Luc E. Anselin

I study methods for the analysis of geospatial data. I have developed new statistical and data visualization techniques, disseminated them in the form of software tools and applied them in a range of empirical settings in the environmental and social sciences, dealing with topics such as the effect of environmental quality on house prices, the spatial distribution of homicides, spatial models of international conflict, and the diffusion of university-generated knowledge. Specifically, I laid out a framework for a field of spatial econometrics, stressing the importance of the concepts of spatial dependence (how observations in nearby locations tend to be similar) and spatial heterogeneity (how processes differentiate themselves across space), and how these are reflected in model specification, estimation of parameters, and diagnostic tests for model performance. I also work in exploratory spatial data analysis, by developing new graphical techniques to visualize and explore spatial autocorrelation, especially in conjunction with the technology of geographic information systems. In my earlier work, I introduced the notion of local indicators of spatial autocorrelation, which allow the detection of hot spots and spatial outliers, and which have found wide application in several scientific fields. These ideas are incorporated into software that I developed and disseminated.