DeGolier, Kimberly S. (1990, ASQC) 3M, St. Paul, MN
In this world of multi-lingual trade there is a common language that can cross country and cultural boundaries: the language of test results. A meaningful test measurement means the same thing in Japan as it does in Kalamazoo.
Implementation of much of the current "state-of-the art" quality technology such as SPC, Pareto analyses and designed experiments runs into roadblocks that are said to be caused by anything from lack of management support to lazy machine operators. One of the most common but most easily forgotten blockades is poor test method systems.
This paper will offer some suggestions to enhance the use of test results for improved communication and for more effective use of quality tools.
A test method system can be roughly divided into four categories:
One useful test method maintenance technique is based on the concept that a test procedure can be viewed as a "process" and statistical process control techniques applied to that process. By using control charting techniques information about all kinds of special cause variability can be uncovered and the test precision and accuracy can be monitored over time in a simple manner. Control chart techniques can also provide information on test discrimination, inter-location consistency and other test systems issues.
To be able to understand the sources of variability in a test program through the use of control charts results in the ability to focus attention on the manufacturing process and on the product rather than spending time debating the reliability of the test. This results in better customer-supplier communication and a more consistent product.