The Washington PostDemocracy Dies in Darkness

Closing the gender citation gap: Introducing RADS

This is the third post in our gender gap symposium (see here and here for the first two.) We are delighted to welcome Daniel Maliniak, a PhD candidate in Political Science at University of California, San Diego, and Ryan Powers, a graduate student in Political Science at the University of Wisconsin, Madison. Daniel and Ryan are co-authors of the citations paper. Both of them graduated from William & Mary, where they were heavily involved in the TRIPS project, which has become the primary documentation of trends and facts about teaching and research in international relations. See here for some very cool visualizations they have created of citation networks.


In our recent paper with Barbara Walter, we show that gender differences in self-citation and self-promotion are a contributing factor to the of the overall gender citation gap in international relations (IR) (see the graph below). The question we are asking ourselves now is what to do about it.

In our paper, we encourage women to self-cite more often as one strategy to narrow the citation gap. As (very) junior scholars, we find recommendations that give authors some individual agency compelling, in part, because we are not in a position to make the needed structural/policy changes for which others are advocating. But our emphasis on self-citation overlooks another – potentially more important – way to help narrow the gender citation gap: encouraging authors to improve the gender balance of their own citation patterns.

How do you do this? We don’t know why people tend to cite along gender lines. That’s a whole other research project in itself. But we do have some potential antidotes. The first is to simply make people aware of the problem and encourage them to be more self-conscious about the citations used in their own research. One senior IR scholar, well known for his mentoring of women in IR, told us that he was astonished when an editor noted the disparity in gender representation in a paper that he had submitted for publication. Most scholars seem simply unaware that they under-represent women in their bibliographies.

The second is to encourage faculty members to be more self-conscious and careful about how they construct the reading lists on their syllabi. Some of the gender bias in citation patterns may be the result of the map of the field imparted to scholars in graduate seminars. For example, if the research assigned on syllabi is mostly produced by men, then a significant gender gap is likely bequeathed to the next generation of scholars.

The challenge is to make it easy for scholars to identify and then cite high-quality, substantively relevant female scholarship. Search tools such as Google Scholar and Web of Science are helpful, but they are only useful insofar as we are able to pick the correct set of keywords. In addition, scholars have to wade through search results filled with work that they are already aware of. Finally, since citation analysis is an integral part of both these services, search results are affected to some degree by the very biases we are hoping to address!

This suggests room for a new research discovery tool. We call it the “TRIP Research Analysis and Discovery System” (RADS) and it is under active development. RADS is a Web service that takes as input a list of citations from the user. This might be some portion of a works cited list or a syllabus. RADS analyzes this input and then returns a list of articles that cover similar substantive issues or topics, use a similar research design or methodology, adopt similar epistemologies or employ similar paradigmatic assumptions. This list can be filtered or sorted in a variety of ways, including the gender of the author. RADS relies on the TRIP project’s Journal Article Database, which now includes approximately 5,700 articles published in 12 top journals between 1980 and 2012. These articles are coded on over two dozen dimensions including methodology, paradigm, substantive focus and epistemology (see a complete list here ). TRIP’s database is by no means the universe of IR literature, but it is large and diverse subset of the field on which we can build a proof of concept.

This tool accomplishes two things. First, it gives scholars a valuable research aide by searching for articles that are similar on a number of pre-coded dimensions. Second, it could help decrease the gender gap in citations by allowing scholars to find substantively relevant work written by women to which they otherwise may not have been exposed.

RADS is not ready for public use. However, we do need volunteers to test this tool with real and diverse projects and to offer suggestions on how to make this tool more useful. If you would like more information about this project or want to get involved in testing, please sign up here.