3.44 Metropolitan doctors. In Example 1.1, we considered a simple linear model to predict the number of doctors (MMDs) from the number of hospitals( Hospitals) in a metropolitan area. In that example, we found that a square root transformation on the response variable, Sqr*MMDsproduced a more linear relationship. In this exercise, use this transformed variable, in County Health, as the response variable. In cryHuh a. Either the number of hospitals (Hospital) or number of beds in those hospitals ( Beds) might be good predictors of the number of doctors in a city. Find the correlations between each pair of the three variables, SquMMDS, Hospitals, Beds. Based on these correlations, which of the two predictors would be a more effective predictor of SqrdMD's in a simple linear model by itself? b. How much of the variability in the SqrowDs values is 412 explained by Hospitals alone? How much by Beds alone? c. How much of the variability in the SqrENDs values is explained by using a two-predictor multiple regression model with both Hospitals and Beds? d Based on the two separate simple linear models ( or the individual correlations), which of Hospitals and/or Beds have significant relationship(s) with SquMDs? e. Which of these two predictors are important in the multiple regression model? Explain what you use to make this judgment