Effectiveness of Sensory Difference Tests: Sequential Sensitivity Analysis for Liquid Food Stimuli

Abstract

Versions of the triangle test in which either one or the other stimulus was maintained as the odd one, were found to differ in their effectiveness for a flavored sparkling water system but not for a wine system. The same was found for different orders of presentation of the R-index sensory difference test. These differences were predicted successfully from the signal detectabilities of the various stimuli using a predictive system called Sequential Sensitivity Analysis (%A). This extended SSA beyond model systems to food systems. Use was made of a technique of rapid alternate tasting of different stimuli, called 'warm-up', to improve judges' performance.

Are three-sample tasks less sensitive than two-sample tasks? Memory effects in the testing of taste discrimination

“…Although it might be a valid assumption for visual and auditory stimuli, it might not hold with chemical stimuli, such as tastants and odorants. Numerous experiments have been conducted in which the influence of sequence effects on discrimination testing performance has been investigated (Dessirier & O'Mahony, 1999;Masuoka et al, 1995;O'Mahony & Goldstein, 1986;O'Mahony & Odbert, 1985;Rousseau & O'Mahony, 1997;Tedja, Nonaka, Ennis, & O'Mahony, 1994). These sequence effects illustrate the fact that a Stimulus A might not elicit the same perceptual sensation if tasted after itself or tasted after the alternative Stimulus B.…”

Comparison of the triangle and a self-defined two alternative forced choice test

“…In two early reports, a greater number of significant triangle tests were seen when a warm-up was used for discrimination of wine samples and of fruit beverages (O'Mahony & Goldstein, 1986) and also for NaCl solutions and for orange juices (O'Mahony et al, 1988). In the latter study it was unclear whether actually naming the difference gave any additional advantage.…”

“…It is possible that preliminary tasting of the two substances, when the difference between them is not specified, would yield a dimension which could be used in the 2-sample method". This statement thus foreshadowed the warmed-up paired comparison method of O' Mahony andGoldstein (1986) andO'Mahony et al (1988) as well as the SD-2-AFC used here. Much has been written concerning the potential advantages of giving warm-up samples before discrimination tests (e.g.…”

“…These results are generally consistent with the predictions of Sequential Sensitivity Analysis (O'Mahony (21) Less sweet (7) Less sour (1) Sweeter (7) Other (5) Tea Sweeter (20) -Sourer (3) Less sweet (2) Other (8) Soup Saltier (9) Stronger (6) Less salty (9) Weaker (1) Other (7) Total atypical responses include inverted plus unique responses for purposes of comparison in Table 6. & Goldstein, 1986;Rousseau & O'Mahony, 1997). The first replicate of each test was compared to the fourth replicate to examine order effects.…”

“…A self-specified directional difference test is similar to some forms of the warmed-up paired comparison (W/U PC) studied by Barbary, Nonaka, Delwiche, Chan, and O'Mahony (1993), O'Mahony and Goldstein (1986), and O'Mahony, Thieme, and Goldstein (1988). In these procedures, participants would taste the pair of items to be tested from 3 to 10 times and in some (but not all) cases, would be asked to specify the nature of the sensory difference.…”