Sensory Evaluation

Sung Eun Choi, PhD, RD © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALEChapter OR DISTRIBUTION objectives NOT FOR SALE OR DISTRIBUTION THe STuDenT wIll be emPowereD To: Ⅲ Identify the sensory characteristics of food. Ⅲ Discuss the factors affecting the outcomes of sensory evaluation. Ⅲ Demonstrate an© understanding Jones & Bartlett of the process Learning, for sensory LLC evaluation tests. © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Ⅲ Formulate an effective sensory evaluation strategy by selecting appropriate test design, panelists, and instruments. Ⅲ Discuss how to analyze and interpret the sensory data and recognize specifi c methodological advances related to sensory evaluation. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 83 29/12/12 11:36 PM 84 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION IntroductionNOT FOR SALE OR DISTRIBUTION Determining how food products affect consumers’ senses is one of the most important goals of the food industry. It also is a primary concern for nutrition- ists and dietitians who develop healthier recipes. Because our five senses act © Jones & Bartlettas the Learning,gatekeeper of LLC our bodies, the benefits of© healthy Jones food & Bartlettwill be reaped Learning, LLC only if our senses accept it. Therefore, consumer reaction, as perceived by NOT FOR SALEthe OR five DISTRIBUTION senses, is considered a vital measure ofNOT food FORdevelopment. SALE ORBecause DISTRIBUTION no apparatus can substitute for the senses in evaluating food, humans are used as test subjects. Such studies are becoming more prevalent, despite the potential biases of humans and the costs involved. sensory© Jonesevaluation & BartlettThe scientific Learning, LLCSensory evaluation is a scientific© Jones method& Bartlett that evokes, Learning, measures, LLC ana- measurement method of food quality lyzes, and interprets responses to products, as perceived through the senses of based NOTon sensory FOR characteristics SALE OR as DISTRIBUTIONsight, smell, touch, taste, and sound.NOT1 ThisFOR widely SALE accepted OR DISTRIBUTION definition is used perceived by the five senses. by sensory evaluation committees within various professional organizations, including the Institute of Food Technologists and the American Society for Testing and Materials.2 Like other scientific methods of taking measurements, © Jones & Bartlett Learning, LLC sensory evaluation© Jones is concerned & Bartlett with precision, Learning, accuracy, LLC and sensitivity and with avoiding false-positive results.3 Reliable sensory evaluation is based on NOT FOR SALE OR DISTRIBUTION the skill of theNOT sensory FOR analyst SALE in optimizing OR DISTRIBUTION four factors: definition of the problem, test design, instrumentation, and interpretation of the results:4,5 • Definition of the problem: The item to be measured must be defined precisely. © Jones & Bartlett• Learning,Test design: LLC Not only must the design take© Jones into account & Bartlett unknown Learning, LLC NOT FOR SALE OR DISTRIBUTIONsources of bias, but it also must minimizeNOT the FORamount SALE of testing OR DISTRIBUTION required to produce the desired accuracy of the results. • Instrumentation: The panelists must be selected and trained to give a reproducible result. • Interpretation of the results: The analyst should select appropriate © Jones & Bartlett Learning, LLCstatistics based on the ©correct Jones statistical & Bartlett assumptions Learning, and draw LLC NOT FOR SALE OR DISTRIBUTIONonly those conclusionsNOT that are FOR supported SALE by OR the DISTRIBUTION data. The benefits of well-performed sensory evaluation can be realized in many ways; however, if the sensory analyst fails to optimize one of the four factors, much time and money is wasted. For effective sensory evaluation, the analyst should duly recognize the purpose of the study, select the appropriate © Jones & Bartlett Learning, LLC experimental design,© Jones use panelists& Bartlett who Learning, fit the purpose, LLC choose the proper NOT FOR SALE OR DISTRIBUTION method for preparingNOT FOR and presenting SALE OR the DISTRIBUTION samples, and analyze the data correctly. A sensory researcher should always consider whether the method is appropriately implemented and whether errors have been introduced at any stage of the experiment. In this chapter, the principles of sensory evaluation will be introduced. © Jones & BartlettExamples Learning, will be providedLLC to demonstrate the© use Jones of sensory & Bartlett evaluation Learning, LLC NOT FOR SALEtechniques OR DISTRIBUTION and the application of the results towardNOT developing FOR SALE and modify OR DISTRIBUTION- ing food recipes.

The Human Senses © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC The characteristics of food are perceived by the five senses: sight, smell, taste, NOT FOR SALE OR DISTRIBUTIONsound, and touch. NOT FOR SALE OR DISTRIBUTION

Sight The eyes perceive the initial quality of food, receiving such information as © Jones & Bartlett Learning, LLC color, size, shape,© Jones texture, &consistency, Bartlett andLearning, opacity. Light LLC entering the lens of NOT FOR SALE OR DISTRIBUTION the eye is focusedNOT on theFOR retina, SALE where OR the DISTRIBUTION rods and cones convert it to neural

03441_CH03_Printer.indd 84 29/12/12 11:36 PM THe Human SenSeS 85

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC 5 NOT FORimpulses SALE thatOR travel DISTRIBUTION to the brain via the optic nerve. PerceptionNOT FOR by the SALE visual OR DISTRIBUTION system of light of wavelengths 400–500 nanometers (nm; blue), 500–600 nm (green and yellow), and 600–800 nm (red) is commonly expressed in terms of the hue, value, and chroma of the Munsell color system.5 Color may accurately indicate ripeness, strength of dilution, and the degree to which the food© Jones has been & heated. Bartlett Color Learning, is used to evaluateLLC a food’s © Jones & Bartlett Learning, LLC desirability and acceptability.NOT FOR Greenish SALE bananas, OR DISTRIBUTIONburnt meat, and dark brown NOT FOR SALE OR DISTRIBUTION avocado send visual signals that can change a person’s choices. Color often triggers certain expectations in the mind; for example, the creamy color of vanilla ice cream evokes an expectation of richness. But, color can be deceiving. The quality of food can be masked by changes in color.© Jones For instance, & Bartlett if yellow Learning, coloring is added LLC to a food without actual© fatJones & Bartlett Learning, LLC havingNOT been FOR added, SALE the quality OR DISTRIBUTION of low-fat products can be improved. ColorNOT FOR SALE OR DISTRIBUTION changes alone can increase a food’s acceptability considerably. Even small visual details such as adjacent or background colors and the relative sizes of areas of contrasting color can affect a consumer’s perception. Parameters of size and shape, such as width, length, thickness, particle size, © Jones &geometric Bartlett shape, Learning, and distribution LLC of pieces, also provide© Jones information & Bartlett on Learning, LLC NOT FORfood SALE quality. OR The DISTRIBUTION dullness, shininess, smoothness, or roughnessNOT FOR of aSALE surface OR DISTRIBUTION and the clarity of liquids evoke preconceptions about the food. See Special Topic 3.1 for more information on some of the latest trends chefs use to create dishes that delight the senses.

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Special TopicNOT 3.1 FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Molecular Gastronomy Sarah Churchill © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC TheNOT concept FOR of molecular SALE gastronomy OR DISTRIBUTION was founded by Hervé This, editor at Pour la ScienceNOT, FOR SALE OR DISTRIBUTION and Nicholas Kurti, a low-temperature physicist. They coined the term molecular molecular gastronomy Form gastronomy to encompass all the physical and chemical changes that occur during of modern gourmet cooking that food production and cooking. In an effort to identify the methods for creating the best seeks to identify all the physical and chemical changes that occur during flavor and texture, they compared existing recipes with old proverbs and old wives’ tales.1 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning,food production LLC and cooking to In 1992, they established the first international Workshop on Molecular and Physical create the best flavor and texture. NOT FOR SALEGastronomy. OR 2DISTRIBUTION The workshop still takes place today; however, theNOT field ofFOR molecular SALE OR DISTRIBUTION gastronomy has moved from the scientific realm into the media limelight, becoming a cultural phenomenon.

Today, popular culture© uses Jones the term &molecular Bartlett gastronomy Learning, to refer to LLC the way modern chefs innovate with© ingredients Jones and& Bartletttechniques Learning, LLC to bring excitement to NOTthe dining FOR experience. SALE The ORpractice DISTRIBUTION of molecular gastronomy also has been referredNOT to as culinary FOR constructivismSALE OR, DISTRIBUTION experimental cuisine, molecular cooking, modernist cuisine, culinary deconstructivism, and progressive cuisine. For example, the modern chefs Ferran Adriá of el Bulli, Heston Blumenthal of The Fat Duck, and Grant Achatz of Alinea are attempting to understand the chemical and physical nature of cooking as well as the best ingredients and techniques to improve upon traditional methods. They accomplish these goals© Jones through the& deconstructionBartlett Learning, of certain recipes, LLC the transformation of the physical© Jonesstates of foods & Bartlett (i.e., gases, Learning,liquids, solids), andLLC the use of different equipment and ingredients to change cooking methods.3 NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Such chefs do not necessarily ascribe their innovations to molecular gastronomy, however. Adriá and Blumenthal say they go beyond the mere scientific exploration promoted by This and Kurtis. They also dispute the claim that the cuisine they are creating is pure novelty. They explain that they view food with a certain openness and embrace new ingredients and cooking methods without forsaking tradition. They © Jones & Bartlettaffirm that Learning, their methods helpLLC them realize the full potential of© each Jones ingredient & Bartlettby questioning Learning, traditions about LLC the best way to cook a food NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 85 29/12/12 11:36 PM 86 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION and finding the best way to maintain flavor. The result is a unique cuisine that gives © Thomas_EyeDesign/iStockphoto.com consumers a chance to engage and confuse all of their senses.4 What innovations are used to create this bold new cuisine? Modern chefs use centrifuges, syringes, freeze© dryers, Jones blast & chillers, Bartlett and carbon Learning, dioxide dispensers. LLC © Jones & Bartlett Learning, LLC Canapés can be created fromNOT centrifuged FOR frozenSALE peas OR to create DISTRIBUTION a buttery spread. NOT FOR SALE OR DISTRIBUTION A syringe can even be used to inject a small amount of unexpected flavor. Sometimes molecular gastronomy turns a traditional dish on its head. For example, instead of pasta with grated cheese, how would you like cheese with grated pasta? Parmesan noodles© can Jones now be made& Bartlett from boiling Learning, the cheese, pressing LLC out the water, and © Jones & Bartlett Learning, LLC passingNOT it through FOR a pastry SALE bag; ORthe cheese DISTRIBUTION noodles can be topped with grated freeze- NOT FOR SALE OR DISTRIBUTION dried pasta.5 New ingredients at the modern chef’s disposal include xantham gum, alginate, calcium salts, soy lecithin, agar-agar, and liquid nitrogen. These ingredients are becoming as © Jones & Bartlett Learning, LLC © Jones & BartlettThe kitchen Learning, of a chef cooking LLC using molecular prevalent in the modern chef’s pantry as spices. They can be used to make gels, foams, gastronomy techniques may look more similar to a NOT FORand SALE spheres OR without DISTRIBUTION adding unwanted flavor.6 Liquid nitrogen can beNOT used to FORfreeze SALEhigh-end OR chemistry DISTRIBUTION or physics laboratory than the something immediately and even make it shatter. Soy lecithin is crucial in making kitchen of the local greasy spoon. A fusion of chemistry and cooking, molecular gastronomy seeks to transform foams because it helps emulsify and hold ingredients together. Spheres resemble the common cooking ingredients into novel forms through texture and appearance of caviar but can be made using any liquid. the use of new technologies and techniques. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC With the help of new technology and new ingredients, these chefs are able to create NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION unique, artistic dishes. Chefs want to amaze consumers, and they really have. Whatever this cuisine should be called, it has become very popular. Of the 50 best restaurants in the world, the top 3 are associated with molecular gastronomy. Diners are becoming more experimental in their choices, even bravely trying tobacco- or crab-flavored ice cream. Innovation in cuisine is always progressing. With increased knowledge and technology, we’ll just have to wait to see what comes next! © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC ReferencesNOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 1 Molecular gastronomy: Food science raises the culinary bar. Env Nutri. 2010;33(9):7. 2 Blanck JF. Molecular gastronomy: Overview of a controversial food science discipline. J Agr Food Inf. 2007;8(3):77–85. 3 Lanchester J. Incredible edibles: The mad genius of “modernist cuisine.” The New Yorker. 2011 March:87, 64. 4 Adria F, Blumenthal H, Keller T, McGee H. Statement on the ‘new cookery.’ The Observer. 2006. Available at: http://www.guardian.co.uk/uk/2006/ © Jones & Bartlettdec/10/foodanddrink.obsfoodmonthly. Learning, LLC Accessed June 28, 2012. © Jones & Bartlett Learning, LLC NOT FOR5 SALE Adler J. OR Extreme DISTRIBUTION cuisine. Smithsonian. 2011;42:60–66. NOT FOR SALE OR DISTRIBUTION 6 Ehrenberg R. What’s cooking? Science News. 2008;173(13):202–203.

