ABILITY TO DISCRIMINATE BETWEEN RELATED ODOR MIXTURES

The purpose of this study was to investigate (i) the ability of human subjects to detect changes in related odor mixtures following omission of components, and (ii) the possible relationship between the number of substances comprising a mixture and the ability to detect changes in its composition. A total of 36 panelists was tested repeatedly in five separate sessions using sniff bottles, and asked to rate the difference between pairs of stimuli using a 6-point verbal scale. Stimuli were 3-, 6- and 12-component mixtures paired with stimuli from which one or more of the components had been omitted or had been left unchanged. Rating behavior was highly consistent, with no significant differences found between sessions or panels in judging the identical control pairs. Although asymmetrical pairs were judged to be different significantly more often than control pairs, correct assignment was considerably poorer than expected. Correlations between complexity of a stimulus and ability to discriminate changes in its composition were weak and possibly confounded by substance-specific, synergistic effects. While factors such as lack of feedback or the arbitrary nature of the stimuli may have affected performance, it remains to be investigated whether the findings indeed reflect our limited capacity to discriminate between related odor mixtures.