Multivariate statistics as a classification tool in the food laboratory

Multivariate statistics Is a powerful tool for answering complex questions like the geographical origin of wine. Instead of expensive tests such as stable isotope measurements, less costly routine analysis data can be used. Problems like the choice of analysis parameters, the size of reference data, and the best suited statistical technique are addressed. Multivariate statistics was used very successfully with real samples in an official Swiss wine-tasting championship. Although the examples presented in the paper deal with wine, the principles, opportunities, and shortcomings may also apply to other food products.