I dunno. It can be fun and all, but the morning after, things with statistics can be kind of... awkward. Often you'll find that you have nothing, really, to say to the statistics, and it can be pretty uncomfortable.
posted by flapjax at midnite at 6:44 AM on August 31, 2011 [9 favorites]

The first bullet of this post reminded me of the existence of the song "New Age Girl" by Dead Eye Dick, and, well... I don't thank you for that.

Also, if Metafilter is a cover for a temperance league, my experiences with Chicago meetups mean that, locally at least, we are failing greatly.
posted by MCMikeNamara at 6:45 AM on August 31, 2011 [1 favorite]

this should be: "I love fucking statistics."

The grammar aside, as that tends to be what OKC does to "statistics" in these sorts of posts, I have to concur.
posted by valkyryn at 6:53 AM on August 31, 2011 [2 favorites]

The first bullet of this post reminded me of the existence of the song "New Age Girl" by Dead Eye Dick, and, well... I don't thank you for that.

Well, you just reminded me of the existence of Dumb and Dumber, and I'm now contemplating having a lobotomy.
posted by dortmunder at 6:57 AM on August 31, 2011

because Metafilter is not the minutes of The Ladies' Temperance League

*spits lemon water all over screen* WHAAA?
posted by codacorolla at 6:57 AM on August 31, 2011 [2 favorites]

Peeling the banana.

Maybe it's just me but I really don't like the image of 'peeling' associated with my penis.
posted by shakespeherian at 7:02 AM on August 31, 2011 [3 favorites]

- Twitter users are more likely to masturbate today

louche_mustachio @0bvious - No, they are just more likely to tell you about it. #TMIorselfreporting
posted by louche mustachio at 7:05 AM on August 31, 2011

Good news for the squeamish: "Boating" apparently actually means riding in a boat. And people who like that also like gentle sex. Motion of the ocean I guess. Any port in a storm. Okay, now I'm ruining it.
posted by Reverend John at 7:08 AM on August 31, 2011 [1 favorite]

9:40 PM: Member-in-standing ricochet biscuit has blown our cover. Ladies, to the battlements. We must accelerate our plans tenfold and lay the siege tonight.
posted by griphus at 8:42 AM on August 31

9:09 AM: Member "griphus" is no lady. Beware of agents provocateur - they will not be so obvious next time. Stand by for further instructions.
posted by louche mustachio at 7:11 AM on August 31, 2011

Motion of the ocean I guess. Any port in a storm. Okay, now I'm ruining it.

If by "it" you mean my breakfast, then yes.
posted by louche mustachio at 7:12 AM on August 31, 2011 [1 favorite]

I'm so glad my guy kayaks. Yeah, that's right KAYAKS.
posted by angrycat at 7:21 AM on August 31, 2011 [4 favorites]

Member "griphus" is no lady.

How DARE you! (Purse-wallop.)
posted by griphus at 7:24 AM on August 31, 2011 [2 favorites]

You may say you love statistics tonight, obvious, but tomorrow morning you'll be quietly gathering your things and leaving out the back door. That's just as well, because statistics doesn't want to fake its way through breakfast, either. Statistics is just as happy to see the back of you, and will be quite content to forget the crazy, drunken fling you two had. And he'll be able to, at least until his next visit to the doctor.
posted by koeselitz at 7:31 AM on August 31, 2011 [5 favorites]

- Christians and Atheists are just as likely to claim they have never masturbated

This one is ok if you also affix "claim" to being Christian or Atheist. These are all just "People who advertise themselves as X also advertise themselves as Y." Not unininteresting, but not evidence of anything about any other behaviour of these people.

"People who are Durn Bronzefist always wear a tie!" No, that would be the interview.
posted by Durn Bronzefist at 7:33 AM on August 31, 2011

Member-in-standing ricochet biscuit

Is this another preference by herbivores?
posted by infini at 7:36 AM on August 31, 2011

According to that graph, if you are a male vegetarian your odds are about 9:1 that you enjoy giving oral sex, and 7:1 if you are an omnivore. If you are a female vegetarian 3.5:1 (7:2), while a female omnivore less than 2:1.

Okay (Cupid), doesn't that mean that they have the conclusion totally bassackwards? 7:1 and 2:1 means *more* likely to enjoy oral sex than 9:1 and 7:2, respectively.
posted by mcstayinskool at 7:39 AM on August 31, 2011 [1 favorite]

The data is all skewed since people are answering those questions and filling out their profiles with the hope of getting laid and/or finding love. I think it's ridiculous that OK Cupid thinks it reflects on society beyond people on that dating site.
But I do enjoy reading the blog.
posted by aabbbiee at 7:51 AM on August 31, 2011

mcstayinskool: The post notes that the odds shown are "in favor." So there's a roughly 90% chance that a vegetarian male would enjoy giving oral, while there's a roughly 70% chance that a non-vegetarian would enjoy it. This is consistent with the headline. It's confusing because usually odds are presented "against," i.e., the way you interpreted them.
posted by brain_drain at 7:52 AM on August 31, 2011

okay fair enough, however now I'm completely shocked by those numbers. I would never have guessed them a) that high, or b) so skewed in favor of men giving oral over women giving oral. Q: does that match up with what others would expect regarding those stats? Maybe this is because an OK Cupid population is not representative of a general population?
posted by mcstayinskool at 8:06 AM on August 31, 2011

