Sensory Analysis of Italian Dry-Cured Sausage : Checking of Panel Performance

Systematic training and full knowledge of food properties are basic requirements for panelists asked to test products by descriptive analysis. In some cases routine quality control is carried out prior to marketing by a less experienced peripheral panel at the plant, with a trained pool at the central laboratory performing occasional checking of the same product. The performances of two panels (a fully-trained and a half-trained one) assessing the same set of Italian-type Milano salami have been compared in this study. The Generalized Procrustes Analysis (GPA) of data showed that the half-trained panel was less consistent and less capable of discriminating between samples. In a subsequent analysis, a data base, made up by the sensory data enriched with 11 chemical variables collected from the same samples was analysed with GPA, showing incorrect perception of two major attributes (acid and pungent) by the same panel. Results from descriptive testing of raw sausage in this study suggest that descriptive testing deserves systematic effort to find and rule out inconsistent or unreliable rating as sources of potential failure in quality assessment. The combined use of descriptive profile and GPA proves a powerful yet simple tool for the checking of panelperformance.