Complex sampling design based inference on familial models for count data
Thesis (M.A.S.)--Memorial University of Newfoundland, 2008. Mathematics and Statistics Includes bibliographical references (leaves 52-53) Consistent and efficient estimation of the parameters of generalized linear mixed models (GLMMs) has proven to be difficult in the infinite population setup. This...
Main Author: | |
---|---|
Other Authors: | |
Format: | Text |
Language: | English |
Published: |
2007
|
Subjects: | |
Online Access: | http://collections.mun.ca/cdm/ref/collection/theses4/id/113902 |
Summary: | Thesis (M.A.S.)--Memorial University of Newfoundland, 2008. Mathematics and Statistics Includes bibliographical references (leaves 52-53) Consistent and efficient estimation of the parameters of generalized linear mixed models (GLMMs) has proven to be difficult in the infinite population setup. This estimation issue becomes more complex in the infinite population setup where the estimation is done based on a sample of a small number of clusters chosen from a finite population with a large number of unequally sized clusters. This practicum examines the role of the sampling designs on the estimation of the parameters of the GLMM based super-population for clustered count data. |
---|