My sense is that cunnilingus is easier for the giver than fellatio. So, men enjoying giving oral sex statistically outnumbering women enjoying giving oral sex makes sense to me.
posted by mistersquid at 8:11 AM on August 31, 2011

mcstayinskool, what's confusing you is that these represent the % of men who think saying they like oral will get them a desirable date vs. the % of women who think saying they like oral will get them a desirable date.
posted by straight at 8:12 AM on August 31, 2011 [1 favorite]

(make that the % of men and women on OK Cupid, of course)
posted by straight at 8:13 AM on August 31, 2011

Lot of frustrated rough-sex lovin' old dudes on Ok Cupid.
posted by nanojath at 8:14 AM on August 31, 2011

Statisticians do it with 95% confidence.
Statisticians do it with large numbers.
Statisticians do it. After all, it's only normal.
Statisticians probably do it.
Statisticians do all the standard deviations.
Statisticians would like to do it with the population, but only get a small sample.
Statisticians do it continuously but discretely.
posted by kmz at 8:15 AM on August 31, 2011 [13 favorites]

I have to disagree with them in the idea that "skinny" is the most positive term for not being overweight. "Thin" is a much more positive word, and skinny has more negative associations. I've known underweight people who have taken "skinny" as a somewhat of an insult, or at least it hurts their self-confidence when they're called it. To me "skinny" means skin and bones, while "thin" means healthy and fit.
posted by gregoryg at 8:23 AM on August 31, 2011

FPP from last month, if you like this sort of thing. Have yet to get a boating/sex question.
posted by umberto at 8:25 AM on August 31, 2011

"This is one of those times when someone fucks with my childhood and a piece of literature."

You should have kept reading to see what it said about the intersection of people who love to fuck with literature and with children.

"self-reporting from a skewed sample is bizarro world science"

I know you're required to comment in every MeFi thread, but statistics != science, Mr. Grinch.
posted by Eideteker at 8:26 AM on August 31, 2011

these represent the % of men who think saying they like oral will get them a desirable date vs. the % of women who think saying they like oral will get them a desirable date.

well put. my conclusion from this thread discussion is that the most salient point of the herbivores vs. ominvore graph is that it's another example of the reported finding of the statistics being very far off what the statistics actually represent.

don't get me wrong, I'm a huge fan of statistical analysis. Just not when the analysis has poor rigor (or perhaps for this study, turgor) as to its conclusions.
posted by mcstayinskool at 8:28 AM on August 31, 2011

That boating thing is really, really not true at all.
posted by By The Grace of God at 8:34 AM on August 31, 2011

Those figures are further affected by the % of men who think saying they like oral will get them too much unwanted attention vs. the % of women who think saying they like oral will get them too much unwanted attention.
posted by straight at 8:46 AM on August 31, 2011 [1 favorite]

Herbivores are generally more empathetic.
posted by Human Flesh at 8:57 AM on August 31, 2011

I'm really confused by their 'odds' charts, as well. Are their graphs wrong, or are their conclusions wrong? One or the other has things flipped.
posted by Gordafarin at 8:57 AM on August 31, 2011

It's weird to see people deriding the statistical power and the integrity of these results. I wonder, have they ever worked in a social psych lab, or seen how psychologists measure things? Have they taken a methods course? I have. Frankly, most of it is a fucking joke and the participant-self-reporting and "instruments" (questions, dozens of them) involved are as crude and flawed, if not worse. And the sample sizes are orders upon orders of magnitude smaller, and consist primarily of undergrads.

If you've got psychological "science" on a pedestal, get it off there and go see what the methods are like. Learn about IAPS, take a basic statistics course, and read some papers. You'll see how far removed the claims are from what was actually performed, how small the sample sizes are, and how subject-to-fashion much of it is.

Don't get me wrong, of course, some people are doing good science in the social psychology world, but it is not the work that most people hear about through popular media.

On the whole, this OKCupid stuff is pretty fantastic as a set of early experiments in learning from dating data, and I hope that the "equivalent" academic research eventually gets its act together and follows suit. Mine your data. Find interesting correlations and test hypotheses based on self-report AND actual behavioral data. Report immediately in a public, open way what you've found and have a forum for criticism right below the results.

What's not to like?
posted by fake at 9:02 AM on August 31, 2011 [5 favorites]

"Frequent tweeters have shorter real-life relationships than everyone else, probably via some bit.ly hack."
posted by danb at 9:16 AM on August 31, 2011 [1 favorite]

have they ever worked in a social psych lab, or seen how psychologists measure things? Have they taken a methods course?

Yes, yes, and yes. There's no pedestal. Good design accounts for bias. Bad design ignores it and bad reporting ignores the flaws in favour of trumpeting conclusions.

