Descriptive analysis Dominique Valentin ENSBANACESG Universit de Bourgogne

Descriptive analysis Dominique Valentin ENSBANA/CESG Université de Bourgogne valentin@u-bourgogne. fr Descriptive analysis 1

What is descriptive analysis ? Descriptive analysis is used to describe the sensory characteristics of a product, and to use these characteristics to quantify differences between products Sensory profile radar Descriptive analysis 2

When should we use descriptive analysis ? Control quality Packaging Descriptive analysis Marketing R&D Descriptive analysis 3

Quality control Does this product match the target specification? What variation in quality is to be expected? What is the normal variation in each attributes How long can this product be stored before the Sensory quality noticeably changes? Descriptive analysis 4

Research & Development What are my product perceived sensory characteristics ? How can it be made more like the target profile? Have the changes brought the product closer to the target? If the recipe/process were changed, how would quality be affected Descriptive analysis 5

Marketing and Packaging What are the attributes of products already on the market? What are the attributes of the most and least succesfull products? What are the key attributes to meet consumers’ Expectations? Have the changes brought the product closer to the target? Descriptive analysis 6

Why should we use descriptive analysis ? Contrary to informal product tasting, descriptive analysis limits: Bias Subjectivity Poor control of variables Using wrong assessors Patchy information Descriptive analysis 7

What sort of panel is required ? Trained panel: more capable of describing the subtle differences between samples 2 types of trained panels: Internal: within the company External: outside the company Descriptive analysis 8

What type of methods should we use ? Conventional Profiling (QDA® Stone et al, 1974) Marketed by the Tragon Corporation in the USA Free Choice profiling (Williams and Langron, 1984) Profil Flash (Siefferman, 2000) Descriptive analysis 9

Conventional Profiling 4 -step process: 1. Panel selection 2. Development of a lexicon for appearance, odor, flavor, texture, mouth feel and aftertaste 3. Training 4. Ratings of the products Descriptive analysis 10

Panel selection Ø Ability to detect differences in characteristics present and in their intensities Ø Ability to describe characteristics using verbal descriptors and scaling methods Ø Capacity for abstract reasoning Ø Interest in participation and availability Ø General good health Prescreening questionnaires Acuity tests Ranking/rating tests Personal interview Descriptive analysis 11

Example of prescreening questionnaire Time: 1) Are there any weekdays that you are not available on a regular basis 2) How many weeks of vacation do you plan to take this year Health: 1) Do you have any of the following Dentures Diabetes Oral or gum disease Hypoglycemia Food allergies Hypertension 2) Do you take any medications which affect your senses, especially taste and smell? Descriptive analysis 12

Flavor, fragrance and texture quiz: 1) If a recipe calls for thyme and there is none available, what would you substitute? 2) What are some other food that taste like yogurt? 3) What is the best one- or two-word description of grated Italian cheese 4) Describe some noticeable flavors in cola 5) Describe some of the textural properties of foods in general 6) What are some textural properties of potato chips 7) Describe some of the noticeable smells in a bakery Meilgaard, Civille & Carr (1999), Sensory evaluation techniques. CRC Press Descriptive analysis 13

Exemples de tests de sélection: jury bières Chollet, 2000 Matching odors / names Descripteurs Composes Concentration (mg/L) Banane Amande amère Citron Vanille Beurre Rose Eau Chou Fleur d’oranger Acétate d’isoamyle Benzaldéhyde Citral Vanilline 2, 3 -butanédione Géraniol Eau distillé Sulfure de diméthyle Linalol 5 2 2 20 5 5 – 0, 05 5 Descriptive analysis 14

Exemples de tests de sélection: jury bières Duo-trio test : • Control sample: Kronenbourg beer • Test sample: Heineken Identification of 4 basic tastes Descriptors Compound Concentration (g/l) Sweet Sour Bitter Astringent Sucrose Citric Acid Caffein Aluminum Sulfate 8 1 0, 5 Descriptive analysis 15

Selection test example: beer panel Intensity ranking: sweet taste in beer Sample Concentration (g/l) 1 2 3 4 0 5 10 15 Intensity ranking: Bitterness in beer Bitterness unit (EBU) K Kronenbourg 1664 Tradition allemande 4 15 21 28 Descriptive analysis 16

Exemples de tests de sélection: jury bières Odor description Composés Descripteurs Camphre Phényl acétaldéhyde Acide octanoique Trans-2 -hexenol Hexanoate d’éthyle -Décanolactone 2, 5 -diméthylepyrazine Isobutyraldéhyde Salycylate de méthyle Carvone Camphre jacinthe, lilas Caprique Herbe coupée, feuille vertes Pomme, fruité, sucré, anisé Pêche, abricot, noix de coco Céréales Banane, amande Pharmacie, pommade Menthe, chlorphylle Descriptive analysis Concentration (mg/l) 1 5 50 50 5 20 20 100 100 17

Examples of scaling exercises Mark on the line at the right to indicate the proportion of the area that is shaded None All Meilgaard, Civille & Carr (1999), Sensory evaluation techniques. CRC Press Descriptive analysis 18

If you want to know more: Guidelines for the selection and training of sensory panel members ASTM special technical publication 758 Sensory analysis -- General guidance for the selection, training and monitoring of assessors -Part 1: Selected assessors, ISO 8586 -1: 1993, Part 2: Experts, ISO 8586 -2: 1994 Descriptive analysis 19

