Discrimination Testing with Multiple Samples
Rousseau, B. and Ennis, D. M. (2002). IFPress, 5(1) 2-3.
The fact that variation occurs in the manufacture of consumer products leads to a challenging problem in difference testing. A product with a particular identity may be composed of different variants due to the fact that it is produced at different locations or on different machines. Difference testing designed to evaluate modifications of this product should consider the fact that the product is composed of variants. Classical difference tests such as the triangular, duo-trio or m-alternative forced choice methods are not capable of accounting for the possibility that a product may exist as one of several variants. In this report, an extension of the duo-trio method is used to quantify additional variation such as that caused by different production facilities.
Figure 1. Similarities of the duo-trio method, Torgerson’s method, the ABX method, preferential choice and two-stimulus identification.