A Comparison between the Variable Sample Size (VSS) and Variable Sampling Interval (VSI) Run Sum Charts for the Coefficient of Variation

Abstract

The run sum chart was proposed to improve the performance of detecting shifts in the coefficient of variation   , by separating the charting region into different regions where scores are assigned to these regions. Decisions about the in-control status of the process is then made based on the cumulative scores. Further improvements were made on the run sum  chart, by varying the sampling interval (VSI) and sample size (VSS) based on the current cumulative scores. Both the VSI and VSS run sum  charts show better performance than the run sum  chart. This paper compares the performance of the VSI and VSS run sum  charts, and found that the VSS run sum  charts show better performance in detecting small shifts in  , while the VSI run sum  charts show better performance for moderate and large shifts.