Parameter Estimation of Spatial Durbin Model (SDM) Using Method of Moment

摘要

Regression analysis is an analysis that aims to show the mathematical relationship between the dependent variable and the independent variable. The spatial regression model is a model that analyzes the relationship between one variable with several other variables including the spatial effect of several locations that are the centre of observation. Spatial Autoregressive Model (SAR) is a model where the dependent variable is influenced by the value of the adjacent dependent variable that is defined accordingly. In its development, the SAR model is not able to overcome the problem if there are spatial interactions in the dependent variable and the independent variable. For this reason, Spatial Durbin Model (SDM) appears. In this article, SDM parameter estimation will be discussed using the Method of Moment approach. The basic principle in the moment method is to choose parameter estimates that are related to sample moments which are equal to zero. The estimation results with this method provide a very reliable estimator, which is an unbiased and consistent estimator for β.