ABSTRACT

The term, n(n - 1)/2, is equal to the total number of possible pair combinations for "n" samples or, in this case, test materials. ND is the number of sample pair reversals observed in the accelerated test compared with the exterior or "natural" exposure. NT is the number of sample pairs with identical rank performance in the accelerated test. As with Spearman's rs, a relatively simple computer program or spreadsheet macro can be written to calculate PE% for any n samples.