ASTM C802-96(2008)e1

Certain criteria need to be met before undertaking an interlaboratory study to determine the precision of a test method. It is not necessary that all of the following conditions described be completely fulfilled in every case; however, if some conditions are not met or are met incompletely, the program will become more complicated and require more work and expense, or may result in impaired information. The recommendations outlined in this section are intended to ensure that the test method is free of technical difficulties to the greatest extent possible before an expensive and time-consuming interlaboratory study is undertaken.

The first requirement is the existence of a valid and well-written test method that has been developed in one competent laboratory (or by cooperative work in a small number of laboratories), and has been subjected to a screening procedure, or to ruggedness testing as described in Practice C 1067. As a result of the screening procedure and some experience with the test method in the sponsoring laboratory and one or two others, a written version of the test method has been developed (but not necessarily published as a standard method) that describes the test procedure in terms that can easily be followed in any properly equipped laboratory. Conditions that affect the test results should be identified and the proper degree of control of those conditions should be specified in the description of the test procedure (see Note 1).

Note 1—The desired degree of control of conditions that affect test results may not always be practically achievable, and tolerances in the test method should recognize this fact. Variations in test results due to variations in such conditions contribute to the total variation which determines the precision of the test method. If the resulting variation is so great that uncertainties in average values obtained by the test method are unacceptably high, then the test method itself is at fault, and efforts should be made to improve it or to replace it by a better one. An expensive and time-consuming interlaboratory study should not be undertaken on such a test method.

Any apparatus required for performing the test should be appropriately designed and available at reasonable cost.

Personnel in participating laboratories should have enough experience with the test method so that they are competent to run the test. The importance of this requirement will vary with the complexity of the method and the degree to which it departs from familiar procedures.

Preliminary knowledge should exist about how changes in materials and conditions affect the test results. There should be a reasonable degree of certainty that the within-laboratory variances are the same in different laboratories, and that troublesome interactions do not exist. These conditions are investigated in the analysis of the data of an interlaboratory study, and are discussed further in 8.2.2, 8.2.3, and Appendix X1.

Facilities and procedures for procurement, preparation, and distribution of samples must be available and should be as simple and free of difficulties as practicable.

Selection of samples must be done by a randomization process, and one person who is familiar with randomization procedures should be responsible for seeing that the procedure is carried out. Refer to Recommended Practice E 105.

Adequate numbers of participating laboratories, operators, and materials must be available. Requirements in these areas are specified in Sections 4 and 5.

The entire interlaboratory test program should be developed from the beginning with the help and advice of persons familiar with statistical procedures and with the materials involved (see Note 2). The same persons who design the experiment should also carry out, or at least have control over, the process of analysis of the data.

Note 2—It may not always be possible to obtain people who are familiar with the materials involved who have a sufficient knowledge of the proper statistical techniques and their proper use. In this case, the committee should obtain the services of a competent statistician who has experience in practical work with data from materials testing, and provide him with an opportunity for learning something about the particular materials and test method involved. Planners of an interlaboratory study should also be warned to avoid the pitfall of assuming that the use of a large computer necessarily results in special expertise in the handling of data or the interpretation of results.

It is important to bear in mind that estimates of the precision of a test method are always based on a particular set of data obtained at a particular time and they need to be kept up-to-date. As materials, apparatus, and conditions change, and operators change or gain more experience, the characteristic precision of the results obtained may change, especially if the test method is new. In some cases, it may even be desirable to conduct more tests at a later date (though not necessarily a repetition of the complete interlaboratory study) in order to provide a check on estimates previously obtained and either verify them or introduce revisions.