Correlating analytical data with sensory evaluations using untrained assessors

Abstract: Traditionally, trained sensory panelists are often required when attempting to correlate sensory measurements with analytical data. However, there are instances where access to a trained panel is not an option, e.g., limited time/funding/resources, difficulty in recruiting/training/retaining panelists, etc. In this presentation, we will show how a small group of untrained assessors can successfully correlate sensory with analytical data and provide direction to a trained panel for more formal in-depth evaluations.
In this research, a small group of scientists (n=6) evaluated expeller-pressed chili pepper extracts for the presence or absence of lipid oxidation notes using aroma only. The samples were blind-coded and presented in a randomized order. Assessors were given fresh sunflower oil to use as a fresh oil reference. Meanwhile, the same samples were analyzed for total aldehydes content using static headspace GC-FID.
Results showed a strong, positive relationship between total aldehydes and the frequency of lipid oxidation detection by the untrained assessors. We will demonstrate how the use of untrained assessors can quickly and efficiently help reduce the number of samples needing to be formally evaluated by a trained panel.