© Jones & BartlettSmell Learning, LLC © Jones & Bartlett Learning, LLC olfactory Relating to theNOT sense of FOR SALEOur OR sense DISTRIBUTION of smell, or olfactory sense, also contributesNOT FOR to our SALE evaluation OR ofDISTRIBUTION smell. food quality. The volatility of odors is related to temperature. Because only volatile molecules, in the form of gas, carry odor, it is easier to smell hot foods than cold ones. For example, hot tea is much easier to detect than iced tea, © Jones & Bartlett Learning,and LLC the odor of a baked item is© more Jones intense & Bartlett than that Learning,of ice cream. LLC Lighter molecules that can become volatile are detected by the olfactory NOT FOR SALE OR DISTRIBUTIONepithelium in the nasal cavityNOT through FOR one SALE of two OR pathways: DISTRIBUTION (1) directly through the nose or (2) after entering the mouth and flowing retro-nasally, or toward the back of the throat and up into the nasal cavity.6 If you drink a carbonated beverage and laugh unexpectedly, you may experience the tingling © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 86 29/12/12 11:36 PM THe Human SenSeS 87

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FORof SALE bubbles OR in the DISTRIBUTION nose, showing how the mouth and noseNOT are FORconnected SALE and OR DISTRIBUTION how molecules can reach the olfactory epithelium by either route. adaptation The gradual decrease in The gradual decrease in the ability to distinguish between odors over time the ability to distinguish between odors is called adaptation. Adaptation occurs to prevent sensory overload. Dairy over time. farmers who are exposed daily to the smell of manure will gradually become gustatory Relating to the sense of unaware of it, whereas© visitors Jones to & the Bartlett farm may Learning, be taken aback LLC by the smell. taste.© Jones & Bartlett Learning, LLC Human subjects haveNOT varying FOR sensitivities SALE ORto odors, DISTRIBUTION depending on hunger, NOT FOR SALE OR DISTRIBUTION satiety, mood, concentration, presence or absence of respiratory infections, taste buds The small parts of and gender (e.g., women who are menstruating or are pregnant may perceive gustatory and supportive cells; usually odors differently).7 Because different people perceive a given odorant differ- found on the upper surface of the ently, identifying a new odor from a food product requires as large a panel tongue. as possible© Jones to get & valid Bartlett results. Learning, LLC © Jones &papillae Bartlett Rough Learning, bulges or LLC NOT FOR SALE OR DISTRIBUTION NOT FOR protuberancesSALE OR in DISTRIBUTION the surface of the tongue, some of which contain taste Taste buds. Taste, or the perception of gustatory input, is the most influential factor in a person’s selection of a particular food. For a substance to be tasted, it

© Jones &should Bartlett be dissolved Learning, in water, LLC oil, or saliva. Taste is perceived© Jones by &the Bartlett taste Learning, LLC © F.C.G./ShutterStock, Inc. NOT FORbuds SALE (see ORFigure DISTRIBUTION 3.1), which are primarily on the surfaceNOT of FORthe tongue, SALE by OR DISTRIBUTION the mucosa of the palate, and in areas of the throat. In the middle of each taste bud lies a pore, where saliva collects. When food enters the mouth, bits of it are dissolved in the saliva pools and come into contact with cilia, small hairlike projections, from the gustatory cells.6 The gustatory cells signal© Jones to the &brain Bartlett through Learning, cranial nerves. LLC The brain, in © Jones & Bartlett Learning, LLC turn, translates the nervousNOT FORelectrical SALE impulses OR DISTRIBUTIONinto sensations that people NOT FOR SALE OR DISTRIBUTION recognize as “taste.”6 Taste buds are found in the papillae of the tongue. Two types of papillae contain taste buds. The mushroom-like fungiform papillae on the sides and tip of the tongue generally contain taste buds, and the circumvallate papillae (elevated,© Jones large papillae& Bartlett in the Learning, form of a “V” LLC toward the back of the tongue)© Jones Figu &Re Bartlett 3.1 On the tongue,Learning, the majority LLC of the taste alwaysNOT contain FOR taste SALE buds. 8OR As peopleDISTRIBUTION get older, the original 9,000 to 10,000NOT FORbuds sitSALE on raised OR protrusions DISTRIBUTION of the tongue surface taste buds begin to decrease in number, so that people older than age 45 called papillae. On average, the human tongue has 2,000–8,000 taste buds. often seek more spices, salt, and sugar in their food.6

Genetic Variation © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Individual variation in taste likely has a genetic component. Studies have NOT FORdemonstrated SALE OR aDISTRIBUTION link between the ability to taste bitter NOTthiourea FOR compounds SALE OR DISTRIBUTION and a newly discovered taste receptor gene, TAS2R38.9 Thus, the ability to taste these bitter compounds—phenylthiocarbamide (PTC) or the safer, chemically related compound 6-n-propylthiouracil (PROP)—may be used as 10–12 a phenotypic marker for© geneticJones differences & Bartlett in perceptions Learning, of LLCtaste. In the © Jones & Bartlett Learning, LLC United States, the frequency of nontasters is estimated to be 20% to 25% of the population.13 The NOTfrequency FOR varies SALE by gender OR DISTRIBUTION and race.13,14 NOT FOR SALE OR DISTRIBUTION One factor that may explain variation in taste and the perception of physical sensations is the anatomy of the anterior portion of the tongue. For example, PROP tasters have the most fungiform papillae (FP).13,15–17 Special Topic© 3.2 Jones provides & Bartlettmore information Learning, on how LLC genetic variations may influence© Jones & Bartlett Learning, LLC how we perceive food. BeyondNOT FOR genetics, SALE variation OR DISTRIBUTIONin taste perception also depends on howNOT per- FOR SALE OR DISTRIBUTION ceptible sweet, fatty, and bitter components are in foods and beverages. It also depends on the value a consumer places on other factors, such as health and convenience.18,19 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 87 29/12/12 11:36 PM 88 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Special Topic 3.2 Nutrigenomics Jill M. Merrigan© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALENutrigenomics OR DISTRIBUTION studies the naturally occurring compounds inNOT the foods FOR and howSALE they OR DISTRIBUTION nutrigenomics Study of how affect our bodies based our individual genetic differences.1 “There is good evidence that food affects our genes and the way nutrition has significant influences on the expression of genes, and, likewise, genetic our bodies respond to nutrients. variation can have a significant effect on food intake, metabolic response to food, individual © Jones & Bartlett Learning,nutrient LLC requirements, food safety, and the© efficacy Jones of disease-protective& Bartlett Learning, dietary factors,” LLC explainsNOT nutrigenomic FOR SALEresearcher OR L. R. DISTRIBUTIONFerguson.2 NOT FOR SALE OR DISTRIBUTION The existence of a particular gene or mutation in many cases indicates a predisposition to a particular disease. Once genetic predisposition has been established, determining whether the disease will progress can be investigated by examining the relationship between the human genome and environmental and behavioral factors. The study of nutrigenomics looks at the expression of the genome © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC with regard to nutrition. According to researcher M. Nathaniel Mead, “although genes are critical for determining function, nutrition NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION modifies the extent to which different genes are expressed, and thereby modulates whether individuals attain the potential established by their genetic background.”3 Today, scientists are exploring nutrigenomics to determine how nutrients may be able protect the genome from damage. Through the study of nutrigenomics,© Jones it may & become Bartlett possible Learning, to develop dietary LLC interventions based on an understanding© Jones of &an indBartlettividual’s Learning, LLC nutritional requirements, nutritionalNOT FOR status, SALE and genotype. OR StudyingDISTRIBUTION how foods affects individual genes and genotypesNOT FOR will make SALE it OR DISTRIBUTION possible to design “personalized nutrition” plants that may prevent and cure chronic disease. Studies have been completed on humans, animals, and cell cultures that reveal that macronutrients (fatty acids and proteins), micronutrients (vitamins), and naturally occurring phytochemicals (such as flavonoids, carotenoids, coumarins, and phytosterols) regulate gene expression© Jones in various & Bartlett ways. Micronutrients Learning, and bioreactiveLLC chemicals in foods are involved© Jones in metabolic & Bartlett reactions Learning,that determine LLC hormonalNOT balances FOR and SALE immune ORcompetence, DISTRIBUTION as well as detoxification processes. AdditionallyNOT some FOR biochemicals SALE ORfound DISTRIBUTIONin foods, such as genistein and resveratrol, act as transcription factors, and therefore alter gene expression. Signal transduction pathways and chromatin structures are altered by other biochemicals, such as choline, and therefore indirectly affect gene expression.3 One example of nutrigenomics is folate and the gene for MTHFR (methylenetetrahydrofolate reductase). MTHFR has a role in supplying © Jones &methionine. Bartlett Methionine Learning, plays a central LLC role in certain metabolic pathways,© Jones including & thoseBartlett involved Learning, in the production LLC of neurotransmitters NOT FORand SALE in the regulation OR DISTRIBUTION of gene expression. Folate is required for MTHFRNOT to function FOR efficiently. SALE MTHFROR DISTRIBUTIONhas a common polymorphism that leads to two forms of protein: the reference version (C), which functions normally, and the thermal-labile version (T), which has reduced activity. When individuals have two copies of the reference sequence gene (CC), they have normal folate metabolism. However, individuals who have two copies of the reduced version (TT) and low dietary folate accumulate homocysteine and have less methionine. This combination puts them© at Jones an increased & riskBartlett for vascular Learning, disease and LLCpremature cognitive decline. By making© Jones these connectio & Bartlettns, Learning, LLC individuals with the unstableNOT (TT) genesFOR can SALE take folic OR acid DISTRIBUTIONsupplements or increase their folate from food sourcesNOT to metaboliFOR SALEze excess OR DISTRIBUTION homocysteine and restore their methionine levels to normal.1 Additional studies have determined that there are nine key nutrients that may have an influence on genomic integrity in a handful

of ways. Six of these nutrients—folate, vitamin B12, niacin, vitamin E, retinol, and calcium—are associated with a reduction in DNA damage.© JonesThe other three—riboflavin,& Bartlett Learning, panthenic acid, LLC and biotin—are associated with ©an increaseJones in &DNA Bartlett damage similar Learning, to that seen LLC upon occupationalNOT FOR exposure SALE to genotoxic OR DISTRIBUTION and carcinogenic chemicals. This suggests thatNOT nutritional FOR deficiency SALE or ORexcess DISTRIBUTION can lead to DNA damage as damaging as that seen with exposure to environmental toxins. Other nutrigenomic studies have shown that many antioxidant nutrients and phytochemicals enhance DNA repair and reduce oxidative DNA damage.3

03441_CH03_Printer.indd 88 29/12/12 11:36 PM THe Human SenSeS 89

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Recent research indicates that nutrigenomics may have the potential to prevent, mitigate, and treat chronic diseases and certain cancers by making small but highly useful changes to an individual’s diet. In the future, scientists, doctors, and dieticians may be able to move forward to be able to identify a patient’s DNA profile for a specific disease and ultimately be able to shape a diet that will reduce their chances of developing© that Jones disease. & Nutrigenomics Bartlett mayLearning, be the answer LLC to the obesity epidemic and improve© Jonesthe way individuals & Bartlett age with Learning, LLC better bone and brainNOT health; FORit may also SALE decrease OR the DISTRIBUTION risk of developing certain cancers. NOT FOR SALE OR DISTRIBUTION References 1 Astley SB. An introduction to nutrigenomics developments and trends. Genes Nutri. 2007;2(1):11–13. 2 Ferguson LR. Nutrigenomics: Integrating genomic approaches into nutrition research. Mol Diagn Ther. 2006;10(2):101–108. 3© Jones Mead MN. Nutrigenomics:& Bartlett The Learning, genome–food interface. LLC Env Health Persp. 2007;115(12):A582–A589.© Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Basic Components of Taste Gastronomy Point For many years, four basic tastes were recognized: sweet, salty, sour, and bitter. Monosodium Glutamate Monosodium © Jones &A fifth,Bartlett umami, Learning, was added LLC more recently. These tastes© can Jones be characterized & Bartlett Learning, LLC as follows: glutamate, better known by its abbreviation MSG, is NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR oneDISTRIBUTION substance that contributes to the perception of • Sweet: Substances that produce sweet taste include sugars, glycols, umami. alcohols, aldehydes, and alternative sweeteners.20 • Salty: The salty taste comes from ionized salts, such as the ions in sodium chloride (NaCl) or other salts found naturally in some foods. • Sour: The sour© tasteJones comes & Bartlettfrom the acids Learning, found in LLCfood. It is © Jones & Bartlett Learning, LLC related to theNOT concentration FOR SALE of hydrogen OR DISTRIBUTION ions (H+) that are found NOT FOR SALE OR DISTRIBUTION in the natural acids of fruits, vinegar, and certain vegetables. • Bitter: Bitterness is imparted by compounds such as caffeine (tea, coffee), theobromine (chocolate), and phenolic compounds (grapefruit).6 Many bitter substances are alkaloids that often are umami Taste category based on found in poisonous plants. © Jones & Bartlett Learning, LLC © Jones &glutamate Bartlett compounds, Learning, which areLLC • Umami: This is a most recently defined component of taste, NOT FOR SALE OR DISTRIBUTION NOT FOR SALEcommonly OR found DISTRIBUTION in meats, mushrooms, which was identified from a study of seaweed broth.8 Umami is soy sauce, fish sauce, and cheese. a Japanese word meaning “delicious”—it is evoked by glutamate compounds, which are commonly found in meats, mushrooms, soy ﬂavor The combined sense of taste, sauce, fish sauce, and cheese. Some taste experts do not recognize odor, and mouthfeel. © Jones & Bartlettumami Learning, as a taste atLLC this time. © Jones & Bartlett Learning,mouthfeel LLC The way that a particular NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTIONtype of food feels in the mouth. Flavor Whereas taste relies on the sensation produced through the stimulation of the taste buds, flavor is a broader concept. Flavor is the combined senses of taste, aroma, and mouthfeel. Mouthfeel encompasses textural and chemical sensations such as astringency,© Jones spice & Bartlett heat, cooling, Learning, and metallic LLC flavor. © Jones & Bartlett Learning, LLC Among the flavorNOT components, FOR SALE aroma isOR especially DISTRIBUTION important; it provides NOT FOR SALE OR DISTRIBUTION approximately 75% of the impression of flavor.21 To get an idea of how the ability to smell affects the perception of flavor, pinch your nose and begin to eat a certain flavor of jellybean. Then, as you are chewing, unpinch your nose: You will clearly sense the difference between when the nose is pinched and unpinched.© Jones Suppose & Bartlett you are Learning,eating a buttered LLC popcorn–flavored jellybean.© Jones Innovation & Bartlett Learning, Point LLC WhileNOT pinching FOR your SALE nose, OR you DISTRIBUTION can only perceive sweetness, but as soonNOT as FORA New SALE Taste? OR Research DISTRIBUTION is underway for a sixth you unpinch your nose you can recognize the buttered popcorn flavor. As taste component relating to our perception of fats. another example, consider when you have a cold with a badly stuffed-up Genetic variation in “fat taste” may help explain our nose. Everything tastes different. (This is why pinching people’s nostrils shut food choices and dietary habits, which, in turn, could is helpful in mitigating the flavor of an unpleasant medicine.) influence our nutritional and health status. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 89 29/12/12 11:36 PM 90 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION Sound NOT FOR SALE OR DISTRIBUTION Sound is another sense used in evaluating food quality. Sounds such as siz- zling, crunching, popping, bubbling, squeaking, dripping, exploding, and crackling can communicate much about a food. Most of these sounds are affected by water content; thus, their characteristics indicate a food’s fresh- © Jones & Bartlettness and Learning, ripeness.6 LLC © Jones & Bartlett Learning, LLC NOT FOR SALE ORSound DISTRIBUTION is detected as vibrations in the local medium,NOT FOR usually SALE air. The OR vibra DISTRIBUTION- tions are transmitted via the small bones in the middle ear to create hydraulic motion in the fluid of the inner ear, the cochlea. The cochlea is a spiral canal covered in cilia that, when agitated, sends neural impulses to the brain.5 © Jones & Bartlett Learning,Touch LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTIONThe sense of touch delivers impressionsNOT FOR of a food’sSALE texture OR DISTRIBUTION to us through oral sensations or the skin. Texture is a very complex perception: The first input is visual; second comes touch, either directly through the fingers or indirectly via eating utensils; the third is the feeling in the mouth (mouthfeel), as detected by © Jones &texture Bartlett The sensory Learning, manifestation LLC the teeth and tactile© Jones nerve cells & Bartlett on the tongue Learning, and palate. LLC Texture is the sensory of the structure or inner makeup of NOT FOR SALE OR DISTRIBUTION manifestation ofNOT the structure FOR SALE or inner OR makeup DISTRIBUTION of products in terms of their reac- products in terms of their feel as tions to stress, which are measured as mechanical properties (such as hardness/ measured by tactile nerves on the firmness, adhesiveness, cohesiveness, gumminess, springiness/resilience, and surface of the skin of the hands, lips, viscosity) by the kinesthetic sense in the muscles of the hands, fingers, tongue, or tongue. jaw, or lips.5 Texture also includes tactile feel properties, which are measured as descriptive panel A panel© Jones & Bartlettgeometric Learning, properties LLC (i.e., grainy, gritty, crystalline,© flaky) Jones or &moisture Bartlett properties Learning, LLC (i.e., wetness, oiliness, moistness, dryness) by the tactile nerves in the surface of commonly used to determineNOT FOR SALE OR DISTRIBUTION 5 NOT FOR SALE OR DISTRIBUTION differences between food samples. the skin of the hands, lips, or tongue. The greater surface sensitivity of the lips, The descriptive panelist is experienced tongue, face, and hands makes easy detection of small differences in particle in the type of food being tested and size and thermal and chemical properties possible among food products. receives extensive training prior to the testing.© Jones & Bartlett Learning,variables LLC Controlled ©During Jones Sensory& Bartlett evaluation Learning, LLC consumerNOT panel FOR A SALEpanel selected OR DISTRIBUTIONDuring sensory evaluation, panelistsNOT FORare typically SALE seated OR DISTRIBUTIONat tables, cubicles, from the public according to the or booths, and the food is presented in a uniform fashion. To obtain valid, demographics necessary to taste test reproducible results during a sensory evaluation, the environment in which the a product. sensory panel evaluates foods or beverages should be carefully controlled, as should variables pertaining to the panelists. This section discusses the many © Jones & Bartlett Learning, LLC variables that should© Jones be considered & Bartlett when Learning, designing a LLC sensory evaluation test. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Panel management Two general types of panels are used in sensory evaluation. A descriptive panel is commonly used to determine differences between food samples. The © Jones & Bartlettdescriptive Learning, panelist isLLC experienced in the type of food© Jones being tested & Bartlett and receives Learning, LLC extensive training prior to the testing. A consumer panel is selected from NOT FOR SALEthe OR public DISTRIBUTION according to the demographics necessaryNOT to FOR taste testSALE a product. OR DISTRIBUTION Panel Selection When assembling a panel, it is preferred to use an equal number of men and © Jones & Bartlett Learning,women. LLC The age distribution of© the Jones panel should& Bartlett also be Learning, considered becauseLLC it may affect test results.6 The sensory analyst must recruit the people who NOT FOR SALE OR DISTRIBUTIONcan make a reliable commitmentNOT of time FOR and SALEwho also OR know DISTRIBUTION what is expected of them during the test. General taste panels usually consist of people who meet the following criteria: • They are in good health and free of illness related to sensory © Jones & Bartlett Learning, LLC properties,© Jones such as & chronic Bartlett colds, Learning, food allergies, LLC or diabetes. NOT FOR SALE OR DISTRIBUTION • They areNOT nonsmokers FOR SALE (smoking OR canDISTRIBUTION dull olfactory and gustatory sensations).

