Internal preference mapping of milk–fruit beverages: Influence of color and appearance on its acceptability

Food Sci Nutr. 2018;6:27–35.   | 27 www.foodscience-nutrition.com Received:17March2017 | Revised:17May2017 | Accepted:22May2017 DOI: 10.1002/fsn3.494 ORIGINAL RESEARCH Internal preference mapping of milk–fruit beverages: Influence of color and appearance on its acceptability Rocío Fernández-Vázquez1 | Carla M. Stinco1 | Dolores Hernanz Vila2 | Francisco J. Heredia1 | Carolina Chaya3 | Isabel M. Vicario1 ThisisanopenaccessarticleunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsuse,distributionandreproductioninanymedium, providedtheoriginalworkisproperlycited. ©2017TheAuthors.Food Science & NutritionpublishedbyWileyPeriodicals,Inc. 1FoodColour&Quality Laboratory,DepartmentofNutrition&Food Science,FacultaddeFarmaciaUniversidadde Sevilla,Sevilla,Spain 2DepartmentofAnalyticalChemistry,Facultad deFarmacia,Sevilla,Spain 3DepartmentofAgricultural Economics,StatisticsandBusiness Management,UniversidadPolitécnicade Madrid,Spain Correspondence IsabelM.Vicario,FoodColour&Quality Laboratory,DepartmentofNutrition&Food Science,FacultaddeFarmaciaUniversidadde Sevilla,Spain. Email:[email protected] Funding information Thisworkwassupportedbyfundingfromthe ConsejeríadeInnovaciónCienciayEmpresa, JuntadeAndalucíabytheprojectP11- AGR-7783. Abstract Theindividualpreferencesof100consumersbetween20and30yearsoldforthe colorof16milk–fruitjuicebeverages(MFJB)wereinvestigatedbypreferencemappingtechnique.Consumerswereaskedtoevaluate,justbylookingatthesamples, howmuchtheylikedthem(from“Extremelydislike”to“Extremelylike”).Thecolorof thesampleswasanalyzedbytwodifferentinstrumentaltechniques.Resultsobtained fromtheinstrumentalcolormeasurementshowedthewidediversityinhuesofthe beveragesavailableinthemarket,andcorrelationsbetweentechniquesprovedthat bothofthemwereappropriatetoanalyzecolor. Resultsshowedthatparticipantspreferredsampleswithorangishappearanceinstead ofthosewithawhiterlook.Anyway,punctuationsgivenbytheconsumerssuggest thatgenerally,coloroftheseproductsisnothighlyevaluatedbyconsumers,asthe bestmeanpunctuationwas6.6. KEYWORDS acceptance,color,consumersstudy,milk–fruitbeverage 1 | INTRODUCTION Beveragesmadefrommilkandfruitjuice(MFJB)haveproliferatedin theEuropeanmarketduringthelastyearsinresponsetoagrowing demandfornaturalproductsthatareperceivedashealthierbyconsumers.Thesebeveragesareconsideredthemostwidelyconsumed functionalfoods(Pszczola,2005); however,therearelittledatarelatedtoquality,safety,oracceptabilityoftheseproducts(Sampedro, Geveke,Fan,Rodrigo,&Zhang,2009). Fromanutritionalpointofview,theydonotreplaceorarenot equivalenttoaglassofmilkoraportionoffruit.Thus,initscomposition,fruitcontentsrangefrom7%to41%andtheyusuallycomefrom differentconcentratedfruits;milkmeansupto30%,andtheycontain alsovitamins,fiber,andsugars. Instrumentalcolormeasurementsaresimple,accurate,versatile, andobjective.Tocharacterizecolor,therearedifferentinstruments, such as colorimeters, spectrophotometers, and spectroradiometers, andindependentlyoftheinstrumentused,itisveryimportanttodefine measurementsconditionscorrectly.CIE(InternationalCommissionon Illumination) recommends observers (CIE, 1991a), illuminants (CIE, 