At each transitional point the number of women decreases at a significantly higher rate than men. Thus, while women made up 37% of the students taking physics in U.S. high schools in 1988, only 22% of those taking the calculus-based introductory physics course in college were women (AIP, 1988, 1991). Women's presence is reduced to 15% of those receiving the bachelor's degree in physics and then to 10% of the share of Ph.D.s. The decline continues in the shift from education to academic employment, with women constituting 7% of assistant professors of physics and only about 3% of full professors.

What are the effects of such small numbers on the women who persist in scientific careers? A key finding in our interviews was that as the number of women faculty members in a department increased, they divided into distinct subgroups that could be at odds with each other. Senior female scientists typically shared the values and workstyles of older men; their narrow focus failed to meet the needs of most younger women. In contrast, some younger women (and a few men) struggled to create an alternative scientific role, balancing work and non-work issues. The scientific role thus divides along generational and gender fault lines. These developments have significant unintended consequences for the socialization of female scientists – for example, the availability of relevant role models.