It's Just Sad When They Cut a Picture Out Of a Magazine And Put It In Their Wallets, Isn't It?

Ah,I'm SO glad to see this finally reported in the newspaper, since it has been bothering me all my life. Taking the population as a whole, men cannot possible have substantially more sex partners than women on average. It's not confusing or mysterious; it's just logically impossible. Note the rather feeble explanatory attempt:

Sevgi O. Aral, who is associate director for science in the division of sexually transmitted disease prevention at the Centers for Disease Control and Prevention, said there are several possible explanations and all are probably operating.

One is that men are going outside the population to find partners, to prostitutes, for example, who are not part of the survey, or are having sex when they travel to other countries.

Another, of course, is that men exaggerate the number of partners they have and women underestimate.

Dr. Aral said she cannot determine what the true number of sex partners is for men and women, but, she added, “I would say that men have more partners on average but the difference is not as big as it seems in the numbers we are looking at.”

Dr. Gale is still troubled. He said invoking women who are outside the survey population cannot begin to explain a difference of 75 percent in the number of partners, as occurred in the study saying men had seven partners and women four. Something like a prostitute effect, he said, “would be negligible.” The most likely explanation, by far, is that the numbers cannot be trusted.

Yeah, I think the explanation is that all those dudes have a girlfriend in Canada. Seriously. You've never met her, but she's really hot.

Comments

[Brock Landers]

Taking the population as a whole, men cannot possible have substantially more sex partners than women on average.

Heteronormist.

[Brock Landers]

Also, the article doesn't even consider the possibility of bestiality. Maybe more men are into it than women.

[Stu]

I love you and all, but this is false.

If there are seven men who only sleep with four girls, then that's how it is.

I think the fallacy in thinking about this comes from the idea that one guy can only sleep with one girl, and one girl can only sleep with that guy.

It is entirely possible for the number of men willing to sleep with random girls and the number of girls to sleep with random men to be widely divergent. In the end, there might be widely different numbers of what men report and what women report, but it doesn't make sense to discount these studies just because the men:women ratio doesn't add up.

If a woman sleeps with ten guys, they guys have each slept with only one woman, and the woman has slept with ten men. And vice versa. These things skew the numbers radically.

[belle waring]

Stu: this might be true for a given woman chosen at random, but on average it still can't work out. one of those women who sleeps with 50 men will be sampled and the average number of sex partners per women will go right back up. so either the sample size for a given study isn't large enough, in which case the results are bogus, or the men are lying to sound good, in which case the results are bogus, or the women are lying so they don't look like sluts, in which case the results are bogus. you can argue with me, but you can't argue with math, son.

[Stu]

I think it actually goes the other way. If fewer women are sleeping with more men, it'll skew the numbers. And vice versa.

And if that's your societal trend, that'll skew it even more.

What I'm saying is is that these things won't necessarily be 1:1. And that's the math. Seriously. I'm not being bitchy. The math will never necessarily balance out.

I do agree with you that we need a large enough sample set for these studies to be meaningful, but, not arguing from any _actual_ study, these things don't have to even out at all.

All it means is that, if the number of men who sleep with women is larger, it means they're sleeping with a fewer number of women. The number of partners on the individual level is not related directly the number of people actually having sex.

[Stiff Mittens]

"Yeah, I think the explanation is that all those dudes have a girlfriend in Canada."

Right, because no woman has ever had sexual partners that she is unwilling to admit to.

[belle waring]

I have no particular investment in whether it's the men or the women who are lying; it's Dr. Aral who invoked the canadaian-girlfriend effect.

[dsquared]

Stu: On average, across the whole population, Belle is right and it's 1:1; it does necessarily balance out[1]. Obviously the kind of patterns that you describe would make it more likely that any given sample would miss one of the very sexually active people and thus give a wrong answer for the mean, but at the whole population, it balances out.

Consider the reductio case; an island of 99 married men, 98 married women who only have sex with their husbands and one woman who has sex with all 99 of the married men.

If you do a sex survey on this island which misses the sexually active woman, you'll get the result that men have 2.0 sex partners and women have 1.0. However, if you do the full population survey, then you'll get the right figures (specifically, you'll get 99 women who have 1 sex partner and 1 woman who has 99, for an average of 1.99 sex partners per woman, and 98 men who have 2 sex partners and 1 man who has 1, for an average of 1.99 sex partners per man).

