Phase II monitoring of binary response profiles

In many situations, the quality of a process can be characterized better by a relationship, known as a profile, between a response variable and one or more predictors. Almost all research efforts assume that response variable is continuous and follows a Normal distribution while there are instances in which the response is a binary variable and methods such as logistic regression are commonly used. In this paper four control schemes namely Hotelling, MEWMA, likelihood ratio test (LRT) and LRT/EWMAare proposed to monitor binary response profiles in phase II. The performance of the proposed control charts are evaluated and compared by simulation experiments for different shift values in the parameters of the profile in terms of the average run length (ARL) criterion. The results show that all methods work well in the sense that they can effectively detect shifts in the process parameters. Based on the results, MEWMA and LRT/EWMA methods display a better performance for small to moderate and large shift values, respectively.