A family of group chain acceptance sampling plans based on truncated life test

• Main Author: Razzaque, Abdur
• Format: Thesis
• Language: English; English; English
• Published: 2018
• Subjects: Q Science (General)
• Online Access: https://etd.uum.edu.my/6885/1/DeporitPermission_s99039.pdf; https://etd.uum.edu.my/6885/2/s99039_01.pdf; https://etd.uum.edu.my/6885/3/s99039_02.pdf; https://etd.uum.edu.my/6885/

Acceptance sampling is a statistical quality control procedure used to accept or reject a lot, based on the inspection result of its sample. For high quality products, zero acceptance number is considered and the life test is often terminated on a specific time, hence called truncated life test. A plan having zero acceptance number is deemed unfair to producers as the probability of lot acceptance drops drastically at a very small proportion defective. To overcome this problem, chain sampling which uses preceding and succeeding lots information was introduced. In ordinary chain sampling plans, only one product is inspected at a time, although in practice, testers can accommodate multiple products simultaneously. In this situation, group chain sampling plan with small sample size is preferred because it saves inspection time and cost. Thus, it is worthwhile to develop the various types of chain sampling plans in the context of group testing. This research aims to develop new group chain (GChSP), modified group chain (MGChSP), two-sided group chain (TS-GChSP) and modified two-sided group chain (TS-MGChSP) sampling plans using the Pareto distribution of the 2nd kind. These four plans are also generalized based on several pre-specified values of proportion defective. This study involves four phases: identifying several combinations of design parameters; developing the procedures; obtaining operating characteristic functions; and measuring performances using both simulated and real lifetime data. The constructed plans are evaluated using various design parameters and compared with the established plan based on the number of minimum groups, and probability of lot acceptance,. The findings show that all the proposed plans provide smaller and lower compared to the established plan. All the plans are able to reduce inspection time and cost, and better at protecting customers from receiving defective products. This would be very beneficial to practitioners especially those involved with destructive testing of high quality products.