Integrated Inferences: Causal Models for Qualitative and Mixed-Method Research(Strategies for Social Inquiry)

There is a growing consensus in the social sciences on the virtues of research strategies that combine quantitative with qualitative tools of inference. Integrated Inferences develops a framework for using causal models and Bayesian updating for qualitative and mixed-methods research. By making, updating, and querying causal models, researchers are able to integrate information from different data sources while connecting theory and empirics in a far more systematic and transparent manner than standard qualitative and quantitative approaches allow. This book provides an introduction to fundamental principles of causal inference and Bayesian updating and shows how these tools can be used to implement and justify inferences using within-case (process tracing) evidence, correlational patterns across many cases, or a mix of the two. The authors also demonstrate how causal models can guide research design, informing choices about which cases, observations, and mixes of methods will be most useful for addressing any given question.

Table of Contents:
1. Introduction; I. Foundations: 2. Causal models; 3. Illustrating causal models; 4. Causal queries; 5. Bayesian answers; 6. Theories as causal models; II. Model-based causal inference: 7. Process tracing with causal models; 8. Process tracing applications; 9. Integrated inferences; 10. Integrated inferences applications; 11. Mixing models; III. Design choices: 12. Clue selection as a decision problem; 13. Case selection; 14. Going wide, going deep; IV. Models in question: 15. Justifying models; 16. Evaluating models; 17. Final words; V. Appendices: 18. Causal Queries; 19. Glossary; Bibliography; Index.

About the Author :
Macartan Humphreys is Professor of Political Science at Columbia University and Director of the Institutions and Political Inequality group at the WZB Berlin, conducting research on post-conflict development, ethnic politics, and democratic decision-making. He has been President of the APSA Experimental Political Science section and Executive Director of the Evidence on Governance and Politics network. Alan M. Jacobs is Professor of Political Science at the University of British Columbia, conducting research on comparative political economy in democratic settings. He has been President of the APSA's Qualitative and Multi-Method Research section, winner of the section's Mid-Career Achievement Award, and a regular instructor at the Institute for Qualitative and Multi-Method Research.

Review :
'An ambitious attempt to leverage the strengths of qualitative and quantitative social scientific approaches by embedding them within a Bayesian framework, this book will give economists, political scientists, and other researchers a lot to chew on for years to come.' Andrew Gelman, Columbia University 'This study will become required reading for researchers seeking to answer causal questions in the social sciences. Humphreys and Jacobs provide a framework that helps researchers assess the relative contribution of different pieces of evidence - either qualitative or quantitative - in answering causal questions.' Isabela Mares, Yale University 'Many books aspire to be useful to both qualitative and quantitative researchers, but few succeed as well as this novel approach to integrating case-level and population-level evidence to make causal inferences. Starting from a focus on causal models, Integrated Inferences explains how to use Bayesian logic to update models based on evidence, and how to select cases and collect data to further probe those models on particular causal questions. In short, it turns the model-dependence of inferences from a bug into a feature. An extremely valuable addition to graduate research methods courses.' Andrew Bennett, Georgetown University 'With this illuminating book, Humphreys and Jacobs expand their pioneering approach to multi-method research, combining the Bayesian foundations of Lindley and Savage, the scaffolding of Neyman, Rubin, and Holland's causal reasoning, and the flexible architecture of Pearl and Glymour's DAG models. Social science scholars will find an engaging and accessible exposition and synthesis of ideas that are otherwise scattered across a sometimes daunting literature in statistics, computer science, philosophy, and political science.' Tasha Fairfield, London School of Economics and Andrew Charman, University of California, Berkeley