INSTRUCTIONS: Use a 5% level of significance. A hospital administrator is studying the relationship between patient satisfaction with patient's gender, age in years, severity of illness and anxiety level. She randomly selected 23 patients and collected the data presented below, where larger values are associated with more satisfaction, increased severity of illness, and more anxiety, respectively. Suppose further that patient 6, 8, 9, 11, 12, 17, 20, 22 are males (i.e., Gender = 0). Assume that the multiple linear regression model with normal random error terms is appropriate. Severity of Anxiety Patient Age Illness Level Satisfaction 51 2.3 48 46 2.3 57 48 2.2 66 4 41 44 1.8 70 5 28 43 1.8 89 6 49 54 2.9 7 42 50 45 48 52 62 50 48 53 Patient I r r r 1 50 2 36 3 40 ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 8 9 10 11 12 29 29 43 24 2.2 2.4 2.9 2.1 2.4 2.4 36 46 54 26 77 89 67 Patient Age 13 38 14 34 15 53 16 36 17 18 29 19 33 20 55 21 29 22 44 23 43 a. Forward Stepwise Procedure b. Backward Stepwise Procedure PP 33 Severity of Anxiety Patient Illness Level Satisfaction 55 2.2 47 51 2.3 51 54 2.2 57 49 2 66 56 2.5 79 46 1.9 88 49 51 52 Are there outliers in the given data? Justify. Are there influential points in the given data? Justify. Independent Variables in Model 58 50 Find the estimated regression (i.e., full) model. Interpret the coefficient for gender. N - Is multicollinearity present among the variables in the given model? Support your answer by calculating relevant statistic(s). Compare the following models by filling-up the table below: Model Adjusted R 2.1 2.4 2.3 2.9 2.3 Determine which among patient's gender, age in years, severity of illness and anxiety level are significant predictors for patient satisfaction using Full Using Forward Stepwise Using Backward Stepwise Which is the best model in predicting patient satisfaction? Why? 60 49 77 52 60 MSE