This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression Classical Linear Regression Model SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression 1. Assumptions A. (CLRM.I) Linear in Parameters : may or may not be linear in the variables. The regression model has an additive error term SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression 1. Assumptions A. (CLRM.I) Linear in Parameters : may or may not be linear in the variables. The regression model has an additive error term B. (CLRM.II) No Endogeneity : All explanatory variables X s are uncorrelated with the stochastic error term. ) ,..., 2 , 1 ( ) , cov( K k X k SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression 1. Assumptions (continued) C. (CLRM.III) E ( ε all X s) = 0. SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression 1. Assumptions (continued) C. (CLRM.III) E ( ε all X s) = 0. D. (CLRM.IV) Homoscedasticity : The variance of each ε is constant. N i i ,..., 2 , 1 ) var( 2 SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression Homoscedasticity Heteroscedasticity SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression 1. Assumptions (continued) C. (CLRM.III) E ( ε all X s) = 0. D. (CLRM.IV) Homoscedasticity : The variance of each ε is constant. E. (CLRM.V) No Autocorrelation (Serial Correlation): The error terms are uncorrelated with each other. N i i ,..., 2 , 1 ) var( 2 j i j i ) , cov( SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression No Autocorrelation Positive Autocorrelation Negative Autocorrelation SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression Topic 17, Page 8 SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression SUNY at Albany  Department of Economics Eco 320 Economic Statistics • Regression 1. Assumptions (continued) C. (CLRM.III) E ( ε all X s) = 0....
View
Full Document
 Spring '11
 Chan
 Economics, Linear Regression, Regression Analysis, Department of Economics, SUNY, Albany, Economic Statistics

Click to edit the document details