The objective of this study was to investigate the predictive validity of l
aboratory sensory tests on consumption. Thirty-six subjects (30 women, six
men; mean age 21 years) participated in a taste-and-spit test, a taste-and-
swallow test, a fixed quantity test in which an amount of 300 g was consume
d, and an ad libitum consumption test. Subjects rated on a 10-point categor
y scale pleasantness and perceived sweetness intensity of yogurts varying i
n sucrose concentration (1, 5.9, 10, 17.6 and 30% w/w). Consumption was mea
sured in the ad libitum consumption test. Results showed that the optimal s
ucrose concentration as determined by the taste-and-spit test (10% w/w) was
higher than that determined from the taste-and-swallow test, the fixed qua
ntity test and the ad libitum consumption test (5.9% w/w). The mean of with
in-subject's correlation coefficients between pleasantness ratings and amou
nt of yogurt eaten in the ad libitum consumption test were 0.45, 0.62, 0.75
and 0.81 for the taste-and-spit test, the taste-and-swallow test, the fixe
d quantity test and the ad libitum consumption test, respectively. The perc
eived sweetness intensity did not change as function of the sensory test us
ed. We conclude that pleasantness ratings collected after the taste-and-swa
llow test give a better prediction of consumption than the taste-and-spit t
est. (C) 1999 Elsevier Science Ltd. All rights reserved.