This course is a more practical, hands-on course in using Process Tracing (PT) methods in one’s own research, complementing the more theoretical PT I ECPR Summer School course held in the first week, which focuses on the research design aspects of the method. The course requires that one has already had some form of introduction to PT, either by taking the week 1 course, the course at the ECPR Winter School, or another introductory course on PT.
The promise of process-tracing as a methodological tool is that it enables the researcher to study more-or-less directly the causal mechanism(s) linking a cause or trigger (commonly known, but not the same as independent variables) to an outcome, allowing us to open up the ‘black box’ of causality. Within political science methodology PT is arguably the only method that allows us to study causal mechanisms giving us an understanding of how an X (or set of X’s) produces Y. It is therefore an ‘…invaluable method that should be included in every researcher’s repertoire.’ (George and Bennett 2005:224).
In this second week (or advanced course for those who only join for the second week) we look specifically at how to study that causal link in practice. In comparison to other research methods such as large-N correlation-based analysis and comparative methods process-tracing as a distinct method involves research where ‘The cause-effect link that connects independent variable [trigger] and outcome is unwrapped and divided into smaller steps; then the investigator looks for observable evidence of each step.’ (Van Evera 1997:64, italics added). This course will thus focus on turning 'empirical observations' into evidence in support of a causal mechanism (the trigger, the steps, the outcome). To do so will follow the Bayesian logic of inference.
The first day introduces the Bayesian logic of inference. We look at Bayes Theorem and show how it allows us to draw a causal inference without variation by prioritising finding new and different evidence in relation to our theory. Taking a practical approach this introduction is followed by an exercise that will illustrate and make explicit how we can draw a causal inference on the basis of a single case: the case of Silver Blaze solved by Sherlock Holmes.
Day 2 introduces contemporary debates on empirical tests in PT and case studies. In line with the Bayesian logic of inference we focus on the inferential strength of (a) piece(s) of evidence in relation to our theory in terms of its certainty and uniqueness (or confirming or dis-confirming power). After a more theoretical introduction of the strength of evidence we analyse Schimmelfennig's article on the EU Eastern enlargement process to look at the inferential power of the evidence he provides for his causal mechanism. Can he really draw the conclusion he draws based on the evidence he presents?
Day 3 introduces source criticism and the practical challenges in working with empirical evidence in PT. Arguing that whether a piece of evidence allows us to make a strong causal inference is not only related to what it tell us about our theory but also depends on the 'trustworthiness' and accuracy of the source we will question the 'accuracy' of evidence. We focus upon archival material, elite interviews and secondary historical sources. This includes questions such as how we should evaluate bias, what a ‘good’ source is, and how we deal with bias in secondary historical material. We will use a compilation of sources and observations regarding the Cuban Missile Crisis to discuss the challenges relating to evidence in PT.
Day 4 turns to a discussion of how we can use empirical material to build theorized causal mechanisms. In relation to empirical tests we will have largely followed a theory-testing logic which assumes we have an hypothesised causal mechanism before we go out and test it. Rather than following this deductive logic, today we will look at the challenges and possibilities of induction for process tracing using Janis’ study of Groupthink as an example.
The course concludes with a discussion of practical challenges in using PT. This discussion will be based on your own research projects and the assignments made during the week. In short; after presentations of your own work we will discuss the particular pit-falls that may come with your research designs and ideas.