Dissertation Information for Warren BilkerNAME:
DEGREE:
DISCIPLINE:
SCHOOL: ADVISORS: COMMITTEE MEMBERS: MPACT Status: Incomplete - Inspected Title: Statistical methods for cross-sectional survival data
Abstract: In this thesis we deal with two type of incompleteness in survival data, censoring and truncation. Right censoring has long been recognized and there has been much methodology developed for handling censored data. On the other hand, only recently has the problem of truncation become widely recognized. Two new methodologies for handling truncated data, one from nonparametric estimation and one from semiparametric testing are developed. In many follow-up studies, the data are subject to left truncation and right censoring. An AZT Prevalent Cohort Study provides an interesting example of such data. A method for obtaining bootstrap confidence intervals for the survival curve and other parameters of interest derived from left truncated and right censored data is presented and shown to be a valid method. In epidemiologic studies, statistical data are sometimes retrospectively ascertained subject to a sampling criterion resulting in pure right truncated data. An example of such data is the CDC Blood Transfusion Data. In classical testing problems, the Mann-Whitney test is used for testing the equality of two distributions. An extension of this test is considered for the case when the two distributions are estimated in the presence of truncation, with the truncation distribution parameterized. |
MPACT Scores for Warren BilkerA = 0 Advisors and Advisees Graph |