Get access

The Tetrad Test: Looking Back, Looking Forward

Authors


Corresponding author. TEL: +1-530-756-549; Fax: +1-530-756-7320; EMAIL: maomahony@ucdavis.edu

Abstract

Until recently, the tetrad test was considered an interesting idea but was seldom used. It was not until Ennis and Jesionka revisited the topic of power for discrimination tests, that it was observed that the tetrad test had, theoretically, considerably more power than the triangle test. This made it a suitable candidate as a replacement for the low powered, yet commonly used, triangle method. This has sparked a lot of activity concerning the tetrad and there will be more to come. Yet, the tetrad had been explored beforehand, for the purposes of testing Thurstonian models. Had the authors had the right tools, could they have produced any evidence to support Ennis and Jesionka's observation?

Practical Applications

The use of difference tests where the attribute change is not described to the judge, is especially useful for testing consumers in normal conditions of consumption. In such normal conditions, unlike those for a two-alternative forced choice (paired comparison), there is no person nearby specifying the nature of the attribute that should be attended to while consuming food samples. The nature of the difference between such food samples has to be discovered by the consumer. Available tests like the triangle or duo-trio are not statistically powerful and therefore require large samples of data, which can incur greater expense. Yet, the tetrad is more theoretically powerful than these tests and is thus worthy of, and is indeed receiving, further investigation.

Get access to the full text of this article

Ancillary