Spatial diffusion and spatial statistics: revisting Hägerstrand’s study of innovation diffusion.

Abstract

Torsten Hägerstrand’s 1953 study of innovation diffusion [1] was pathbreaking in many ways. It was based on an explicit micro-model of information spread, and on Monte Carlo simulation of the hypothesised spatial process. Using the original aggregated data and Hope-type tests of the ability of the simulations to capture the observed adoptions, (author?) [2] and (author?) [3] and others found problems. This study attempts to examine the extent to which we may be able to "do better" with a range of approaches drawn from spatial statistics, including using a SAR lattice model, geostatistical modelling, Moran eigenvectors, and other approaches.