1991b),andcolorspaces(CIE,1978),inordertonormalizeobjective colormeasurement.Itmustbealsotakenintoaccountsampleposition,systemgeometry,lightsourceintensity,sampledeepness,blank measurement, background, and surrounding (Meléndez-Martínez, Vicario,&Heredia,2006;Stincoetal.,2012).Digitalimageanalysis (DIA)isbeingusedincreasinglyforinstrumentalfoodcolorspecificationalthoughmoretraditionaltechniqueslikespectroradiometry(SPE) willalsoprovideanappropriatecharacterizationofcolorinbeverages (Fernández-Vázquez,Stinco,Hernanz,Heredia,&Vicario,2013,2014; Stinco,Fernámdez-Vázquez,Heredia,Meléndez-Martínez,&Vicario, 2014). Furthermore,colorisoneofthemostimportantattributesrelated toquality,affectingchoiceofpurchase(Baker&Günter,2004;Calvo, Salvador, & Fiszman, 2001). For the consumers, color is, alongside 28 | FERNÁNDEZVÁZQUEZ Et al. TABLE1 Compositionofthecommercialbeveragesanalyzed Samples Main ingredients Colorant Fiber Vitamins Jelling agents Sugar added Other S1aSkimmedmilk,concentrateorange 5% B-Carotene 1%inulin – – 7.6%sugar Corndextrose,ascorbicacid,citricacid, aromas,milkproteins,guargum,lactobacillus acidophilus,bifidobacterium,plantsterols, andsterolsesters S2 Water,juicefromconcentrate: pineapple,carob,appleandgrape, orangeandlemon(12%),skimmed milk(4.5%) B-Carotene – – Pectin Sugar,sucralose,acesulfame-K Milkprotein,citricacid,aromas,ascorbicacid S3aPartiallyskimmedmilk,orangejuice fromconcentrate(5%) B-Carotene 1%inulin – – 7.6%sugar Corndextrose,sodiumcitrates,ascorbicacid, guargum,milkproteins,lactobacillus acidophilus,milkenzymes,bifidobacterium S4aSkimmedmilk,6%concentrate (orange,pineapple,mango) B-Carotene GumArabic – – Sugar Bifidobacterium,milkenzymes,aromas S5 Water,fruitjuice:orange,apple, pineappleandlemon(15%), skimmedmilk(10%) – – – Pectin Sugarandsucralose Citricacid,ascorbicacid,aromas S6 Water,fruitjuice:pineapple,orange, appleandlemon(15%),skimmed milk(10%) – – – Pectin Sugar,acesulfame-K,neohesperidinDC Citricacid,ascorbicacid,aromas S7 Water,fruitjuice:orange(25%), apple(18%),pineapple(5%),lemon (1%),skimmedmilk(10%) –Fiber A,C,E Pectin Glucoseandfructose(syrup), sugar Citricacid,aromas S8 Water,fruit,andvegetablejuice: orange,grape,pineapple,carrot, peach,andpassionfruit(25%), skimmedmilk8% B-Carotene – – Pectin Sucralose,acesulfame-K Citricacid,ascorbicacid,aromas S9 Water,fruit,andvegetable concentratejuice:orange(8%), carrot(5%),lemon(2%),pineapple (2%),passionfruit(2%),skimmed milk(10%),pureepeach(4%) B-Carotene Dietaryfibers C,E Pectin,xanthan gum Fructose(syrup),sucralose Citricacid,aromas S10 Water,skimmedmilk(10%), Concentrate:pineapple,mango(7%) B-Carotene –A,C,E Pectin Sugar,sucralose Citricacid,aromas S11 Water,fruit,andvegetablejuices: orange,carrot,peach(15%), skimmedmilk(10%) B-Carotene – – Pectin Saccharose Citricacid,ascorbicacid,aromas S12 Water,concentratefruitjuice: orange,pureepeach,carrot(18%), skimmedmilk(10%) B-Carotene –A,C,E Pectin Sucralose Citricacid,aromas (Continues) | 29 FERNÁNDEZVÁZQUEZ Et al. freshness,oneofthemaincriteriaforfoodselection,andakeyfactorin sensoryacceptance(Fernández-Vázquez,Stinco,Melendez-Martinez, Heredia,&Vicario, 2011; Quitao-Teixeira,Aguiló-Aguayo, Ramos,& Martín-Belloso, 2008).Traditionally, this kind of products has been