This two-day workshop outlines the case study method of process tracing and uses discussion and examples to enable students to use process tracing in their research. The course is applicable to research in political science, sociology, economics, education, business, criminology, and other social sciences. The audience for this course is PhD students, faculty, researchers in government or think tanks, and M.A. students who plan to write an MA thesis or apply to PhD programs. No specialized prior coursework is necessary for students to successfully participate in the workshop, although prior study of research methods and the philosophy of science will be helpful.
Process tracing is one of the most common techniques used in qualitative case studies. It is a within-case method of analysis, much like detective work, that involves both deductive and inductive assessment of evidence from individual cases. Deductively, process tracing uses Bayesian logic to analyze evidence from an individual case in order to assess the likelihood that alternative explanations of the outcome of the case are true. Inductively, process tracing uses "soaking and poking" in information about a case to develop new alternative explanations of the case.
After first introducing the Bayesian logic that underlies process tracing, the workshop offers practical advice on how to do process tracing and illustrates this advice through exercises and discussions of published research. The course also addresses case selection for process tracing, and how to combine process tracing with cross-case comparisons and typological theorizing. It concludes with constructive critiques of students' plans for process tracing in their own research projects.
Students interested in presenting their own research project for feedback on their use of process tracing should email Professor Bennett (BennettA@Georgetown.edu) with a very brief (2-3 pages at most) outline of: 1) their research question or puzzle; 2) a list of the cases in which they will use process tracing and a description of the population of which these cases are members; 3) the candidate alternative explanations for the outcomes in the cases to be studied; 4) the observable implications of these explanations to be assessed through process tracing; 5) the kinds of evidence (archives, media reports, memoirs, interviews, etc.) that will be used in the process tracing. If case selection has not yet been finalized, that is fine -- just say so or list some possible cases for study.