chapter  8
18 Pages

Sensometrics: the application of multivariate analysis to sensory data

I. Introduction.................................................................................................. 206 II. The use of multivariate analysis in examining

sensory data: application examples ......................................................... 207 A. Cinnamon-flavored cereal.................................................................. 207

1. Materials and methods ................................................................ 207 2. Results............................................................................................. 208 3. Conclusions.................................................................................... 208 4. Looking to the future ................................................................... 210

B. Cheese sauce ........................................................................................ 210 1. Materials and methods ................................................................ 210 2. Results............................................................................................. 211 3. Conclusions.................................................................................... 213 4. Looking to the future ................................................................... 214

C. Chocolate .............................................................................................. 214 1. Materials and methods ................................................................ 214 2. Results............................................................................................. 215 3. Conclusions.................................................................................... 220 4. Looking to the future ................................................................... 220

References ............................................................................................................ 221

Multivariate analysis is a tool used by the sensory professional to simplify and aid in the interpretation of descriptive, consumer, and analytical data. Multivariate techniques are mainly used in data reduction (Principal Component Analysis [PCA], Factor Analysis [FA], Correspondence Analysis, etc.), for classification (FA, hierarchical cluster analysis, discriminant analysis, etc.), and data relationships (multiple regression, principal component regression, partial least square analysis (PLS)) (Piggott, 1986). In the industry, it can be used in product development to guide R&D reformulations, aid in understanding process variation (e.g., plant-to-plant, line-to-line, or within line or scale-up), monitor and identify panelist performance issues, shelf life studies, etc. Multivariate analysis can also be used to understand and identify key attributes driving consumers’ liking, segmentation (e.g., age, gender, geographic region), and reducing the number of attributes on a ballot. The opportunities are endless. In practice, sensory professionals have ample opportunities every day to use multivariate techniques to understand the sensorial world around them.