Economics (i) Use OLS to estimate a model relating colGPA to hsGPA, ACT, skipped, and PC. Report estimates, usual standard errors, and robust standard errors, but not need to write them in an equation. Obtain the OLS residuals for use in the next part. (ii) Conduct a White test for heteroskedasticity. Is there evidence of heteroskedasticity? (iii) Use FGLS to estimate the model of part (i). Report estimates and standard errors. (iv) Regarding statistical significance of skipped and PC ownership, is this sensitive to the estimation procedure used? In other words, is the choice of standard errors important here? (v) What concern might we have with using FGLS to estimate this model? (vi) Test for group heteroskedasticity with respect to PC ownership and report your findings. Hint: Use the regression of log(u? ) on the other variables and do a test of individual significance on PC