Skip to main navigation menu Skip to main content Skip to site footer

Articles

Vol. 37 No. 3 (2017)

Process tracing: Induction, deduction, and causal inference

DOI
https://doi.org/10.4067/S0718-090X2017000300659
Submitted
December 11, 2019
Published
2017-12-11

Abstract

Process tracing is a method to obtain solid causal inferences. Concerned by the fragmentation of the growing literature distinguishing different types of process tracing, we highlight the elements that unite this method: the construction, through different routes, of narratives that provide plausible, persuasive explanations of the outcomes of interest. Our argument draws on the comparison of two studies with different entry points to the research process: the first one starts inductively due to the novelty of the outcome of interest, while the second study starts deductively since previous theories were available. We show how to conduct process-tracing analysis rigorously and highlight that these studies converge to produce narratives structured around hypotheses and causal mechanisms that explain the outcomes.