Handout4-W09

Handout4-W09 - -y j j ! 2 = 1 2 X i s X j s i j- ij ij y i...

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

View Full Document Right Arrow Icon
STAT 454/854 Winter 2009 Handout #4 Unequal Probability Sampling and the HT Estimator Let π i and π ij be respectively the first order and the second order inclusion probabilities under a general sampling design. The Horvitz-Thompson (HT) estimator for the population total t y = N i =1 y i is defined as ˆ t HT = X i s y i π i = X i s d i y i , where d i = 1 i are often referred to as the basic design weights. (1) The HT estimator is unbiased, i.e. E ( ˆ t HT ) = t y . (2) The theoretical variance of ˆ t HT is given by V ( ˆ t HT ) = N X i =1 1 - π i π i y 2 i + N X i =1 N X j 6 = i ( π ij - π i π j ) y i π i y j π j = N X i =1 N X j =1 ( π ij - π i π j ) y i π i y j π j . Here we used π ii = π i in the last expression. (3) An unbiased variance estimator is given by v ( ˆ t HT ) = X i s 1 - π i π 2 i y 2 i + X i s X j 6 = i π ij - π i π j π ij y i π i y j π j = X i s X j s π ij - π i π j π ij y i π i y j π j . (4) The Yates-Grundy-Sen variance formula when the sample size n is fixed. V ( ˆ t HT ) = N - 1 X i =1 N X j = i +1 ( π i π j - π ij ) y i π i - y j π j ! 2 = 1 2 N X i =1 N X j =1 ( π i π j - π ij ) y i π i - y j π j ! 2 v ( ˆ t HT ) = X i s X j>i π i π j - π ij π ij y i π i
Background image of page 1
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: -y j j ! 2 = 1 2 X i s X j s i j- ij ij y i i-y j j ! 2 (5) Techniques useful for proofs: i) Let a i = 1 if i s and 0 otherwise, i = 1 , 2 , ,N . Then E ( a i ) = i , E ( a i a j ) = ij for i 6 = j , V ( a i ) = i (1- i ), and Cov ( a i ,a j ) = ij- i j . ii) To nd out the expectation (or variance) of A = i s Z i and B = i s j s Z i Z j , we re-write A and B as A = N X i =1 a i Z i and B = N X i =1 N X j =1 a i a j Z i Z j . iii) When the sample size n is xed, we have N X i =1 i = n ; N X j 6 = i ij = ( n-1) i and N X i =1 N X j 6 = i ij = n ( n-1) ....
View Full Document

Ask a homework question - tutors are online