Nsowah-Nuamah, N. N. N.Nortey, E.N. N.Iddi, S.Turkson, E.H.University of Ghana, College of Basic and Applied Sciences, School of Physical and Mathematical Sciences, department of Statistics2017-01-272017-10-132017-01-272017-10-132015-06http://197.255.68.203/handle/123456789/21447Thesis(PHD)-University of Ghana, 2015The objective of this research is to apply appropriate statistical techniques to model the length of time to successive births (birth spacing) within the reproductive health of mothers and identify the socio-demographic and sociocultural factors that cause variation in the length of birth interval. The data used in this study were collected using questionnaire from Gomoa East District 2014, where multi-stage cluster sampling was used as the sampling design. The clusters were defined as small enumeration areas (EAs) and it was assumed that birth intervals for the women in the same cluster (children) were correlated because they share the same environment. This study used: Conventional Methods (Cox regression), Acceleration Failure Time (AFT); Exponential, Weibull, Log-logistic Lognormal Frailty modelling techniques. In the analysis, the covariates mother's age at first birth, mother's educational level, birth experience, and survival status of the index child were found to have significant effect on the length of birth interval for Gomoa East women. The observations within specific women were found to be correlated and the Cox's proportional hazards model was used to underestimate the standard errors of the parameter estimates. The analysis with the frailty model showd that parity may not be an important covariate for analyzing birth intervals, especially when mother's age at birth is already in the model.Xvi, 278p: illenStatistical ModellingBirth SpacingReproductive HealthThesisUniversity of Ghana