Size and Power of Tests of Hypotheses on Parameters When Modeling Time-to-Event Data with the Lindley Distribution

<p>Covariates on subjects are collected in addition to times-to-event in time-to-event studies. Such data are often analyzed by choosing a model that allows the covariate information to be utilized in the analyses. The analysis proceeds by estimating parameters in the model and testing hypotheses about the parameters based on their estimates; validity of inferences from tests of hypotheses about the parameters depends on size and power of the tests. The Lindley model is considered, in this dissertation, as an alternative model facilitating the analysis of time-to-event data with or without covariates for complete or incomplete data. Covariate information is incorporated using the form of Cox's proportional hazard's model with the Lindley model as the time dependent component (called Lindley-Cox model). Results suggest that size of tests on parameters arising from their maximum likelihood estimates (MLEs) in the Lindley-Cox model is -level and power of tests on parameters arising from their MLEs in this model compares to that from MLEs in Cox's.</p>

Name thesis:degree_name

Doctor of Public Health in Biostatistics (Dr.P.H.)

Level thesis:degree_level

Dissertation (restricted to Georgia Southern)

Discipline thesis:degree_discipline

Department of Biostatistics (COPH)

Year dc:date.available

2012

Author dc:creator

• Okwuokenye, Macaulay

Contributors dc:contributor

• Broderick Oluyede
• Lili Yu
• Kao-Tai Tsai

• ETD
• Size
• Power
• False Positive Rate
• Lindley Model
• Lindley-Cox
• Biostatistics

Repository record dc:identifier

https://digitalcommons.georgiasouthern.edu/etd/791

OAI identifier oai:identifier

oai:digitalcommons.georgiasouthern.edu:etd-1796