top of page


Thurstonian Scaling for Difference Tests

Ennis, D. M. (1998). IFPress, 1(3) 2-3.



Suppose that an ingredient supplier has changed a key ingredient used in the manufacture of your company’s chocolate chip cookies. Based on the baking chemistry of this new ingredient, there is reason to suspect that the new ingredient will make your cookies harder. Additionally, recent market research data has connected cookie hardness with consumer liking. For these reasons, you decide to conduct a test to determine what effect the change in ingredient will have on your cookies. After some deliberation, you decide to use the duo-trio method to determine whether cookies made using the new ingredient are perceptibly different from cookies made using the current ingredient.


Figure 1. Psychometric functions for the 2-AFC and duo-trio.

bottom of page