oftenconsumedintheirsingle-usepackages,meaningthatconsumers couldnotperceivethecolor,whichmightbethereasonwhyindustries havenotputtoomuchattentiontothischaracteristicofthebeverages. However,currently,duetotheincreasingintheconsumption,manyof thecompaniesarechangingthepackagesofthebeveragestosome withmorevolume,whichmeansconsumerswilldrinktheminglasses, andtherefore,theywillperceiveandevaluatetheircolor. Despitetherelevanceofmilk–fruitbeveragesMFJBinthemarket, uptonowthereisalackofinformationaboutthecolorpreferences fortheseproductsandits consumeracceptance.Thus,the aimsof thisstudywere(1)tocharacterizethecolorofMFJBbeveragesusing differenttechniques,(2)tostudytheconsumers’acceptanceoftheir color,and(3)to identifyand characterizedifferent consumerssegmentsintermsofcoloracceptancebymeansofinternalpreference mapping. 2 | MATERIALS AND METHODS 2.1 | Samples SixteencommerciallyavailableMFJB formulatedwithmilk or dairy products and fruit juice were purchased in different supermarkets inSpain,andtheircompositionsareshowninTable1.Threeofthe samplesaresoldatrefrigerationtemperature(4±2°C)astheywere pasteurized,whiletheother13samplesaresoldatroomtemperature (20±2°C)sincetheyweresubmittedtoaUHTprocess. 2.2 | Color measurements 2.2.1 | Spectrophotometry (SPE) The color of the beverages was measured in a spectrophotometer CM5(KonicaMinoltaSensingAmericas,Inc.,NY).Eachsamplewas containedin75-mlcapacitytransparentplasticbottles.ThecolorparametersoftheuniformcolorspaceCIELABL*; a*; and b*wereobtaineddirectlyfromtheapparatus.Colordataobtainedwereaverages of three measurements. 2.2.2 | Digital image analysis (DIA) TheDigiEyeimagingsystemwasusedtocapturethedigitalimages (Luo,CUI,&LI,2001).Thelattersystemincludesacalibrateddigital camerawith10.2-megapixelNikonD80(NikonCorporation,Tokyo, Japan)andanobjectiveNikkor35-mmf/2D(NikonCorporation),a colorsensorfordisplaycalibration,andanilluminationboxdesigned byVeriVideLtd.(Leicester,UK). Thesampleswereplacedin75-mlcapacitytransparentplasticbottles,illuminatedbyadiffuseD65simulator,andmeasuredagainsta graysurround(L*=50)andwhitebackground. Samples Main ingredients Colorant Fiber Vitamins Jelling agents Sugar added Other S13 Water,fruit,andvegetablejuice: orange,carrot,pineapple,passion fruit,mango,guava,papaya,and apricot(25%),skimmedmilk(10%) –Fiber A,C,E Pectin Sugar Water,citricacid,aromas S14 Water,juicefromconcentrate: orange,pineapple,apple,papaya, carrot,mango,guava(20%), skimmedmilk(10%) B-Carotene –C,E Pectin Sugar,glucose,andfructose (syrup),sucralose Citricacid S15 Water,juicefromconcentrate: orange,apple,andmango, skimmedmilk(10%) B-Carotene –A,C,E Pectin Sucralose,acesulfame-K Citricacid,aromas S16 Water,juicefromconcentrate:8% orange,1%mango,1%pineapple, skimmedmilk(10%) B-Carotene – E Pectin Sugar Citricacid,sodiumcitrates aSamplesstoredunderrefrigeration. TABLE1 (Continued) 30 | FERNÁNDEZVÁZQUEZ Et al. TocalculatetheCIELABcoordinatesfromRGBcolorspace,the DigiFood software was used (Heredia, González-Miret, Álvarez, & Ramírez,2006). FromtheCIELABuniformcolorspace,thepsychophysicalparameterschroma(C*ab)andhue(hab)aredefinedasfollows: Chroma( C∗ab )isusedtodeterminethedegreeofdifferenceofa hueincomparisonwithagraycolorwiththesamelightnessandis consideredthequantitativeattributeofcolorfulness.Hue(hab)isthe attribute according to which colors are usually defined as reddish, greenish,etc.andisusedtodefinethedifferenceofacolorwithreferencetoagraycolorwiththesamelightness.Thisattributeisrelatedto thedifferencesinreflectanceatdifferentwavelengthsandisconsideredthequalitativeattributeofcolor. 