Document Details

Document Type : Thesis 
Document Title :
On Some Goodness of Fit Tests for Burr Type X Distribution
عن بعض اختبارات جودة التوفيق لتوزيع بير من النوع العاشر
 
Subject : Statistics department 
Document Language : Arabic 
Abstract : Goodness of fit tests is used when the form of the population (the distribution function) is in question with the hope that the null hypothesis will be accepted. These tests are designed for check the acceptance of the null hypothesis, which is a statement about the form of the distribution function of the population from which the sample is drawn. Ideally, the distribution function is completely specified (Simple Null Hypothesis) if it includes known values for the parameters. Otherwise, it contains unknown parameters which must be estimated using the sample data to perform any inference for the population and testing hypotheses. The tables of critical values of goodness of fit tests are used when the distribution is well defined (contains known parameters) and the data is complete. When the data is censored or the distribution contains unknown parameters, then the tables of critical values are inappropriate for use in goodness of fit tests. In this thesis, we used Monte Carlo simulation to construct tables of critical values for some goodness of fit tests based on the Burr type X distribution when the data are censored and the parameters of the distribution are estimated by maximum likelihood method. These tests are Kolmogorov-Smirnov (KS), Cramér–von Mises (CVM), and Anderson Darling (AD). Burr family of distributions assumed a position of importance in the field of life testing. Our goal in this thesis is to construct tables of critical values for the three test statistics mentioned above. Moreover, we use different sample (Small – Medium – Large) of Burr type X distribution to study the sampling distributions of the test statistics numerically. The sampling distributions are obtained using Pearson' family of distributions. The sampling distributions are used to construct tables of critical points for the test statistics. Also in this thesis it is a position of importance to study the power performance among the three tests with some alternative distributions using Monte Carlo simulation. 
Supervisor : Prof.: ABDULLAH MOHAMAD ABDELFATTAH 
Thesis Type : Master Thesis 
Publishing Year : 1433 AH
2012 AD 
Added Date : Wednesday, March 21, 2012 

Researchers

Researcher Name (Arabic)Researcher Name (English)Researcher TypeDr GradeEmail
عبد العزيز صالح الغامديALGHAMDI, ABDUL AZIZ SALEHResearcherMaster 

Files

File NameTypeDescription
 32680.pdf pdf 

Back To Researches Page