Question 1 - Compensation of Senior Executives A large international bank is reviewing its compensation policy for their senior executives. They want to closely examine the relationship between senior executives' salaries and the growth in their business portfolio. The dataset Executive Salaries contains last year's annual compensation of the bank senior executives [in thousands of dollars], including performance pay, and the annual rate of growth of their respective business portfolio [as a percentage]. These variables are called Salary_raw and ROG_raw, respectively. a] b] g] h) Draw a boxplot of the salary of senior executives [variable Salary_raw] and produce a scatter plot of this variable against the annual rate of growth [variable ROG_raw]. Are there any apparent outliers in the data? Are there high leverage points? Use statistical software to estimate the model below and report your results. Salary_rawi = 30 + 131 X ROG_rawi + 5. Create the corresponding residual plot [residuals against predicted values] and normal probability plot of the residuals. Looking at these two plots, does the estimated model appear satisfactory? Use statistical software to estimate a new model, this time by using the adjusted variables Salary_adj and ROG_adj which exclude all the data points for which the salary of the senior executive appears extraordinarily large considering the ROG of their business portfolio. Report your results. Produce a histogram and a normal probability plot of the residuals of your regression model in part (1] above. Does this regression appear to meet the condition of near normality? What are the units of the slope coefcient b1 in this last regression equation? What is the average impact on the salary of senior executives whose ROG increases by 1% point? Calculate and interpret the value of the R2 for this regression. Calculate a 95% confidence interval for the regression slope B1 in part i]. Use your results to calculate a 95% interval to estimate the mean salary of senior executives with an ROG of 20 per cent