Constructing Two-Sided Group Chain Acceptance Sampling Plans for Non-Symmetrical Data

Abstract

In developing acceptance sampling plan, the mean is often the choice due to the fact that it is a good estimator. However, it is only a good estimator when dealing with symmetrical data. If the data is not symmetrical, then the mean is no longer a good estimator, but the median is. The recent two-sided group chain acceptance sampling plans (TSGChSP) has only been applied to Pareto distribution of the 2^nd kind by using the mean to cater for symmetrical data. In reality, some data are actually non-symmetrical. Hence, motivated by such scenario, the aim of this study is to produce the TSGChSP for Pareto distribution of the 2^nd kind by using the median. The performances of TSGChSP are measured based on minimum number of groups,  and probability of lot acceptance, . The  is calculated by satisfying the consumer’s risk while the  is obtained at several values of median ratio. Finally, the number of million revolutions for 23 ball bearings is shown in order to show the application of the TSGChSP in the industry.