This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Economics 120B Name: _________________________ Professor Yongil Jeon March 18, 2005 Student ID#: _________________________ Answer to Exam 3: Econometrics 1. (5 points) If an additional independent variable, however irrelevant, is added to a multiple regression model, a smaller sum of squared errors will result. Explain why this is so, and discuss the consequences for the interpretation of R 2 . Answer Suppose we estimate two models, where the second model includes more explanatory variables. Then the following holds: SST=SSR 1 +SSE 1 = SSR 2 +SSE 2 . Since the second regression has more explanatory variables SSR 1 ≥ SSR 2 , which implies that SSE 1 ≤ SSE 2 . As a consequence, R 2 will only provide a meaningful measure of the explanatory power of a regression if the coefficients for the explanatory variables are significantly different from zero. 2. (5 points) A dependent variable is regressed on two independent variables. Is it possible that the hypotheses : 1 = β H and : 2 = β H cannot be rejected at low significance levels, yet the hypothesis : 2 1 = = β β H can be rejected at a very low significance level. In what circumstances might this result arise? Answer This occurs when the explanatory variables 1 x and 2 x are highly correlated. 2 EXAM #3, ECON 120B, Winter 2005 2 3. (40 points) Based on 107 students’ scores on the first examination in a course on managerial cost accounting, the following model was estimated by least squares (Note that the standard errors are in parentheses) y = 2.178 + 0.469 x 1 +3.369 x+3....
View Full Document
This note was uploaded on 04/23/2008 for the course ECON 120B taught by Professor Jeon during the Winter '08 term at UCSD.
- Winter '08