THE SENSORY WHEEL OF VIRGIN OLIVE OIL

During a 3-year FLAIR study extra virgin olive oils, varying in specie s, degree of ripeness and extraction method, were evaluated by 6 diffe rent institutes according to QDA or COl-methods in order to identify p arameters related to the quality of extra virgin olive oil. The curren t COl-method yields a poor between-panel reproducibility. This could w ell be caused by a difference in the perception of positive quality as pects. Whereas the QDA-method is especially suitable for determining s ensory profiles according to the perception of the consumer, the COl-m ethod should be tailored to detect possible defects only. In order to cluster all attributes to one condensed set of sensory attributes for describing virgin olive oil, the COl and QDA data of all panels were p ooled and analyzed separately for appearance, texture and flavour. Thi s approach resulted in a set of 3 appearance, 3 texture and 12 flavour descriptors which can be conveniently represented graphically in the form of a ''sensory wheel''. On the basis of the findings it is recomm ended to base the ''extra virgin'' qualification for olive oils solely on the absence of defects. The between-panel reproducibility of such a simplified COl-test can be assessed by means of ring tests and impro ved by training with reference products. When an oil passes this scree ning it can be profiled subsequently using the attributes of the senso ry wheel. Such a profile can be linked to preferential profiles derive d from consumer studies enabling the production of most preferred oliv e oils.