Information Loss from Type I versus Type II Censoring Lim, Jo-Han; Song, Hyun-Seok; Lee, Sung-Im;
If the completely observed data are assumed to have full information, the censoring causes the loss of information. Previous studies have introduced the indices of information loss via measuring relative changes between the data with censoring and without censoring. In this paper, the comparisons are made for the information loss between type I and type II censoring in two sample problems.
Information loss;Kendall's tau;Partial likelihood;
Brooks, R.J. (1982), On the loss of information from censoring. Biometrika, Vol. 69, 137-144
Chaloner, K. and Verdinelli, I. (1995), Bayesian Experimental Design: A Review. Statistical Science., Vol. 10, 237-304