i
Distribution and properties of variance estimators for complex multistage probability samples; an empirical distribution
-
March 1975
-
Details:
-
Personal Author:
-
Corporate Authors:
-
Description:This report presents results from an empirical investigation of the behavior of the replication and Iinearization techniques for measuring variance. The study utilizes data collected in the Health Interview Survey. Most scientific sample surveys are based on complex multistage probability designs, with design components that include unequal probabilities of selection of elements in the population, stratification, and several stages of clustering. The estimation procedures usually involve nonresponse and poststratification adjustments. These actions cause concern in the methods of estimation of standard errors and their subsequent use in constructing confidence intervals and testing hypotheses. There are several ways to study the characteristics of variance estimates. The method in this report is an empirical investigation. Data that were collected in a national sample survey become the universe and repeated samples are drawn from it. For each sample the statistics under study were calculated and sampling distributions of the estimates were generated. The material in this report was taken, in part, from Dr. Bean’s doctoral dissertation.
-
Content Notes:[by Judy A. Bean]
Includes bibliographical references.
-
Subjects:
-
Series:
-
Document Type:
-
Place as Subject:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type: