The following table contains computer outputs after estimating the following two models for the same 20...

 

Transcribed Image Text:

The following table contains computer outputs after estimating the following two models for the same 20 observations. The first model (Model 1) is y=B+x+B3w+e, while the second model (Model 2) is y = B + Bx+E. B. B, and B3 are unknown parameters and & is the random error term. Coefficients Intercept W Model 1 Model 2 3.636 -5.838 (2.763) (2.000) -0.998 4.107 (1.235) (0.338) 0.498 (0.117) Table 1: Estimation results with standard errors in the parentheses. The Regression Equation Specification Error Test (RESET) suggests augmenting an ex- isting model with the squares of its predictions, or with their squares and cubes. Applying RESET to the second model yields F-values of 17.98 (for 2) and 8.72 (for and ). The correlation between x and w is 0.975. Answer the following questions. [13] (i) Should w be included in the above model? (ii) What can you say about the existence of collinearity in the above model and its possible effect on model results? (iii) What would happen if we use RESET to augment the Model 2 with the predicted value of y, i.e., ? The following table contains computer outputs after estimating the following two models for the same 20 observations. The first model (Model 1) is y=B+x+B3w+e, while the second model (Model 2) is y = B + Bx+E. B. B, and B3 are unknown parameters and & is the random error term. Coefficients Intercept W Model 1 Model 2 3.636 -5.838 (2.763) (2.000) -0.998 4.107 (1.235) (0.338) 0.498 (0.117) Table 1: Estimation results with standard errors in the parentheses. The Regression Equation Specification Error Test (RESET) suggests augmenting an ex- isting model with the squares of its predictions, or with their squares and cubes. Applying RESET to the second model yields F-values of 17.98 (for 2) and 8.72 (for and ). The correlation between x and w is 0.975. Answer the following questions. [13] (i) Should w be included in the above model? (ii) What can you say about the existence of collinearity in the above model and its possible effect on model results? (iii) What would happen if we use RESET to augment the Model 2 with the predicted value of y, i.e., ?