Résumé
Two four-way multiblock methods, Generalized Orthogonal Multiple Co-Inertia Analysis (GOMCIA) and DOuble-Analyse Conjointe de Tableaux (DO-ACT) are used as a tool to compare the agreement of two experts or consumer panels. GOMCIA is a PLS-based method, while DO-ACT is a generalization of STATIS. Both methods are capable of analysing four-way multiblock data tables (i.e. two multiblock data tables) and take into account the block structure of the data. The main advantage of these two methods is that they can evaluate the agreement between the panels and the agreement between individual assessors within each panel simultaneously. These two methods are used to compare two consumer panels of children who evaluated eight chocolate bars on 14 sensory descriptors. With these methods, all the original data set is used without averaging over one direction of the multiblock. The results show great potential of both GOMCIA and DO-ACT for sensory data analysis.