Consider the cascade process data in Table 11.5. In fitting regression models to both y1 and y2 you will find that not all of the process variables are required to obtain a satisfactory regression model for the output variables. Remove the nonsignificant variables from these equations and obtain subset regression models for both y1 and y2. Then construct individuals control charts for both sets of residuals. Compare them to the residual control charts in the text (Fig. 11.11) and from Exercise 11.18. Are there any substantial differences between the charts from the two different approaches to fitting the regression models?