Identification of confusable odours including wines: appropriate labels enhance performance

This study examined how the effectiveness of training people to identify confusable odour samples depends on the kind of label used in the task. In Experiments 1–3 participants were required to identify three compound odour samples (AX, BX and CX) containing a common element, Citral (X), using a multiple-choice test (A vs B vs C) on each trial. Participants given appropriate labels for the stimuli (Appropriate condition) performed better than participants who produced their own labels, but only in the first training block (Experiment 1). Participants given appropriate labels also performed better than participants trained to give inappropriate (Experiment 2) or irrelevant labels (Experiment 3). Subsequently, wine samples of similar discriminability to the compound odour samples (Experiment 4) were used as stimuli in the same kind of identification task. Novices were able to apply descriptors and grape names, e.g. ‘Shiraz’, at a level above chance (Experiment 5), but not as well as the Appropriate groups in Experiments 1–3. The procedure provides a measure of the degree to which a particular label or description is appropriate.