Research in the field of sensometry makes particular use of so-called multivariate methods, which consider large volumes of data and all variables simultaneously. Multivariate methods are very important for understanding the totality of the data and for being able to predict quality. They cover for example quality control of panels of assessors, segmentation of consumers and connections between sensory perception and consumer data.
Describing consumers
Food products are perceived by consumers both as properties (sour, soft, sweet) and as preferences (big, good, nasty). Put together, these create the basis for the perceptions we form (must have, can wait, uninteresting) and our actions (buy, throw away, eat). All along this entire chain of events - from who we are, via our intended behaviour to our actual behaviour - there lie some fascinating connections. What makes the statistical modelling of this so exciting is that the connections are not always quite clear, because the exceptions and variations are many.
In sensometry we are developing:
- methods for understanding the interaction between products' sensory properties, product information and consumer surveys.
- methods for segmenting consumers and identifying individual differences
- analyses of sensory descriptions
Sensometry is a fast-growing technical field that has been developed since the 1980s; it has gained ever more significance as competition in the market has increased and requirements for quality documentation have become more stringent.