2.3 | Consumer study OnehundredSpanishconsumerswererecruitedfromstaffandstudents at the University of Sevilla. Information regarding demographics and consumptionhabitswascollectedviaaquestionnairepriortothesensoryassessmentofthesamples.Alltheconsumerswere20–30yearsold (30%malesand70%females),whichisespeciallyinterestingasitisthe groupofpopulationwhicharepotentialconsumersoftheseproducts. Thetestwascarriedoutindesignedindividualsensorybooths, underNorthernHemispherelightingconditions.Samples(75ml)were presentedmonadicallyinthesamebottlesusedfortheothermeasurements,labeledwiththreedigitsrandomcodes,inarandomizedorder. Consumers were asked to evaluate how much they liked the appearance of the beverages (from “Dislike extremely” to “Like extremely”)usingthe9-pointhedonicscale,justbylookingatthesamples.Theratingdecisionwasbasedonlyontheappearance,without further information. 2.4 | Data analysis Thestatisticalanalysisofinstrumentalcolordatawasperformedby one-way analysis of variance (ANOVA), and statistically significant differences(p < 0.05)weredeterminedusingtheTukeymultiplecomparisontest.Correlationanalysiswasdonebetweenthecolorimetric parametersmeasuredbybothinstrumentaltechniquesandbetween instrumentalanalysisandconsumerstudyresults. Consumer data first underwent normality testing (Shapiro–Wilk test) and were subsequently analyzed using nonparametric tests (Kruskal–Wallis)toidentifydifferencesamongsamples.Then,these datawere further examined using hierarchal cluster analysis, using SquaredEuclideanDistancesandWardscriterion,andinternalpreferencemapping. Independence between demographic variables and consumer clusterswereanalyzedbyχ2 test. AllstatisticalanalyseswereperformedusingtheprogramStatistica 8 for Windows (StatSoft, 2007) and XLStat (Version 2009.6.03, Addinsoft,USA). C ∗ ab = √ (a∗)2+(b∗)2,hab =arctan(b∗∕a∗ ) TABLE2 Colorcoordinatesofthebeveragesmeasuredbythedifferenttechniques:spectroradiometer(SPE)andbydigitalimageanalysis (DIA) Sample Spectrophotometry Digital image analysis L*C*hab L*C*ab hab S1* 85.75±0.02a21.53±0.02a75.01±0.03a90.91±0.12a22.05±0.20a77.90±0.40a S2 54.67±0.09b9.73±0.04b99.56±0.08b73.31±0.13b21.97±0.20a87.91±0.18b S3* 81.19±0.01c27.60±0.03c74.44±0.03c86.67±0.07c27.69±0.11b75.38±0.09c S4* 83.06±0.02d28.64±0.08c80.34±0.03d89.57±0.09d30.55±0.12c81.80±0.22d S5 55.64±0.12e21.69±0.83a86.48±0.47e72.31±0.05e30.66±0.31c78.59±0.13ae S6 49.03±0.08f14.63±0.06d94.14±0.08f70.40±0.65f28.80±1.09b86.83±0.41f S7 60.08±0.07g25.90±0.09e88.67±0.14g68.79±0.11g33.10±0.14d84.10±0.06g S8 50.86±0.02h51.84±0.04f62.08±0.09h61.75±0.22h50.48±0.19e54.24±0.07h S9 51.73±0.18i43.50±0.10g68.40±0.03i67.53±0.02ik 51.40±0.07e62.47±0.12i S10 58.25±0.02j26.66±0.12ce 83.12±0.04j76.15±0.00j42.93±0.00f79.21±0.00e S11 53.45±0.14k37.40±0.15h64.31±0.05k66.95±0.00i45.86±0.00g59.75±0.00j S12 53.04±0.19l36.36±0.44h61.95±0.08h67.76±0.08kl 45.18±0.08g57.40±0.15k S13 48.69±0.06f43.91±0.12g61.79±0.06h62.01±0.05h51.54±0.08e52.59±0.15l