Dr. Hackney STA Solutions pg 105

# Dr. Hackney STA Solutions pg 105 - 7-8 Solutions Manual for...

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: 7-8 Solutions Manual for Statistical Inference 7.19 a. L(|y) = i 1 2 2 exp - 1 (y - xi )2 2 2 i 1 2 2 2 = (2 2 )-n/2 exp - (2 2 )-n/2 exp - Yi2 , (y i -2xi yi + 2 x2 ) i i 2 = 2 i xi 2 2 exp - 1 2 2 2 yi + i 2 xi yi i . By Theorem 6.1.2, ( b. i i xi Yi ) is a sufficient statistic for (, 2 ). 2 2 2 2 logL(, 2 |y) = - For a fixed value of 2 , n 1 n log(2) - log 2 - 2 2 2 2 2 yi + xi yi - i x2 . i i logL 1 = 2 Also, xi yi - i 2 set ^ x2 = 0 = i i i xi yi 2 . i xi 2 logL 1 = 2 2 x2 < 0, i i ^ ^ so it is a maximum. Because does not depend on 2 , it is the MLE. And is unbiased because xi E Yi xi xi ^ E = i = i = . 2 2 i xi i xi ^ c. = i ai Yi , where ai = xi / tributed with mean , and ^ Var = i j ^ x2 are constants. By Corollary 4.6.10, is normally disj 2 a2 Var Yi i = i xi 2 j xj 2 = ( x2 2 i = x2 )2 j j i 2 2. i xi 7.20 a. E b. Var Because i i Yi = i xi 1 i xi i E Yi = 1 i xi xi = . i Yi i xi = ( 1 2 i xi ) Var Yi = i i ( 2 n 2 2 = 2 2 = . 2 n x n2 x i xi ) i i x2 - n2 = x i i (xi - x)2 0, x2 n2 . Hence, x i i ^ Var = 2 2 = Var 2 n2 x i xi Yi i xi . ^ (In fact, is BLUE (Best Linear Unbiased Estimator of ), as discussed in Section 11.3.2.) ...
View Full Document

Ask a homework question - tutors are online