10_unequalprob

# 10_unequalprob - Sampling with Unequal Probabilities...

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: Sampling with Unequal Probabilities Lecture 10 STAT 651: Survey Sampling Methods, Kaizar – p.1/8 Lecture 10: Sampling with Unequal Probabilities Reading: Lohr Chapter 6, sections 1 - 6 Connections to Standard Sampling Plans Horvitz-Thompson Estimation STAT 651: Survey Sampling Methods, Kaizar – p.2/8 Connections to Standard Sampling Plans STAT 651: Survey Sampling Methods, Kaizar – p.3/8 Example: Sample Clusters Proportional to Size STAT 651: Survey Sampling Methods, Kaizar – p.4/8 Horvitz-Thompson Estimation ˆ tHT = wi yi i∈S ˆ yHT = ¯ wi yi i∈S wi i∈S STAT 651: Survey Sampling Methods, Kaizar – p.5/8 Variance of HT Estimators One-Stage Cluster Sample: N ¯ V [tHT ] = i=1 1 = 2 1 − πi 2 ti + πi N N N i=1 k =i,k =1 N (πi πk − πik ) i=1 k =i,k =1 ˆ [tHT ] = 1 V¯ 2 (πi πk − πik ) i∈S k =i,k ∈S πik − πi πk ti tk πi πk ti ti − πi πk ti ti − πi πk 2 2 STAT 651: Survey Sampling Methods, Kaizar – p.6/8 Variance of HT Estimators Two-Stage Cluster Sample: N ¯ V [tHT ] = i=1 1 = 2 ˆ [tHT ] = 1 V¯ 2 1 − πi 2 ti + πi N N N i=1 k =i,k =1 N (πi πk − πik ) i=1 k =i,k =1 (πi πk − πik ) i∈S k =i,k ∈S πik − πi πk ti tk + πi πk ti ti − πi πk i=1 N 2 ti ti − πi πk N 2 + i=1 + i∈S ˆ V ti πi ˆ V ti πi ˆˆ V ti πi STAT 651: Survey Sampling Methods, Kaizar – p.7/8 Example: Random Digit Dialing STAT 651: Survey Sampling Methods, Kaizar – p.8/8 ...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online