Nonparametric ram-order tests for the right censored and grouped data in linear model

We propose a class of nonparametric rank-order tests for the right censored and grouped data in l inear model. For deriving the test statistics, the likelihood ratio principle will be used. Based on the permutat ion principle, these procedures can be seen as conditionally nonparametric ones. In view of the excessive computational work involved in implementing the test for large sample case, we proceed to examine the limiting distributions of the test statistics. Finally, we discuss the scoring system.