Homojen olmayan varyans altında varyans analizi için bir aşamalı ve iki aşamalı yöntemler


Thesis Type: Postgraduate

Institution Of The Thesis: Gazi University, Fen Bilimleri Enstitüsü, İSTATİSTİK ANABİLİM DALI, Turkey

Approval Date: 2005

Thesis Language: Turkish

Student: ÖZLEM TUNCER

Supervisor: HAMZA GAMGAM

Abstract:

IV SINGLE-STAGE AND TWO-STAGE METHODS FOR ANALYSIS OF VARIANCE UNDER HETEROSCEDASTICITY (M.Sc.Thesis) Özlem TUNCER GAZI UNIVERSITY INSTITUTE OF SCIENCE AND TECNOLOGY August 2005 ABSTRACT Usual F - test used in ANOVA problems requires the assumptions of normality, independence and homogeneous of the error variances. When these assumptions are not supplied, it is not suggested to use F - test In these cases alternative methods are recommended for F - test. Transformations of the data, Nonparametric tests, Single - stage method and Two-stage method are these methods. In this study, Single - stage method offered by Chen and Chen (1998) and Two - stage method offered by Bishop and Dudewicz (1978) are described. In second chapter Single - stage method, in third chapter Two - stage method, in fourth chapter results of comparing with Single - stage methods and Two-stage method and in fifth chapter, the application of Single - stage method is made with data derived from simulation program. Science Code : 212 Key Words : Heteroscedasticity, One-way Layout, Two-way Layout. Page Number : 64 Adviser : Prof.Dr. Hamza GAMGAM