One of the aims of this study is to bring methodological structure and rigor to the question of civilian agency. In this chapter, I outline my multimethod empirical strategy for studying the question of civilian autonomy. The chapter serves as a guide for the next four empirical chapters. I start by describing the broader research design of the study and how the different methods and pieces of evidence fit together. I also discuss the different choices and trade-offs of particular methods and their benefits and limitations. The design deals with the issues of reverse causality between the impact of civilians and armed groups as well as possible bias in case selection. I then describe the research process and preview the data sources I collected during eleven months of fieldwork in Colombia spread over four years.

The research design seeks to understand the central counterfactual question of whether armed groups would have used more violence if not for civilian autonomy strategies and the organizations that enable them. The empirical work is useful for theory building, but its main purpose is to test clearly stated and falsifiable implications of theory. The use of multiple methods in a subnational study in a single (post-)conflict country is beneficial for providing deep and comparative understanding. While no single test or method alone provides conclusive answers for the outcomes being investigated, each method plays an important role in inquiry. The combination of quantitative analysis, purposive case selection, and fieldwork aims to push the methodological boundaries of civil war studies. Taken together, the approaches tend to point toward the same conclusions and paint a coherent picture of civilian behavior.

This research embodies an interplay between inductive theory building and deductive theory testing. When I began this project, I was principally engaged in building theories. I had little initial knowledge about the potential security concerns that could have made this work infeasible or the details or effectiveness of civilian organizational processes. I began my fieldwork interviewing a broad cross-section of subjects from many communities and with diverse experiences to gain background on civilian choices and formulate my research questions.