A New Approach to Sensory Evaluation

Flavour is the significant criterion of evaluation in beer tasting, which depends on many primary and secondary causes. This includes desirable and un-desirable aromatic compounds, but also haptic and physical issues like viscosity, the CO2-level and colour. A beverage for example with a nontypical colour presented in a wrong drinking vessel is always evaluated differently to the “correct” form without any modifications.

Next to these problems taste in general is a subjective matter and is valued individually quite different. Furthermore the influences of process changes on taste are difficult to evaluate. In order to get representative and fast results in the brewing and beverage technology, a sensory test developed by Sommer was used [1]. The test consists of a dual testing, therefore one sample is the product with the original taste and the second one is the product with a changed aromatic profile. The proband has to detect the changed sample. For the case he is not able to determine it, he has to guess. This test is repeated several times with a changing taste difference of both samples. New about this procedural is the interpretation of the testing results by using a probabilistic evaluation form. Thereby an intensity curve is given which indicates the relation between rising taste difference and consumer taste perception.

The needed requirements are the same as for every taste test. The relative number of test persons, who can detect a difference in change, remains almost constant in a group and more or less independent from external influences. Individual errors follow statistic behaviour and can be assessed if the number of tasters is big enough. The significance of results from a smaller group of experienced tasters in a sensory panel can be improved by increasing the number of test persons even if they are less experienced.

The results of this work are based on differentiation trials with beer of different concentrations of Benzaldehyde as a typical off-flavour compound [2]. The usage of this compound is caused by its concentration depended nontoxity and the well known marzipan taste. In a group of inexperienced tasters, mainly young food technology students, qualified differentiation showed high conformance and significance. This test indicates that the method with the incorporation of statistical methods can be used efficiently for the evaluation of process changes and their effect on beer taste. For the performed Benzaldehyde beer mixing process the developed functional coherence can be applied for economic process optimisation. So for the possibility of creating a new and atypical marzipan beer which can taste 99 % of the consumers the needed dosage of Benzaldehyde could be calculated.

BrewingScience - Monatsschrift für Brauwissenschaft, 63 (November/December 2010), pp. 128-132