« Back Up and Ask Whether, Not Why | Main | Recognizing Intelligence »

November 07, 2008


Because cheating in women is regarded as worse than cheating in men? That shouldn't be hard to measure...

I don't know that men are actually more honest than women in cheating surveys. It's not that I don't trust the researchers' "complex statistical analysis of the data", but hinting something isn't even close to proving it, is it?

If it is true, I can think of many reasons. One has to do with social pressure, another with evolution, another with the researchers not being careful in the way they conducted the interviews, another with women being evil creatures... but I have no reason to choose one of them, although I think the last one is the least credible.

Men that can cheat are have higher status than those that can't.

Surely this is an extension of mating strategies? I.e. opportunities to reproduce for females (1 egg, long gestation) compared to males (billions of sperm).

A female needs to maintain the appearance of fidelity so that a prospective mate would be more likely to protect her and her offspring, assuming them to be his own.

For a male, admission of cheating suggests that he is desirable and therefore worth considering as a mate. For other males, it's useful to associate with someone powerful in order to maximise their chances of reproduction.

Of course this doesn't say why females cheat, but perhaps two donors are more likely to produce a viable sperm than one.

As math requires, men and women cheat in equal numbers:

How does the math require this? I wouldn't be surprised if it's already been explained on this blog, but, if so, I missed it. I don't even see how the math requires that men and women reproduce in equal numbers. I can see that they should have the same number of expected offspring, but that's different, isn't it? This article seems to claim that 87% percent of women reproduce, while only 81% of men do, but I haven't looked at it closely.

How does math require that men and women cheat in equal numbers? Fewer women could cheat, but those who do could cheat with more men.

Whoops. Tyrrell's comment wasn't there when I posted.

Your math also assumes that cheating is always heterosexual.

A man who cheats is not necessarily cheating with a female cheater, so it is not even necessarily true that, if more men cheat, women who cheat cheat more. In fact, a man who cheats is not necessarily cheating with a female, period. The study talks about heterosexual couples, not heterosexual individuals...

So I guess Robin's math is more complicated than all that...

A married man is also cheating when he has sex with an unmarried woman, but the woman is not (and vice versa).

It should be entirely possibly that unmarried women are more willing than unmarried men to have sex with a married person.

So why are men more honest than women in cheating surveys?
It goes back to the archaic mindset that women control what goes into them, ergo promiscuous women are whores but promiscuous men are not.

Daniel Reeves has merely set the explanation back a step. Does anyone have a theory on what might underly that mindset?

Math does not require the incidence of cheating to be equal across genders.

It is a well known curiosity that women report fewer total number of opposite-sex partners than men, though math requires the means to be equal (for a 50/50 population). Perhaps you are confusing cheating with this effect.

For math, I did have in mind average hetero cheat partners with also-coupled others.

Daniel, there is also the matter of skill/value.

To have sex, a woman needs to show up and say yes (unless she is morbidly obese or something). Revealing she cheated == revealing she is dishonest (bad).

For a man to have sex, he needs to demonstrate some degree of skill (game) and intrinsic value. Revealing he cheated == revealing he is dishonest (bad) but also of high enough quality to get laid (good).


"So why are men more honest than women in cheating surveys?"

Because boy monkeys, when they catch cheating girl monkeys, beat them with tree branches to within an inch of their lives - literally.


"A female needs to maintain the appearance of fidelity so that a prospective mate would be more likely to protect her and her offspring"

It's true that this is a common theory, but isn't it in stark conflict with evidence about how monkey societies work? Primates are largely homosocial and live with their extended families. Girl monkeys raise their children primarily in their female network, with the help of grandmothers, aunts, and unmarried younger relatives or associates ("teenaged baby sitters"). It takes a village.

Boy monkeys really don't hang around the hearth too much - they have to go to "work" doing things that create and maintain their place in the male status network. So why would a monkey girl work to appear to be faithful to a monkey boy who really won't be there, who she doesn't expect to be there, esp. when she has all these great more-reliable relatives to help her?

"Of course this doesn't say why females cheat"

Perhaps because we can unconsciously but literally "smell" who has the best genes, the most awesome testerone, and who most wants to knock us up?

