Conceptualizing, measuring and developing indicators: Deliverable D4.1

In this report we develop several core variables the study of gender bias in grant allocation. Firstly, an alternative approach is proposed and tested for measuring scholarly performance with bibliometric data. It avoids using individual indicators but is based on the latent dimensions underneath the indicator space. Secondly, we further developed the indicators for measuring researcher independence indicator, based on earlier work. However, the available indicators are very resource intensive and therefore not useful for large scale samples. We here show how independence can be measured in large samples. Thirdly, we investigated whether gender differences in awards and prizes may be a form of gender bias, and the case investigated confirmed that. The implication is that awards and prizes are not only an important measure of merit, but also can lead to indirect bias in grant decision-making. It is important to include a variable representing awards and prizes in the model also as a mediator. Fourthly, we explored how the role of team characteristics, and sketch how the new collected data can be used to include team characteristics in the model. Finally, as the development of quality indicators has proceeded slower than expected, we focus in this deliverable on the conceptual and measurement aspects of indicator development.