Statistical consultant, National Measurement Laboratory, U.S. Department of Commerce, National Bureau of Standards, Washington, D.C.,
In designing an interlaboratory study for the determination of the precision and accuracy of a test method, detailed attention must be given to a number of items. Foremost among these is the selection of laboratories and materials. The materials should cover the range of levels to which the test is intended to apply. The laboratories should have acquired thorough familiarity with the method, obtained if necessary in the course of a preliminary round-robin test. It is good practice to analyze the data level by level and to subsequently examine the within- and between-laboratory components of variance as functions of the level. The paper includes a comparison of this method of analysis with one based on two-way analysis of variance.
Paper ID: CCA10407J