Conventional Profiling 4 -step process: 1. Panel selection 2. Development of a lexicon for appearance, odor, flavor, texture, mouth feel and aftertaste 3. Training 4. Ratings of the products Descriptive analysis 20

Development of a lexicon Goal: Describe the product category completely without overlapping Procedure: Ø Word generation Ø Reduction of the list into a working list Ø Choice of referents and definitions for the descriptors Descriptive analysis 21

Word generation 1. Collect commercially available samples which represents as many as possible of the attribute differences likely to be encountered in the product category 2. Present 4 samples to the panelists and ask them to write down independantly as many terms as they can to describe the sensory characteristics fully 3. Compile the terms given by all the panelists to form a list of descritors Descriptive analysis 22

Reduction of the list 1) Rearrange the list in family of terms (appearance, aroma, textures …) 2) Discuss with panelists to suppress: Hedonic and quantitative terms Redundant terms (synonyms) Non relevant terms Descriptive analysis 23

Example of lexicon List of original terms pomme fruit pomme cuite chimique poire caramel fermenté coing vert-pas mûr Final list cidre vanille gras pomme verte pruneau pomme citron colle blanche Not relevant Not discriminant Synonyms Descriptive analysis pomme fraîche coing cidre pomme verte caramel vanille pruneau poire métallique gras 24

Examples of referents and definitions: Soy yogurts Texture Thickness: consistency of the mass in the mouth Rate of Melt: amount of product melted after a certain pressure of the tongue Graininess: amount of particle in mass Mouth coating: amount of film left on the mouth surfaces Basic tastes Sweet: Sucrose Sour: lactic acid Bitter: caffeine Salty: sodium chloride Trigéminal Astringent: Shrininking or pukering of the tongue surface caused by tanins or alumn Arôme Water: taste like water down Wood: cutting from pencil sharpening Milk: whole milk Cream: crème fraiche Earthy: earth Flour: 1 spoon of flavor mixed in water Chalk: smecta Raw pie crust: commercial raw pie crust Hazelnut: : hazelnut powder Mushroom: dry mushrooms soaked in water Descriptive analysis 25

Procedure for manipulation of the samples 1) First compression (e. g. thickness) Place ½ spoon of sample in mouth and compress between tongue and palate 2) Manipulation (e. g. rate of melt, graininess) Compress sample several more times (2 to 3 times) 3) After feel (e. g. mouth coating) Swallow or expectorate Descriptive analysis 26

Conventional Profiling 4 -step process: 1. Panel selection 2. Development of a lexicon for appearance, odor, flavor, texture, mouth feel and aftertaste 3. Training 4. Ratings of the products Descriptive analysis 27

Training Goal: assessors need to be able to detect and repeatedly quantify the sensory characteristics corresponding to the lexicon Need to be exposed to a wide range of products Need to be familiarized with the use of the scale Need to receive feed back to rectify any problem Descriptive analysis 28

For each attribute: 1) Ranking tests Panelists receive 4 samples covering the range of differences likely to be encountered in the product category and are asked to rank them according to the attribute Sweetness Descriptive analysis 29

For each attribute: 2) Rating tests Panelists receive 4 sample and are asked to score them on an intensity scale Sweet Very sweet Not sweet Start with large differences, then collect samples which represent smaller differences within the product category Encourage panelists to refine the procedures for evaluation and terminology with definitions and references Descriptive analysis 30

3) Final practice should approach the testing situation in which the panel will be used: Same type of products Same experimental conditions Several replicates Panelists should receive feed back on their performance compare to that of the panel Samples 275 391 856 912 057 Average 6. 3 5. 7 1. 2 3. 6 2. 5 Descriptive analysis Panelist ## 12 Panelist 2. 3 5. 7 1. 8 4. 2 4. 6 2. 3 2. 4 3. 7 3. 5 1. 9 31

Conventional Profiling 4 -step process: 1. Panel selection 2. Development of a lexicon for appearance, odor, flavor, texture, mouth feel and aftertaste 3. Training 4. Ratings of the products Descriptive analysis 32

Code: 592 Épais Pas du tout Très Pas du tout Très Gras Fondant Sucré Acide Amer Salé Astringent Descriptive analysis 33

FAQs How many assessors do we need ? How many samples are there ? How many replicates are necessary ? Ø Normally 3 assessments per sample/assessors, however this is very expensive in time and resource. Ø If the panel is fully trained and constantly evaluating the same type of products then a single assessment is likely to provide reliable data otherwise 1 or 2 replicates are advisable. Descriptive analysis 34

How many training sessions? Ø The number of training sessions depends on the complexity of the product, on the number of attributes to be covered, on the requirement for validity and reliability Ø A more experienced panel will provide greater detail with greater reproducibility but a short version with fewer and simpler attributes is acceptable for control quality for example Descriptive analysis 35

Many references exist to learn about sensory profiles These include: Sensory Evaluation Techniques, Third Edition Meilgaard, Morten C. Civille, Gail Vance Carr, B. Thomas CRC Press Sensory Evaluation Practices Herbert Stone, Joel L Sidel Academic Press Sensory Evaluation of Food Principles and Practices Harry T. Lawless, Hildegarde Heymann Aspen press Descriptive analysis 36