A methodological development repository for abductive process tracing with Bayesian inference
This site documents the design process of an evidence database for abductive process tracing with Inference to the Best Explanation (IBE), following the Bayesian framework of Fairfield & Charman (2022). The substantive case is a reform process in Brazilian tertiary education; the methodological contribution is the schema itself.
Este site documenta o processo de design de um banco de evidências para process tracing abdutivo com Inferência à Melhor Explicação (IBE), seguindo o arcabouço bayesiano de Fairfield & Charman (2022). O caso substantivo é um processo de reforma da educação superior brasileira; a contribuição metodológica é o esquema em si.
The available theoretical explanations for the outcome under study are individually insufficient, underspecified, or misspecified. Under these conditions, no single theory can be “tested” in the classical sense. The appropriate response is abductive inference: reasoning from systematically organized evidence to the best available explanation.
This repository records the process of thinking through how to build such a database — the design decisions, the dead ends, the methodological reversals, and the lessons learned across four LLM-assisted conversations between May 2 and May 26, 2026.
| Page | What you’ll find |
|---|---|
| Methodology | The epistemological premises guiding database design: Bayesian inference, verbal probability scales, decibel weight of evidence, symmetry of hypotheses |
| Database Manual | Complete technical documentation of the v7 working spreadsheet — every table, column, formula, and convention |
| Decision Log | Structured changelog of design decisions: what changed, why, and which files were affected |
| Narrative History | The full story of how the database was conceived, broken, rebuilt, and broken again — a research memo, not documentation |
| Repository Plan | Why the repository is organized the way it is |
| Formulas | Google Sheets formulas for auto-populating the analysis table |
The project sits at the intersection of four ideas:
Fairfield, T., & Charman, A. E. (2022). Social Inquiry and Bayesian Inference: Rethinking Qualitative Research. Cambridge University Press.
This database design is part of a FAPESP-funded MA thesis on the political economy of tertiary education in Brazil, supervised by Prof. Marta Arretche (USP/CEM).