Rubrics to eliminate gender bias in faculty hiring less objective than thought

In a Policy Forum, Mary Blair-Loy and colleagues discuss the use of candidate evaluation rubrics to curb gender bias in faculty hiring. Through a multi-year case study of rubric use and its impact on hiring in an R1 research university’s engineering department, Blair-Loy et al. show that while using rubrics can improve diversity in hiring, they require additional strategic frameworks to be effective, particularly in the evaluation of short-listed candidates. Although academic faculty hiring aspires to be based on one’s accomplishments and achievements, a growing body of research clearly illustrates a gender bias against women, particularly in the science, technology, engineering, and mathematics (STEM) fields. Hiring rubrics – in which faculty evaluators systematically rate each candidate on a set of standardized criteria – have been widely implemented as tools to mitigate hiring bias. Many academic policymakers and equity, diversity and inclusion specialists strongly recommend using rubrics as a best practice in faculty hiring activities. However, according to Blair-Loy et al., there is a general lack of evidence that they truly work in combating gender bias. Through their study, the authors show that gender bias remains endemic in some scoring categories and evaluators’ comments in this seemingly objective evaluation process. Blair-Loy et al. do not suggest abandoning rubrics as they can help mitigate bias and create a more equitable hiring process when used strategically. However, to be most effective, hiring processes must also include department-level strategies addressing individual and interactional bias during the discussion of potential candidates. “In light of our findings, that gender bias remains endemic even in this seemingly objective evaluation process, it is vital that rubric usage be accompanied by strategic application in departmental meetings to counteract individual bias and check interactional bias during the discussion of candidates,” Blair-Loy et al. write.