My research question as it stands—what explains the lack of women in peace processes—does not fit very well with large-n statistical research because there are very few cases where women are present or included in any form in the peace process.1 To address this problem, I visited Clement Ho, the research librarian available to the program, and we talked about how different questions work for different types of research, and that sometimes it is necessary to change the question to fit the methodology. Consequently, for this methodology sketch, my question is changing to “How is the role of women in politics affected by the presence of a UN Peacekeeping force in the resolution of a conflict?” The indicator I will use for my dependent variable, the role of women, will be the number of women elected to parliaments. I found a fantastic dataset for this variable through the American University library website that catalogues the number of women elected to national parliaments in over 150 countries beginning 20 years ago in 1997.2 Clement also suggested a dataset that lists and provides detailed information on UN Peacekeeping missions, organized by region and dating from 1948 to the present.3 One of the obvious limitations of the of the women in parliaments dataset is the time range—because it is limited to 20 years, the peacekeeping missions would also have to be within the 20 years, or it would not be possible to measure their affect. Conversely, the issue could also be the Peacekeeping missions. Because they have such a large range over time, there are not enough cases within the 20-year range to do a valid statistical analysis. Consequently, my independent variable would need to change to accommodate that range and still have enough cases to test.
1.Sarah Taylor, “A Better Peace? Including Women in Conflict Negotiations” (PhD diss. The New School, 2015), 3.
2.“Women in National Parliaments,” accessed October 25, 2017, http://archive.ipu.org/wmn-e/classif.htm.
3.United Nations, “United Nations Peacekeeping,” accessed October 25 2017, http://www.un.org/en/peacekeeping/operations/current.shtml.
Julia — a good job here in thinking about how you might tweak your research question due to data limitations while also thinking about how to maximize the leverage offered by large-n analysis. The data and variables that you identify are all excellent, but the level of measurement for the *specific* indicator that you would choose to capture your DV is not discussed here as is required. Keep thinking about these things as you work towards Research Design Sketch #1 as thinking about operationalization is part and parcel of thinking about data sources!