S14 49.64±0.01m39.93±0.11i63.10±0.06l64.97±0.22m47.58±0.22h56.20±0.14m S15 57.28±0.04n53.97±0.09j81.28±0.05m65.71±0.4m57.87±0.14i81.69±0.12d S16 58.30±0.00j40.03±0.04i76.09±0.03n68.49±0.03gl 57.52±0.08i70.63±0.06n a–oDifferentsuperscriptswithincolumnsforeachparameterindicatestatisticallysignificantdifferences(p < 0.05). *Samplesstoredunderrefrigeration. | 31 FERNÁNDEZVÁZQUEZ Et al. 3 | RESULTS AND DISCUSSION 3.1 | Instrumental color measurements Table2showsthecolorcoordinatesofthebeveragesmeasuredby thetwotechniques.Lightness(L*)valuesrangedfrom61.75to90.91 inDIAandfrom48.69to85.75inSPE,thussamplesseemedlighter whenSPEwasconsidered.Chroma(C*ab)valuesrangedfrom21.55 to57.52inDIAbutfrom9.73to53.97inSPE,showingagainlower values,andwiderrangewhenSPEisconsidered.Finally,hue(hab)valuesrangedfrom52.58to87.91inDIAandfrom61.79to99.56in SPE,whichmeansthat,whenDIAisconsidered,samplesseemedmore reddish. Zulueta, Esteve, and Frígola (2007) reported similar values, rangingfrom53.05to80.62;17.67to63.59;and35.47to115.57;forL*, C*ab and hab,respectively,indifferentcommercialMFJBanalyzedwith a Hunter Labscan II Colorimeter (Hunter Associates Laboratory, Inc., Reston,VA,USA). Thesewiderangesinthevaluesshowthebroadvarietyofcolorin thebeveragesavailableinthemarket.Thus,ANOVA(Table2)showed thatmostofthesamplesweresignificantlydifferentinallthecolorimetricparameters. Thefruit/milkcompositionofthebeveragesisrelatedtothecolor andthustothecolorimetricparametersandcouldexplaincolordifferencesamongsamples.Inthissense,samplesstoredunderrefrigeration(S1,S3,andS4),whichwerethesampleswiththehighestvalues oflightness,hadasmainingredientmilkanditsproportionoffruit wasminor(5%–6%ofconcentratedfruit).Thiscouldexplainthehigh valuesoflightnessinthosesamples. However, generally, beverages which contain mayor proportion of fruits with reddish-yellowish colorations, such as orange, peach, mango, or carrot, presented hue values minor of 70°, while fruits with yellowish-greenish colorations, such as apple,pineapple,orlemon,areassociatedwithvaluesofhueabove 90°. Regardingchrome,nosignificantrelationbetweenthecompositionandthecolorimetricparameterscouldbeargued. 3.1.1 | Relationship between instrumental techniques Figure1 shows the samples in the CIELAB space (Diagram a*b*) measuredbySPE(Figure1a)andDIA(Figure1b).Itcanbeobserved thatsamplesweresituatedinthesameareaoftheCIELabdiagram, independentlyofthetechniqueusedalthoughwithslightmodificationsinthepositions.Theseslightvariationsobservedinthecolor coordinatesvaluesmeasuredbydifferenttechniqueswerealsofound inapreviousstudy(Fernández-Vázquezetal.,2011),wherecolorof differentorangejuiceswasmeasuredbyDIAandSPE. Instrumentalmeasurementsareconsideredanaccurateandsuitablemethodforevaluatingcolorinfood,asithasalreadybeenreported for wine and orange juice (Fernández-Vázquez etal., 2011; Martínez,Melgosa,Pérez,Hita,&Negueruela,2001).However,there might be differences among the measurements, depending on the technique used. They may be related mainly to differences in the thicknessofthemeasuredsample,othersecondaryfactorssuchas illuminationconditions,orthegeometryofthesystem.Inthissense,it isinterestingtocheckifallthetechniquesaresignificantlycorrelated. Inthiscase,allthecorrelationcoefficientswerehigh(>0.85)and statisticallysignificant(α=0.05).Thus,itconfirmedtherelationship betweenbothtechniquesinthe measurementsofthe colorofthis kindofbeverages. 