Academic article

The magic water test. An affective paired comparison approach to evaluate taste sensitivity in pre-schoolers.

Vennerød, Frida Felicia; Hersleth, Margrethe; Nicklaus, Sophie; Almli, Valérie Lengard

Publication details

Journal: Food Quality and Preference, vol. 58, p. 61–70, 2017

Publisher: Elsevier

International Standard Numbers:
Printed: 0950-3293
Electronic: 1873-6343

Open Access: none


Children in pre-school age present a limited cognitive development, a restricted vocabulary and a short attention span, which do not allow conducting classical taste sensitivity measures as developed for adults. The aim of this study was to develop and validate an age-appropriate protocol for sensitivity testing of the five basic tastes with pre-schoolers. One hundred and forty children aged 3–4 years (mean age 46.3 months, SD 3.4) performed a paired comparison task consisting of discriminating water from “magic water” by relying on their affect. In order to accommodate the age group, the task was gamified and did not require verbal responses. Sweet (sucrose, ranging from 0.94 to 4.32 g/l), sour (citric acid monohydrate, 0.20–0.38 g/l), umami (monosodium glutamate, 0.17–0.49 g/l), salty (sodium chloride, 0.34–0.98 g/l), and bitter (quinine hydrochloride dihydrate, 0.0014–0.0038 g/l) water dilutions were tested at four levels. Subsets of about 20 children per taste participated in retest sessions. Individual sensitivity scores were derived in two manners, following the experimenter’s success criterion (water with taste is classified as magic) and the child’s criterion for “magic water” (children’s individual magic-labelling strategies). Results show that the protocol was highly engaging, that the children performed the task consistently, and that the children’s performances were not dependent on cognitive level differences, thus validating the protocol for this study. On average, the children were clearly more sensitive to sour stimuli than to sweet and bitter stimuli, suggesting that the selected dilutions were not equivalent across tastes. Different sensitivity distributions of the children were obtained across taste modalities and scoring criteria. The key elements of an age-appropriate protocol for pre-schoolers are discussed.