Abstract
Several options in designing a Shewhart-type control chart for individual observations are discussed. A number of possible estimators of the standard deviation are considered. A two-stage procedure is suggested for retrospective testing. Finally, it is shown that displaying a moving range chart has no real added value and is, therefore, ill-advised.
Additional information
Notes on contributors
Kit C. B. Roes
Mr. Roes is a Senior Statistician. He is a Member of ASQC.
Ronald J. M. M. Does
Dr. Does is a Management Consultant in Statistics and a Professor in Industrial Statistics. He is a Member of ASQC.
Yvonne Schurink
Ms. Schurink is a Statistician.