Regular linear models are usually applied when a data set consists of observations which are influenced by changes in the appurtenant independent variables. In reality many data sets consist of observations which are influenced by changes in the independent variables of neighbor observations in which case the observations are referred to as spatial dependent. Two econometric models can be applied to spatial dependent and approximately normally distributed observations: The Spatial Autoregressive Model also referred to as the SAR-model and the Conditional Autoregressive Model known as the CAR-model. This paper describes the theory behind each model and gives two methods for estimating the parameters of the models. The first method is maximum likelihood estimation and the second is the Bayesian approach where the latter of the two incorporate prior knowledge of the parameters. Most of the theory is translated into practice using data from the real estate agent HOME.