Table 1 Selection of methods in research impact assessment and how they may be used to investigate and address gender bias

Logic modelling describes and graphically represents the chains of results associated with all stages of the research process; theories of change are formulated on the basis of these chains of results to explain how to achieve the desired results

To provide a framework to systematically investigate gender equity at all stages of the research process, and investigate how gender equity can help maximise desired research outputs and impacts

Scientometrics pertain to the statistical analysis of articles and citations in academic journals (bibliometrics) and other research-based outputs, e.g. patents, commercialisation

To measure gender-based differences in research outputs, their academic impacts, including gender-biased citation behaviour, and non-academic impacts

Altmetrics measure the online attention to academic articles on social media, mainstream news websites, blogs and social bookmarking websites

To measure gender-based differences in more immediate academic and non-academic impact of research

Surveys, curriculum vitae data, narrative case studies and text-mining algorithms collect quantitative and qualitative information that may not already have been captured as part of the research process

To investigate gender-based differences in perceptions of and approaches to the research process, outcomes and impacts

Information and management science approaches allow developing balanced scorecards and other performance management systems

To assess and monitor gender equity in research organisations

Economic modelling categorises and cost-benefit analysis assesses different types of benefits from research

To examine how gender equity can maximise economic returns on investment in research