The 2-out-of-5 and the triangle test have similar sensitivity, but the 2-out-of-5 test has much greater statistical power

Abstract

Our study addresses a long-standing assumption that multiple tastings in the 2-out-of-5 test induce sensory adaptation, consequently diminishing participants’ discrimination ability compared to a triangle test. Our objectives were to compare the number of tastings, the degree of taste adaptation, and the sensitivity of these two tests. Fourteen trained participants evaluated two concentrations of sugar in iced tea using each test. We counted the number of times each person tasted each sample, calculated the proportion of discriminators for each of the two tests, and measured adaptation to sweet, sour, and bitter tastes. Participants tasted more times during the 2-out-of-5 tests than the triangle tests, however both tests produced similar proportions of discriminators (0.40 vs 0.36) and similar amounts of adaptation. The much smaller chance probability of the 2-out-of-5 test compared to the triangle test and thus its much greater statistical power was not outweighed by any impact of multiple tasting.