Assignment: Multivariate Regression In this assignment, students will build on the bivariate analysis by conducting multivariate analysis using OLS regression. Dataset: states_REG Dependent Variable: Cigarette consumption (cigarettes) Using the cigarette consumption variable as your dependent variable and the cigarette tax (cig_tax) as your key explanatory variable, do the following: 1. Run a bivariate regression between cigarette consumption (cigarettes) and cigarette taxes (cig_tax). [NOTE: This should be the same model complete in the Bivariate Statistics Assignment]. 2. Run a multivariate regression model with the following control variables (include the results in your homework). NOTE: For the multivariate regression, you should include these variables in the model as control in addition to the cig_tax variable from the bivariate model a. Percent of population with college degree (college) b. Percent of population age 18-24 (pop_18_24) c. Per capita income (prcapinc) d. Percent urban population (urban) 3. Rerun the model, but this time, also control for Southern states (south). NOTE: this variable should be added to the model above (i.e. you should have a total of 6 IVs in this model). Answer the following questions: 1. Is the hypothesis that higher cigarette tax rates are associated with lower consumption supported? (provide an answer for all three models models). NOTE: Discuss both substantive and statistical significance. 2. What variables are statistically significant in the third model? 3. Based on the third model, if taxes in a state were to increase by $0.50, what would the expected impact be on cigarette consumption? 4. How well do these models explain variation in cigarette consumption? (i.e. R2) 5. Which model do you consider to be the best model? Why?