Recent research from Caltech reveals that the confidence gender gap, often implicated in women’s workplace underachievement, is “contagious.” When assessing a worker’s performance based on self-evaluations, reviewers tend to reward apparent self-confidence and, conversely, penalize a lack of confidence, rather than solely focusing on actual performance.
Consider a scenario: You’ve been in your job for three years, your performance reviews are good, and you have every right to ask for a pay raise. Do you go for it? Gender plays a role in such decisions. Statistics show that men are more likely to confidently assert their worth and ask for a raise and promotion, while women are more likely to downplay their achievements, imagining room for improvement.
Confidence gap
The consequence of this “confidence gap” is notable: Even when men and women perform equally and are equally trained, men, on average, receive higher pay and have greater upward mobility than women.
Traditionally, women have been advised to “lean in,” a concept popularized by Sheryl Sandberg’s 2013 book. The idea is for women to emulate successful men, value themselves, and confidently communicate their worth. However, some argue that the responsibility should be on employers to recognize gender differences in self-assessment and interpret self-reports accordingly.
To investigate this, the researchers divided participants into workers and evaluators, and they each took a math and science quiz, followed by incentivized self-assessment questions. The results confirmed the confidence gap: While actual test scores showed no gender difference, significant variations emerged in participants’ confidence levels.
“On every single one of the self-assessment questions, women report more pessimistic beliefs about their performance than men,” the researchers explain. “Basically, we replicated the expected finding that there is a gendered confidence gap between equally performing individuals.”
Different scenarios
In the second phase, the focus shifted to the evaluators. How would they respond to the differences in confidence observed among the workers whose results they were reviewing?
Evaluators were initially presented with a random worker whose gender was disclosed, but no additional information was provided. They were then asked to estimate the percentage likelihood that this worker’s performance was subpar. Interestingly, evaluators gave similar estimates for any random person, regardless of gender. This ruled out the possibility of taste-based discrimination, indicating that evaluators did not start with the belief that women were more likely to perform poorly on a math and science quiz.
However, when evaluators were given access to the workers’ self-assessments and asked once again to specify the percentage chance of poor performance, a shift occurred. The relative lack of confidence exhibited by women was found to be contagious, influencing evaluators to now conclude that women performed worse.
To gauge evaluators’ prior awareness of the gender gap in confidence, they were tasked with estimating the likelihood of workers being overconfident or lacking confidence. Evaluators accurately guessed that male workers were more likely to be overconfident, and female workers were more likely to lack confidence, indicating an awareness of the confidence gap. Surprisingly, even when presented with these questions, evaluators did not show improved recognition of the impact of confidence on workers’ self-reports.
“We thought that maybe if we just asked them about gender and confidence, this would nudge the evaluators to take that into account,” the researchers explain. “But that had no effect on their assessments about workers’ performance. They continued to guess that women actually performed worse than men after learning about the workers’ self-reports.”
Real-world implications
The implications of this research extend into real-world scenarios. Some believe that implementing gender-blind applications or reports could address gender imbalances, but this study suggests that such an approach might exacerbate the situation. Without knowledge of applicants’ or workers’ gender, evaluators would be unable to consider the gender gap in confidence, even if they wished to.
Moreover, the cognitive bias uncovered in this research may not be exclusive to gender but could extend to other groups, such as those whose cultural norms lead them to project more humility and less confidence. The study tested whether evaluators would exhibit the same bias towards non-gendered groups by labeling them as “group A” and “group B,” and the results mirrored those observed in the gender-specific scenario.
While the findings might be discouraging for individuals with lower self-confidence, the silver lining is that we understand a lot about cognitive biases and know ways to mitigate them. Evaluators form biased assessments because they struggle to incorporate the information provided, not because they actively discriminate against women or other groups. This suggests that interventions aimed at addressing these biases could be promising in fostering fairer evaluations.
