Parable of the Polygons
Interactive simulations of how slight bias can affect a very simple system.
Blind Orchestra Auditions
Goldin and Rouse's 1997 work on how gender-blind auditions contributed to a dramatic increase in the fraction of female musicians in top symphony orchestras. It's on the old side, but a very well-known result. One subtle point is that overall women who auditioned after gender-blind policies were established were LESS likely to be hired, mainly because more women started auditioning (perhaps because they thought they finally had a chance at getting the job)! However, within the "quality-controlled" group of musicians who participated in both blind and non-blind auditions, the women had more success in the blind auditions.
Co-author penalty for women and NYT summary
In this December 2015 preprint Heather Sarsons examines tenure statistics over four decades. Tenure rates for men and women seem comparable...but only for women who write all their papers individually or with other women. On average, papers with mixed-gender co-authors gave women less of a boost towards tenure, and that boost dropped to effectively zero for papers with entirely male co-authors. Sarsons did the obvious checks, such as whether women tend to write papers with "better" (more senior, more established) men. They don't; just the opposite, in fact.
Which students are perceived as smartest and Washington Post summary
This study of several biology classes found that when students were asked to identify other students who were on top of the material, men had a strong bias towards naming other men. On average, a female student needed a course grade more than 0.5 points higher than that of a male student (on a 4-point scale!) to be named as often as a strong student. In this study, female students showed much less bias in their evaluations. The authors threw in the scary thought that if this is how things work in biology, it may bode ill for physical science fields that are more often perceived to need math skills and brilliance.
Gendered Language in Teacher Reviews
and NYT description
An interactive created by a professor at Northwestern lets the user search for how often words are used to describe men or women on Rate My Professor. Those used more often for men include "smart" (and even more so, "brilliant" and "genius"), "great," and "best." "Competent" was about equal, but women were more often described as "incompetent." Women also were more often called "disorganized," "awful," and "ridiculous." Then again, "jerk" showed up far more for the men. There's plenty of procrastination potential here!
This is a description of a workshop in a biology department. It contains various links to sources used, as well as comments on the activities during the workshop and how it went. As a footnote to the piece by John Johnson referenced in the "Prep" section, the grad student he mentions, Gwen Rudie, and her partner are now in Staff Astronomer positions at the Carnegie Institute; they were also both offered faculty positions at a major research university. This is not to say that Rudie's concerns about two-body issues were unfounded. Unfortunately, that remains a very difficult problem. But I was happy to learn that things worked out for someone who had the gumption as a grad student to explain to a professor exactly why he was wrong.
Slides from Presentation at 2014 CUWiP with some stats on women in science and references to research on bias. The "leaky pipeline" slide(s) are quite interesting, since they actually show that the leaky pipeline model is incorrect. There is only minor leakage at the graduate level and none beyond that, at least judging simply from the percentage of women at each level. The big drop is from high school to college, and from other data I've seen it happens mainly in the first year of college. Serious efforts to get more women into physics really need to focus on that step. (The leaky pipeline idea was that women dropped out at a higher rate than men at all levels. Perhaps that was once true, but it seems not to be now. The good news is that it should be easier to focus on a single problematic step than on a gradual disappearance of women at many different stages.)
Gender Bias Bibliography by Danica Savonick and Cathy Davidson, with all sorts of good stuff in it.