This paper proposes and discusses criteria for determining whether or not a test is useful in fulfilling its intended purpose. Distinctions are drawn between those criteria which are desirable but not absolutely necessary and those which are essential to a useful test. Key criteria include correlation of test results to performance characteristics, applicability to a wide range of products, discrimination between different products, and sufficient precision and accuracy. In addition, some of the basic philosophy behind test method development and the compromises necessary in that process are discussed.
ASTM states that a test method is “a definitive procedure for the identification, measurement, and evaluation of one or more qualities, characteristics, or properties of a material, product, system or service that produces a test result.”(1) This statement describes what a test method is, but does not define the characteristics distinguishing a useful and appropriate test from one with little or no practical value. As a major supplier of trim products for the automotive industry, Davidson/Textron expends considerable resources each year on testing. These tests are conducted to demonstrate compliance of our products to customer requirements, to assure the quality of chemicals and components we purchase, to predict the field performance of our products and to aid in the development of new and/or improved products and processes. Through the efforts of such organizations as SAE, IFAI and ASTM, many excellent tests are available. Unfortunately, there are also many tests in common use which are virtually useless.
Because there are no generally agreed upon criteria defining a useful test method, it is often difficult to identify poor test procedures and correct them. One purpose of this paper, therefore, is to suggest a list of such criteria. This list is probably not all-inclusive. However, it is our hope these criteria will provide the starting point for fruitful discussions which will ultimately result in a consensus list of criteria of value not only to the automotive industry but to all those who conduct tests in the normal course of their endeavors. Diligent application of such criteria to new and existing test methods should do much to minimize the resources expended on useless testing.