Though the US National Science Foundation introduced a broader impacts criterion to their merit review process in 1997, policy evaluations remain still scarce. Reactions from different scientific fields varied. This paper aims to quantitatively compare the proposed broader impacts of 360 funded abstracts from biology, engineering, and mathematical/physical sciences. Specifically, it considers whether or not certain fields are more likely to propose certain types of broader impacts activities, whether women principal investigators are more likely to propose broader impacts, and the effect of grant size. This study demonstrates that cultural differences exist between scientific fields and also supports existing policy recommendations that encourage the creation of organizations and partnerships at university level to allow scientists to more easily participate in activities with broader impacts. Emphasizing broader impacts activities may also attract a more diverse scientific workforce, as many individuals do not pursue science because of a perceived lack of impact.
