Overview
For this project, you will take the result of your OLS model from Project 2 further to assess and improve the models. Following the steps below: 1. Based on your resulting OLS model in Project 2, run the residual diagnosis by testing for the model assumptions of normality, equal variance and spatial dependence for the regression residuals. The diagnosis on regression residuals should at least include a residual plot (residuals vs. fitted values of the response variable), QQ plot, and Moran’s I on the regression residuals. You may also opt-in for other statistics such as Jacque-Bera test (for normality) and Breusch-Pagan test (for homoscedasticity). Reference to the previous handouts in R for the residual analyses that we have done. The only test we have not practiced here is the Jacque-Bera test (optional here); the function is available in the tsereis library as jacque.bera.test( ). 2. Generate a spatial lag model (SLM) AND a spatial error model (SEM) for your dataset. The variable selection and the spatial weights matrix should be based on the results of your multiple linear regression (OLS) in Project 2.