03441_CH03_Printer.indd 90 29/12/12 11:36 PM varIableS ConTrolleD DurIng SenSory evaluaTIon 91

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE• They OR are DISTRIBUTION not color blind. NOT FOR SALE OR DISTRIBUTION • They have no strong likes or dislikes for the food to be tested.

Panelist Preparation The level of training for descriptive panels and consumer panels is quite different, given the differences in© purpose Jones of & the Bartlett evaluations Learning, in which theyLLC participate. © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Descriptive Panels Because the investment in a descriptive panel is large in terms of time and human resources, it is wise to conduct an exhaustive screening process rather than train unqualified panelists.2 If the ability to detect subtle differences is essential,© Jones the sensory & Bartlett analyst may Learning, need to screen LLC the sensory acuity of potential© Jones & Bartlett Learning, LLC panelistsNOT on FOR key properties SALE OR of the DISTRIBUTION product(s) that will be tested. NOT FOR SALE OR DISTRIBUTION Descriptive panels can be selected through a series of tests that may include a set of prescreening questionnaires, a set of acuity tests, a set of ranking/rating tests, and a personal interview. However, it is not necessary to have only the most highly discriminating panelists because the average © Jones &panelists Bartlett will Learning,improve markedly LLC with training and some© people Jones may & beBartlett very Learning, LLC NOT FORdiscriminating SALE OR DISTRIBUTIONin general but just have one or two problemNOT areas. FOR2 SALE OR DISTRIBUTION The amount of training required is determined by the task and the level of sensory sensitivity desired. For most descriptive panels, expensive and in-depth training is necessary.2 During the training, the trainer must make sure the panelists realize that sensory testing work is difficult and requires

attention and concentration.© Jones If team & Bartlett spirit can Learning,be developed LLC by the panelists © Jones & Bartlett Learning,© Lisa F. Young/ShutterStock, Inc. LLC during the training sessions,NOT FORthis will SALE smooth OR the DISTRIBUTION way for the main evaluation NOT FOR SALE OR DISTRIBUTION and facilitate panelist performance. The performance of trained panelists used over long periods of time may fluctuate because of a loss of focus and a lack of motivation during the evaluation sessions.

Consumer© Jones Panels & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC In contrastNOT FORto descriptive SALE panels,OR DISTRIBUTION consumer panels typically require a largerNOT FOR SALE OR DISTRIBUTION number of panelists and may range from 200 to 500 people (see Figure 3.2). Consumer panelists can be screened on a test criteria; for example, demographics or potential use of product. The questions asked of consumer panels should be answerable by untrained panelists. © Jones &Other Bartlett Considerations Learning, LLC © Jones & Bartlett Learning,FiguRe 3.2 Manufacturers, LLC scientists, food NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR technologists,DISTRIBUTION and marketers can use consumer panels Other considerations also should be taken into account to optimize panelist to gain a clear perception of what ordinary consumers performance during a sensory evaluation: may experience when tasting a particular food item. • It is wise to schedule the evaluation of certain product types at the time of day when that product is normally used or consumed.5 For example, breakfast© Jones cereals & would Bartlett be better Learning, tested in LLC the morning. © Jones & Bartlett Learning, LLC In contrast, itNOT would FOR not be SALE recommended OR DISTRIBUTION to test highly flavored NOT FOR SALE OR DISTRIBUTION or alcoholic products in the early morning. • Midmornings or midafternoons (such as 11 am or 3 pm) are considered the best times for testing because at these times people are not usually overly hungry or full.6 • © JonesPanelists & should Bartlett not ingestLearning, any other LLC food for at least 1 hour before© Jones & Bartlett Learning, LLC NOTtesting FOR and SALE should OR not DISTRIBUTIONchew gum immediately before testing.6 NOT FOR SALE OR DISTRIBUTION • The instructions provided to the panelists should be very clear and concise. It is frequently desirable to give the instructions on how to perform the sensory evaluation verbally, before the panelists enter the booth area, and then also in written form on the score sheet.2 © Jones & Bartlett• Incentives Learning, usually LLCare given as a token of appreciation© Jones to motivate& Bartlett Learning, LLC NOT FOR SALEpeople OR DISTRIBUTION to participate voluntarily.2 Common incentivesNOT FOR include SALE OR DISTRIBUTION snacks or small gifts.

03441_CH03_Printer.indd 91 29/12/12 11:36 PM 92 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION environmentalNOT Controls FOR SALE OR DISTRIBUTION Physical and chemical factors present at the location of the sensory evaluation must be carefully controlled so that any possible extraneous effects of the surroundings on the test results are minimized and each panelist experiences the food in the same environment. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALETemperature, OR DISTRIBUTION Humidity, and Air CirculationNOT FOR SALE OR DISTRIBUTION The ambient temperature should be comfortable, and the surroundings should be quiet and odor-free. The temperature and relative humidity for the sensory evaluation area should be 72–75°F (22–24°C) and 45–55%, respectively.5 The use of replaceable active carbon filters in the ventilation system ducts is © Jones & Bartlett Learning,recommended. LLC A slight positive© pressure Jones should & Bartlett be maintained Learning, in the LLC booth NOT FOR SALE OR DISTRIBUTIONareas to prevent odor contamination.NOT FOR5 The sensorySALE scientistOR DISTRIBUTION should check if any unnecessary odors are detected in sensory testing areas. Color and Lighting The color and lighting in the sensory lab should be planned to permit adequate © Jones & Bartlett Learning, LLC viewing of samples© Jones while minimizing & Bartlett distractions. Learning,22,23 LLCThe walls of the sensory NOT FOR SALE OR DISTRIBUTION evaluation areaNOT should FOR be off-white;SALE OR the DISTRIBUTION absence of hues of any color will prevent unwanted effects on food appearance.5 Illumination in the booths should be uniform, shadow-free, and at least 300–500 lx at the table surface.2 An ideal lighting system is controllable with a dimmer switch to a maximum of 700–800 lx, the common illumination intensity in offices.2 Incandescent © Jones & Bartlettlights Learning,can control both LLC the light intensity and the© Joneslight color, & Bartlettbut they genLearning,- LLC NOT FOR SALEerate OR heat, DISTRIBUTION which will require cooling. FluorescentNOT lights FOR generate SALE less OR heat DISTRIBUTION and allow for choice of whiteness; for example, cool white, warm white, or simulated north daylight.5 Colored lights are used to mask visual differences among samples, calling for the subject to determine by flavor or texture only. A choice of low-intensity red, green, and/or blue lights using colored bulbs © Jones & Bartlett Learning,or LLCspecial filters is a common ©feature Jones of sensory & Bartlett booths. Learning,5 LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Product Controls Variables pertaining to the product samples themselves must also be controlled. Sample Preparation © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Food samples must be of the same size (usually enough for two bites or NOT FOR SALE OR DISTRIBUTION sips) and fromNOT the same FOR portion SALE of OR the DISTRIBUTIONfood (e.g., middle versus outside). The sensory analyst should determine and control the amount of product to be used in all the tests, including the amount of each added ingredient, the holding time The minimum and preparation process, and holding time, which is defined as the minimum maximum time after preparation© Jones that a & Bartlettand maximum Learning, time LLCafter preparation that a product© Jones can be used & Bartlett for a sensory Learning, LLC product can be used for a sensory test. test. For instance, suppose the test sample is a pumpkin muffin. The sensory NOT FOR SALEanalyst OR needsDISTRIBUTION to decide the size of the muffin, NOTexactly FOR how theSALE muffins OR DISTRIBUTION will be baked, the appropriate holding time, and at which temperature and on what plate the sample will be served. The sensory analyst should be very careful to standardize all serving pro- © Jones & Bartlett Learning,cedures LLC and sample preparation© techniquesJones & except Bartlett the variableLearning, under LLC evalu- NOT FOR SALE OR DISTRIBUTIONation. If the appearance of theNOT sample FOR is not SALE the variable OR DISTRIBUTION under evaluation, then the samples should appear identical. Samples should be blind-labeled with random three-digit codes, and the sample order should be randomized to avoid bias due to order of presentation. A reasonable number of samples, say two to four, should be tested at a time to avoid taste fatigue. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 92 29/12/12 11:36 PM meaSuremenT THeory 93

