ST16
Using SAS(R) to Control Multistream Binomial Processes with a Chi-Squared Control Chart
Invited
Peter Wlydyka
University of North Florida
,Dan Cavey
Bank of America
,Brett Friedlin
University of North Florida Abstract:
A process control scheme for multistream binomial processes is described. A multistream process in one in which the streams are uncorrelated or very weakly correlated, as might arise for example when J units are in a single frame that proceeds along a prod
uction line. The chi-squared control chart, which can be used for either homogeneous (the streams have identical rates at which nonconforming units arise) or nonhomogeneous (the streams have different rates at which nonconforming units arise) processes, wi
ll be described and SASŪ programs provided that can be used to estimate the quantiles of the control limit distribution, estimate the ARL distribution for process and stream shifts, and produce control charts for process control. That is, a complete tool k
it will be provided for initiating and maintaining process control using the SASŪ system.
Biography:
Peter Wludyka is an associate professor of Statistics and Director of the Center for Research and Consulting in Statistics at the University of North Florida. He has presented papers at SESU, SSU, and SUGI.
Dan Cavey is a graduate student at University of North Florida working on his Master's degree in Mathematics and Statistics. He received his undergraduate degree in Statistics from Virginia Tech in 1980. He is a manager in the Application Software Engineer
ing department at Bank of America.
Brett Friedlin is a graduate student in the Department of Mathematics and Statistics at the University of North Florida.