Political methods
Answer the following question in a single file that you upload. Show work when necessary.
I estimated a bivariate regression model using the data from each U.S. state, with percentage of women in the state legislature as the dependent variable (womleg), and percentage of state residents with a college degrees (college) as the independent variable. The estimated equation was:
LaTeX: womleg_i=-1.002+0.938collegew o m l e g i = - 1.002 + 0.938 c o l l e g e
a) interpret the parameter estimate for the effect of a state’s level of education on average percentage of women in the state legislature.
b) In the dataset described above, the value of college for Kansas equals 28.7. What is the model’s predicted percentage of women in the state legislature for Kansas?
c) The actual observed percentage of women in the state legislature is 29.1%. What can we say about ui for this individual case? What does ui represent in the regression model?
d) The estimated standard error for the slope parameter in the model described above was 0.19. Test the hypothesis that the parameter for college is greater than zero. Show your work. What does this tell you about the estimated relationship?
e) The R-squared statistic for the model is .34 and the root MSE=6.03. What do these numbers tell us about the model? Interpret both statistics.