This paper argues that spatial analysis tools can enhance mixed-methods research by informing case selection. It introduces two specific methods—spatial autocorrelation in outcomes and residual autocorrelation—to integrate quantitative findings with qualitative insights.

New Research on Case Selection:

Geographic techniques like Local Indicators of Spatial Association (LISA) help identify clusters or patterns not visible through standard approaches, improving the selection of cases for mixed-methods studies. This approach provides a systematic way to use geography in political science research.

Two Key Strategies Highlighted: Using LISA Statistics:

* Spatial Autocorrelation in Outcome Data: Helps pinpoint areas where outcomes are clustered or spatially dependent.

* Spatial Analysis of Regression Residuals: Uncovers geographic patterns missed by traditional regression models, guiding qualitative research focus.

Practical Value for Political Scientists: Why Geographical Patterns Matter:

The LISA methods enable scholars to better understand scope conditions, determine appropriate analysis units (like regions or districts), examine causal mechanisms through space, and uncover previously omitted variables. This improves the overall quality of political science inquiry.