Research suggests that men have accurate perception of women's sexual interest, overconfidence is not adaptive

Overconfidence sounds like an inherently bad trait to have, but when it comes to natural selection, some evolutionary psychologists have suggested it could be advantageous in finding a mate. Males who overestimate their chances of having sex with a given female partner would fare better in passing on that trait, as the occasional success would outweigh the consequences of being wrong the rest of the time, or so the theory has traditionally gone.       

New research from the University of Pennsylvania and Texas State University is calling this theory into question, showing that men might be accurate when estimating their female partner's sexual interest.   

The research, published in the journal Psychological Science, was conducted by Carin Perilloux, a lecturer in Texas State's Department of Psychology, and Professor Robert Kurzban of the Department of Psychology in Penn's School of Arts & Sciences.

"Much of the literature in evolutionary psychology suggests that men overestimate women's interest in sex," Kurzban said. "This falls under a theory known as error management theory, which is a very broad starting point for the argument that it is a good idea to make systematic mistakes in social perception. If there's a really good potential outcome, say, a highly valuable mate, you want to make a mistake and think that's she really interested in you even when she's really not. Our study undermines the overconfidence argument as it pertains to sex."

While emphasizing that these results in no way can be understood as an excuse for bad behavior of any kind, Kurzban noted that the results also "undermine the overconfidence argument in general. We're saying you're just better off being right as much as possible."

Unable to directly observe this dynamic, the researchers conducted three studies that asked heterosexual men and women to report their perceptions of what various behaviors might indicate while on a date. The researchers drew these "signals" from a standardized survey known as the "dating behavior scale." These 15 behaviors included items like "holding hands," "kissing" and "inviting back to apartment for a drink."

In the first study, they asked 271 men to say what those behaviors would mean if a woman they were on a date with did them, rating them on a seven point scale ranging from "extremely unlikely" to "extremely likely" that it meant she was interested in sex. The researchers also asked 213 women to rate what they would mean by those actions when it came to indicating their own interest in sex.           

"When you combine all 15 responses into an average score," Perilloux said, "men's scores are higher than women's. Men are essentially saying that these actions are more indicative of sexual interest than women are self-reporting."

While these data seemed to confirm the idea that men over-perceive women's sexual interest, it opened the door for a second study that delved deeper into this discrepancy.

The second study recruited a different set of men and women, in roughly the same numbers, and presented them with the same 15 behaviors. This time, however, they asked both men and women to guess how the women in the first study responded to each question.   

"Here, we see that men's answers are basically the same in both studies. They're essentially saying, 'We think that our perceptions are pretty accurate," Perilloux said. "Women's scores, on the other hand, go up in study two, when asked what they think other women mean by the behaviors they start answering more like men. They are saying 'other women intend more by these behaviors than I do.'"

A third study, with the same general set-up and numbers of participants, attempted to get at the distinction between what people say and what they mean. As in study two, the researchers asked men and women to guess how the women in the first study responded to each of the 15 questions, but they also asked them to rate what they thought the women really meant.   

"For men, if you ask what women say, they are consistent with the responses with the other two studies," Perilloux said, "But, for both men and women, when you ask them to say what women mean, their answers go up in contrast to what they think they will say."

While there are inherent limitations to survey-based studies when it comes to human behavior, the method provides a counter-argument to the idea that men overestimate women's interest in sex.   

"If a woman touches your thigh on a date, we don't know what the actual likelihood is that she wants to have sex with you," Perilloux said. "All we can say is that women's responses move around, whereas men's are very stable. The most parsimonious explanation is that men are pretty accurate, since their responses more closely match what women say when asked to think about other women. Everyone is converging on the truth."

The researchers stressed that their findings have implications beyond sexual behavior; they speak to the overall role of accuracy when it comes to the evolution of social behavior.   

"We're adding credence to the idea that, by and large, people are generally pretty good at the estimations necessary for the social tasks they have to do," Kurzban said.