Photo by David Hurley on Unsplash
Ideas for Leaders #727

Gender Bias Against Women Leaders Is Higher Than We Think

This is one of our free-to-access content pieces. To gain access to all Ideas for Leaders content please Log In Here or if you are not already a Subscriber then Subscribe Here.

Key Concept

While men are more prejudiced than women against women leaders, a new study demonstrates that when surveyed, women are less likely than men to admit their prejudice.

Idea Summary

Women face an uphill battle in achieving high leadership positions because of the prejudice of many men — and women — against women leaders. Too many men and women buy into the gender stereotypes of women as emotional, caring and gentle, who don’t have the active, competitive, independent and self-confident traits of men needed in a leader.

Because such prejudices can undermine women’s careers and encourage gender bias in the workplace, an accurate assessment of such prejudices is important to measure the full extent of the barriers women face. In the past, such assessments are based on self-reported results: respondents are asked to state their beliefs about women leaders.

The problem with such assessments is that there may be a ‘social desirability bias’ influencing the results. In other words, people may be hesitant to express their true opinions about sensitive subjects — for example, that women don’t make good leaders — because they know that such opinions are not socially acceptable. As a result, statistics on the extent of prejudice against women, while already alarming, may even be understated.

Two researchers from the University of Dusseldorf’s Department of Psychology tested this concern by using a combination of two very different questioning techniques. One was the direct question (DQ) technique, in which respondents are asked to respond ‘true’ or ‘false’ to questions. The second, the Crosswise Model (CWM), used an indirect questioning technique that randomized responses to ensure that survey responses could not be traced back to the individuals giving them; CWM’s built in anonymity characteristic was expected to significantly reduce the need for respondents to adapt their answers to what was socially acceptable.

For example, the question about gender bias in the DQ format would be:

‘I think that women possess fewer leadership qualities than men.’ Please answer true or false.’

The question about gender bias in the CWM format included a second question to randomize the answers, as follows:

Statement 1: ‘I think that women possess fewer leadership qualities than men.’ Please answer true or false.’
Statement 2: ‘I was born in November or December.’

Answer options:

  • A) ‘Both statements or neither of the two statements are true.’
  • B) ‘Exactly one of the two statements is true, irrespective of which one.’

Because the prevalence of statement 2 was known — that is, the researchers would know based on statistics what percentage of the respondents would have been born in November or December — the researchers would be able to statistically determine the respondents answer on the gender issue.

To confirm that CWM did in fact reduce the social desirability bias, the researchers asked a non-sensitive control question — does your surname begin with one of the letters, K, L, or M? — in addition to the sensitive experimental question on gender bias. The researchers combined the two methods for each participant, with some answering the sensitive gender bias in the CWM format and the control question in the DQ format, while others answered the sensitive gender bias question in the DQ format and the control question in the CWM format (i.e., with a randomizing question).

The 1529 participants in the survey were recruited from three German universities. The statistical analysis of their answers revealed the following results:

  • As expected, whether respondents answered the non-sensitive control question (about their surnames) in the DQ format or CWM format made no difference: the results (22% replied yes) were the same.
  • For the sensitive experimental question, 23% expressed gender bias when replying to the question in the DQ format, while 37% expressed gender bias when replying to the question in the CWM format. 
  • By gender, 36% of men expressed gender bias when replying to the question in the DQ format, while 45% of men expressed gender bias when replying to the question in the CWM format.
  • By gender, 10% of women expressed gender bias when replying to the question in the DQ format, while 28% of women expressed gender bias when replying to the question in the CWM format.

In summary, women were less prejudiced than men against women leaders; however, the number of women willing to express their gender prejudice increased much more than men when they were given greater anonymity. In short, women were more influenced by the social desirability bias than men.

Business Application

The research results are unequivocal: gender bias against women leaders is high, with 45% or men and 28% of women expressing gender bias when their anonymity is protected. (In other words, nearly 1 out of 2 men and nearly 1 out of 3 women believe men make better leaders!) The increase in bias, especially among women, when asked indirectly rather than directly about their bias indicates that other surveys on gender bias, using direct questioning techniques, are underestimating the extent of the problem — and thus underestimating the challenge of overcoming this bias.

The first step for business leaders and other initiating measures to reduce gender bias is to be aware of the impact of the social desirability bias. You may believe that your employees, managers and senior leaders are generally unbiased, while in fact, some may be hiding their bias — and this is, according to this research, especially true for the women in your organization! You may also believe that education programs and other initiatives to reduce gender bias are working… when in fact, people are only saying what they believe you want them to say. 
Leaders need to be more diligent and resolute in the battle against gender bias, no matter what they may be hearing.

Contact Us

Authors

Institutions

Source

Idea conceived

  • December 2018

Idea posted

  • January 2019

DOI number

10.13007/727

Subject

Real Time Analytics