The Dos and Don ts of Control Charting Part II John McConnell, Brian Nunnally, and Bernard McGarvey Analysis and Control of Variation is dedicated to revealing weaknesses in existing approaches to understanding, reducing, and controlling variation and to recommend alternatives that are not only based on sound science, but also that demonstrably work. Case studies will be used to illustrate both problems and successful methodologies. The objective of the column is to combine sound science with proven practical advice. Reader comments, questions, and suggestions will help us fulfil our objective for this column. Case studies illustrating the successful reduction or control of variation submitted by readers are most welcome. Please send your comments and suggestions to column coordinator John McConnell at john@wysowl.com.au or journal coordinating editor Susan Haigney at shaigney@advanstar.com. KEY POINTS DISCUSSED The following key points are discussed: There is a right way and a wrong way to control chart. control chart. calculated correctly. INTRODUCTION control charting: EXAMPLE FOUR we reason that the people who are looking at the data [ For more Author information, go to gxpandjvt.com/bios ABOUT THE AUTHORS John McConnell is owner and director at Wysowl Pty Ltd. in Queensland, Australia. He may be reached at john@wysowl.com.au. Brian K. Nunnally, Ph.D., is in charge of process validation at Pfizer in Sanford, NC. He may be reached at nunnalb@wyeth.com. Bernard McGarvey, Ph.D., is Process Modeling Group Leader, Process Engineering Center at Eli Lilly and Company in Indianapolis, IN. He may be reached at mcgarvey_bernard@lilly.com. PROCESS VALIDATION Continued Process Verification 21
Figure 1: Conversion time monthly average. X Chart for Conversion Time Monthly Average Individual Values 8 7 6 5 3 2 1 Feb-1 Mar-1 Apr-1 May-1 Jun-1 Jul-1 Aug-1 sep-1 Oct-1 Nov-1 Dec-1 Jan-2 Feb-2 mr Chart for Conversion Time Monthly Average Moving Range Values 35 3 25 2 15 1 5 Feb-1 Mar-1 Apr-1 May-1 Jun-1 Jul-1 Aug-1 sep-1 Oct-1 Nov-1 Dec-1 Jan-2 Feb-2 this is no longer the case. 22 PROCESS VALIDATION Continued Process Verification
Figure 2: Product conversion time on a lot basis. Individual Values 9 8 7 6 5 3 2 1 X Chart for Conversion Time by Lot mr Chart for Conversion Time by Lot Moving Range Values 35 3 25 2 15 1 5 Lot 1 Lot 3 Lot 5 Lot 7 Lot 9 Lot 11 Lot 13 Lot 15 Lot 17 Lot 19 Lot 21 Lot 23 Lot 25 Lot 27 Lot 29 Lot 31 Lot 33 Lot 35 Lot 37 Lot 39 Lot 1 Lot 3 Lot 5 Lot 7 Lot 9 Lot 11 Lot 13 Lot 15 Lot 17 Lot 19 Lot 21 Lot 23 Lot 25 Lot 27 Lot 29 Lot 31 Lot 33 Lot 35 Lot 37 Lot 39 the last three to four lots of these data are trending sharply upward. EXAMPLE FIVE rational for this is typically a lack of understanding of Percent completed training. The only PROCESS VALIDATION Continued Process Verification 23
reasons. Percent analytical variation. is trying to detect changes in the production where σ and σ to display a lack of statistical control on the suspect. Raw materials management. Asset management. EXAMPLE SIX exposed to the use of control charts and was he was enthusiastic to apply his learning and wanted 2 PROCESS VALIDATION Continued Process Verification
Figure 3: The right and wrong way to control chart individuals data. 6 Individual Measurement of Metric 9 Signal 2-2 Metric 85 8 UCL=85.36 Avg=79.81-75 LCL=7.26 YY -6 1 9 85 8 75 7 2 3 5 6 Lot # Lot # Signal 1 2 3 5 6 Lot # Metric Mean LCL UCL Incorrect Control Limits Moving Range of Metric 8 7 6 5 3 2 1 8 Moving Range of Metric 8 12 16 2 2 28 32 36 8 12 16 2 2 28 32 36 8 Lot # Correct Individuals / Moving Range Control Chart UCL=6.82 Avg=2.9 LCL=. σ σ and σ dataset. easily calculated. This is not the correct approach. of the data into account. In practice this is done σ learning and application. PROCESS VALIDATION Continued Process Verification 25
of what is happening in the process. It is easy to that the change in the process that led to the out of data. CONCLUSIONS presented as follows: control chart insight into the process calculated correctly. there are the correct ways to do things and the series. REFERENCES Journal of Validation Technology Understanding Statistical Process Control Out of the Crisis JVT The Journal of Validation Technology 26 PROCESS VALIDATION Continued Process Verification