Using Process Tracing: Case Study Methods and Practice

Professor Derek Beach, University of Aarhus, Denmark

The aim of this course is to improve your use of in-depth case study methods in relation to your current research project (Master’s, Ph.D., post doc or other). Participants will ideally have attended the introductory course on comparative historical analysis and the introductory process-tracing course in week two of the IPSA-USP Summer School. Students should come prepared with their own research as many of the exercises will based on student research projects. Students who completed week 2 of the process-tracing course in the IPSA-USP Summer School will be able to significantly improve their process-tracing project throughout the course of the third week of this module.

In the first two days of the course, we will be working with translating theories from your own research into process theories. On day 1, we will be exploring how the causes and outcomes in your research can be defined to make the compatible with process tracing case studies. During day 2, we will be working on the theories that you have developed in your own research, attempting to specify what activities link parts of your causal process theory together. This often involves translating a set of particular historical events into a causal process that in theory could be present in other historical cases.

Days 3 and 4 deal with how to trace historical causal processes using empirical evidence. We discuss what empirical evidence looks like in process tracing, discussing the differences between evidence of difference-making used in variance-based designs and mechanistic evidence (or traces) left by the operation of causal processes used in process tracing. We also discuss how to evaluate the probative value of individual pieces of mechanistic evidence, and how we can aggregate what they tell us. Here we draw on tools developed in historical analysis relating to source criticism and evaluation.

On day 5 we discuss mixing methods, and in particular how process tracing case studies can be nested within comparative analyses of bounded populations of cases.