3.2 | Consumer study Results from the demographic and consumption questionnaire are showninTable3.Resultsindicatethat85%ofconsumersconsidered theseproducts as beneficial,and 70% consumedthem quite often (39%didsotwoormoretimesperweek).Participantswhodidnot usetoconsumethesebeveragesstatedthattheydidnotlikethem (63%),ortheypreferredotherbeverages,suchasnaturaljuices(20%). Figure2 shows mean values given by consumers to the beverageappearances.Itisworthytohighlightthatthepunctuations suggestedthatcoloroftheseproductswasnothighlyevaluated,as thebestmeanpunctuationwas6.6.Anyway,generally,participants FIGURE1 a*b*colordiagramfor juice–milkbeveragesmeasuredby spectrophotometer(a)anddigitalimage analysis(b) 10 20 30 40 50 60 0º 10º 20º 30º 40º 50º 60º 70º 80º 90º a* 0102030405060 0 a* 0102030405060 0 10 20 30 40 50 60 0º 10º 20º 30º 40º 50º 60º 70º 80º 90º (a) (b) 32 | FERNÁNDEZVÁZQUEZ Et al. significantlypreferredsampleswithorangishandmorevividappearanceinsteadofthosewithmorewhitishlook(Table4).Thiscould beduetothefactthatconsumersexpectedtofindcolorfulbeveragesastheyaresupposedtobemadewithfruits,andwhenthey observedbeverageswithwhitishcolor,theymayassociateitwith alowerfruitcontent.Thisapparentlylowerpreferencecouldalso beduetothelowmeanconsumptiononthegroup.However,the consumerssegmentwith highest rateofdislikers (Cluster1, 32% dislikers,seeTable3)arethebestdiscriminatorsbetweensamples intermsofcolorappearance,reachingthehighestmeanscoresin somesamples. Relationbetweenconsumeracceptanceandcolorparameterswas explored, andtheyshowedthathue andchromaweresignificantly (p<0.05)correlatedwithconsumeracceptance,withthehighestcorrelation coefficientfor consumeracceptance and huemeasuredby DIA(r=0.97). 3.2.1 | Cluster analysis Tofindoutifthereweregroupsofconsumersdifferingintheirpreferencesforcolor,asegmentationofthepanelgroupwasdoneby Cluster analysis (Vigneau, Qannari, Punter, & Knoops, 2007). The Frequency response (%) p- Value (χ2)a Cluster 1, n = 19 Cluster 2, n = 29 Cluster 3, n = 52 Gender 0.107 Male 11 38 33 Female 89 62 67 Consumption 0.202 Yes 63 83 67 Frequency 0.300 <Onceaweek 42 28 29 Onceaweek – 14 14 Twiceaweek 5 10 16 >Twiceaweek 16 31 8 No 37 17 35 Reasons – Highprice – 3 – Itisnotbeneficial – – 1 Don’tlike 32 3 22 Other 5 11 12 Opinionaboutiftheseproductsarebeneficial – Yes 84 93 80 No 16 7 20 aWhen suitable, χ2 test was applied to test independence between clusters and demographic characteristics. TABLE3 Demographiccharacteristics andconsumptionhabitsforeachcluster and χ2testresults FIGURE2 Meanvaluesappearance scoresgivenbyconsumersandsignificant differencesidentifiedbyKruskal–Wallis test | 33 FERNÁNDEZVÁZQUEZ Et al. threegroupsofconsumersclearlyidentified,andtheirdemography andconsumptionhabitsareshowninTable3.Resultsfromχ2 test showed that there were no significant differences in demographic characteristics or consumption frequency among clusters, indicatingthatthesevariablesdidnotinfluencecoloracceptancepatterns. Previousstudiesonconsumer’scoloracceptanceofdifferentproducts like strawberry nectar from puree and orange juices showed