A hierarchical approach to optimizing descriptive analysis multiresponse experiments

The design and improvement of products and processes often calls for experi ments where several response variables are analyzed simultaneously. Frequen tly, some of these variables are sensory attributes that can only be measur ed subjectively, through sensory evaluation panels or using expert opinion. In this paper we apply a multiresponse optimization procedure presented in Fogliatto and Albin 1997, to optimize a military food product where 24 sen sory attributes are evaluated through descriptive analysis techniques. Our objective is to select the best design and operating control factors consid ering all attributes simultaneously.