The method is generally destructive. A sample of a population is tested. The accuracy of inferences that are drawn depends on the sampling frequency and the degree to which the sample is representative. The probability curve is presented in the format of a Weibull distribution. The validity of a Weibull assumption should@ however@ be tested. Such tests are outside the scope of this method. The strength distribution can be inferred from the failure stress data with certain assumptions about the flaw growth mechanism and the parameters that describe it. These inferences are illustrated in the annexes@ but are outside the scope of the method. This test method applies to a population characteristic that is not specifiable@ at least at the present time. Therefore@ this test method shall not be called out in any TIA specification document.