similarresults,whereneithergendernorageorconsumptionhabits hadsignificantimpactoncoloracceptance(Fernández-Vázquezetal., 2011;Gossingeretal.,2009). Mean appearance scores given by the clusters are shown in Figure3.Thefirstcluster(19%)showedaclearpreferenceforthose sampleswithorangishappearance,givingaveragepunctuationseven higherthan7(“Likemoderately”)andalsoadeeplydislikingforthose with a whiter look, with punctuations lower than 2 (“Dislike very much”). TABLE4 Meanacceptancescoresandappearanceofthesamples(picturestakenwithDigiEyeSystem) Samples Acceptance scores Samples Acceptance scores Samples Acceptance scores Samples Acceptance scores S1 3.5 S2 3.7 S3 3.9 S4 4.0 S5 4.2 S6 4.5 S7 4.7 S8 4.9 S9 5.3 S10 5.4 S11 6.1 S12 6.2 S13 6.3 S14 6.4 S15 6.4 S16 6.6 FIGURE3 Meanappearancescoresgivenbyeachoftheclusters 34 | FERNÁNDEZVÁZQUEZ Et al. However,cluster2(29%)didnotdifferentiatesomuchthesamples.Thus,theworstvaluatedbythisgroupofconsumerswasSample 2(averageof4.7)andthebestvaluatedwasSample14(averageof 7.6),whichimpliedaminorrangethantheothertwosegments. Finally,thethirdcluster(52%)wasthemostnumerous.Theseconsumersgavepunctuationslowerforallthesamples,andonlySample 16,withanaverageof6.2,wasvaluatedabove6(“Likeslightly”). These observations give additional information to the general resultsdiscussedearlier.Forinstance,itseemsthatwhileforsome consumers(Cluster1),colorofthesamplesisimportant,sincethey are capable of differentiating samples in terms of acceptance, just bylookingatthem;otherconsumers(Cluster2)didnotfounddifferencesamongthe beverageswhen theyonlyevaluated theirappearances.Moreover,fromthisanalysisitcanbeobservedthatfor agroupofconsumers(Cluster3),noneofthecolorsofthesamples wereappreciated. 3.2.2 | Internal preference mapping Internalpreferencemappingreferstotheanalysisofpreferencedata only,anditwasconductedtovisualizethegeneralbehaviorofthe clusters of consumers (Figure4). Two preference dimensions accountedfor95.18%ofthetotalvariance,sothethirdpreferencedimension was not considered. Alltheclustersappearrepresentedinthepositivevaluesofthe firstdimension.Cluster2issituatedinthehigherhalfofthesecond dimension,clearlyseparatedfromtheothersclustersandoppositeto samplesS2,S6,S7,andS8,indicatingarelativelylowerpreferencefor them(asconfirmedinFigure3).However,clusters1and3arelocated inthelowerhalfoftheseconddimension,oppositetosamplesS1,S3, S4,andS5,indicatingarelativelylowerpreferenceforthosesamples inbothclusters.Inaddition,cluster1ismuchclosertosamples12,13, and14,indicatingahigheracceptanceofthosesamplesascompared withcluster3(Figure3). Furthermore,samplesaredividedbythe first dimensionintwo well-defined groups. In the part of negative values, samples with slightlywhite and gray appearancearelocatedclearlyfar fromthe threeclusters.However,alltheorangishandreddishsamples(with theexceptionofonesample,S8)appearedinthepositivevaluesofthe firstdimension.Thisfactshowedgraphicallytheconsumertendency togivehigherpunctuationstothesampleswithmorevividcolors. 