Robyn Price, Bibliometrics Manager, Library Services
Studies suggest that our perceptions of research quality maybe affected by unconscious biases based on the country that the research was produced in. Many things could contribute to these biases; for example, the financial resources of a country to fund research, reputation of a country’s universities, the Global North publishing market, implications of English as the dominant language of science, the consequences of colonialism and more. There are also parallels with the broader ‘decolonise the curriculum’ movement, which many UK universities are pursuing in both response to explicit student demand as well as part of strategies to reduce BAME attainment gaps and other EDI initiatives. To facilitate understanding of whether Imperial curriculum might be affected by any of these biases, we created an interactive dashboard of geographic author data from Imperial reading lists.
Thousands of citations make up the reading lists for Imperial modules and we had no way of identifying the author countries at this scale. Other universities had explored manually looking up author country for individual references, but we wanted to automate this process to be able to show the data for hundreds of modules over different years. Our team of Library, ICT and research staff from the Department of Primary Care and Public Health created a novel software method to combine author countries, a socioeconomic rank of countries and the Imperial reading list management system. The data is made available in a dashboard for course leaders to view and includes the number of reading list articles found, the number of authors, the number of countries they represent, and the income status of these countries.
For example, the dashboard can display visually the spread of author affiliated countries. The below images taken from the dashboard represent the authors referenced on one real Imperial reading list, anonymised here as Module A. In this visual, countries where an author of a text referenced on the reading list found are indicated by a blue dot. The blue dot is in the geographic centre of the country and is size-relative, e.g. the larger the blue dot, the greater the number of authors found.
Affiliated countries of authors cited by Module A reading list in 2018/19
Affiliated countries of authors cited by Module A reading list in 2020/21
I hope that providing easily accessible data on the geographic diversity of authors on a reading list might inform course leaders and students by providing insights for them to begin examining their curricula. Because it is scalable for modules to be studied over time, course leaders could use it as a tool to support understanding of how their courses change over time. It is important that this quantitative data must be used to inform understanding rather than guide it, particularly with the risk of tokenistic intervention on a reading list and with awareness that country affiliation to author is a limited indicator. At present, the University of Leeds has funded a project to replicate our model and the University of Bristol’s School of Psychology and University of Sussex’s School of Life Sciences are investigating adapting it.
Course leaders or students interested to access the dashboard and look at their modules are warmly encouraged to register interest through this form. We have data for over 1,700 Imperial modules taught since 2015 – although not all modules are suitable for this analysis due to the types of references on their reading lists. Our EDU workshop runs once a term and we have facilitated discussions and workshops on diversity in curricula and unconscious bias in research for both internal Imperial groups as well as external groups, available on request by email. Some may already be actively pursuing inclusive curricula, and others may be at the very beginning of thinking about this. We would be interested to talk to groups or individuals at any stage of this journey of reflection, and more information can be found on our webpage.
This project has received support from Imperial College President’s Excellence Fund for Learning and Teaching Innovation and NIHR ARC NW London.