Control charts are the most popular monitoring tools used to distinguish between special (assignable) and common causes of variation and to detect any changes in processes. The time that a control chart gives an out‐of‐control signal is not the real time of change. The actual time of the change is called the change point. Knowing the real time of the change will help and simplify finding the assignable causes of the signal, which may be the result of a shift in the process mean or change in process variability. This article gives an overview of change point estimation in control charts, provides a classification scheme, and describes the research that has previously appeared in the literature. In addition, a gap analysis in this area provides direction for future research. Copyright © 2011 John Wiley & Sons, Ltd.

Control charts are one of the most applicable tools in statistical process control. The time in which the control chart signals an out-of-control alarm is not the actual time in which the change has occurred. In other words, control chart detects the change with some delay. The actual time of the change taking place is referred to as change point. Change point estimation facilitates the identification of cause(s) of change and reduces the corresponding time and cost. There are many processes in which the control of two correlated attributes is necessary. Multiattribute control charts are used in such cases due to correlation between attributes. In this paper, two methods including maximum likelihood estimation (MLE) and clustering are proposed for estimating change point in nonconformity ratio vector of a process with bivariate binomial distribution. Also, the performances of these methods are evaluated and compared by Monte Carlo simulations.