Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 41 of the 41 references it cites were found in a registry or an index; none of the 16 sources its claims cite is checked for support yet.

A Dialogue with the Data: The Bayesian Foundations of Iterative Research in Qualitative Social Science

Open access
Fairfield, Tasha, Charman, Andrew · 2019 · Perspectives on Politics

Abstract

We advance efforts to explicate and improve inference in qualitative research that iterates between theory development, data collection, and data analysis, rather than proceeding linearly from hypothesizing to testing. We draw on the school of Bayesian “probability as extended logic,” where probabilities represent rational degrees of belief in propositions given limited information, to provide a solid foundation for iterative research that has been lacking in the qualitative methods literature. We argue that mechanisms for distinguishing exploratory from confirmatory stages of analysis that have been suggested in the context of APSA’s DA-RT transparency initiative are unnecessary for qualitative research that is guided by logical Bayesianism, because new evidence has no special status relative to old evidence for testing hypotheses within this inferential framework. Bayesian probability not only fits naturally with how we intuitively move back and forth between theory and data, but also provides a framework for rational reasoning that mitigates confirmation bias and ad-hoc hypothesizing—two common problems associated with iterative research. Moreover, logical Bayesianism facilitates scrutiny of findings by the academic community for signs of sloppy or motivated reasoning. We illustrate these points with an application to recent research on state building.

Overview

CoChat Summary

TL;DR

This article argues that logical Bayesianism provides the missing methodological foundation for iterative qualitative research, because prior/posterior probabilities are logical rather than temporal and “new” evidence has no special status over “old” evidence. It concludes that separating exploratory from confirmatory research, and tools such as time-stamping or preregistration, are unnecessary for qualitative inference guided by Bayesian logic.

Key findings

  • Within logical Bayesianism, prior and posterior probabilities refer to logical states of knowledge, not to when evidence was learned; evidence can be incorporated in any order without changing posterior probabilities.
  • The paper argues that time-stamping and preregistration are not useful tools in qualitative research, because they do little to deter fraud and may burden honest scholars; it notes that the retracted LaCour-Green study was pre-registered.
  • Bayesian safeguards are claimed to mitigate two pitfalls: likelihood ratios force comparison with rival hypotheses and help guard against confirmation bias, while priors and Occam penalties help guard against ad-hoc hypothesizing.
  • In the illustrative state-building example, E1 moderately favors the resource-curse hypothesis over the labor-repressive agriculture hypothesis and strongly favors labor-repressive agriculture over the warfare hypothesis; E2 very strongly favors labor-repressive agriculture over each alternative, and the overall likelihood ratio strongly favors labor-repressive agriculture over both alternatives.

Methods

  • Conceptual and methodological article: it reviews frequentist versus Bayesian approaches, introduces logical Bayesianism, and derives implications for iterative qualitative research.
  • Uses a heuristic Bayesian application to Marcus Kurtz’s published state-building research on Peru, comparing hypotheses HR (resource curse), HW (warfare), and HLRA (labor-repressive agriculture) using evidence E1 and E2.
  • The authors state they make no claims about how Kurtz’s research process actually unfolded; the example is illustrative rather than a new empirical test.

Limitations

  • The paper states that logical Bayesianism is an aspirational ideal that usually cannot be fully realized without approximations, that some subjectivity inevitably enters probability assignments, that there is no mechanical procedure for translating qualitative information into precise probabilities, and that analytical errors may still occur.
  • The paper states that systematically incorporating all background information is infeasible in social science, so the illustrative analysis uses simplified prior choices such as equal odds or moderate penalties.
  • Evident from the text: the state-building application is not an empirical test and does not establish how the original research proceeded; the heuristic example does not assign numerical probabilities.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1017/s1537592718002177

Cites

41 of 41 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 Valid · 2 Not valid · 13 Not checked · 129 Own results and interpretations

54 more claims

Create a free account to see all 144 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 65
  • It may state two separate claims as one. 45

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

Show all 39 references

Research works printed but not linked: 6

  1. Monogan, 2015, “Research Preregistration in Political Science.”, PS: Political Science and Politics, 48, 425
  2. Jefferys, 2003, “Bayes’ Theorem,”, Journal of Scientific Exploration, 17, 537
  3. Hunter, 1984, Political/Military Applications of Bayesian Analysis
  4. Büthe, 2015, “Conclusion: Research Transparency for a Diverse Discipline.”, Qualitative and Multimethod Research: Newsletter of the American Political Science Association’s QMMR Section, 13, 50
  5. Bennett, 2015, Process Tracing in the Social Sciences: From Metaphor to Analytic Tool
  6. Jacobs, 2019, The Production of Knowledge

Cited by

See all 79

Metadata

Authors
Fairfield, T., & Charman, A.
Affiliations
  • London School of Economics and Political Science (GB)
  • University of California, Berkeley (US)
Corresponding author
Fairfield, Tasha
Venue
Perspectives on Politics
Published
2019-02-13
Volume
17
Issue
1
Pages
154-167
ISSN
  • 1537-5927 (ISSN-L)
  • 1541-0986
OpenAlex
W2970085226
Cited by
79
References
39
Citations by year
2026: 20 · 2025: 6 · 2024: 7 · 2023: 5 · 2022: 5 · 2021: 8 · 2020: 25 · 2019: 3
FWCI
27.63
Type
journal article
Language
en
Open access
Locations
  • Perspectives on Politics (publishedVersion)
  • London School of Economics and Political Science Research Online (London School of Economics and Political Science) (acceptedVersion)
  • London School of Economics and Political Science Theses Online (London School of Economics and Political Science) (acceptedVersion)
  • LSE Research Online Documents on Economics (submittedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
Topics
  • Social Sciences › Education › Qualitative Research Methods and Applications
  • Social Sciences › Sociology and Political Science › Qualitative Comparative Analysis Research
  • Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
Keywords
Bayesian inference, iterative research, qualitative research methodology, theory development, confirmation bias, motivated reasoning, state building, research transparency

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-09-29, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-09-29 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: A Dialogue with the Data: The Bayesian Foundations of Iterative Research in Qualitative Social Science. https://app.cochat.ai/papers/10.1017/s1537592718002177
The paper
Fairfield, T., & Charman, A. (2019). A Dialogue with the Data: The Bayesian Foundations of Iterative Research in Qualitative Social Science. Perspectives on Politics. https://doi.org/10.1017/s1537592718002177

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.