SOLN_3b

# 1 2 so we must have e 1 e 1 e 1

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: β1 Þ This meaningless prediction results from the linearity of the regression, which can give negative results for small sales values. ˆ (1) ˆ ( 2) ˆ 5. (i) β 1 , β 1 and β 1 are all least-squares estimates based on samples from the same linear model: Yi = β0 + β1 xi + εi . ˆ (1) ˆ ( 2) ˆ So we must have E ( β 1 ) = E ( β 1 ) = E ( β 1 ) = β 1 . Also (1) ~ E ( β1 ) = n1 n2 ˆ (1) ˆ ( 2) E(β1 ) + E ( β1 ) = β1 n1 + n2 n1 + n2 ˆ (1) ˆ (2) By definition, β1 and β1 are linear estimators, so (ii) (2)  β1 = n1 n1 + n2 where w1i = (å n1 i =1 ) w1iYi + n2 n1 + n2 (å xi − x1 and w2 j = å k ( xk − x1 )2 n2...
View Full Document

## This homework help was uploaded on 02/16/2014 for the course ESE 302 taught by Professor Ese302 during the Fall '08 term at UPenn.

Ask a homework question - tutors are online