Gender and the Evaluation of Research

Panel research assessment and the use of metrics in evaluating the quality of academics and the institutions in which they work has become an increasingly ubiquitous activity. The assessment of new knowledge derived via research is controversial, and how that evaluation occurs may have important impacts on individuals and groups, with the impact on women being a key element of that debate. That women are not disadvantaged is a pressing policy issue in higher education, especially since legislation requires the public sector to take proactive steps to promote equality rather than simply prevent discrimination. Using data from the UK we find that while the proportion of men in a school has some effects on peer review derived measured research quality at the aggregate level, there is evidence that women are disadvantaged by the use of journal ratings rather than peer-based assessment. Women who are able to utilise networks to co-author with individuals outside their institution do better than those who do not. Additionally, women who are attributed with “individual staff circumstances” (e.g. maternity leave or part-time working) are negatively impacted by metrics while men are not. We also observe important differences in the quality of journals that men and women publish in across the sub-disciplines within the field. Finally, we show that when we allow for a range of relevant factors there is no evidence of panel biases in favour of leading universities. Overall, our results imply that the decision to eschew metrics in the Research Excellence Framework in favour of peer review to evaluate output quality was a sensible one