Sergio Rey, San Diego State University, San Diego, USA, Julie Le Gallo, IERSO (IFReDE-GRES) Université Montesquieu-Bordeaux IV , Pessac, France
Properties of Markov Based Tests in the Presence of Spatial Dependence (assigned to theme D3)

This paper extends earlier research on exploratory spatial data analysis (ESDA) methods in the context of regional income dynamics. The linkages between a spatial Markov matrix and the Markov transition matrix for local indicators of spatial association (LISA) are explored. To date these methods have been used independently of one another and in a largely descriptive fashion to study regional evolutions. Here we suggest some alternative inferential frameworks for the summary statistics. We evaluate the sampling behavior of these new metrics across a range of experimental data sets containing different forms and magnitudes of spatial dependence.

Paper not on CD

Conference organized through conf-vienna (copyright Gunther Maier)

<