Procrustes multiple factor analysis to analyse the overall perception of food products

We propose a methodology to analyse sensory data collected in the form of sensory differences between products. This methodology provides a synthesis of all the assessors’ configurations integrating both advantages of Generalized Procrustes Analysis (GPA) and Multiple Factor Analysis (MFA). It consists in calculating a MFA consensus configuration by procrustes rotations of the individual configurations (hence the name of Procrustes Multiple Factor Analysis: PMFA).This methodology is applied to highlight the relative typicity of white wines of the Val de Loire made of Chenin or Sauvignon.