Replicate Methods for Variance Estimation
Replicate methods for variance estimation are commonly used in large sample surveys with many variables. The procedure uses estimators computed on subsets of the sample, where subsets are selected in a way that reflects the sampling variability. Replication variance estimation is an appealing alternative to Taylor linearization variance estimation for nonlinear functions. Replicate methods have the advantage of transferring the complexity of variance estimation from data set end users to the statistician working on creating the output data set. By providing weights for each subset of the sample, called replication weights, end users can estimate the variance of a large variety of nonlinear estimators using standard weighted sums. Jackknife, balanced half-samples, and bootstrap methods are three main replication variance methods used in sample surveys. The ...
Looks like you do not have access to this content.
Reader's Guide
Ethical Issues In Survey Research
Measurement - Interviewer
Measurement - Mode
Measurement - Questionnaire
Measurement - Respondent
Measurement - Miscellaneous
Nonresponse - Item-Level
Nonresponse - Outcome Codes And Rates
Nonresponse - Unit-Level
Operations - General
Operations - In-Person Surveys
Operations - Interviewer-Administered Surveys
Operations - Mall Surveys
Operations - Telephone Surveys
Political And Election Polling
Public Opinion
Sampling, Coverage, And Weighting
Survey Industry
Survey Statistics
- All
- A
- B
- C
- D
- E
- F
- G
- H
- I
- J
- K
- L
- M
- N
- O
- P
- Q
- R
- S
- T
- U
- V
- W
- X
- Y
- Z