Statistics are everywhere: They grin at us from newspaper graphics, clog our e-mail inboxes and stare from
children's report cards. Despite their ubiquity, we have--like fish, which spend their lives surrounded by water without ever becoming cognizant of it--little awareness of the statistics that
surround us. Quality managers, though, can't become oblivious fish. A manager's success demands vigilance and analytical savvy. Data entered into a computer for charting must
ultimately be scrutinized for meaning using an entire tool set of correct analytical approaches. Let's consider one Hartford Simsack, who, unfortunately, has become one of
those fish who fails to appreciate the water. As quality manager for Greer Grate and Gate (GG&G), he's responsible for quality improvement for the extruded metal products that his
brother-in-law's firm produces. He secured his position through family influence shortly after being released from prison, where he had served a sentence that hadn't allowed him to build an
appropriate résumé to launch a career. He believes his most important task is covering up the fact that he knows nothing about SPC and doesn't understand why the company makes such a fuss about
quality. The plant manager, another family member, sees ISO 9000 certification as the company's highest priority, and has enrolled Simsack in training through a local
technical college to prepare him to lead the efforts that will contribute to successful registration. In a burst of enthusiasm, he encouraged Simsack to take a class in statistical methods as
part of his training.
Figure 1: A Portion of GG&G's Control Chart |
|
A technician, Franklin Benjamin, has been responsible for GG&G's SPC program for several months, after learning everything he needed to know about statistics from the previous
quality manager, who had been fired for incompetence. Benjamin produces a control chart that he can't figure out by himself. The chart (Figure 1) shows a series of ranges below the R-bar--a
pattern that Benjamin's former boss had emphatically endorsed, saying that it indicates decreasing variability in the process. Benjamin believes that new control
limits should be applied that would reflect the new level of variability. Because he believes that coursework will have made his boss a statistician
overnight, Benjamin takes the chart to Simsack. Poor Hartford Simsack doesn't want to acknowledge his ignorance, so he tells the technician that he's too busy to
look at the chart but will get back to him on the matter, grumbling meanwhile about all the demands that are made on a statistician these days.
Simsack takes the chart to his instructor at the technical college. What will the instructor say about the chart? Answer Benjamin is partially correct in his assessment of the control limits. Professor Stan
Deviation points out that one of the indicators of an out-of-control condition is a run of seven above or seven below the process average.1 In this case, the situation is
technically out of control, but because it indicates that the process variation has diminished, the condition is actually positive.
However, Benjamin didn't go far enough in his assessment. Deviation reminds Simsack of the four steps that must be taken before control limits should be changed:
1. Determine assignable cause for the series of points below the R-bar. This may take a while, but it must be done before going to Step 2.
2. Once the cause has been determined, standardize the process. If the improvement has resulted because of a particular supply source, for example, that source should be maintained.
3. Observe the process to ensure that its variability has indeed been reduced. 4. Once enough data has been collected, calculate new control limits to reflect the improvement. Simsack tells Benjamin what he's learned--without, of course, revealing that it has
not come from the depths of his own understanding of variability. "If you have any other questions, don't hesitate to ask," he magnanimously adds. References 1. Total Quality Tools. Dayton, OH: PQ Systems Inc., 1995. About the author
Michael J. Cleary, Ph.D., is founder and president of PQ Systems Inc. He is a noted authority in the field of quality management and a professor emeritus of
management science at Wright State University in Dayton, Ohio. A 29-year professorship in management science has enabled Cleary to conduct extensive
research and garner valuable experience in expanding quality management methods. He has published articles on quality management and statistical process control in a
variety of academic and professional journals. E-mail Cleary at mcleary@qualitydigest.com . |