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FORSample SALE Temperature OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Samples must be presented at the same temperature, which must be specified in the test protocol. For example, ice cream should be tempered at 5–9°F (–15°C to –13°C) for at least 12 hours before serving because scooping is not easy if the ice cream is colder.2 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Presentation NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Samples should be presented in containers or on plates that are the same size, shape, and color. White or clear containers are usually chosen so as not to influence panelists’ perceptions of the food’s color. The sensory analyst should choose the container that is most convenient. However, the choice of container© Jones should & not Bartlett negatively Learning, affect the LLCflavor characteristics of the© food Jones & Bartlett Learning, LLC product.NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Carriers Carriers refer to materials that form a base or vehicle for the food being tested but may more broadly be considered as any other food that accompanies the © Jones &one Bartlett being tested Learning, so they are LLC ingested, too.2 Examples include© Jones spaghetti & Bartlett sauce Learning, LLC NOT FORon SALE a spaghetti OR DISTRIBUTIONnoodle, cream fillings in pastries, butterNOT on bread, FOR chips SALE with OR DISTRIBUTION salsa, and carrots with ranch dips. A carrier can mask or disguise differences or minimize the panelist’s abilities to perceive the difference due to the addition of other flavors and modifications to texture and mouthfeel characteristics. However, for a product that is rarely consumed alone and almost always involves a carrier, the artificial© Jones situation & Bartlett where Learning,the carrier is LLCnot provided may © Jones & Bartlett Learning, LLC affect test results, especiallyNOT inFOR consumer SALE testing. OR 2DISTRIBUTION Therefore, whether a carrier NOT FOR SALE OR DISTRIBUTION is used should be carefully determined. Palate Cleansers Room temperature water or plain bread is made available for panelists to eat between© Jones samples &to Bartlettprevent carryover Learning, tastes. ALLC rest period of at least 30 seconds© Jones & Bartlett Learning, LLC is scheduledNOT FOR between SALE samples. OR DISTRIBUTIONPaper towels or napkins are provided, NOTand, FOR SALE OR DISTRIBUTION because swallowing the food or beverage influences the taste of subsequent samples, small containers into which samples may be spit are provided. One study that evaluated the effects of a range of palate cleansers (i.e., chocolate, pectin solution, table water crackers, warm water, water, © Jones &whole Bartlett milk) onLearning, foods representing LLC various tastes and mouthfeels© Jones concluded& Bartlett Learning, LLC NOT FORthat SALE table ORwater DISTRIBUTION crackers were the only palate cleanser effectiveNOT FOR across SALE all rep -OR DISTRIBUTION resentative foods.24 These foods included jelly beans (sweet), coffee (bitter), smoked sausage (fatty), tea (astringent), spicy tortilla chip (pungent), mint (cooling), and applesauce (nonlingering). © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC measurement TheoryNOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION measurement Hierarchy Four levels of measurement are commonly used in sensory evaluation: nominal, ordinal, interval, and ratio.25 It is important to recognize that there is a hierarchy© Jones in & the Bartlett level of measurement.Learning, LLC At lower levels of measurement,© Jones & Bartlett Learning, LLC assumptionsNOT FOR tend SALEto be less OR restrictive DISTRIBUTION and data analyses tend to be less sensiNOT- FOR SALE OR DISTRIBUTION tive. At each level up the hierarchy, the current level includes all of the qualities of the one below it and adds something new. In general, it is desirable to have a higher level of measurement (such as interval or ratio) rather than a lower one (such as nominal or ordinal).26 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 93 29/12/12 11:36 PM 94 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION Nominal NOT FOR SALE OR DISTRIBUTION With nominal measurement, the numbers simply identify unique attributes; they are not ordered. For example, gender may be coded by assigning a “1” to males and a “2” to females. With nominal data, common descriptive statistics such as range, mean, or standard deviation are not appropriate. Instead, © Jones & Bartlettfrequency Learning, counts, numberLLC of categories, or mode© Jonescan be used& Bartlett to get some Learning, LLC NOT FOR SALEidea OR of theDISTRIBUTION distribution of nominal data. NOT FOR SALE OR DISTRIBUTION Ordinal With ordinal measurement, the attributes can be ordered, but the difference between levels is not equal. It is not normally distributed and often is skewed. © Jones & Bartlett Learning,For LLC example, cakes can be ordered© Jones by rank & for Bartlett perceived Learning, overall sweetness. LLC NOT FOR SALE OR DISTRIBUTIONThe interval between values is notNOT interpretable FOR SALE for ordinal OR DISTRIBUTION measurement. In this case, the rank number can tell us where the cake falls in order of sweetness; however, we cannot draw conclusions about the differences among the products. © Jones & Bartlett Learning, LLC Interval © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION With interval measurement,NOT FOR thereSALE are ORordered DISTRIBUTION levels, and the difference between levels is equal. However, there is no true zero. For example, when we measure the oven temperature in degrees Fahrenheit, the distance from 100°F to 200°F is the same as the distance from 300°F to 400°F. Because the interval between the values is interpretable, it makes sense to calculate the average of the interval © Jones & Bartlettvariable. Learning, However, inLLC interval scaling, ratios do not© Jones make sense: & Bartlett 200°F is Learning,not LLC NOT FOR SALEtwice OR as DISTRIBUTION hot as 100°F, although the value is twiceNOT as large. FOR SALE OR DISTRIBUTION Ratio With ratio measurement, there are ordered levels in which the difference between levels is equal, and there is a true zero. For example, weight is a © Jones & Bartlett Learning,ratio LLC measurement. We can say© that Jones 200 £s Bartlett of sugar Learning, weighs twice LLC as NOT FOR SALE OR DISTRIBUTIONmuch as 100 pounds of sugar,NOT and zero FOR pounds SALE of ORsugar DISTRIBUTION means that there is no sugar.

Common Scales used in Testing © Jones & Bartlett Learning, LLC Category Scales© Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION Category scalingNOT may FORbe the SALEoldest method OR DISTRIBUTION of scaling; it involves the choice of discrete response alternatives to signify increasing sensation intensity in terms of degrees of liking and/or preference. The most popular category scale used in hedonic scale A scale with which sensory testing is the hedonic scale, which measures the extent of like or dislike judges indicate the extent of their for the sensory characteristics of food. Examples of category scales are shown like or dislike for the sensory© Jones & Bartlettin Figure Learning, 3.3. Due LLCto their simplicity, category scales© Jones are well & suitedBartlett for conLearning,- LLC characteristics of food. NOT FOR SALEsumer OR panels. DISTRIBUTION In addition, they offer some advantagesNOT FOR in data SALE coding OR and DISTRIBUTION tabulation (for speed and accuracy) because they are easier to tabulate than line markings or the more variable magnitude estimates, described below.

Line Scales © Jones & Bartlett Learning,Line LLC scales may also be referred© to Jones as graphic & Bartlettratings or Learning,visual analog LLC scales. NOT FOR SALE OR DISTRIBUTIONExamples of line scales are shownNOT in FOR Figure SALE 3.4. With OR lineDISTRIBUTION scales, the test participant’s response is recorded as the distance of the mark from one end of the scale, usually whatever end is considered “lower.” Line scaling dif- fers from category scaling in the sense that the person’s choices seem more continuous and less limited. Stone et al. recommended the use of line scaling © Jones & Bartlett Learning, LLC for Quantitative© DescriptiveJones & AnalysisBartlett (QDA), Learning, then a relativelyLLC new approach NOT FOR SALE OR DISTRIBUTION to specifying theNOT intensities FOR SALEof all the OR important DISTRIBUTION sensory attributes.27

03441_CH03_Printer.indd 94 29/12/12 11:36 PM meaSuremenT THeory 95

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE ORHedonic DISTRIBUTION Scale for Children NOT FOR SALE OR DISTRIBUTION Maybe good Very or Very good Good maybe bad Bad bad © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

(a)

Chewiness © Jones &1 Bartlett23456789 Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR WeakSALE OR DISTRIBUTION Strong NOT FOR SALE OR DISTRIBUTION (b)

Sourness

Fatty Flavor Number Word Anchor

0 None ½ © Jones & BartlettVery slight Learning, LLC © Jones & Bartlett Learning, LLC 1 NOT FOR SALE OR DISTRIBUTIONSlight NOT FOR SALE OR DISTRIBUTION 1½ Slight–moderate 2 Moderate 2½ Moderate–strong 3 Strong (d)

FiguRe© 3.3 Jones Examples & of Bartlettcategory scales. Learning, (a) Hedonic scale LLC for children. (b) Chewiness. (c) Sourness.© Jones & Bartlett Learning, LLC (d) FattyNOT flavor. FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

Brownness 110

Weak Strong © Jones & Bartlett(a) Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Sweetness 110

(b) © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Roughness NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 110

X X Smooth Rough (c) © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC

NOTOverall FOR Acceptance SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 1 10

Strongly Strongly dislike like (d) © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FORFigu SALERe 3.4 ORExamples DISTRIBUTION of line scales. (a) Brownness. (b) Sweetness. (c) Roughness.NOT (d) FOR Overall acceptance.SALE OR DISTRIBUTION

03441_CH03_Printer.indd 95 29/12/12 11:36 PM 96 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION Magnitude EstimationNOT FOR Scales SALE OR DISTRIBUTION Magnitude estimation scaling is a popular technique in psychophysical studies. In this procedure, the panelists are instructed to assign numbers to their sensations in proportion to how strong each sensation feels.2 In the analysis, the ratios between the numbers are supposed to reflect the ratios of sensation © Jones & Bartlettmagnitudes Learning, that have LLC been experienced. For example,© Jones if sample & Bartlett A is given Learning, the LLC NOT FOR SALEvalue OR of DISTRIBUTION 15 for bitterness intensity and sample BNOT seems FOR three SALE times as OR bitter, DISTRIBUTION then B is given a magnitude estimation of 45. Two variations of the magnitude estimation technique are available. In one method, a standard stimulus is given to the panelist as a reference with a fixed value, and all subsequent stimuli are rated relative to the reference. Values © Jones & Bartlett Learning,of zeroLLC are allowed in this method,© Jones but the &rating Bartlett of zero Learning, should not beLLC used NOT FOR SALE OR DISTRIBUTIONas a reference. In the second variation,NOT FOR no standard SALE stimulus OR DISTRIBUTION is given, and the panelist is free to choose any number for the first sample. All samples are then rated relative to the first intensity. For the second variation, because the panelists choose different ranges of numbers, the data have to be manipulated to bring all panelists into the same range. This adds an extra step to the analysis. © Jones & Bartlett Learning, LLC Panelists are© cautionedJones & to Bartlett avoid falling Learning, into previous LLC habits of using only NOT FOR SALE OR DISTRIBUTION category scalesNOT that they FOR are SALE used to OR using. DISTRIBUTION2 This may be a difficult problem with previously trained panels that have used a different scaling method because people like to stick with a method with which they are familiar.

© Jones & BartlettTypes Learning, of Sensory LLC Tests © Jones & Bartlett Learning, LLC analytical tests SensoryNOT tests used FOR SALESensory OR DISTRIBUTION tests may be analytical or affective. AnalyticalNOT FOR tests SALE are based OR onDISTRIBUTION to detect discernible differences. discernible differences, whereas affective tests are based on individual accept- affective tests Sensory tests ability or preferences. Analytical tests are divided into two types of tests: used to determine differences in difference (discriminative test) and descriptive. Affective tests have two acceptability or preference between categories, depending on the main task of the test: acceptance or preference. products.© Jones & Bartlett Learning,The LLC primary task of an acceptance© Jones test is “rating,”& Bartlett whereas Learning, the primary LLC task NOT FOR SALE OR DISTRIBUTIONof a preference task is “choice.”5NOT FOR SALE OR DISTRIBUTION difference tests Sensory tests designed to detect discernible analytical Difference Tests differences. Difference tests are testing samples for their differences from each other. Dif- triangle test A difference test in ference tests can be used to test the sensitivity of judges as well as to perform © Jones &which Bartlett three samples Learning, are presented LLC a practical function© Jones such as & determining Bartlett Learning, whether a food LLC company should buy NOT FORsimultaneously SALE OR (two DISTRIBUTION of which are an inexpensiveNOT ingredient FOR to SALE replace OR a more DISTRIBUTION expensive one in formulating a the same), and the judge is asked to food product. The two types of difference tests are overall difference tests and identify the odd sample. attribute difference tests.

© Jones & BartlettOverall Learning, Difference LLC Tests © Jones & Bartlett Learning, LLC NOT FOR SALEThe OR question DISTRIBUTION of interest for an overall difference testNOT is, “DoesFOR aSALE sensory OR differ DISTRIBUTION- ence exist between samples?” Overall difference tests are the simplest sensory tests and include the triangle and the duo-trio test. In a triangle test, three number-coded samples are presented simultaneously. The panelist is asked to indicate which one is odd (different from the © Jones & Bartlett Learning,other LLC two). The difference in this© methodJones of& presentation Bartlett Learning, reduces the LLCchance NOT FOR SALE OR DISTRIBUTIONof guessing the right answer toNOT 33.3% FOR (1 in SALE 3). This OR method DISTRIBUTION is particularly effective in situations where treatment effects may have produced product changes that cannot be characterized simply by one or two attributes. It also can be useful for selecting panelists and monitoring their performance to discriminate given differences. © Jones & Bartlett Learning, LLC In a triangle© Jonestest, six &possible Bartlett serving Learning, orders (AAB, LLC ABA, BAA, BBA, NOT FOR SALE OR DISTRIBUTION BAB, ABB) areNOT counterbalanced FOR SALE across OR all DISTRIBUTION panelists. It should be conducted with 20 to 40 participants who have been screened for their sensory acuity to common product differences and who are familiar with the test procedure.

03441_CH03_Printer.indd 96 29/12/12 11:36 PM TyPeS of SenSory TeSTS 97

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FORIf SALEdifferences OR are DISTRIBUTION large and easy to identify, as few as NOT12 panelists FOR SALEmay be OR DISTRIBUTION employed.5 The following is an example application of the triangle test: • Situation: A food service management director wishes to confirm whether there is a significant sensory difference between a canned product and a paper carton product before changing the product. She© is Jonesconsidering & Bartlett changing Learning, the chicken LLCbroth product © Jones & Bartlett Learning, LLC from cans to NOTcartons FOR because SALE of theOR clients’ DISTRIBUTION desires for “greener” NOT FOR SALE OR DISTRIBUTION packaging. • Test objective: The objective is to determine if the change of product packaging causes an overall difference in sensory reactions Please taste the samples from left to right. Two samples to the chicken broth. are the same, and one is different. Circle the number of the sample that is different. • © JonesTest design: & Bartlett The test design Learning, is shown LLC in Table 3.1. There are © Jones & Bartlett Learning, LLC NOT48 panelists FOR SALE with noOR replication. DISTRIBUTION NOT FORSample SALE code ______OR DISTRIBUTION ______• Score sheet: An example score sheet is shown in Figure 3.5. Thank you! • Results and analysis: The actual number of panelists who correctly identified the odd sample from the triangle test is counted. Out of FiguRe 3.5 Example of a triangle score sheet. 48 panelists, 25 correctly chose the odd sample. When a statistical © Jones & Bartletttest isLearning, applied, it is LLC determined that the panelists© Jonescould detect & Bartlett a Learning, LLC NOT FOR SALEdifference OR DISTRIBUTION between samples. NOT FOR SALE OR DISTRIBUTION • Conclusion: A significant overall difference was found between the canned product and paper carton product. For detailed statistics regarding this example, see the Appendix at the end of this chapter. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC The duo-trio testNOT is another FOR testSALE of overall OR DISTRIBUTION difference. In this test, the NOT FOR SALE OR DISTRIBUTION reference sample is presented first; it is followed by two other samples, one of which is the same as the reference. The panelists are requested to duo-trio test A difference test in identify which of the latter two samples is the same as or different from which three samples are presented the reference. With the duo-trio test, there is a 50% chance of being right at the same time. A reference is by chance© Jones alone. & BartlettThe following Learning, example LLC shows an application of© the Jones &designated, Bartlett and Learning, the judge is asked LLC duo-trio test: to select the one most similar to the NOT FOR SALE OR DISTRIBUTION NOT FOR SALEreference. OR DISTRIBUTION • Situation: A catering manager is faced with two very similar fluid egg products. Product A has been used for years; product B is a new product offered at a cheaper price. He wants to confirm whether there is an overall difference in perception between the two products. © Jones & Bartlett• Test objective:Learning, The LLCtest objective is to determine© whether Jones the & sourceBartlett Learning, LLC NOT FOR SALEof OR the DISTRIBUTIONfluid egg causes any overall difference inNOT sensory FOR perception SALE OR DISTRIBUTION of the scrambled egg.

