Optimal Design of Interviewer Variance Experiments in Complex Surveys (Q3692602)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 3917431
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Optimal Design of Interviewer Variance Experiments in Complex Surveys |
scientific article; zbMATH DE number 3917431 |
Statements
1985
0 references
linear models approach
0 references
response variance
0 references
survey nonsampling error
0 references
survey evaluation
0 references
measurement error
0 references
interviewer variance estimation
0 references
multistage sampling designs
0 references
decomposition
0 references
interpenetrations
0 references
linear combinations
0 references
simple random sampling
0 references
Optimality
0 references
combined estimator
0 references
Optimal Design of Interviewer Variance Experiments in Complex Surveys (English)
0 references
0.7418471574783325
0 references
0.7376847267150879
0 references
0.7361465692520142
0 references
The paper is based on the first author's unpublished Ph. D. dissertation and develops some new knowledge on interviewer variance estimation. The estimates and their properties are studied for general multistage sampling designs, the main idea being a decomposition of the next-to- last-stage units into two blocks, one of them used for interpenetrating interviewer assignment and the other for standard assignment. The method allows for a substantial generalization of the usual concept of equally large interpenetrations. Provided normality, two independent estimates of the interviewer variance are studied and some advantages of their linear combinations are established. A special paragraph deals with the most elementary case - simple random sampling within areas. Optimality of the combined estimator which is more efficient than any of the independent ones is discussed both analytically and empirically, but no way of construction of an optimal combination is offered.
0 references