http://www.pnas.org/content/111/28/10107 - This is the exact sort of study I was referencing. It only shows that there is a different in result, not opportunity. It further shows that as the baseline competency standard increases (up to labs being operated by Nobel Laureates) - so does the "bias". It proposes explanations for this being either self selection by women, or bias by men. It ignores the most likely explanation which is that though the pool is split about 50/50 by gender, competencies are not.
----
http://www.pnas.org/content/109/41/16474.full.pdf - I was familiar with this study, and it's a good example of the ongoing issues with social psychology toy study. For reference the replication rate in social psychology is now at around 25%. Put another way, if a social psychology study tells you something - you'd generally be vastly more well informed if you assumed the opposite, or at least assumed what was stated, was not true!
This study offers a demonstration in a number of ways this has occurred. One major issue is that there was no effort to manage a response bias, other than in broad characteristics (race/gender) of applicants. Corinne Moss-Raucin [1] personally mailed a number of faculty asking them to respond and rate a variety of potential students. One glance at her faculty page will tell you what she's actually doing. So who voluntarily opts into this? In total just around 30% of contacted faculty chose to. I think there is a 0% chance that this is not a biased sample.
The questions were also framed in a context that seems to imply a potential personal "affinity" for an individual. One important nuance here is that the students offered up for consideration were all low quality. The questions to demonstrate bias included:
- "How likely would you be to encourage the applicant to continue to focus on research if he/she was considering switching focus to teaching?"
- "Would you characterize the applicant as someone you want to get to know better?"
Do you think you'd try to keep low performing Jennifer in your office, even if she was looking into teaching instead? Would you like to get to know her better? I mean come on this is just absurd, and a reason that the social sciences and especially social psychology is imploding in on itself. It's like if the "biases" went in the opposite direction our researcher was ready to write up an article about unhealthy professional attitudes towards females and female independence.
[1] - https://www.skidmore.edu/psychology/faculty/moss-racusin.php