Measurement uncertainty in wine appreciation

Alemán. In an accredited laboratory measurement uncertainty must be considered within the evaluation of the assessment of relevant parameters. The crucial point as to determine for wine appreciation is whether a value does or does not differ significantly - with due regard being given to measurement uncertainty - from the preset limit that is regulated by law or acknowledged by experience. If the measured analytical value is really exceeding this crucial point - the decision limit -, this may give rise to complaints against the product. Several concepts for estimating the measurement uncertainty and its further consideration in wine appreciation by defining the decision limit are discussed, e.g. the expanded measurement uncertainty, the decision limit CC\alpha\ according to Commission decision 2002/657/EC and the critical difference according to the ISO 5725 norm. In principle, decrease of measurement uncertainty will be obtained by performing repetitive successive measurements. The recommendations of ISO 5725 specify the role of the critical difference for the determination of the decision limit. The examination of the critical difference depending on the ratio of repeatability/reproducibility limit and on the number of successive measurements shows that distinct reduction of the critical difference can be obtained by carrying out the n-repetition of measurement with approximating precision limits only. In the area of wine analysis, however, the accentuation of measurement uncertainty proved to be slight in fast, persistently slight even in the case of multiple measurement repetition. The differences found among decision limits comparison of the various concepts having been used respectively to pay tribute to measurement uncertainty, could be considered as relatively minor shown by the chosen examples. From the point of view of wine control, the application of a one-tailed distribution approach for instance results in accentuating the limits, but the basic problem formulation and research hypothesis, includes specific requirements for the analysis of natural / native components must be taken into account. [AZ 35] (Auth. abstract)