Vol. 5, No. 5
Black belt not attainable without knowledge
Because of her failure to understand calculation of control limits for X-bar and S charts, Dinah Thyrst failed in her attempt to garner a black belt last month. Still longing for this enhancement to her wardrobe, she has purchased the white uniform she will wear with her belt, if she ever gets one. After her bumbling attempts to improve the defect rate for part # 5290F, however, that rate has actually gone up rather than down. The cpk is now .4. Even Dinah realizes that things are worse, not better.
To accomplish her projected savings goal of $100,000, Dinah is forthright in taking steps toward that goal. To begin with, she deletes last month’s data so the new base for the project will reflect a cpk of .4 and a loss to the company of $200,000. Further, she makes some minor adjustments in the process, then undertakes a new control chart and capability study. She has carefully filtered the data to be used, so it is no miracle that the charts indicate improvement in the process. The chart indicates upper control limit of 12.99 and a lower control limit of 10.20, with a cpk of .52. Creative accounting produces cost savings of $122,425.12 from the “improved” process.
“This calls for a PowerPoint presentation,” Dinah Thyrst decides. Using the following control chart, she prepares a presentation for Cliff Clymer, a master black belt. The data (carefully filtered) has sample sizes that vary from two to five. Cliff questions her about this variation, and she realizes that she has overlooked sample size entirely.
“Well,” she begins, “with sample size varying from two to five, the chart will reflect reality, which is not always laced with certainty, as we know. It’s important to use an X-bar and S chart to reflect this slippery data.” Is she right?
Copyright 2003 PQ Systems.
Please direct questions or problems regarding this web site to the Webmaster.