Dealing with spatial heterogeneity in entrepreneurship research

In quantitative research, analyses are generally made using a geographically defined population as the study area. In this context, the relationships between predictor and response variables can differ within the study area, a feature that is known as spatial heterogeneity. Without analyzing spatial heterogeneity, a global model may not be correct, and there may be unclear spatial boundaries in the generalizability of the findings. The authors discuss how the method of geographically weighted regression (GWR) can be used to identify the study area, and illustrate the utility of GWR for empirical analyses in entrepreneurship research. Future entrepreneurship research can benefit from analyzing whether conflicting evidence may be due to spatial heterogeneity and from applying GWR in an exploratory way.