What the doctor is saying is that for plausible levels of variance and reasonable sampling errors, the "prostitutes effect" can't explain the observed discrepancy.

[1]I'm assuming that every sex act involves one of each sex - god, there's going to be some bloody site on the internet which undermines this assumption isn't there.

[dsquared]

(strictly I should add that the ratio has to be equal to the ratio of men to women in the population, but we're talking about normal societies here, not oil rigs or other oddball situations).

[David Moles]

Mean, median, or mode?

[David Moles]

A quick back-of-the-envelope example with seven women and seven men:

Ruth: Elvis, Ringo
Sarah: Elvis, Ringo
Mary: Elvis, Ringo
Hannah: Paul, George
Deborah: Paul, George
Naomi: Paul, George
Rachel: George, Ringo
Women's mean: 2
Women's median: 2
Women's mode: 2

Matthew: (no partners)
Mark: (no partners)
Luke: (no partners)
Elvis: Ruth, Sarah, Mary
Paul: Hannah, Deborah, Naomi
George: Hannah, Deborah, Naomi, Rachel
Ringo: Ruth, Sarah, Mary, Rachel
Men's mean: 2
Men's median: 3
Men's mode: 0

The article seems to confuse mean and median on purpose.

[woof]

On that island with the 99 married men? The one woman who had sex with all the married men? That's my ex-girlfriend!

[LizardBreath]

Stu would be right about the math if the population of men was substantially larger or smaller than the population of women. But if the populations are the same size (as they are, or at least close enough for government work), the math is as described in the article, and the averages have to work out to be the same.

[LizardBreath]

Sorry, the means have to be the same. Median and mode can, as David Moles said, be all over the place.

[ajay]

The obvious conclusion is that one sex or the other has really bad memories. Either women are sleeping with men and then forgetting they've done it, or men are sleeping with women and forgetting they've slept with the same woman before. The latter effect could be due to fatigue, or a radical new haircut or something.

[Chris]

The article talks about medians, and there's no need for the medians to be the same, but in studies like these over the years, the means are off too. That is, the mean number of sexual partners reported by men is almost always higher than the mean number reported by women. So you're stuck with the same conundrum (and the obvious explanation is over and under-reporting).

It's too bad that the reporter writing this article hasn't taken an intro stats course, though. The median thing is annoying.

[Stu]

Ugh.

Yes. I am wrong.

And stupid.

I blame the wine.

I swear, it made so much sense in my head at the time.

The other thing _could_ be that men report more things as sex than women do. That is, the exact opposite of the Clinton problem.

[The Modesto Kid]

men report more things as sex than women do

Sort of a subset of the "men exaggerate, wome undercount" thing, right?

[bitchphd]

Dear god, the prostitute explanation is so, so offensive. Prostitutes: not actual women!

[The Modesto Kid]

Right -- does "prostitutes, who were not part of the survey" mean prostitutes were intentionally excluded from the sample set? -- Or just that the putative prostitutes visited by these men did not happen to be in it? Because the former explanation seems like it could skew the results a lot.

[Badtux]

There is another possibility. If men have more sexual partners on average than women... what does that say about the gender of these sexual partners?

Hmm, one wonders...

[The Modesto Kid]

the gender of these sexual partners

Or, as Brock points out above, their species.

[Gary Farber]

If a sufficiently large number of transvestites, who self-identify as men while convincing a lot of men that they're women, exist, that could explain it.

I don't believe this to likely be the case, but I offer it in the spirit of open-mindedness.

Ockham's Razor would tend to suggest the obvious reporting problem probability: that more men like to over-claim and more women like to under-claim; still, it's an unproven hypothesis.

Yet another unerestimated possibility is that Dr. Doom is behind it.

[monkey.dave]

If a woman gets religion, she can declare secondary virginity and reset her partner count back to zero. That will depress the mean.

[zuzu]

As Jordan Ellenberg points out in Slate, the Times did mix up median and mean. However, the figures given for the British study (12.7 and 6.5, IIRC) were means, not medians.

So the conclusion is accurate, if the terminology is confused.

Also, someone mentioned that Dr. Gale looked at the raw data from the CDC study, not just at the median figure, and came to the same conclusion: this is a mathematical impossibility.