Abstract
A procedure for testing an industrial process for control with respect to defects-per-unit is to compare the number of defects observed in each of k samples with an upper control limit. The control limit is calculated using a multiplier of the standard deviation which may be different than the multiplier three, which is usually used for control charts. Using the variation in the probability of a type I error and power for different values of the multiplier and k, procedures are described to increase the effectiveness of the control chart for number of defects.
Additional information
Notes on contributors
Edmund B. McCue
Edmund B. McCue is Associate Professor in the Department of Mathematics, Statistics and Computer Science at The American University, Washington, D.C.