Advisor
Shafie, Khalil
Committee Member
Mundfrom, Daniel
Committee Member
Schaffer, Jay
Committee Member
Heiny, Robert
Department
Applied Statistics & Research Methods
Institution
University of Northern Colorado
Type of Resources
Text
Place of Publication
Greeley (Colo.)
Publisher
University of Northern Colorado
Date Created
12-1-2009
Genre
Thesis
Extent
144 pages
Digital Origin
Born digital
Description
Longitudinal studies are commonly encountered in a variety of research areas in which the scientific interest is in the pattern of change in a response variable over time. These observations are traditionally scheduled prospectively and therefore common fixed time interval models for repeated measurements are adequate. Conversely, in informative schedule studies in which subsequent observations are scheduled on the basis of prior response outcomes, time between observations now becomes informative in the longitudinal process. Traditional fixed time approaches, however, are unable to utilize the informative nature of the data lessening the inferences achieved by these approaches. Therefore, the purpose of this research was the development of a joint model of a longitudinal process and informative time schedule data. Maximum likelihood estimates (MLE) for two special cases of the proposed model were obtained from Monte Carlo simulated data by employing the Multivariate Newton-Raphson optimization routine implemented in a SAS/IML call statements. Parameter estimates were determined for a few select cases of subject and observation length and included parameter estimates for rectangular and nonrectangular observation matrices. Finally, estimates obtained from PROC MIXED and from the proposed model were compared for accuracy and efficiency by examining their bias, variance, mean square error (MSE), and relative efficiency.
Degree type
PhD
Degree Name
Doctoral
Language
English
Local Identifiers
Bronsert_unco_0161N_10023
Rights Statement
Copyright is held by author.
