Truth and consequences: using the bogus pipeline to examine sex differences in self-reported sexuality

Journal of Sex Research, Feb, 2003 by Michele G. Alexander, Terri D. Fisher

Women's reports of sexual experiences fluctuated more than did men's as a function of testing condition. This is not altogether surprising, given the different expectations for the sexes regarding sexual behavior, with more constraints placed on women (Schwartz & Rutter, 1998). Gender expectations consistent with the sexual double standard may be responsible for heightening women's sensitivity to the degree of privacy or pressure to respond honestly more so than men's, especially in the exposure threat condition. Men have a history of enjoying and expressing sexual freedom, autonomy, and liberation, and therefore may be more comfortable than women expressing their sexuality on self-report measures. Because men do not face the same negative consequences for expressing their sexuality as do women, they may not experience the need to inhibit these responses to the same degree.

The lack of significant effects of testing condition on sex differences in erotophilia and sexual attitudes is interesting and requires further explanation, in light of some of the significant findings related to behavior. One plausible explanation is that reports of sexual attitudes and opinions are not as influenced by normative expectations for men and women as are reports of sexual behaviors. This would account for fluctuations found in self-reported behavior but not attitudes across testing conditions. A second possible explanation is that individuals, particularly women, experience more constraints placed on their sexual behaviors than on their sexual attitudes, which may pressure them to falsely accommodate to behavioral norms more so than to attitudinal norms for sexuality. Thus, sexual behaviors may be more susceptible to social desirability responding and self-presentation strategies than are sexual attitudes. If this is the case, findings on sex differences in self-reported sexual attitudes may indicate real differences between the sexes whereas the typical patterns found in self-reported sexual behavior may not accurately reflect true sex differences.

It is well known that response bias can weaken the credibility and validity of findings obtained with the traditional survey approach (Catania, Gibson, Matin, Coates, & Greenblatt, 1990). As the present study suggests, self-presentation strategies relevant to gender role norms also affect self-reports of sexual behavior. Much of the data reported on sexuality are collected in settings more similar to our exposure threat condition than either the bogus pipeline or the anonymous conditions. Thus, in sex research based on self-reports, sex differences may be exaggerated due to false accommodation to gender role norms. These differences may reflect respondents' ideas of what they are expected to report rather than their actual experience. Although it is not practical to use the bogus pipeline technique in all sex research, our results illustrate the need for researchers to do everything possible to minimize the likelihood that participants' responses are tainted by social expectations.