TABLE 3.1 Triangle Test Design © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Sample identification NOT FORCodes SALEin Sets with OR Two DISTRIBUTION As Codes in Sets with Two Bs NOT FOR SALE OR DISTRIBUTION A: Canned product 624, 738 325* B: Paper carton product 199* 801, 514 Panelist© Jones Number & Bartlett Learning,Serving Pattern LLC Codes in Order © Jones & Bartlett Learning, LLC 1, 7,NOT 13, 19, FOR 25, 31, SALE 37, 43 OR AAB DISTRIBUTION 624, 738, 199* NOT FOR SALE OR DISTRIBUTION 2, 8, 14, 20, 26, 32, 38, 44 ABA 624, 199*, 738 3, 9, 15, 21, 27, 33, 39, 45 BAA 199*, 624, 738 4, 10, 16, 22, 28, 34, 40, 46 BBA 801, 514, 325* © Jones &5, Bartlett 11, 17, 23, Learning,29, 35, 41, 47 LLC BAB 801,© Jones325*, 514 & Bartlett Learning, LLC NOT FOR 6,SALE 12, 18, OR24, 30, DISTRIBUTION 36, 42, 48 ABB 325*,NOT 801, FOR 514 SALE OR DISTRIBUTION

* Odd sample/correct answer.

03441_CH03_Printer.indd 97 29/12/12 11:36 PM 98 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION • Test design:NOT The FOR test SALE is conducted OR DISTRIBUTION with 40 subjects who regularly eat eggs. Each of the two samples is used as the reference in half (20) of the evaluations. Scrambled eggs are made using the same method and equipment. The only variable that changes is the fluid egg product. Samples are presented without any condiment and at Please taste the samples from left to right.© Jones The left & Bartlett Learning,the same temperature. LLC © Jones & Bartlett Learning, LLC sample is a reference. Circle the numberNOT of the sampleFOR SALE OR• DISTRIBUTIONScore sheet: An example score sheet is shownNOT FORin Figure SALE 3.6 .OR DISTRIBUTION that matches the reference. If no difference is apparent • Results and analysis: Thirteen out of 40 panelists correctly between the two unknown samples, you must guess. matched the sample to the reference. Therefore, the panelists could Reference ______not detect a difference between samples. Sample code ______Sample code ______• Conclusion: The manager can conclude that there is no significant © Jones & Bartlett Learning, LLCoverall difference between© Jones the two &fluid Bartlett egg products. Learning, LLC Thank you! NOT FOR SALE OR DISTRIBUTIONFor detailed statistics regardingNOT thisFOR example, SALE see OR the DISTRIBUTION Appendix at the end of this chapter. FiguRe 3.6 Example of a duo-trio score sheet. Attribute Difference Tests Attribute difference taste tests focus on a single sensory attribute such as © Jones & Bartlett Learning, LLC sweetness or moistness.© Jones Attribute& Bartlett tests Learning, often are administered LLC to evaluate NOT FOR SALE OR DISTRIBUTION qualitative differencesNOT FOR in taste, SALE color, OR and DISTRIBUTION texture. The paired comparison test is a test of difference in which a specific characteristic is designated. The subject is asked to test the two samples presented to identify the sample with the greater amount of the characteristic paired comparison test being measured. With this type of test, the subject has a 50% chance of being A difference test in which two© Jonessamples & Bartlettright byLearning, chance alone. LLC The following example shows© Jones an application & Bartlett of Learning,the LLC are presented, and the judgeNOT is asked FOR SALEpaired OR comparisonDISTRIBUTION test: NOT FOR SALE OR DISTRIBUTION to select the one that has more of a • Situation: A cookie manufacturer receives reports from the particular characteristic. market that her cookie (cookie A) is deemed insufficiently sweet, ranking test A difference or so a test cookie (cookie B) is made using more sweetener. She preference© Jones test in which & Bartlett more than two Learning, LLCwants to produce a cookie© Jones that is perceptibly& Bartlett sweeter, Learning, but not LLC samplesNOT are presented FOR SALE and all samples OR DISTRIBUTION excessively so. NOT FOR SALE OR DISTRIBUTION are compared by ranking them from • Test objective: The test objective is to compare cookie A with lowest to highest for the intensity of a cookie B to determine whether a small but significant increase in specific characteristic. sweetness has been attained. • Test design: A paired comparison test is chosen because the © Jones & Bartlett Learning, LLC characteristic© Jones of interest & Bartlett is only Learning, sweetness, nothing LLC else. The NOT FOR SALE OR DISTRIBUTION sensoryNOT analyst FOR codes SALE the cookies OR DISTRIBUTION “796” and “308” and offers them to a panel of 30 subjects with proven ability to detect small changes in sweetness. The panelists are not asked, “Is 308 There are two samples in each of the two paired sweeter than 796?” but rather, “Which cookie is sweeter?” so as to comparison sets for you to evaluate. Taste each of the coded samples in the sequence presented from left to right. eliminate the potential for bias. Please write down the code of the sweeter© sample.Jones & Bartlett• Learning,Score sheet: LLC An example score sheet is shown© Jones in Figure & Bartlett 3.7. Learning, LLC • Results and analysis: Significantly more panelists (22) identified Sample code ______Sample codeNOT ______FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION test cookie B as being sweeter. Which sample is sweeter? ______• Conclusion: There is a significant sweetness difference between cookies A and B. The test cookie was successful. Thank you! For detailed statistics regarding this example, see the Appendix at the FiguRe 3.7© Example Jones of a &paired Bartlett comparison Learning, end LLC of this chapter. © Jones & Bartlett Learning, LLC score sheet.NOT FOR SALE OR DISTRIBUTIONThe ranking test is valuableNOT when FOR several SALE samples OR need DISTRIBUTION to be evaluated for a single characteristic. Ranking test procedures have the advantage of simplicity in instructions to panelists, ease of data handling, and minimal assumptions about the level of measurement because the data are ordinal. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 98 29/12/12 11:36 PM TyPeS of SenSory TeSTS 99

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FORAlthough SALE ORranking DISTRIBUTION tests are most often applied to hedonicNOT data, FOR they SALEalso are OR DISTRIBUTION applicable to questions of sensory intensities. The following example shows an application of the ranking test: • Situation: A sandwich maker wishes to compare the saltiness among different types of cheese to select a cheese for a new sandwich menu.© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC • Test objective:NOT The test FOR objective SALE is ORto determine DISTRIBUTION whether there is NOT FOR SALE OR DISTRIBUTION a significant difference in saltiness among three types of cheese (mozzarella, cheddar, and provolone). • Test design: The ranking test is suitable because it is simple to carry out and does not require much training. The three samples © Jonesare tested & with Bartlett a panel Learning, of 28 students. LLC The panelists receive the © JonesPlease & Bartlett taste each of the Learning, coded samples in LLCthe set in the NOTsamples (coded FOR SALE with OR three-digit DISTRIBUTION numbers) in balanced, random NOT FORsequence SALE presented, OR from DISTRIBUTION left to right. Rank the three samples in descending order of saltiness. You may re-taste order. any of the samples while ranking for intensity of the • Score sheet: An example score sheet is shown in Figure 3.8. saltiness. No ties are allowed in the ranking. Rinse your mouth with water between samples, and wait for • Results and analysis: When compared to the minimum critical 30 seconds before you taste the next sample. Remember differences, no significant differences are found among the samples that the most intense sample should be ranked 1. © Jones & Bartlettbecause Learning, all of the LLCdifferences are less than 18.© Jones & Bartlett Learning, LLC Saltiness NOT FOR SALE• Conclusion: OR DISTRIBUTION There is no significant difference inNOT saltiness FOR among SALE OR DISTRIBUTION the three types of cheese. Sample code ______123 For detailed statistics regarding this example, see the Appendix at the Thank you! end of this chapter. The rating difference© Jones test &among Bartlett multiple Learning, samples LLCis used when a FiguRe 3.8© ExampleJones of a& ranking Bartlett score sheet. Learning, LLC rating scale is applied.NOT The followingFOR SALE example OR showsDISTRIBUTION an application of the NOT FOR SALE OR DISTRIBUTION rating difference test: • Situation: A baker is producing a new low-fat pound cake prepared by replacing butter with applesauce. He is making four samples with four levels of applesauce substitution (0%, 33%, rating difference test Test to © Jones66%, and & 100%).Bartlett He Learning,discovers that LLC moistness is the sensory © Jones & Bartlett Learning, LLC differentiate among multiple samples NOTcharacteristic FOR SALE that OR shows DISTRIBUTION the most noticeable difference amongNOT FOR SALE OR DISTRIBUTION that uses a rating scale. Products are samples; thus, it is the variable most likely to affect product ranked using a rating scale to assess acceptability. for differences between samples. • Test objective: The test objective is to compare the moistness of four pound cakes in order to determine which product has no © Jones & Bartlettsignificant Learning, difference LLC in moistness from the control© Jones product & Bartlettand Learning, LLC NOT FOR SALEhas OR the DISTRIBUTION lowest fat content. NOT FOR SALE OR DISTRIBUTION • Test design: A completely randomized design is used, in which all samples are compared together using a rating scale. Twelve subjects evaluate the moistness for four samples on scale of 1 to 9. • Score sheet: An© exampleJones score& Bartlett sheet is Learning,shown in Figure LLC 3.9. © Jones & Bartlett Learning, LLC • Results and analysis:NOT FOR There SALEare significant OR DISTRIBUTION differences in moistness NOT FOR SALE OR DISTRIBUTION between 100% applesauce pound cake and the other three products (control with butter and no applesauce, and 33% applesauce, and Sample code ______66% applesauce). Moistness • Conclusion: The 66% applesauce pound cake should be 12345 678 9 © Joneschosen because & Bartlett it is theLearning, product with LLC the lowest fat content © Jones Weakest& Bartlett Learning, LLCStrongest NOTthat shows FOR SALE no significant OR DISTRIBUTION difference from the control in NOT FORThank SALE you! OR DISTRIBUTION moistness. For detailed statistics regarding this example, see the Appendix at the FiguRe 3.9 Example of a rating difference end of this chapter. score sheet. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 99 29/12/12 11:36 PM 100 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION analytical DescriptiveNOT FOR SALE Tests OR DISTRIBUTION The descriptive sensory test is the most comprehensive and informative test used in sensory evaluation. The question for the descriptive test is, “How do descriptive tests Sensory tests products differ in specific sensory characteristics?” Descriptive tests enable designed to provide information on the researchers to characterize their products through selective, critical scoring of specific sensory characteristics© Jones of food & Bartlettspecific Learning, attributes LLCof each product. The descriptive© Jones techniques & Bartlett are frequently Learning, LLC samples and to quantify theNOT sensory FOR SALEused OR for DISTRIBUTION developing new products and for qualityNOT assurance. FOR SALE The informa OR DISTRIBUTION- differences. tion from these methods can be especially valuable for dealing with sensory problems that consumers may detect. proﬁling A group of highly trained panelists work together to develop the Descriptive tests require a well-trained panel and tend to be expensive. vocabulary needed to provide specific Descriptive tests should never be used with consumers because consistent and © Jones & Bartlett Learning,reproducible LLC data are an essential© partJones of descriptive & Bartlett tests. Learning,2 Descriptive LLCtesting descriptionsNOT of FOR food samples; SALE used OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION to detail the specific flavors (flavor is usually conducted by using a scorecard containing precise word descriptions profiling) or textures (texture profiling) that structure the form of the responses. Each of the characteristics of a sample of a food or beverage. to be evaluated is described over a range, and the panelist selects the specific description matching the sample for each item on the scorecard. The respon- sibility for selecting the appropriate vocabulary to elicit an accurate picture © Jones & Bartlett Learning, LLC of the samples© rests Jones with the& Bartlettresearcher Learning, who has developed LLC the scorecard; a NOT FOR SALE OR DISTRIBUTION well-constructedNOT scorecard FOR willSALE give ORthe desired DISTRIBUTION information.2 Innovative Point Profiling is another approach to descriptive testing. To conduct profiling, Approaches to Descriptive Tests The four a group of highly trained panelists work together to develop the vocabulary major approaches of descriptive tests are flavor profile, needed to provide specific descriptions of food samples. This technique is used to detail the specific flavors (flavor profiling) or textures (texture profiling) quantitative descriptive analysis, texture© Jones profile, and & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC sensory spectrum. of a food or beverage. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Flavor Profile Analysis® Flavor Profile Analysis (FPA) is based on the concept that flavor consists of identifiable taste, odor, and chemical-feel factors as well as an underlying complex of sensory impressions that are not separately identifiable.28 Scientists at © Jones & Bartlett Learning,Arthur LLC D. Little developed this ©technique Jones in & the Bartlett late 1940s Learning, and early LLC1950s, NOT FOR SALE OR DISTRIBUTIONand the name and the techniqueNOT were trademarked FOR SALE to ArthurOR DISTRIBUTION D. Little and Co.2 The method is a qualitative descriptive test that involves formal procedures for describing and assessing the aroma, flavor, and aftertaste of a product in a reproducible manner. FPA is a consensus technique.2 The vocabulary used to describe the sample and the sample evaluation itself are achieved by panel © Jones & Bartlett Learning, LLC members who ©work Jones together & Bartlett to come to Learning, an agreement. LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Quantitative Descriptive Analysis ® Quantitative Descriptive Analysis (QDA) was developed during the 1970s to correct some of the perceived problems associated with FPA.1,27 Unlike © Jones & BartlettFPA data, Learning, QDA data LLC are not generated through© consensus Jones & discussion, Bartlett andLearning, LLC panel leaders are not active participants. Unstructured line scales are used NOT FOR SALEto describeOR DISTRIBUTION the intensity of rated sensory attributes.NOT Stone FOR et SALEal.27 chose OR the DISTRIBUTION linear graphic scale, a line that extends beyond fixed verbal endpoints, based, in part, on earlier studies.29 In QDA, 10 to 12 panelists begin their training by generating a consensus 27 © Jones & Bartlett Learning,vocabulary. LLC They are exposed© to Jones many possible & Bartlett variations Learning, of the product LLC to facilitate the acquisition of an accurate concept, and the panel leader acts only NOT FOR SALE OR DISTRIBUTIONas a facilitator by directing discussionNOT FOR and supplying SALE OR materials DISTRIBUTION such as references and other samples required by the panel.2 The actual product evaluations are performed by each panelist individually, usually in an isolation booth.

