Regression analysis for longitudinal hemoglobin data for premature infants with outcomes subject to non-response

• Main Author: Neupane, Rajendra, 1973-
• Other Authors: Memorial University of Newfoundland. Dept. of Mathematics and Statistics
• Format: Thesis
• Language: English
• Published: 2002
• Subjects: Regression analysis; Premature infants > Nutrition > Statistical methods; Newfoundland studies; University of Newfoundland
• Online Access: http://collections.mun.ca/cdm/ref/collection/theses3/id/202222

Thesis (M.A.S.)--Memorial University of Newfoundland, 2002. Mathematics and Statistics Bibliography: leaves 51-52 In analyzing longitudinal hemoglobin data for low-birth-weight infants, it is of interest to examine the effects of iron fortification and other covariates such as gender and gestation weeks on the hemoglobin status of the infants over the months. As the hemoglobin data are collected repeatedly over a period of time, the longitudinal correlations of the responses must be taken into account in finding the covariate effects. Further problems get mounted when some of the responses are missing. In this practicum, we conducted a regression analysis after accomodating the longitudinal and missingness nature of the data into account. Also, several non-parametric tests were applied to examine any possible monotonic trend in the longitudinal hemoglobin data.