People are more prejudiced against women leaders than the statistics might indicate. This could be because participants in surveys investigating attitudes towards men and women in leadership positions may not answer honestly unless they are guaranteed confidentiality of their answers. These are the findings of a new study by Adrian Hoffmann and Jochen Musch of the Heinrich-Heine-University Düsseldorf in Germany, which is published in Springer's journal Sex Roles.Hoffmann and Musch used an indirect questioning technique to gather information on people's true feelings about women leaders. The technique showed that people are not always honest when directly asked their opinion on socially sensitive questions because they prefer to give answers that they think will be accepted by others.

Gender stereotypes and gender-oriented prejudice pose a serious threat to women's careers and facilitate gender bias in the workplace. According to theorists, prejudice against women leaders emerges from an incongruity between their gender role and the more masculine social role of a leader. Self-reports are often used to study people's prejudice against women leaders, and these have helped to gather important information about the form intolerance and stereotyping against women in power takes. Such methods may, however, be biased because of the unwillingness of some people to disclose information that could be viewed by others as violating social norms, and therefore reflect unfavourably on the respondent.

To counter this effect, Hoffmann and Musch collected responses from 1529 German students with either a conventional direct question, or an indirect question in the format of the so-called Crosswise Model. The randomization of individual answers which lies at the core of this advanced questioning technique grants full confidentially to respondents answering questions of a sensitive nature. It therefore allows respondents more freedom to answer honestly without feeling the need to provide answers that would put themselves in a better light. Estimates of the prevalence of prejudice obtained using the Crosswise Model were further compared with ones obtained using a conventional direct question.

Results from the Crosswise Model show that people are significantly more prejudiced against women (37 per cent) than results from direct questions indicate (23 per cent). This could be because more participants were willing to admit to being prejudiced against women leaders if they were granted full confidentiality in their answers.

When granted full confidentiality, 28 per cent of women and 45 per cent of men in the sample indicated that they considered women to be less qualified for leadership positions than men. Across the two study methods, men showed more prejudice than women. However, the increase in the estimated prevalence of prejudice from a conventional direct question to the Crosswise Model was higher in women (from 10 per cent to 28 per cent) than in men (from 36 per cent to 45 per cent), indicating that women responded more strongly to being granted full confidentiality of their answers.

"This pattern suggests that women are much more reluctant than men to express their prejudice against women leaders. Perhaps because women feel obligated to solidarize with members of their in-group," explains Hoffmann.

"Given that even many women have reservations against women leaders, the societal and political promotion of gender equity has obviously not been successful at changing the attitudes of every potential future leader. It therefore does not seem unreasonable to expect the further persistence of workplace bias," adds Musch.