All Words
Exact Phrase
Title Search Only
advanced search
Digital Archives Initiative
Memorial University - Electronic Theses and Dissertations 5
Anthropology
Aquaculture
Archaeology
Biochemistry
Biology
Biopsychology
Chemistry
Classics
Community Health
Computational Science
Computer Science
Counselling Centre
Earth Sciences
Economics
Education
Educational Administration
Educational Psychology
Engineering
English
Environmental Science
Folklore
French and Spanish
Geography
German and Russian
History
Human Kinetics and Recreation
Linguistics
Marine Studies
Mathematics and Statistics
Medicine
Nursing
Pharmacy
Philosophy
Physics and Physical Oceanography
Political Science
Psychology
Religious Studies
Social Work
Sociology
Toxicology
Women's Studies
home
browse
preferences
my favorites
about/feedback
recent uploads
help/search tips
Français
menu off
add document to favorites
:
add page to favorites
:
reference url
back to results
:
previous
:
next
Search this object:
0
hit(s) ::
previous hit
:
next hit
View:
document description
page description
page & text
previous page
:
next page
Document Description
Title
Generalized
quasilikelihood
method
for
misclassified
longitudinal
binary
data
Author
Tao
,
Yi
,
1981-
Description
Thesis
(M.A.S.)--Memorial
University
of
Newfoundland
,
2010.
Mathematics
and
Statistics
Date
2010
Pagination
vii, 70 leaves : ill.
Subject
Biometry;
Linear
models
(Statistics);
Logistic
distribution;
Regression
analysis
Degree
M.A.S.
Degree Grantor
Memorial University of Newfoundland. Dept. of Mathematics and Statistics
Discipline
Mathematics and Statistics
Language
Eng
Notes
Bibliography:
leaves
68-70.
Abstract
In this
practicum
we
develop
the
generalized
quasi-likelihood
approach
to
analyzing
longitudinal
binary
data
with
misclassification
in
response.
We
utilize
the
method
of
Monahan
and
Stefanski
(1992)
to
approximate
the
expectation
of an
unknown
function
involved
in the
calculation
of the
means
and
covariances
,
which
are
further
used
to
develop
GQL
estimating
functions.
The
results
of an
intensive
simulation
study
show
that the
proposed
method
works
very
well
in
all
the
preselected
settings.
The
efficiency
gain
as
compared
to the
naive
method
is
remarkable.
The
method
is
robust
in the
sense
that the
performance
varies
just
slightly
when
model
parameters
change
in the
simulation.
--
Keywords:
Logit
link;
longitudinal
binary
response;
GQL;
Misclassification.
Type
Text
Resource Type
Electronic
thesis
or
dissertation
Format
Image/jpeg;
Application/pdf
Source
Paper copy kept in the Centre for Newfoundland Studies, Memorial University Libraries
Local Identifier
a3497997
Rights
The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.
Collection
Electronic
Theses
and
Dissertations
Scanning Status
Completed
PDF File
(3.14
MB)
--
http://collections.mun.ca/PDFs/theses/Tao_Yi.pdf
CONTENTdm file name
10762.cpd
