Incorporating Space in Multimethod Research: Combining Spatial Analysis with Case-Study Research
by Imke Harbers, University of Amsterdam and Matthew C. Ingram, University at Albany, SUNY
Most outcomes that social scientists care about are not distributed randomly across geographic space. Similar units are often located near one another so that phenomena of interest tend to cluster. This clustering is no accident and should matter in how scholars understand and explain these outcomes. Nevertheless, the spatial dimensions of political data rarely receive explicit attention in how multimethod scholars design research. This article outlines why integrating insights from spatial analysis and multimethod research can lead to stronger conclusions than using either approach in isolation. Specifically, we argue that (1) without integrating insights from spatial analysis, multimethod designs can be self-defeating because one method may undermine the logic of another; and (2) without integrating insights from multimethod research, spatial statistics and econometrics can fall short by assuming rather than demonstrating both (a) the mechanisms underlying key spatial processes and (b) the proper unit or level of analysis. Incorporating the spatial nature of data can enrich multimethod research by providing a new set of geographic tools to analyze data. This article shows how basic, exploratory, univariate diagnostics of spatial autocorrelation can help scholars to (1) understand the boundedness of phenomena, and (2) select interesting cases for further analysis.