Abstract
In this paper a procedure is considered for a new modified goodness of fit test for the normal type2 censored population. Samples of sizes 10(10)60 are chosen and censored at a specified percentage. Cramer von Mises and Anderson Darling test statistics are used with a nonparametric density estimator in place of the empirical distribution function. Tables of critical values for the two tests are generated. The power of the tests for nine alternative distributions is shown. Results show high power of discriminating the alternatives.
Highlights
An important problem in statistics is to find information about the form of the population from which a sample is drawn
Goodness of fit tests are given for the normal distribution with unknown mean and unknown variance from type 2 censored samples
THE NORMAL DISTRIBUTION The normal distribution is without a doubt the most important and most widely used continuous probability distribution
Summary
An important problem in statistics is to find information about the form of the population from which a sample is drawn. Goodness of fit tests are given for the normal distribution with unknown mean and unknown variance from type 2 censored samples. A Monte Carlo procedure is used to develop and compare the modified goodness of fit (GOF) tests from censored samples. Critical values for different sample sizes n are generated. The modified CvM and AD test statistics are calculated for the given values of n. This procedure is repeated 10000 times for each test statistic. Tables of critical values for the two modified test statistics from type 2 censored samples for the normal model are found. The power study of the modified tests to compare the efficiency of the CvM and AD tests for different sample sizes is discussed
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