Texture Profile Analysis® © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION Texture profileNOT analysis FOR (TPA) SALE was developed OR DISTRIBUTION by scientists working for General Foods during the 1960s and was subsequently modified by several sensory specialists.30–35 The TPA uses a standardized terminology to describe textural

03441_CH03_Printer.indd 100 29/12/12 11:36 PM TyPeS of SenSory TeSTS 101

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC 2 NOT FORcharacteristics SALE OR byDISTRIBUTION both their physical and sensory aspects.NOT Definitions FOR SALE and list -OR DISTRIBUTION ing order of the terms are determined through consensus by the TPA panelists. The reference scales anchor both the range and the concept for each term.35 The full range of a specific parameter by reference products helps panelists confirm the intensity increments within each scale. The use of the same reference frame is the key ©to Jonesa successful & Bartlett TPA. Sample Learning, preparation, LLC presentation, © Jones & Bartlett Learning, LLC and evaluation shouldNOT be strictly FOR controlled. SALE OR Panelists DISTRIBUTION should also be trained NOT FOR SALE OR DISTRIBUTION to bite, chew, and swallow in a standardized way. Sensory Spectrum ® Gail Civille, who became a TPA expert at General Foods, subsequently created© Jonesthe Sensory & Bartlett Spectrum Learning, (SS) technique. LLC2 This technique is an expan© Jones- & Bartlett Learning, LLC sion NOTon descriptive FOR SALE analysis OR techniques. DISTRIBUTION The unique characteristic of theNOT SS FOR SALE OR DISTRIBUTION technique is that panelists do not generate a panel-specific vocabulary to describe sensory attributes of products, but rather use a standardized word list (lexicon).36 The terminology used to describe a particular product is chosen a priori and remains the same for all products within a category © Jones &over Bartlett time.2 Because Learning, panelists LLC are trained to use the scales© Jones in an & identical Bartlett Learning, LLC NOT FORmanner, SALE useOR of DISTRIBUTION the SS technique should allow data fromNOT experiments FOR SALE that OR DISTRIBUTION include only one sample to be compared against data from different samples used in other studies. Panelist training for the SS technique is much more extensive than that for QDA, and the panel leader has a more directive role than in QDA.2 Similar to the TPA, panelists© Jones are provided& Bartlett lexicons Learning, that are LLCused to describe © Jones & Bartlett Learning, LLC perceived sensations associatedNOT FOR with SALE the samples. OR DISTRIBUTION The panelists use a numeri- NOT FOR SALE OR DISTRIBUTION cal intensity scale—usually a 15-point scale—and they are supplied with reference standards.

affective Tests © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC For a food product to be successful in the marketplace, consumers must preferNOT it over FOR other SALE products. OR Therefore, DISTRIBUTION consumer panels often are usedNOT to FOR SALE OR DISTRIBUTION indicate preference of one sample over another. The panelist rates his or her preference for one of the samples on a specific quality on the score sheet. Hedonic rating scales can be used to measure the degree of pleasure experienced with each sample. Sometimes, the frequency that a panelist might © Jones &desire Bartlett to eat theLearning, sample is LLCmeasured as a way to determine© Jones the acceptability & Bartlett Learning, LLC NOT FORof SALE the various OR DISTRIBUTIONsamples. NOT FOR SALE OR DISTRIBUTION Acceptance Tests Acceptance tests involve rating the difference in acceptance between two samples. The following example shows an application of a rating acceptance test: © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC • Situation: A restaurant manager wishes to compare acceptance between two NOTtypes ofFOR chocolate SALE candies OR DISTRIBUTION (products A and B). She NOT FOR SALE OR DISTRIBUTION wants to give the more-liked chocolate product as a customer appreciation present at the end of a meal. • Test objective: The test objective is to determine Please taste each sample, and indicate how well you like it. which chocolate product is more liked. © Jones & Bartlett Learning, LLC 318© (A)Jones & Bartlett Learning,442 (B) LLC • NOTTest FOR design: SALE 20 panelists OR DISTRIBUTION are asked to evaluate NOT FOR SALE OR DISTRIBUTION acceptance for the two chocolate products using ______Like very much (5) ______Like very much (5) 5-point hedonic scales. Two samples (coded ______Like moderately (4) ______Like moderately (4) with random three-digit numbers) are presented ______Neutral (3) ______Neutral (3) ______Dislike moderately (2) ______Dislike moderately (2) simultaneously at the same temperature. Half of ______Dislike very much (1) ______Dislike very much (1) the panelists test product A first; the other half © Jones & Bartlett Learning, LLC © JonesThank you!& Bartlett Learning, LLC NOT FOR SALEtest OR product DISTRIBUTION B first. NOT FOR SALE OR DISTRIBUTION • Score sheet: An example score sheet is shown in Figure 3.10. FiguRe 3.10 Example of a 5-point acceptance score sheet.

03441_CH03_Printer.indd 101 29/12/12 11:36 PM 102 Chapter 3 SenSory evaluaTIon

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION • ResultsNOT and dataFOR analysis: SALE Product OR DISTRIBUTION A has a higher acceptance rate than product B. The calculated t-value for the difference values exceeds the reference t-value and is statistically significant. • Conclusion: Product A is selected as a consumer appreciation present because product A is significantly more acceptable than product B. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC For detailed statistics regarding this example, see the Appendix at the NOT FOR SALEend OR of thisDISTRIBUTION chapter. NOT FOR SALE OR DISTRIBUTION Sometimes, rating acceptance tests involve more than two samples. The following is an example application of a rating acceptance test that uses multiple samples: © Jones & Bartlett Learning, LLC• Situation: A dietitian is© developing Jones & gluten-free Bartlett chocolate Learning, chip LLC NOT FOR SALE OR DISTRIBUTIONcookies prepared by replacingNOT FOR all-purpose SALE flour OR DISTRIBUTIONwith brown rice and chickpea flours. • Test objective: The test objective is to determine whether the gluten- free chocolate chip cookies (100% brown rice flour, 50% brown rice/50% chickpea flour, and 100% chickpea flour) are sufficiently © Jones Taste& Bartlett each sample, and Learning, indicate how well LLCyou like it. acceptable© Jones against & 100% Bartlett all-purpose Learning, flour cookies.LLC NOT FOR SALE OR DISTRIBUTION • Test design:NOT AFOR completely SALE randomized OR DISTRIBUTION design that compares all Sample code ______samples together using a rating scale is used. One hundred subjects Like extremely evaluate the overall acceptability of four samples on a 9-point Like very much hedonic scale. Like moderately • Score sheet: An example score sheet is shown in Figure 3.11. Like slightly Neither like nor dislike © Jones & Bartlett• Learning,Results and LLC analysis: The 100% chickpea© flourJones cookie & Bartlett product Learning, LLC Dislike slightly has a significantly higher acceptance rate than other gluten-free Dislike moderately NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Dislike very much cookies (50% brown rice/50% chickpea flour cookies and 100% Dislike extremely brown rice flour cookie products); however, there is no significant difference among the control (all-purpose flour) and 100% Thank you! chickpea flour cookies. © Jones & Bartlett Learning, LLC• Conclusion: The 100% ©chickpea Jones flour & Bartlett cookies areLearning, acceptable LLC gluten- FiguRe 3.11NOT Example FOR of a 9-point SALE verbal OR hedonic DISTRIBUTION free alternatives to the NOTconventional FOR SALEall-purpose OR flourDISTRIBUTION cookie. scale sensory sheet. For detailed statistics regarding this example, see the Appendix at the end of this chapter.

Preference Tests © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC The question of interest for the preference test is, “Which sample do you pre- NOT FOR SALE OR DISTRIBUTION fer?” The followingNOT example FOR SALE shows anOR application DISTRIBUTION of a paired preference test: • Situation: A caterer wants to know determine which brand of soft drink she should use. She wants to compare the preferences for two prospective products. © Jones & Bartlett• Learning,Test objective: LLC The test objective is to determine© Jones which & Bartlett soft drink Learning, LLC NOT FOR SALE OR DISTRIBUTIONproduct is preferred over the other product.NOT FOR SALE OR DISTRIBUTION • Test design: One hundred subjects who are soft-drink drinkers are invited to a central location where the company caters. Two products (A and B) coded with three-digit random numbers Taste the sample on the left rst and the sample on the right second. Which one do you prefer? Please choose one. are presented simultaneously at the same temperature. Half the © Jones & Bartlett Learning, LLC participants receive the© soft Jones drinks & in Bartlett the order Learning,A–B, and the LLC other 337NOT FOR SALE198 OR DISTRIBUTIONhalf receive them in theNOT order FOR of B–A. SALE The serving OR DISTRIBUTION temperature and Thank you! serving time after opening the containers are carefully controlled. (Both are particularly important variables to control for carbonated FiguRe 3.12 Example of paired preference score beverages.) sheet. • Score sheet: An example score sheet is shown in Figure 3.12. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 102 29/12/12 11:36 PM CHaPTer revIew 103

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE• Results OR DISTRIBUTION and analysis: The number (66) that choseNOT sample FOR SALE OR DISTRIBUTION A is larger than the critical value (61) at a level of statistical significance. • Conclusion: There is a significant preference difference between soft drink products. The analyst recommends that the caterer serve soft drink product© A. Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC For detailed statisticsNOT regarding FOR SALE this example, OR DISTRIBUTION see the Appendix at the NOT FOR SALE OR DISTRIBUTION end of this chapter. The final example in this chapter details theapplication of a ranking preference test: • © JonesSituation: & A Bartlett school dietitian Learning, wishes LLC to com pare preferences for © Jones & Bartlett Learning, LLC NOTthree FOR different SALE varieties OR DISTRIBUTIONof apples. NOT FOR SALE OR DISTRIBUTION • Test objective: The test objective is to determine whether there is a significant difference in preference among three types of apples (products A, B, and C). • Test design: The three samples are ranked by 100 panelists. Each © Jones & Bartlettpanelist Learning, receives three LLC samples coded with three-digit© Jones numbers & Bartlett and Learning, LLC NOT FOR SALEserved OR DISTRIBUTIONin balanced, random order. NOT FOR SALE OR DISTRIBUTION • Score sheet: An example score sheet is shown in Figure 3.13. • Results and analysis: When compared to the minimum critical differences at α = 0.05 (34), product B is preferred over products Please taste each of the coded samples in the set in the A and C. There is no significant preference difference between sequence presented from left to right. Rank the three samples in descending order of preference. You may products A and© JonesC. & Bartlett Learning, LLC re-taste ©any ofJones the samples & while Bartlett ranking for the Learning, LLC • Conclusion: ItNOT is recommended FOR SALE that OR the DISTRIBUTIONschool serve apple B, preference.NOT No ties FORare allowed SALE in the ranking. OR Remember DISTRIBUTION which is preferred over apples A and C. that the most preferred sample should be ranked 1. For detailed statistics regarding this example, see the Appendix at the Sample code ______123 end of this chapter. Thank you! © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Chapter review FiguRe 3.13 Example of a preference ranking NOT FOR SALE OR DISTRIBUTION NOT FORscore SALEsheet. OR DISTRIBUTION Sensory evaluation is a scientific testing method for accurate measurement of human responses as perceived by the five senses. Sensory evaluation is a vital part of food development because it is the essential means of determining how consumers will react to a food. Reliable sensory evaluation can be © Jones &performed Bartlett by Learning, optimizing fourLLC steps: definition of the© problem, Jones test& Bartlett design, Learning, LLC NOT FORinstrumentation, SALE OR DISTRIBUTION and interpretation. People evaluate a particularNOT FOR food SALE primar -OR DISTRIBUTION ily based on how it looks, smells, tastes, sounds, and feels. The food attributes that are typically perceived through the human senses are appearance, odor, taste, flavor, consistency, and texture. The environment in which the sensory test is conducted should be carefully controlled, and samples© Jones must & beBartlett prepared Learning, and presented LLC in a uniform © Jones & Bartlett Learning, LLC fashion so as not to NOTinfluence FOR panelists’ SALE ORperception DISTRIBUTION of the food’s quality. NOT FOR SALE OR DISTRIBUTION Panelists who are well suited to the purpose of the sensory test should be selected and trained appropriately. The two types of sensory tests are analytical and affective. Analytical tests are based on discernible differences, whereas affective tests are based on individual© Jones acceptability & Bartlett or Learning, preferences. LLC Analytical tests are divided© into Jones & Bartlett Learning, LLC two typesNOT of FOR tests: SALE difference OR tests DISTRIBUTION (discriminative tests) and descriptive NOTtests. FOR SALE OR DISTRIBUTION Depending on the main task of the test, affective tests are either acceptance tests or preference tests. The primary task of acceptance tests is to rate the degree of liking, whereas with preference tests the goal is to identify the item that is more liked. © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 103 29/12/12 11:36 PM 104 Chapter 3 SenSory evaluaTIon

