Gender Bias in the Perception of Gender Bias: Is It a Myth?
New research challenges the claim that men are less receptive to evidence of gender bias
Women are underrepresented in many areas of STEM (that is, science, technology, engineering, and math). More precisely, they’re underrepresented in inorganic STEM fields, such as physics, engineering, and computer science. The most common explanation for the imbalance is gender bias, and the most common interventions aimed at narrowing the gaps target this ostensible cause.
According to an influential 2015 paper by Ian Handley and colleagues, a major roadblock for such interventions is the fact that many people - men in particular - are skeptical that bias is a problem. In a series of studies, Handley et al. found that men were less persuaded than women by evidence of gender bias. This was widely interpreted as gender bias in the perception of gender bias.
Now, however, the conclusion is being called into question. Several months ago, I reviewed a paper by David Shanks, Hollie Coles, and Nadia Yeo that set out to replicate Handley’s findings. The results of the paper - recently published in Royal Society Open Science - are easy to summarize: The researchers found no evidence at all that men are more skeptical than women of evidence of gender bias.
This adds to a growing body of research suggesting that earlier claims about the ubiquity of bias against women in STEM were overstated. This is good news, of course - but it’s also news that few may get to hear, as academics and media seem determined to paint as bleak a picture as possible of the situation for women in STEM.
The Original Study: A Closer Look
Are men less receptive than women to research demonstrating gender bias? That’s the question that Handley and colleagues set out to answer in a series of experiments destined to become classics in the field.
The researchers presented participants with the abstract of a famous and highly cited study by Corinne Moss-Racusin and her team. The key finding of that study was that science faculty rated a résumé more favorably when it was attributed to a male applicant than a female one (although see this paper for evidence of the opposite). Handley and colleagues’ hypothesis was that their male participants would rate the Moss-Racusin study as lower in quality than their female participants would.
Sure enough, in the first two of their three experiments, men did indeed rate the study as lower in quality. In the third, the sexes didn’t differ in their ratings. Nonetheless, the study is generally taken as strong evidence for gender bias in the perception of gender bias.
We shouldn’t be too quick to accept this conclusion, however, because the Handley studies had serious flaws.
The most glaring is that the first two experiments lacked a control condition - that is, an abstract unrelated to gender bias that could serve as a point of comparison. Without that condition, the findings of the first two studies are basically uninterpretable, for at least two reasons.
First, it’s impossible to know whether men were underrating the Moss-Racusin study or women were overrating it. Almost everyone took the former view, of course, but this was assumed rather than demonstrated.
Second, without a neutral control condition, we can’t say whether men were more skeptical of gender-bias research specifically, or just more skeptical of research in general. The findings of the first two studies were entirely ambiguous with respect to these options.
Handley et al.’s third study did include a control condition. In effect, then, it was their first real test of the hypothesis. But this study didn’t find a sex difference in ratings, or lower male ratings for the gender-bias abstract than the control abstract. Thus, although the Handley paper is usually treated as convincing evidence for gender bias in the perception of gender bias, its findings were actually rather weak.
Note as well that, even if Handley and colleagues had found lower male ratings for the gender-bias abstract in Study 3, the finding would still have been somewhat ambiguous. This is because the researchers didn’t include a bias-against-men condition. Without that, we wouldn’t know whether men’s downgrading of evidence of anti-female bias reflects a unique bias on the part of men, or is just one half of a more general phenomenon, namely that both sexes are skeptical of research finding bias against the other.
In short, the Handley paper raised an interesting question - that of gender bias in the perception of gender bias - but left it largely unanswered.