I suggest we seriously consider that in fact it's the boy monkeys who are primarily interested in female fidelity and create all kinds of incentives and social structures to get it - including gifts, manipulation, bribery, and violence.

I will point it is a fairly well-known result that men are forgiven for cheating far more often than women are.

From a (sigh) evolutionary psychology perspective, this makes sense. If you are a man who cheated and your mate finds out, the loss has really just been that you might have fathered a child out there somewhere. The tangible harm is somewhat restricted, so it makes sense that the woman won't necessarily dump his lame bum.

On the other hand, if a woman cheats, the guy now has to worry about whether the children he has been taking care of are really his own. He is potentially using all his resources on another guy's genetic material.

Predictably, today even with genetic testing, a man is far more likely to leave a woman when he finds out she has cheated than a woman is to leave a cheating man.

Given this asymmetry, it makes sense that men would be less hesitant to tell the truth.

Robin, I can't see why math requires men and women to cheat in equal numbers. Here's a possible scenario: men want to cheat with younger women who accordingly have a smaller chance of being married, while women don't seem to have age standards for cheating.

Frelkins, you deny that single mothers have it harder than married ones, on average?

Cheating (rather, being caught cheating) has a much higher cost for women. Woman cheats -> man provides less or nothing for the kiddies. Man cheats -> woman... what, accepts less help with the kiddies? There's no credible threat there!

"Complex statistical analysis of the data hinted that a further 10 per cent of the women in the study had cheated on top of the 18.5 per cent who admitted to it in the questionnaires, whereas the men had been honest about their philandering."

Complex statistical analysis of this sentence hints that there's a 93% chance it's a load of crap.

For math, I did have in mind average hetero cheat partners with also-coupled others.

I'm sorry, but I don't follow.

Hypothetical situation: many men wish to cheat, few women wish to cheat, but each cheating woman has affairs with many cheating men while each cheating man has affairs with only one woman. We can reverse the relative position of the genders in that scenario if we wish.

The point is that math does not require that equal numbers of men and women cheat, even if we limit discussion to heterosexual relationships where both partners are otherwise married.

because maternity is easier to determine than paternity.

As it's been pointed out there's no math rule here. Married men can and do sleep with unmarried women, which are sometimes unsuspicious. However women will report less partners, and that's inconsistent with math, even when allowing some homosexuality. Again as it's been pointed out, promiscuity in women is not balanced by an accomplishment.

I do believe however it's not dishonesty people are ashamed to point out but infidelity or promiscuity. And it's a damn shame. From what I've seen one of the following is true

a) women are generally incredibly naive
b) women generally prefer adulterers to be dishonest

Weird ethical preferences if you ask me.

I guess I should elaborate.

Men need to be both jealous and perceptive about there partner's cheating in order to pass on their jeans. A male who spends his life raising another mans children is waisting a lot of his resources on something that doesn't help his inclusive fitness. Women only need their partner to not change his loyalty to another partner during the early years where she is unable to gather as much food and requires more nourishment for nursing. Clearly there is a larger evolutionary benefit to male suspicousness. Women have a higher cost to suspiciousness as well as they could always end up raising the kid on their own so they will tend to overshoot less and have fewer false positives.

Women cheat because it is difficult to determine paternity and they want good jeans for their child or perhaps a higher probability of getting pregnant. Men cheat because it is difficult to determine paternity and they want to pass on their jeans. Men have the larger evolutionary incentive to cheat so they cheat more. Cost of cheating is again higher for women who have to carry and nurse the baby.

"promiscuity in women is not balanced by an accomplishment."

I offer you one of the world's most successful women, Jenna Jameson. This could lead into an interesting discussion of why many - perhaps even most - human societies have had various arrangements for sacred whores.

Better genes are my significant reward for sluttishness. I've gotta raise this kid for 13-15 years+ - I'm breast-feeding probably for 2 or 3 years, so that's gonna suppress my fertility - when I have a kid every 4th year, half of them are gonna die - I seriously want the best genes I can get myself on. Child survival is the greatest "accomplishment." But dammit, those boy monkeys with sticks make it hard to arrange.


Jenna Jameson is not celebrated for her promiscuity but for her talent as a porn actress, which is an accomplishment. Any woman can be promiscuous, not any woman can be a porn star.

i love that someone commenting on evolutionary biology doesn't know that it's genes, not jeans (error made twice, so unlikely to have been inadvertent misuse).

