Multi-Method Social Science: Combining Qualitative and Quantitative ToolsReflecting the rising popularity of research that combines qualitative and quantitative social science, Multi-Method Social Science provides the first systematic guide to designing multi-method research. It argues that methods can be productively combined using the framework of integrative multi-method research, with one method used to carry out a final causal inference, and methods from other traditions used to test the key assumptions involved in that causal inference. In making this argument, Jason Seawright considers a wide range of statistical tools including regression, matching, and natural experiments. The book also discusses qualitative tools including process tracing, the use of causal process observations, and comparative case study research. Along the way, the text develops over a dozen multi-method designs to test key assumptions about social science causation. |
Contents
Causation as a Shared Standard | 19 |
Using Case Studies to Test and Refine Regressions | 45 |
Case Selection after Regression | 75 |
Combining Case Studies and Matching | 107 |
Combining Case Studies and Natural Experiments | 124 |
Embedding Case Studies within Experiments | 150 |
Other editions - View all
Multi-Method Social Science: Combining Qualitative and Quantitative Tools Jason Seawright Limited preview - 2016 |
Multi-Method Social Science: Combining Qualitative and Quantitative Tools Jason Seawright No preview available - 2016 |
Common terms and phrases
alternative analysis approach argue argument assumptions attention average case-study case-study research causal effect causal inference causal pathways causal process causes central chapter choose combination comparative comparison components condition confounding connected consider context contribute control variables course decision dependent variable discovering discussed economic estimate evidence example existence experimental extreme fact finding Furthermore given goals Hence ideas important included independent integrative interest involves issues kind least less matching measurement error methods multi-method designs multi-method research natural experiments necessary needed observed omitted variables outcome overall path Political possible potential-outcomes framework practice problem process-tracing produce qualitative quantitative question randomization reason receive regarding regression relationship relevant represent requires research design rule scholars scores selection similar simply social sciences sources statistical steps studies takes techniques theory treatment assignment true unusual