Published in Science earlier this week (h/t Dan Hicks):
Although the representation of women and racial or ethnic minorities within the scientific community has increased in recent decades, the overall pace of diversification remains relatively slow (1). A number of factors may be involved (2), but one possible explanation for this limited progress is that gender and racial or ethnic biases persist throughout academia (1, 3).
In response, we propose a scientific approach to the design, assessment, and broad implementation of diversity interventions. We review evidence of positive and negative outcomes of existing interventions relevant to academic scientists. We then offer an evidence-based framework identifying elements of successful interventions (see the table). Finally, we discuss research needed to define success more rigorously and policy changes to encourage widespread adoption of successful programs.