Homework Five

  1. Data is in angell.data. Construct a dummy regressor to represent the East (i.e. 1 if region is "E" and 0 otherwise -- use the command in.east <- ifelse(region=='E',1,0) to do this, though you can change the name in.east to something more snappy if you like) and construct appropriate models with hetero as the X-variable and moral as the Y-variable. Determine whether the effect of hetero on moral is the same for the East as for other regions, and if so, whether there is, in general, a different level of moral for a specified level of hetero in the East when compared to the rest of the country.
  2. Data is in angell.data. Construct enough dummy regressors to represent each of the four regions, East, Midwest, West, and South; regress moral integraion on ethnic heterogeneity, geographic mobility, and region. Use an incremental F test to test the null hypothesis of no region effects. Write out the regression equation for each region.
  3. Continuing the analysis of Angell's data on moral integraion of U.S. cities, include interactions between region and each of the other two independent variables. Write out the fitted regression equation for each of the four regions. Test the statistical significance of each of these interactions with an incremental F test. Then fit a separate regression of moral integraion on heterogeneity and mobility for the 14 southern cities, and confirm that the fitted equation for the South is the same as that obtained from the combined regression.