Eric.ed.gov – Unconscious Bias in the Classroom: Evidence and Opportunities, 2017

The underrepresentation of women and racial and ethnic minorities in computer science (CS) and other fields of science, technology, engineering, and math (STEM) is a serious impediment to technological innovation as well as an affront to fundamental notions of fairness and equity. These gaps emerge in the early grades and tend to persist, if not widen, throughout the secondary and postsecondary years. The unconscious biases (UB) of teachers, school administrators, and fellow students may contribute meaningfully to the persistence of these gaps. Fortunately, a nascent literature on targeted interventions that directly address UB suggests there may be compelling opportunities to promote broader engagement in CS and STEM education and employment. The fields of neuroscience, social psychology, economics, and sociology articulate the many possible origins of UB and the ways in which UB can harm stereotyped groups, particularly in educational settings. This interdisciplinary literature yields two troubling, important insights: (1) Humans consciously “and unconsciously” store experiences in our brains and those experiences (memories) later influence instantaneous, automatic decision-making, which is critical to cognitive functioning and “cannot be turned off”; and (2) Exposure to UB can trigger self-fulfilling prophecies by changing stereotyped groups’ behaviors to conform to stereotypes, “even when the stereotype was initially untrue”. These insights provide specific guidance for mitigating the negative consequences of UB via interventions that disrupt the channels through which UB influences individuals and that highlight the insidiousness of UB, respectively. In particular, the following design insights should be considered when addressing UB systematically: (1) Asking individuals to “suppress biases” is likely to be “counterproductive”, as this requires a great deal of mental effort and can cause UB to eventually rebound above pre-intervention levels; and (2) Teachers and classroom climate moderate the impact of UB, suggesting that teacher-facing interventions that carefully leverage the relevant psychological mechanisms (e.g., awareness, motivation, individuation, and empathy) have substantial promise to reduce teachers’ UB and improve student outcomes. In sum, UB is a nontrivial problem in education, especially in CS and STEM education, and it is not easily addressed via traditional educational policies and interventions. However, interventions that identify and alter the frequently unconscious psychological processes that harm individuals’ outcomes are currently being developed and piloted. Teacher-facing interventions, which can be administered to both pre- and in-service teachers, are particularly promising. In part, this is because by addressing UB among teachers, we can help shape the entire classroom context in supportive ways. Furthermore, teacher-facing interventions are potentially cost-effective and scalable, because infrastructure for teacher training is already in place. Still, much remains to be learned about the scalability, external validity, and optimal design of such interventions. Given the scope and complexity of the problem, interdisciplinary and inter-sector partnerships between public schools, universities, researchers, and industry will likely play a pivotal role in meeting these objectives.