Completely inadvertant. That is freaking embarassing. Is there any way I can get that changed?

There are actually a huge amount of typos in that post ("jeans" being the most embarassing well ahead of "waisting".) I honestly have no idea how I did that (three times by my count).

Sure am glad you "love" it though, superflat, you card, you.

Aside from any other reasons for why the numbers of men and women cheating don't have to be equal, there's prostitution.

The math doesn't say that the same number of heterosexual men and women have to cheat. I think it only says that they have to have the same average number of partners.

"The math doesn't say that the same number of heterosexual men and women have to cheat. I think it only says that they have to have the same average number of partners."

Bit of a tautology, isn't it? And if that's the case, "men and women cheat in equal numbers" is highly misleading. It's pretty obvious that, if you narrow the sample to two heterosexual and married partners, there will be, on average, the same number of partners for the average man and woman. I think the confusion in the comments stems from the implication that the same number of unique women in a sample cheat as unique men. That is, the wording makes it seem that in a universe of 100 married couples, 30 different men will cheat with 30 different women. [Avg of 1.3 partners]. However, if the actual claim is simply that the average number of partners will be the same, then it could easily be that 30 different men cheat with 1 married [and busy] woman. The average is the same at 1.3, but the number of unique individuals who did the cheating is unbalanced between the genders. Robin, perhaps it's time to clarify exactly what you meant by the intro so we can move to the more substantive issues.

Also, let's get over the instinctive distrust of "complex statistical analysis." If you have an objection to the methodology, find out what it is and then critique.

Daniel Reeves has merely set the explanation back a step. Does anyone have a theory on what might underly that mindset?
I underscored a possibility in my wording: people control that which goes into their bodies.

Also, the fact that women are physically weaker is another possibility. Physical strength may facilitate societal views of men as the domineering ones and women as the submissive ones. Therefore, a man who has lots of sex is expressing his dominance; a woman who has lots of sex is expressing her lack of will to abstain.

Another problem is that men and women sometimes have different ideas about what constitutes "cheating" and whether or not they are in a monogamous relationship.

Robin, it is not sufficient to acknowledge such an important assumption in the comments. You should add something to the post to make it clear.

What analysis allows them to infer that the women were untruthful? Not that I would find it hard to believe.

I couldn't access the article so perhaps reading it would change my thoughts on the matter.

Robin's "as math requires..." comment is quite troubling to me.

Up until now everything I've seen from Robin has been impressive, and he has appeared to be amazingly rational and intelligent. How could such a person make the assumption that people only cheat with others who are also in committed relationships? (Are there similarly unjustified assumptions plagueing Robin's other writings? Am I missing these errors along with him when I accept his arguments?). And even with that weird assumption, did he really not see that all the male cheaters could be cheating with the same woman (the article talks about % of people who cheat, not # if cheating incidents)? Or is there some miscommunication?

When evaluating Alcor vs. The Cryonics Institute I was giving a lot of weight to Robin's selection of Alcor, assuming that he did a lot of investigation and used the good judgment that I thought I saw previously in him. Now I fear his choice was based on various flawed assumptions and reasoning errors, and that my brain would have been accidentally destroyed had I signed up with Alcor, while all the brains at CI would live happily ever after..

Robin, will you please clarify this? Or give some account explaining your post (you were drunk, in a hurry, etc)?

There's no credible threat there!

That ought to be really important, but it's not clear to me what it should cause. eg, it's not obvious that it should cause people to care more about the cases where they have a credible threat.

"For math, I did have in mind average hetero cheat partners with also-coupled others. "

Why would you assume "with also-coupled others"? That seems absurd.

Would it help if someone said "ceteris paribus?" Damn.

I second Bob V!


No, it wouldn't. "Ceteris paribus" is used to isolate a causal relationship, not to enable people to make false assumptions.

Robin's thinking Cousin, Cousine here.

Clearly my attempt to introduce this news piece via three word reference to equal-cheating math theorems failed, as it distracted from the main point. Yes the actual data mentioned in this news piece is not of the right form to directly test equal-cheating theorems.

Robin, what are these "equal-cheating theorems" of which you speak :) ? Googling that phrase only brings up your comment here.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30