Practitioner Reflections on Bayesian Reasoning for Qualitative Research

Session Submission Type: Roundtable

Session Description

This roundtable panel invites prominent practitioners of qualitative research from different subfields of the discipline to reflect on Fairfield & Charman’s (2017, 2019, forthcoming) Bayesian approach to inference. These authors contend that Bayesianism is much closer than alternative methodologies to the way that scholars naturally approach case studies. They argue that learning the basic principles of Bayesian updating can help leverage and improve intuition, while also helping scholars avoid cognitive biases that can lead to sloppy reasoning. In addition, they emphasize that Bayesianism facilitates knowledge accumulation by providing a clear framework for scrutinizing inferences and pinpointing sources of disagreement. The authors view a substantial component of their methodological project as an effort to explicate and lend legitimacy to commonly used, intuitively sensible practices in qualitative research that are often questioned by disciplinary norms that come from an orthodox statistical framework.

The panel participants will discuss prospects for applying Bayesian reasoning in light of their own experiences of conducting qualitative research and engaging with alternative methodological perspectives. Kenneth Roberts and Margaret Pearson will reflect on how various aspects of Fairfield & Charman’s core guidelines resonate with research they have conducted and how these guidelines might inform their work in progress. Lauren MacLean will address the authors’ perspective on iterative research, in light of her own experience conducting fieldwork and delineating associated practices and principles. Saadia Pekkanen will in turn discuss the authors’ Bayesian guidelines for case selection and how they resonate with her medium-N research (with Margaret Pearson) on the politics of controversy over Chinese infrastructure investment. Stephan Haggard will reflect on his experience of incorporating Bayesian reasoning in his graduate qualitative methods course and will offer some thoughts on Fairfield & Charman's approach vis a vis his work on large-N qualitative research. Finally, Rudra Sil will situate the Bayesian approach with respect to his cautionary perspective on multi-method research and will contemplate which styles of qualitative research can most benefit from it. We hope the panel will stimulate broader discussion and debate about the current status of qualitative research within the discipline and the role that case studies play in advancing knowledge—not only within multi-method designs, but also in their own right.

Sub Unit

• Division 46: Qualitative Methods

Chair

• Andrew Bennett Georgetown University

Presenters

• Kenneth M. Roberts Cornell University

• Margaret M. Pearson University of Maryland

• Lauren M. MacLean Indiana University, Bloomington

• Saadia M. Pekkanen University of Washington

• Stephan Haggard University of California, San Diego

• Tasha A. Fairfield London School of Economics