Understanding Process Tracing Author(s): David Collier Source: PS: Political Science and Politics, Vol. 44, No. 4 (October 2011), pp. 823-830 Published by: American Political Science Association Stable URL: https://www.jstor.org/stable/41319974 Accessed: 26-11-2018 23:42 UTC 2019 Addendum To accompany David Collier’s 2011 Process-Tracing Article Since the publication of this article (Collier 2011), the methodology of process tracing has seen two key innovations: (1) a major stride forward in the application of Bayesian tools; and (2) a more complete conceptualization of different types of rival explanations. The ideas presented by Collier (2011) become more useful if applied in conjunction with these advances. First, with regard to Bayesian tools, Humphreys and Jacobs (2015) offer a major systemization of Bayesianism as it applies to process tracing and multi-methods research more broadly. Collier (2011: 825) had proposed, as a framework for evaluating process tracing tests, the two dichotomies around which Table 1 (below) is organized. That is, he focused on whether (or not) a given process tracing test is necessary and/or sufficient for affirming a causal inference. However, he argued that these dichotomies “are a useful heuristic, but should not be taken rigidly." Humphreys and Jacobs (2015) greatly extend this framework. They formulate a methodology for viewing these two dimensions as continua, rather than in terms of necessary and sufficient conditions, thereby yielding a far more precise formulation of the inferential leverage derived from process tracing tests. For an excellent summary of their approach, see Bennett (2014) and Bennett and Checkel (2015: appendix). Further innovations in Bayesian tools are introduced by Bennett and Checkel (2015), and Fairfield and Charman (2017, 2019) add major refinements to this framework. Second, with regard to rival explanations, Zaks (2017) creatively broadens the evaluation of altern- ative hypotheses to encompass a better understanding of "relationships among rivals." She demonstrates that, in addition to sometimes being mutually exclusive, as in Collier (2011), three other types of rival explanations should be considered: coincident, congruent, and inclusive (Zaks 2017: 348). The value of process tracing tests is far greater when this wider range of options is considered. In sum, since the publication of Collier (2011), the methodology of process tracing has been greatly strengthened by these innovations. Hence, Collier’s article should be seen as offering an initial framing, to be supplemented as appropriate by these new perspectives. References • Bennett, Andrew. 2014. “Process Tracing with Bayes: Moving beyond the Criteria of Necessity and Sufficiency.” Qualitative and Multi-Method Research 12(1): 46–51. • Bennett, Andrew, and Jeffrey Checkel, (eds). 2015. Process Tracing in the Social Sciences: From Metaphor to Analytic Tool. Cambridge, UK: Cambridge University Press. • Collier, David. 2011. "Understanding Process tracing." PS: Political Science and Politics. 44(4): 823–30. • Fairfield, Tasha, and Andrew E. Charman. 2017. “Explicit Bayesian Analysis for Process Tracing: Guidelines, Opportunities, and Caveats.” Political Analysis 25(3): 363–80. • Fairfield, Tasha, and Andrew E. Charman. 2019. “A Dialogue with the Data: The Bayesian Foundations of Iterative Research in Qualitative Social Science.” Perspectives on Politics, 17(1): 154-167. • Humphreys, Macartan, and Alan Jacobs. 2015. “Mixing Methods: A Bayesian Approach.” American Political Science Review 109(4): 653–73. • Zaks, Sherry. 2017. “Relationships Among Rivals (RAR): A Framework for Analyzing Contending Hypotheses in Process Tracing.” Political Analysis 25(3): 344–62.