Page Content
Sensory panels are used routinely in the food and drink sectors to evaluate new products, to assess products in process, and to ensure the quality and consistency of existing brands. This study reports a novel postprocessing procedure for numerical sensory panel data that takes account of the unique features of data provided by individual panelists so that these individuals’ scores can be corrected for prior to deriving panel means. This treatment results in improved resolution between aggregated scores for products, and it is particularly relevant to situations in which specific products are routinely assessed by a pool of experienced tasters. Critically, this approach does not require any modification to existing sensory experiments, which can impact on the reliability of historical data.
