In this week's edition of GrantWeek:
As reported in Science, a new study using computerized textual analysis has uncovered probable gender bias in grant awarding by at least one funder. The study from the National Bureau of Economic Research was commissioned by the Bill & Melinda Gates Foundation to review their own life science grant awarding metrics in the hopes of uncovering any previously unknown biases. Sure enough, the study indicates that female applicants are being scored lower on proposal reviews than male applicants.
The bias centers around differences of language used in applications, termed by the study as "broad" language versus "narrow" language. Male applicants tend to use "broad" language- terminology that appears across the whole range of topic areas. Women tend to use "narrow" language- terminology that appears in one or a few topic areas. The chosen terminology describes the effect the proposal topic might have on a reviewer; where a broadly-phrased proposal might gesture at ultimately finding a cure for a disease, a narrowly-phrased proposal might suggest experimenting to understand an aspect of a disease, even if both proposals are only incremental steps to a cure or treatment.
It is unclear whether this bias exists outside of the Bill & Melinda Gates Foundation. Other studies have failed to turn up similar biases in the National Institutes of Health, which has a significantly different review process. The same cannot be said of NIH for other biases, such as race and ethnicity.
Studies utilizing computerized textual analysis are fairly new, and further research is both necessary and forthcoming.