What OKC has going for it is an absolutely huge subject group. A group dedicated to a particular goal. One doesn't remedy the other.
posted by Durn Bronzefist at 9:17 AM on August 31, 2011 [1 favorite]

The correlation between men who prefer gentle sex & use of the word 'boating'

Inconclusive unless we can correlate other maritime references to other sex styles. What about:

'Yacht' & preference for a legal but much younger companion

'No Wake Zone' and Somnophilia

'Canoeing' and preference for one to be in front of/behind the other

'Piracy' and amusement by way of the partner's infidelity

and the low-hanging fruit, the easiest to correlate of them all:

'Water Sports' and well, water sports
posted by chambers at 9:34 AM on August 31, 2011 [1 favorite]

men enjoying giving oral sex statistically outnumbering women enjoying giving oral sex makes sense to me.

Me too. Totally jives with my experience.
posted by desjardins at 9:59 AM on August 31, 2011

Kayaking and wet entry?

sorry
posted by desjardins at 10:00 AM on August 31, 2011 [1 favorite]

OK I am feeling really stupid right now but...what is boating?
posted by naoko at 10:47 AM on August 31, 2011

Me too. Totally jives with my experience.

Jibes. Ye gods.
posted by ricochet biscuit at 10:53 AM on August 31, 2011 [1 favorite]

I am now envisioning "retro dancing in boats" as a new euphemism for -- well, something naughty.
posted by maudlin at 10:58 AM on August 31, 2011 [1 favorite]

Jibes. Ye gods.

oh deer.
posted by desjardins at 11:15 AM on August 31, 2011 [1 favorite]

OK I am feeling really stupid right now but...what is boating?

Going around in a boat. Paddling, sailing, motoring, fishing from a boat… It's an unexpected correlation, not a euphemism. Unless you add motor to the front.
posted by JiBB at 11:28 AM on August 31, 2011 [1 favorite]

People aren't herbivores.
posted by John Cohen at 11:42 AM on August 31, 2011

I love these OKC postes, but every time, MeFites line up to see who can yell "self-selected sample" and "correlation != causation." Yes, we get it, this is not "real" statistics. It's sexy fluff, voyeuristic data mining, titillating factoids.

C'mon, spoilsports, the article even addresses this right in the eighth sentence:

Other than the unifying theme, sex, there's no big point or thesis to this post: just comparisons, correlations, and quirky trends.

So OKC knows this isn't of deep, meaningful sociological importance. You know this isn't of deep, meaningful sociological importance. We all know this isn't of deep, meaningful sociological importance.

Now let's all have fun and giggle about "boating."
posted by lekvar at 12:09 PM on August 31, 2011 [3 favorites]

love these OKC postes, but every time, MeFites line up to see who can yell "self-selected sample" and "correlation != causation."

Ugh. This reminds me of people who make unfunny jokes and then complain WHATSA MATTER, YOU DON'T LIKE HUMOUR? You might want to pay closer attention to what people are saying in this thread, because no one is arguing self-selection (this would be a factor, but is nowhere close to the biggest problem here) nor correlation /= causation (which would make no sense whatsoever -- the whole thing is presented as entertaining correlation).

Take the first conclusion, as fun or fluffy as it may be intended:

We found this by crossing the match questions Do you like to exercise? and Is it difficult for you to have an orgasm?, and, as you can see, women who don't like working out report twice the orgasm problems of women who do.

Well, at least they use the word "report", but no, I don't see that. I see "Women who report that they don't like working out report twice the orgasm problems of women who report that they do like working out when advertising themselves in a dating service." This would be an important (but generally unstated) distinction in any confidential survey, where there can be a tendency to show some concern for the opinion of an unknown researcher, a person who is and shall remain a stranger. Here, we're talking about self-reporting behaviours, experiences, and attitudes in the context of advertising yourself to people you ideally want to meet, date, fuck, whatever. That's a whole other kettle of fish. It can still be fun and interesting. I like pop psychology, myself. All anyone is saying in this thread is let's be clear about what this data represents and what it does not.
posted by Durn Bronzefist at 12:43 PM on August 31, 2011

Actually, you could frame this problem as "self-selected" sample, but it goes way beyond the usual "people who like to answer surveys" or otherwise "people in a population subset who are disposed to answer your questions." They're not answering your questions for your purposes. They're answering them for theirs. Would someone lie about whether they like to exercise? Possibly. On a dating site? Heh.
posted by Durn Bronzefist at 1:00 PM on August 31, 2011

Going around in a boat. Paddling, sailing, motoring, fishing from a boat… It's an unexpected correlation, not a euphemism. Unless you add motor to the front.

Oh. Well I already knew "motorboating." Feeling less old now, thanks!
posted by naoko at 7:01 PM on August 31, 2011

Oh god. The Sarah Lawrence College tagline.
posted by Clandestine Outlawry at 7:22 PM on August 31, 2011

Twitter users are more likely to masturbate - today?

The hell? How can the even know that?
posted by Smedleyman at 3:23 AM on September 1, 2011