Key Terms Study Points 1. Sensory evaluation is a scientifi c testing method for accurate measurement of © Jonespage & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC human responses as perceived by the fi ve senses. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION adaptation 87 2. The food attributes that are typically perceived through the fi ve senses are affective tests 96 appearance, odor, taste, fl avor, consistency, and texture. 3. The environment in which the sensory test is conducted, as well as sample analytical tests 96 preparation and presentation, should be carefully controlled so as not to bias consumer panel 90 the results of the test. Panelists who are well suited to the purpose of the sensory © Jones & Bartlett Learning, LLCtest should be selected and trained© Jones appropriately. & Bartlett Learning, LLC descriptiveNOT panel FOR SALE OR DISTRIBUTION90 4. The two types of sensory testsNOT are analytical FOR SALE and affective. OR DISTRIBUTIONAnalytical tests are descriptive tests 100 based on discernible differences, whereas affective tests are based on acceptability or preferences. difference tests 96 5. Analytical tests include difference tests (discriminative tests) and descriptive duo-trio test 97 tests. The two types of difference tests are overall difference tests and attribute © Jones & Bartlett Learning, LLC difference tests.© Jones & Bartlett Learning, LLC NOT FOR fl avor SALE OR DISTRIBUTION89 6. The triangleNOT test is anFOR overall SALE difference OR test. DISTRIBUTION In the triangle test, three samples are gustatory 87 presented simultaneously; two samples are alike and one is different. Panelists are asked to indicate the odd sample. The chance of obtaining a correct answer hedonic scale 94 by guessing is 33.3% (1 in 3) in the triangle test. holding time 92 7. The duo-trio test is an overall difference test. In the duo-trio test, the reference © Jones & Bartlettsample Learning, is presented LLC fi rst; it is then followed by two© Jonesother samples, & Bartlett one of which Learning, LLC molecular gastronomy 85 NOT FOR SALE ORis the DISTRIBUTION same as the reference. The judge is asked toNOT identify FOR which SALE of the last OR two DISTRIBUTION mouthfeel 89 samples is same as or different from the reference. There is a 50% chance of being right by simply guessing in a duo-trio test. nutrigenomics 88 8. The paired comparison test is an attribute difference test in which a specifi c char- olfactory 86 acteristic is designated. The subject is asked to test the two samples presented to identify the sample with the greater amount of the characteristic being measured. paired comparison© Jones test & Bartlett Learning,98 LLCThe subject has a 50% chance© of Jonesbeing right & byBartlett chance alone. Learning, LLC papillae NOT FOR SALE OR DISTRIBUTION87 9. Descriptive sensory tests are theNOT most comprehensiveFOR SALE and OR informative DISTRIBUTION tools used in sensory evaluation. Descriptive tests require a well-trained panel and should profi ling 100 never be used with consumers because consistent and reproducible data are rating difference test 99 an essential part of descriptive tests. Examples of descriptive tests include the Flavor Profi le Analysis (FPA), Quantitative Descriptive Analysis (QDA), Texture © Jones ranking & Bartlett test Learning, LLC 98 Profi le Analysis© Jones (TPA), &and Bartlett the Sensory Learning, Spectrum (SS) LLC technique. NOT FOR sensory SALE evaluation OR DISTRIBUTION84 10. Affective testsNOT include FOR acceptance SALE testsOR and DISTRIBUTION preference tests. The primary task of acceptance testing is to rate the degree of liking, whereas the preference task taste buds 87 seeks to identify the product that is liked more. texture 90 Issues for Discussion triangle test 96 © Jones & Bartlett1. Are Learning, sensory preferences LLC due to genetic or environmental© Jones causes? & Bartlett Learning, LLC umami NOT FOR89 SALE 2. OR Should DISTRIBUTION the food industry provide foods with preferredNOT tastes FOR that SALE may be OR sugar DISTRIBUTION and salt based or try to provide healthier foods, knowing that the taste sensations may not be what most consumers prefer? Is there an ethical choice? research Ideas for Students © Jones & Bartlett Learning,1. LLCTaste preferences of different© ethnic Jones groups & Bartlett and possible Learning, reasons for LLC such differences NOT FOR SALE OR DISTRIBUTION2. Taste preferences of cancer patientsNOT FOR SALE OR DISTRIBUTION references 1. Stone H, Sidel JL. Sensory Evaluation Practices. 3rd ed. San Diego, CA: Academic Press; 2004. © Jones & Bartlett Learning, LLC 2. Lawless HT,© Heymann Jones H. & SensoryBartlett Evaluation Learning, of Food: LLCPrinciples and Practices. NOT FOR SALE OR DISTRIBUTION 2nd ed. NewNOT York: Springer;FOR SALE 2010. OR DISTRIBUTION

03441_CH03_Printer.indd 104 29/12/12 11:36 PM learnIng PorTfolIo 105

3. Meiselman HL. Critical evaluation of sensory techniques. Food 21. De Roos KB. How lipids infl uence food fl avor.Food Tech. Qual Pref. 1993;4:33–40. 1997;51(5):60–62. 4. Pfenninger HB. Methods© Jones of quality & control Bartlett in brewing. Learning, Schweizer LLC22. Amerine MA, Pangborn RM,© RoesslerJones EB. & PrinciplesBartlett of SensoryLearning, LLC Brauerei-Rundschau. NOT 1979;90:121. FOR SALE OR DISTRIBUTIONEvaluation of Food. New York:NOT Academic FOR Press;SALE 1965. OR DISTRIBUTION 5. Meilgaard MC, Civille GV, Carr BT. Sensory Evaluation 23. International Organization for Standardization (ISO). Sensory Techniques. 4th ed. Boca Raton, FL: CRC Press; 2007. analysis-general guidance for the design of test rooms. In: Inter- 6. Brown A.Green Understanding Point Food: Principles and Preparation. national Standard ISO 8589. Geneva, Switzerland: International 3rd ed. Belmont, CA: Thompson-Wadsworth; 2008. Organization for Standardization; 1998. 7. Maruniak © JonesOrganic JA. The& Bartlett sense Sugar of smell. SugarLearning, In:sold Piggott in the U.S. LLCJR (ed.). Sensory 24. Lucak CL,© JonesDelwiche JF.& EffiBartlett cacy of variousLearning, palate cleansers LLC with representative foods. Chemosensory Perception. 2009;2:32–39. AnalysisNOT labeledFORof Foods. “organic” SALE 2nd ed. must OR London: meet DISTRIBUTION the Elsevier; requirements 1988. NOT FOR SALE OR DISTRIBUTION 8. McWilliams of the M. USDA. Foods: There Experimental are specifi c requirementsPerspectives . 6th ed. Upper 25. Stevens SS. Mathematics, measurement and psychophysics. In: Saddle River,for how NJ: organic Pearson/Prentice-Hall; sugar cane is raised, without2008. Stevens SS (ed.). Handbook of Experimental Psychology . New 9. Kim UK,using Jorgenson chemical E, pesticides Coon H, or et herbicides. al. Positional Organic cloning of the York: Wiley; 1951: 1–49. human quantitativesugar will perform trait identically locus underlying to refi ned sugartaste sensitivity to 26. Trochim WM. Research Methods: The Concise Knowledge Base . phenylthiocarbamide.and is available Science in virtually . 2003;299(5610):1221–1225. any conventional Cincinnati, OH: Atomic Dog Pub; 2005. © Jones &10. Bartlett Blakesleesugar AF. Learning, Genetic product: ofgranulated sensory LLC sugar, thresholds: brown sugar, taste for phenylthio-© Jones27. Stone & Bartlett H, Sidel J, OliverLearning, S, et al. Sensory LLC evaluation by quantitative NOT FOR SALEcarbamide. ORetc. ProcNatlAcadSciDISTRIBUTIONThe demand for organic USA. sugar 1932:120–130. has increased NOT FORdescriptive SALE analysis. OR DISTRIBUTION Food Tech. 1974;28:24–29, 32, 34. 11. Fox AL. Thebecause relationship the major between organic food chemical categories constitution use and taste. 28. Hootman RC (ed.). Manual on Descriptive Test for Sensory Proc Natlsugar Acad in Sci their USA. beverages, 1932:115–120. dairy products, cereal, Evaluation . ASTM Manual Series MNL 13. Philadelphia, PA: 12. Lawless chocolate,HT. A comparison confectionary, of and different preserves. methods for assessing American Society for Testing and Materials; 1992. sensitivitySource: to the http://www.usda.gov/oce/forum/2008_Speeches/ taste of phenylthiocarbamide PTC. Chem Senses. 29. Anderson NH. Functional measurement and psychological 1980;5:247–256.PDFPPT/Willerton.pdf © , Jonespublished 2008.’ & Accessed Bartlett March 3, Learning, LLCjudgment. Psych Rev. 1970;153–170.© Jones & Bartlett Learning, LLC 2011. 13. Bartoshuk LM, DuffyNOT VB, Miller FOR IJ. SALEPTC/PROP OR tasting: DISTRIBUTION anato- 30. Brandt MA, Skinner EZ, ColemanNOT JA. FOR The texture SALE profi OR le method.DISTRIBUTION my, psychophysics, and sex effects. Physiol Behav. 1994;56(6): J Food Sci. 1963;28:404–409. 1165–1171. 31. Civille GV, Liska IH. Modifi cations and applications to foods 14. Guo SW, Reed DR. The genetics of phenylthiocarbamide percep- of the general foods sensory texture profi le technique.J Texture tion. Ann Hum Biol. 2001;28(2):111–142. Studies. 1975;6:19–31. 15. Tepper B, Nurse R. Fat perception is related to PROP taster status. 32. Szczesniak AS. Texture measurements. Food Tech. 1966;20: Physiol© Jones Behav . & 1997;61(6):949–954. Bartlett Learning, LLC 1292–1298.© Jones & Bartlett Learning, LLC 16. Reedy NOT F FORJr, Bartoshuk SALE LM, OR Miller DISTRIBUTION I, et al. Relationship among 33. SzczesniakNOT AS. FORGeneral SALE foods texture OR profi DISTRIBUTION le revisited—Ten years papillae, taste pores, and 6-n-propylthiouracil (PROP) supra- perspective. J Texture Studies. 1975; 6:5–17. threshold taste sensitivity. [abstract]. Chem Senses. 1996;21:616. 34. Szczesniak AS. Classifi cations of textural characteristics. J Food 17. Hosako NY, Lucchina LA, Synder DJ, et al. Number of fungiform Sci. 1963;28:385–389. papillae in non-tasters, medium tasters and supertasters of PROP 35. Szczesniak AS, Brandt MA, Friedman HH. Development of © Jones & Bartlett(6-n-propylthiouracil). Learning, [abstract]. LLC Chem Senses. 1996;21:616.© Jonesstandard & Bartlett rating scales Learning, for mechanical LLC parameters of texture and 18. Duffy VB, Bartoshuk LM. Food acceptance and genetic variation correlation between the objective and the sensory methods of NOT FOR SALEin taste. OR J Am DISTRIBUTION DietAssoc. 2000;100:647–655. NOT FORtexture SALE evaluation. OR DISTRIBUTION J Food Sci. 1963;28:397–403. 19. Duffy VB, Peterson JM, Dinehart ME, Bartoshuk LM. Genetic 36. Civille GV, Lyon B. ASTM Lexicon Vocabulary for Descriptive and environmental variation in taste. Top Clin Nutr. 2003;18(4): Analysis. Philadelphia, PA: American Society for Testing and 209–220. Materials; 1996. 20. Godshall MA. How carbohydrates infl uence food fl avor.Food Tech. 1997;51(1):62–67.© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 105 29/12/12 11:36 PM 106 Chapter 3 SenSory evaluaTIon

Sensory evaluation © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Data analysis for the Triangle Test NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Step 1: Refer to the reference minimum number of correct responses for the triangle test.1 • Application : Because the number of panelists (trials) in this example is 48, the minimum number of correct responses © Jones & Bartlett Learning, LLC corresponding to ©the Jones probability & Bartlett level of 0.05 Learning, is 22. 1 LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Step 2: Count the number of panelists who correctly identifi ed the odd sample. • Application : Out of 48 panelists, 25 correctly chose the odd sample. © Jones & Bartlett Learning, LLC Step 3: ©If Jonesthe value & (step Bartlett 2) is same Learning, or larger LLCthan the reference NOT FOR SALE OR DISTRIBUTION value (stepNOT 1), FORwe can SALE say that OR the DISTRIBUTION panelists could detect a signifi cant difference between the samples at a probability of 5%. • Application : The value (25) from step 2 is larger than the reference value. Therefore, the panelists could detect a © Jones & Bartlett Learning,difference LLC between samples. © Jones & Bartlett Learning, LLC NOT FOR SALE Data OR DISTRIBUTIONanalysis for the Duo-Trio Test NOT FOR SALE OR DISTRIBUTION Step 1: Refer to the critical number of correct responses in the duo- trio test.1 • Application : Because the number of panelists (trials) in © Jones & Bartlett Learning, LLC this example is 40,© theJones critical & numberBartlett of Learning,correct responses LLC NOT FOR SALE OR DISTRIBUTION corresponding toNOT the probability FOR SALE level ORof 0.05 DISTRIBUTION is 26. 1 Step 2: Count the number of panelists who correctly matched the sample to the reference. • Application : Out of 40 panelists, 13 correctly matched the © Jones & Bartlett Learning, LLC sample© Jones to the & reference.Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION Step 3: NOTIf the FORvalue (stepSALE 2) isOR the DISTRIBUTIONsame or larger than the table value (step 1), we can say that the panelists could detect a signifi cant difference between the samples at a probability of 5%. • Application : The value (13) from step 2 is smaller than the © Jones & Bartlett Learning,reference LLC value. Therefore, the panelists© Jones could & not Bartlett detect a Learning, LLC difference between samples. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Data analysis for the Paired Comparison Test Step 1 : Refer to the reference minimum number of correct judgments for the paired comparison test.1 © Jones & Bartlett Learning, LLC • Application : In our© Jonescase, the &number Bartlett of trials Learning, is 28 and LLC the NOT FOR SALE OR DISTRIBUTION probability level NOTis 0.05. FOR The referenceSALE OR value DISTRIBUTION is 19.1 Step 2: Count the number of panelists who chose sample A or B, respectively. • Application : Six panelists chose sample 796 (cookie A, © Jones & Bartlett Learning, LLC the© Jonesoriginal) & as Bartlett the sweeter Learning, sample and LLC 22 panelists chose NOT FOR SALE OR DISTRIBUTION sampleNOT FOR308 (cookie SALE B, OR the DISTRIBUTIONnew one).

