Part I Instructions: Suppose we are interested in studying the per capita educational expenses (in dollars) of the 50 US states and Washington, D.C. We know that these expenses are likely related to the per capita income of a state. Using the Anscombe dataframe contained in the carData package, and assuming the educational expenses are being predicted by income, please do the following: Fit a simple linear regression model. Test and report the assumptions of normality and constant variance using both a graphical and testing method. Are the assumptions (including approximate linearity) reasonably met? If not, perform a Box-Cox transformation and re-evaluate the assumptions (HINT: remember, the tests are very sensitive to even small departures from normality and constant variance. Take both pieces of information into consideration before deciding to transform the data)..