Sensory Evaluation and Applied Statistics

This à la carte module enables students to deepen their skills in sensory analysis, experimental design, mixture design and advanced statistics. It will be useful for students wishing to become engineers in research and development or in sensory evaluation.
Indeed, sensory evaluation research enables the global analysis of a product through the use of complex statistics. Another key skill is the use of data collection software. During this module, students will learn how to program a sensory analysis session on Fizz software, which is widely deployed in the food industry.
Design of experiments and mixtures are essential methodologies in food companies when developing new products. The course shows how to develop an optimal experimental study strategy in practice, and how to make reliable use of experimental results.
Statistics is a vital engineering tool today, particularly in the fields of biology and agri-food. The aim of this course is to provide engineering students with a theoretical and practical knowledge of the tools needed to understand large panels of data. These tools will enable them to understand the main information provided by the dataset and thus guide their statistical approach (selection of statistical tests according to objectives and interpretation of results).
This module is compulsory for the INH specialization and is of interest for the LAI specialization. (max. 15 students per group).

Mastery of the basic tools of univariate and bivariate descriptive statistics on quantitative and categorical variables.
PORES/PHYME and PLEXP courses

Factors influencing tasting (S. Tempère: 2 CM)

Sensory evaluation (P. Lafenetre: 4CM and 5TD and 1 TP of 4h)

In-depth study of sensory analysis tests
Global analysis of sensory profile results
Programming of Fizz sensory evaluation software
Design and implementation of a sensory evaluation experiment


Experimental designs (R. Savoire: 3 CM + 2 TD)

Second-order designs (Box-Behnken and centered composite)
Taguchi designs
Multi-response optimization

Mixture design (F. Arnal: 4CM + 5TD)

Implementation problems, characteristics of mixtures
Mixture designs without constraints
Mixture designs with upper and lower constraints
Mixture designs with upper and lower constraints
Mixture designs with relational constraints

Data analysis (L. Bombrun: 3 CM, 5TD)

Introduction and generalities on data analysis.
Factorial methods: Principal Component Analysis, Correspondence Analysis (Simple and Multiple).
Unsupervised classification methods: Hierarchical Ascending Classification, Moving Center Method.

Project progress (P. Lafenetre F. Arnal L. Bombrun R. Savoire : 3 TD)

Regular project progress reports.

References
Sensory analysis. Recueil des normes AFNOR. 7th edition.
Probabilités, Analyse des Données et Statistiques, Gilbert Saporta, Editions Technip, June 2006
Approche pragmatique de la classification, Josiane Confais, Jean-Pierre Nakache Editions Technip, November 2004
L'analyse des données tome 2 : l'analyse des correspondances, Jean-Paul Benzecri Bordas, Bordas, 1980
Les plans d'expériences : les mélanges Jacques Goupy Dunod 2000
Les plans d'expériences Gilles et Marie-Christine Sado AFNOR Technique 2002
Pratiquer les plans d'expériences Jacques Goupy Dunod 2005