Sampling Plan Using EWMA Statistic of Regression Estimator

Abstract

It is well-known that a regression estimator of the mean quality characteristic would be more informative when a correlated auxiliary variable is available. In this manuscript, we propose a new lot acceptance sampling plan based on the exponentially weighted average (EWMA) statistic of the regression estimator. It is assumed that the two variables of quality characteristic of interest and an auxiliary variable follow a bivariate normal distribution. Two cases of sigma known and sigma unknown are considered separately. We provide the tables for industrial use for various values of the smoothing constant. The efficiency of the proposed plan is discussed over the existing acceptance sampling plan based on EWMA.