For anyone else curious about the published paper, it’s freely available. Annoyingly, it doesn’t seem to say the standard deviation of number of sex partners for women under each condition, only averages, so it’s hard to do an independent statistical check. The authors did do their own test:
Number of sexual partners. The two-way ANOVA on self-reports of the number of sexual partners yielded no significant effects, F < 1, but the data did strongly favor the predicted pattern (see Figure 2). That is, men reported more sexual partners than did women in the exposure threat condition (3.7 vs. 2.6, η² = .03), where gender expectations are most salient. The magnitude of the sex difference decreased in the anonymity condition (4.2 vs. 3.4, η² = .01), and the direction of the difference actually reversed in the bogus pipeline condition, with men reporting fewer partners than women (4.0 vs. 4.4, η² = .001).
The ‘bogus pipeline condition’ is one where the women were hooked up to a (not working) lie detector.
A much smaller one, which may well be within statistical bounds. Good point about people knowing that lie detectors don’t work. That may account for any remaining difference.
Apparently, there was no experiment with the men hooked up to a (bogus) lie detector, and I think there should have been.
Sure there was:
Men who thought they were attached to a polygraph reported an average of 4.0 sexual partners, compared to 3.7 partners for those who thought their answers might be seen.
Thanks—goes to show I should have read the linked material.
I’m surprised. I’d have expected that men would lie about having more partners rather than fewer, but that might be mere stereotyping on my part.
The other possibility I can think of is that people who think they’re hooked to a lie detector don’t just say what they immediately think is true, they check their memories more carefully.
It’s interesting to note that the difference goes away when the women are told they’re hooked up to a lie detector.
That’s a cute result!
For anyone else curious about the published paper, it’s freely available. Annoyingly, it doesn’t seem to say the standard deviation of number of sex partners for women under each condition, only averages, so it’s hard to do an independent statistical check. The authors did do their own test:
The ‘bogus pipeline condition’ is one where the women were hooked up to a (not working) lie detector.
However, there’s still a difference.
Apparently, there was no experiment with the men hooked up to a (bogus) lie detector, and I think there should have been.
I also have no idea what proportion of the population is cynical about lie detectors. I don’t even have a strong prior on that one.
A much smaller one, which may well be within statistical bounds. Good point about people knowing that lie detectors don’t work. That may account for any remaining difference.
Sure there was:
Thanks—goes to show I should have read the linked material.
I’m surprised. I’d have expected that men would lie about having more partners rather than fewer, but that might be mere stereotyping on my part.
The other possibility I can think of is that people who think they’re hooked to a lie detector don’t just say what they immediately think is true, they check their memories more carefully.
Or maybe the sample’s not representative of most men—the sample was of Midwestern psychology undergraduate students.