Answer Happy

Judge whether below statements are right or not by writing down T (true) or F (false) on your answer sheet. 1.1) To examine the adequacy of the model, we usually detect the underlying model assumptions by examination of the standard summary statistics, such as t or F statistics, or R2 [1 mark] 1.2) The normal probability plot below indicates the distribution of externally studentized residuals is associated with negative skew. [1 mark] Probability 05 1.3) Plotting the residuals against the corresponding values of each regressor variable is helpful in residual analysis. However, in the simple linear regression case, it is not necessary to plot residuals versus both fitted value and the regressor variable. The reason is that the fitted values are linear combinations of the regressor values, so the plots would only differ in the scale for the abscissa. [1 mark] 1.4) The ridge estimator of the regression coefficients in a multiple linear regression model is a nonlin- ear transformation of the least-square estimator. [1 mark] 1.5) For a qualitative variable with m levels involved in a multiple linear regression, one can represent it by m - 1 indicator variables. [1 mark] 1.6) To deal with the multiple linear regression model with qualitative predictor variables, we will use indicator variables or dummy variables. One of the advantages of using indicator variables is that we can use the extra-sum-of-square method to conduct hypothesis tests directly. For example, we can directly use the extra-sum-of-square method to test whether or not the regression lines have a common slope but possibly different intercepts. [1 mark] 1.7) Akaike information criterion places a greater penalty on adding regressors as the sample size increase than Bayesian information criterion. [1 mark] 1.8) Stepwise method in variable selection guarantees that the best subset regression model of any size will be identified. [1 mark] 1.9) Once a regressor has been added, it can not be removed at a later step in stepwise method for variable selection [1 mark] 1.10) The deviance residuals and Pearson residuals are commonly used for conducting model adequacy checking in logistic regression. Tools in graphical analysis of residuals for multiple linear regres- sion such as normal Q-Q plots are still useful in checking the fit of the model at individual data points and in checking for possible outliers for logistic regression model. [1 mark]