A wine is poured, two people taste it – and may report different impressions. Even the same person may judge it differently at a later tasting. That does not make taste arbitrary. Rather, it shows that perception depends on conditions and calls for questions that can be tested: Can assessors detect a difference? Do they describe it in similar terms? Do they reach comparable judgements when the test is repeated?
The 15th German Sensory Days at Hochschule Geisenheim University on 5 and 6 November 2026 provide an opportunity to examine these questions more closely. Sensory analysis is not about establishing a single, universally valid sense of taste. It investigates what people perceive through their senses – and how reliably those perceptions can be recorded under specific conditions.
What exactly is being tested?
A sensory test begins not with a verdict of “good” or “bad”, but with a clear question. Is there a perceptible difference between two wines? How pronounced is a particular characteristic? Or which product do consumers prefer? These are different tasks: a preference tells us something about liking, while a discrimination test tells us whether a difference can be perceived. The two should not be confused.
The choice of assessors also depends on the question. Trained assessors can, for example, describe which characteristics they perceive and how intense they judge them to be. Consumer surveys, meanwhile, investigate how products are received by a particular target group. Hochschule Geisenheim University describes both analytical tests and consumer testing as activities carried out in its sensory laboratories. The appropriate method depends on what you want to find out.
Training makes impressions more comparable
Training is not intended to teach all assessors to have the same taste. Rather, it helps them distinguish between perceptions and judgements, align their terminology and apply scales consistently. References – such as clearly defined comparison samples for a particular characteristic – can help clarify what a term means. This does not turn a description such as “intense” into an objective truth, but it does give it a clear frame of reference.
Even experts are not measuring instruments free from variation. People differ in their sensitivity and in how they name their impressions. That is why quantitative descriptive analysis looks beyond the mean: it also matters whether individual assessors can distinguish between products, how closely they agree within the group and whether they can repeat their own judgements. The ISO 11132 standard sets out criteria for assessing the performance of a trained panel and its members.
Test design: small details, big effects
Knowing what is in the glass can allow expectations to influence perception. In a blind test, the wine’s identity and brand are therefore concealed. Random codes and varying the order of presentation make it harder to identify a sample from its position or label. Standardised glasses, serving sizes and temperatures reduce other influences. Breaks are important because tasting several samples in succession can cause fatigue or affect perception.
Such precautions do not eliminate bias and variation entirely. They limit them and make the test fairer. It is equally important to establish the question and the method of analysis in advance: a difference picked out retrospectively from many possible impressions is less convincing than an expectation tested specifically.
Repetition puts reliability to the test
Testing the same sample again under similar conditions allows repeatability to be assessed: does an individual or a panel arrive at comparable results? This is not the same as reproducibility. To assess that, one would also need to consider whether the results hold under changed conditions – for example, with a different panel or at a different time. Even good repeatability does not prove that a judgement applies to everyone or that a preference is “right”.
This reveals the limits of what sensory analysis can tell us: its findings describe people’s perceptions within a particular testing procedure. They can provide reliable evidence of differences, characteristics or acceptance, provided the question, training and test design are appropriate. The result nevertheless remains tied to those conditions. A personal tasting impression is therefore neither a mere whim nor an objective measurement. Scientific sensory analysis makes it more testable – and more precise about what it can and cannot tell us.
Original reviewed by the editors · 5 November 2026
