Question: Sketch a plot of the standardized residuals graphed against Y‾ that would produce a value of SPREAD-RATIO that is very close to 1, and values for rH1 and rH2 that far exceed 0.1. Then produce a sketch where rH1 is approximately 0 and SPREAD-RATIO is much greater than 1. What does this suggest about the need to use more than one numerical measure of heteroscedasticity?