All 7 articles in this section
The pipeline metaphor implies steady seepage. The data shows something different: two large discontinuities and a long stretch where almost nothing is lost.
Women’s share of computer science degrees peaked in the mid-1980s and then fell for twenty years. The reversal is unique among technical fields and the leading explanations are testable.
It is a measurable state that degrades working memory in real time. The most striking finding is how reliably it passes from adults to the specific children they help.
Most STEM outreach is never evaluated. Among the programmes that have been, a short list of features separates the ones that move numbers from the ones that produce photographs.
The "you cannot be what you cannot see" claim has some support. The research also shows the effect depends heavily on similarity, contact duration, and how the model presents her own success.
A new technology arrives and within a few years access differences harden into assumed aptitude differences. There is a well-documented precedent, and we are early enough to watch it happen.
Academic science loses women disproportionately after the doctorate, at the point where careers become dependent on funding, mobility and unstructured evaluation.