1Department of Statistics, Bangalore University, Bengaluru-560056
2Department of Statistics, SJMV's Arts, Commerce and Science College for Women, Hubli-580021
*E-mail: panditpv12@gmail.com, panditpv@bub.ernet.in
Online published on 18 October, 2017.
Nonparametric control charts have the advantage that they are not based on a possibly wrong and misleading probability distribution, but are fairly independent of the unknown process probability distribution. In this paper, a distribution-free control chart for shift in location of the process distribution is developed, when the process distribution is symmetric. The statistic used for monitoring the process shift is the test statistic developed for one-sample location problem due to Madhava Rao (1990). The performance of the new control chart procedure is evaluated by computing Average run length (ARL) and the procedure is compared with popular distribution-free control charts. It is observed that the new control chart is better in terms of ARL.
Asymptotic relative efficiency, One-sample location problem, Distribution-free, Symmetric distributions, U-statistics