03441_CH03_Printer.indd 106 29/12/12 11:36 PM aPPenDIX 107

Step 3: If the value (step 2) is the same or larger than the reference value (step 1), we can say that the sample that was chosen more often is signifi ©cantly Jones different & Bartlett (sweeter) Learning, than the other LLC sample. © Jones & Bartlett Learning, LLC • Application NOT : The FOR value SALE (22) from OR stepDISTRIBUTION 2 is larger than the NOT FOR SALE OR DISTRIBUTION table value (19). Therefore, the panelists could detect a difference in sweetness between samples. Data analysisGreen for Point the ranking Test © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC 2 TheNOT ranking OrganicFOR data SALE can Sugar be OR analyzed Sugar DISTRIBUTION sold eitherin the U.S. by using the Basker’s tablesNOT or FOR SALE OR DISTRIBUTION those by Newelllabeled “organic”and MacFarlane. must meet the 3 Thisrequirements analysis used a reworked table 1 from Newell andof theMacFarlane’s. USDA. There are 3specifi Table c requirements A shows how the results are organized. The ranksfor are how summed, organic sugar and cane the is raised, differences without between sums are compared to the criticalusing valueschemical inpesticides the table. or herbicides. From Organicthe table in the reference, the cor- respondingsugar number will perform of identicallysamples to onrefi nedthe sugar horizontal axis (in our example, © Jones & Bartlettand Learning, is available in virtually LLC any conventional © Jones & Bartlett Learning, LLC 3 samples) and the number of panelists from the vertical axis (in our exam- NOT FOR SALE ORsugar DISTRIBUTION product: granulated sugar, brown sugar, NOT FOR SALE OR DISTRIBUTION ple, 28 panelists)etc. The demand are selected. for organic Thesugar number has increased where the two points cross will be the criticalbecause values the major for organic the difference. food categories In use our case, the critical value is 18. If the differencesugar in of their sum beverages, of their dairy rank products, between cereal, each pair of samples is greater than the criticalchocolate, value, confectionary, it is andconsidered preserves. that the samples are signifi cantly different inSource: saltiness. http://www.usda.gov/oce/forum/2008_Speeches/© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC PDFPPT/Willerton.pdf , published 2008.’ Accessed March 3, 2011. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION TABLe A Results of the Ranking Test for the Saltiness of Three Types of Cheeses Panelist No. A (183, Mozzarella) B (479, Cheddar) C (862, Provolone) 1 © Jones3 & Bartlett Learning,2 LLC 1 © Jones & Bartlett Learning, LLC 2 NOT FOR3 SALE OR DISTRIBUTION2 1 NOT FOR SALE OR DISTRIBUTION 3 2 3 1 4 2 1 3 5 1 2 3 © Jones & Bartlett6 Learning,3 LLC 1 ©2 Jones & Bartlett Learning, LLC NOT FOR SALE7 OR DISTRIBUTION3 2 NOT1 FOR SALE OR DISTRIBUTION 8 2 1 3 9 2 1 3 10 2 3 1 11 2 © Jones & 1Bartlett Learning,3 LLC © Jones & Bartlett Learning, LLC 12 2 NOT FOR SALE1 OR DISTRIBUTION3 NOT FOR SALE OR DISTRIBUTION 13 2 1 3 14 1 3 2 15 2 1 3 16 © Jones 2& Bartlett Learning,1 LLC 3 © Jones & Bartlett Learning, LLC 17 NOT FOR3 SALE OR DISTRIBUTION2 1 NOT FOR SALE OR DISTRIBUTION 18 3 2 1 19 2 1 3 20 2 3 1 © Jones & 21Bartlett Learning,3 LLC 1 ©2 Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR( continuesSALE )OR DISTRIBUTION

03441_CH03_Printer.indd 107 29/12/12 11:36 PM 108 Chapter 3 SenSory evaluaTIon

TABLe A Results of the Ranking Test for the Saltiness of Three Types of Cheeses (continued) © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Panelist No. A (183, Mozzarella) B (479, Cheddar) C (862, Provolone) NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 22 1 2 3 23 1 2 3 24 3 1 2 25 1 2 3 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC 26 3 1 2 NOT FOR SALE OR DISTRIBUTION27 3 NOT1 FOR SALE OR2 DISTRIBUTION 28 3 2 1 Column sum 62 46 60 Differences A vs. B = 16 B vs. C = 14 A vs. C = 2 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Data analysis for the rating Difference Test The data is evaluated by one-way analysis of variance (ANOVA). There are 4 treatments and 12 observations per treatment (see Table B). Because the © Jones & BartlettF-value Learning, (26.65) is veryLLC signifi cant P( < 0.0001)© (seeJones Table & CBartlett), there areLearning, LLC NOT FOR SALEsignifi OR cantDISTRIBUTION differences between treatments. NOT FOR SALE OR DISTRIBUTION

TABLe B Results of the Rating Test for the Moistness of Four Types of Pound Cake © Jones & Bartlett Learning,Panelist LLC No. © JonesTreatment & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION 0% Applesauce 33%NOT FOR SALE66% OR DISTRIBUTION100% (Control) Applesauce Applesauce Applesauce 1 9 8 8 5 2 8 8 7 6 3 9 7 7 4 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC 4 6 5 5 3 NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 5 7 8 6 5 6 8 9 8 5 7 9 7 7 5 8 9 8 8 4 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC 9 8 7 8 4 NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 10 7 8 7 5 11 7 6 6 5 12 7 8 8 4

© Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTIONTABLe C NOT FOR SALE OR DISTRIBUTION Completely Randomized Design One-Way Analysis of Variance of Results in Table B Source of Sum of Squares Degree of Mean F-value P-value Variation Freedom Square © Jones & Bartlett Learning, LLC Treatments© 77.06 Jones & Bartlett3 Learning,25.69 LLC 26.65 < 0.0001 NOT FOR SALE OR DISTRIBUTION Error NOT42.42 FOR SALE44 OR DISTRIBUTION 0.96 Total 119.48 47

03441_CH03_Printer.indd 108 29/12/12 11:36 PM aPPenDIX 109

To determine which samples are signifi cantly different, perform a Tukey Honestly Signifi cant Difference (HSD) multiple comparison test. The results are as follows: © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Control, 0% applesauceNOT FOR 7.83bSALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION 33% applesauce 7.42b 66% Green applesauce Point 7.08b 100% applesauce 4.58a © JonesOrganic & Bartlett Sugar SugarLearning, sold in the U.S. LLC © Jones & Bartlett Learning, LLC Data analysislabeled “organic” for the must meetaffective the requirements Test: Chocolate Candies NOT ofFOR the USDA. SALE There are OR specifi DISTRIBUTION c requirements NOT FOR SALE OR DISTRIBUTION The result of the acceptance test for two types of chocolate candies and the for how organic sugar cane is raised, without t -value calculationusing chemical are pesticides shown or in herbicides. Table OrganicD . In this case, the critical t -value (degree of sugarfreedom will perform = 19) identically from the to Student’srefi ned sugar t -distribution table at α = 0.05 is 2.093 (a student’sand is available t-distribution in virtually any table conventional can be found in most statistics books or © Jones &online Bartlett [see sugarhttp://www.itl.nist.gov/div898/handbook/eda/section3/eda3672. Learning, product: granulated LLC sugar, brown sugar, © Jones & Bartlett Learning, LLC NOT FOR htmSALE]). Because ORetc. DISTRIBUTIONThe the demand calculated for organic t-value sugar (2.62)has increased is larger thanNOT the reference FOR SALE t -value, OR DISTRIBUTION a signifi cantbecause acceptance the major organicdifference food categories exists betweenuse the samples. sugar in their beverages, dairy products, cereal, TABLe D chocolate, confectionary, and preserves. Results of theSource: Acceptance http://www.usda.gov/oce/forum/2008_Speeches/ Test for Two Types of Chocolate Candies PDFPPT/Willerton.pdf , published 2008.’ Accessed March 3, © Jones & Bartlett Learning, LLC2 © Jones & Bartlett Learning, LLC Panelist2011. Original (A) Modifi ed B( ) Difference (A – B = D) D 1 4 NOT4 FOR SALE 0OR DISTRIBUTION0 NOT FOR SALE OR DISTRIBUTION 2 5 3 2 4 3 3 3 0 0 4 4 2 2 4 5© Jones 4 & Bartlett3 Learning, 1LLC 1 © Jones & Bartlett Learning, LLC 6NOT FOR 4 SALE OR3 DISTRIBUTION1 1 NOT FOR SALE OR DISTRIBUTION 7 3 4 −1 1 8 4 4 0 0 9 5 3 2 4 10 4 4 0 0 © Jones & 11Bartlett 3Learning, LLC3 0 © Jones0 & Bartlett Learning, LLC NOT FOR SALE12 OR 5 DISTRIBUTION4 1 NOT FOR1 SALE OR DISTRIBUTION 13 4 2 2 4 14 3 3 0 0 15 4 4 0 0 16 5 © Jones4 & Bartlett1 Learning, LLC1 © Jones & Bartlett Learning, LLC 17 4 NOT5 FOR SALE−1 OR DISTRIBUTION1 NOT FOR SALE OR DISTRIBUTION 18 5 4 1 1 19 3 3 0 0 20 5 5 0 0 Sum 81 70 11 23 © Jones & Bartlett Learning,2 2 LLC © Jones & Bartlett Learning, LLC Sum of D = 11, Mean of D = 0.55, Sum of D = gD = 23 NOT2 FOR2 SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION gD 2 gD /N

N 2 1 1 2 StandardÅ deviation (SD) of D = 23 2 (121/20) 23 2 6.05 16.95 5 5 5 0.892 5 0.94 © Jones & Bartlett20 2 1 Learning,19 LLC 19 © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION " NOT FOR SALE OR DISTRIBUTION ÅStandard error (SE) of D Å= SD of D/ N = 0.94/Å 20 = 0.94/4.47 = 0.21 t = Mean of D/SE of D = 0.55/0.21 = 2.62 ! " © Jones & Bartlett Learning, LLC. NOT FOR SALE OR DISTRIBUTION.

03441_CH03_Printer.indd 109 29/12/12 11:36 PM 110 Chapter 3 SenSory evaluaTIon

Data analysis for the affective Test: Cookie formulations © Jones & BartlettThe result Learning, of one-way LLC analysis of variance for© the Jones overall & acceptance Bartlett Learning,of LLC NOT FOR SALEgluten-free OR DISTRIBUTION cookies is shown in Table E . NOT FOR SALE OR DISTRIBUTION

TABLe e Completely Randomized Design One-Way Analysis of Variance for the Overall Acceptance of gluten-Free Cookies © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC Source of Sum of Squares Degree of Mean F-value π-value NOT FOR SALE OR DISTRIBUTIONVariation FreedomNOT FORSquare SALE OR DISTRIBUTION Treatments 968.84 3 322.95 439.23 < 0.0001 Error 291.16 396 0.74 Total 1260.00 399 © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Because the F-value is statistically signifi cant, there are signifi cant differences among treatments. Therefore, the Tukey Honestly Signifi cant Difference (HSD) Post-hoc test was performed. The results of the Tukey HSD test are as follows: © Jones & Bartlett Control,Learning, 100% LLC all-purpose fl our cookies© Jones & 7.87c Bartlett Learning, LLC NOT FOR SALE OR 100% DISTRIBUTION brown rice fl our cookies NOT FOR 4.21aSALE OR DISTRIBUTION 50% brown rice/50% chickpea fl our cookies 6.57b 100% chickpea fl our cookies 8.15c

Data analysis from the Paired Preference Test © Jones & Bartlett Learning, LLCStep 1 : Refer to the reference© Jones minimum & Bartlett number ofLearning, agreeing LLC NOT FOR SALE OR DISTRIBUTIONjudgments for the paired NOTpreference FOR test. SALE2 OR DISTRIBUTION • Application: In our case, the number of trials is 100, and the probability level is 0.05. The critical reference value is 61.1 Step 2: Count the number of subjects who chose sample A or B, © Jones & Bartlett Learning, LLC respectively.© Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION • Application NOT FOR : 66 SALE subjects OR chose DISTRIBUTION sample 337 (soft drink A) as the preferred sample and 34 panelists chose sample 108 (soft drink B). Step 3: If the value (step 2) is same or larger than the critical value © Jones & Bartlett(step Learning, 1), we can LLC say the sample that was chosen© Jones more & is Bartlettsignifi cantly Learning, LLC preferred over the other sample. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION Data analysis for ranking Preference Test Ranks are added, and the differences between the sums are compared to the critical reference value. 1 Similar to the data analysis for the difference © Jones & Bartlett Learning,ranking LLC test, select the corresponding© Jones number & Bartlett of samples Learning, on the horizontal LLC NOT FOR SALE OR DISTRIBUTIONaxis and then select the numberNOT of panelists FOR SALEfrom the OR vertical DISTRIBUTION axis in the reference table.1 Find the number where the two points cross. In this example, because 100 panelists ranked three products, the critical value is 34.1 If the difference of the rank sum between each pair of samples is greater than the critical value, then the samples are signifi cantly different in preference. © Jones & Bartlett Learning, LLC Because the rank© Jonesscale used & wasBartlett 1 = preferred Learning, most andLLC 3 = preferred least, NOT FOR SALE OR DISTRIBUTION the smallest rankNOT sum FOR means SALE that thatOR theDISTRIBUTION product is the most preferred.

03441_CH03_Printer.indd 110 29/12/12 11:36 PM aPPenDIX 111

In Table F , because the differences of rank sum between product B and the other products (A and B, B and C) are larger than the critical value (34), product B is signifi cantly© Jones preferred & Bartlett over the Learning,other products. LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION TABLe F Results of theGreen Ranking Preference Point Test for Three Varieties of Apples 908 (A) 144 (B) 862 (C) Rank© sum JonesOrganic & Bartlett 216bSugar SugarLearning, sold in the164a U.S. LLC 220b © Jones & Bartlett Learning, LLC Differences labeled “organic”A vs. Bmust = 52 meet the requirementsB vs. C = 56 A vs. C = 4 NOT ofFOR the USDA. SALE There are OR specifi DISTRIBUTION c requirements NOT FOR SALE OR DISTRIBUTION for how organic sugar cane is raised, without references using chemical pesticides or herbicides. Organic sugar will perform identically to refi ned sugar 1. Lawless HT, Heymann H. Sensory Evaluation of Food: Principles and Practices. and is available in virtually any conventional 2nd ed. New York: Springer; 2010: 563, 565, 566. © Jones & Bartlettsugar Learning, product: granulated LLC sugar, brown sugar, © Jones & Bartlett Learning, LLC NOT FOR 2.SALE Basker OR etc.D. CriticalDISTRIBUTIONThe demand values for organic of differences sugar has increasedamong rank sumsNOT for multiple FOR SALEcompari- OR DISTRIBUTION sons. Foodbecause Tech. the 1988major organic July:88–89. food categories use 3. Newell sugarGJ, MacFarlane in their beverages, JD. Expanded dairy products, tables cereal, for multiple comparison procedures in the analysischocolate, of confectionary, ranked data. and Jpreserves. Food Sci.1987;52:1721–1725. Source: http://www.usda.gov/oce/forum/2008_Speeches/ PDFPPT/Willerton.pdf© , Jonespublished 2008.’ & Accessed Bartlett March 3, Learning, LLC © Jones & Bartlett Learning, LLC 2011. NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 111 29/12/12 11:36 PM © Jones & Bartlett Learning, LLC © Jones & Bartlett Learning, LLC NOT FOR SALE OR DISTRIBUTION NOT FOR SALE OR DISTRIBUTION

03441_CH03_Printer.indd 112 29/12/12 11:36 PM