Modelling a binary response of irvingia cuttings to a rooting hormone

Abstract

Modelling is a statistical method using mathematical equation to describe the relationship between one or more dependent variables and a set of explanatory variables which can be experimental conditions. In this project, the logistic and Cox regression models are used to study the effects of the mother trees and hormone doses on the probabilities and rates of rooting of Irvingia cuttings. Chapter I gives a general introduction on model based analysis and some background information on vegetative propagation. The problematic, the objectives and the significance of the study are then stated. Reasons for the used of logistic and Cox model are presented in section 1. 5. In section 1. 6, previous studies on Irvingia gabonensis add on the application of logistic or Cox regression models to the analysis of experimental data are reviewed. Chapter II describes the statistical methodology used in the study. For the two mocfe'"1s't-h;e' procedures for parameters estimation, significance tests and checking model adequacy are presented. The different postulated models for the Irvingia cuttings data are defined and the methods for selecting the most appropriate presented. In chapter III, the main results of the data analysis are presented, interpreted and commented. The estimated parameters of the appropriate model corresponding to the effects of different treatments are given. Then the predicted probabilities of rooting and the relative rooting rates for each treatment are presented and compared. The last chapter deals with the concluding remarks based on the important results obtained and discuss how these findings can be beneficial in the development of a practical methodology of vegetative propagation of Irvingia gabonensis. Some recommendations are also made for future study in the area.