4 | CONCLUSIONS Inthisstudy,colorofcommercialMFJBwasmeasuredbytwodifferenttechniquesconfirmingthatboth,SPEandDIA,areappropriateto analyzetheappearanceoftheseproducts.However,consumerstudy showedthatthoughgenerallyparticipantssignificantlypreferredsampleswithorangishappearanceinsteadofthosewithawhiterlook,the lowpunctuationsgiventothesamples(mean=5.1)suggestedthat coloroftheseproductswasnothighlyevaluated.Thisfactshouldbe takenintoaccountbyindustriesastheappearanceoffoodproducts hasademonstratedinfluenceinfoodacceptance. ACKNOWLEDGMENTS ThisworkwassupportedbyfundingfromtheConsejeríadeInnovación CienciayEmpresa,JuntadeAndalucíabytheprojectP11-AGR-7783. CONFLICT OF INTEREST Nonedeclared. REFERENCES Baker,R.,&Günter,C.(2004).Theroleofcarotenoidsinconsumerchoice andthelikelybenefitsfromtheirinclusionintoproductsforhuman consumption.Trends in Food Science & Technology,15,484–488. FIGURE4 Internalpreferencemapping definedbythetwofirstpreference dimensionsfromtheconsumers’cluster | 35 FERNÁNDEZVÁZQUEZ Et al. Calvo,C.,Salvador,A.,&Fiszman,S.(2001).Influenceofcolourintensity ontheperceptionofcolourandsweetnessinvariousfruit-flavoured yoghurts.European Food Research and Technology,213,99–103. CIE. (1978). Recommendations on Uniform Color Spaces, Color- Difference Equations, Psychometric Color Terms, CIE Publication No. 15 (E-1.3.1) 1971, Supplement 2. Bureau Central de la CIE, Vienna. CIE.(1991a).CIEstandardcolorimetricobservers.ISO/CIE10527:1991(E). CIE.(1991b).CIEstandardscolorimetricilluminants.ISO/CIE10526:1991(E). Fernández-Vázquez,R.,Stinco,C.M.,Hernanz,D.,Heredia,F.J.,&Vicario, I.M.(2013).Colourtrainingandcolourdifferencesthresholdsinorangejuice.Food Quality and Preference,30,320–327. Fernández-Vázquez,R.,Stinco,C.M.,Hernanz,D.,Heredia,F.J.,&Vicario, I. M. (2014). Digital image analysis and visual evaluation of orange juice:influenceofdifferentmeasurements’conditions.Food Analytical Methods,7,157–164. Fernández-Vázquez,R.,Stinco,C.M.,Melendez-Martinez,A.J.,Heredia,F. J.,&Vicario,I.M.(2011).Visualandinstrumentalevaluationoforange juicecolor:Aconsumers’preferencestudy.Journal of Sensory Studies, 26,436–444. Gossinger,M.,Mayer,F.,Radochan,N.,Höfler,M.,Boner,A.,Grolle,E.,… Berchofer,E.(2009).Consumer’scoloracceptanceofstrawberrynectarsfrompuree.Journal of Sensory Studies,24,78–92. Heredia, F. J., González-Miret, M. L., Álvarez, C., & Ramírez, A. (2006). DigiFood®[Nº.SE-01298]. Luo, M. R., CUI, C. G., & LI, C. (2001). British Patent (Application Nº 0124683.4). Entitled apparatus and method for measuring colour (DigiEyeSystem).DerbyUniversityEnterprisesLimited. Martínez,J.A., Melgosa, M., Pérez, M. M., Hita, E., &Negueruela,A. I. (2001).Notevisualandinstrumentalcolorevaluationinredwines.Food Science and Technology International,7,439–444. Meléndez-Martínez,A.J.,Vicario,I.M.,&Heredia,F.J.(2006).Influenceof whitereferencemeasurementandbackgroundonthecolorspecificationoforangejuicesbymeansofdiffusereflectancespectrophotometry.Journal of AOAC International,89,452–457. Pszczola, D. E. (2005). Ingredients: Making fortification functional. Food Technology,59,44–61. Quitao-Teixeira, L., Aguiló-Aguayo, I., Ramos, A., & Martín-Belloso, O. (2008). Inactivation of oxidative enzymes by high-intensity pulsed electricfieldforretentionofcolorincarrotjuice.Food and Bioprocess Technology,1,364–373. Sampedro,F.,Geveke,D.J.,Fan,X.,Rodrigo,D.,&Zhang,Q.H.(2009). Shelf-lifestudyofanorangejuice-milkbasedbeverageafterPEFand thermalprocessing.Journal of Food Science,74,S107–S112. STATSOFT,I.(2007).STATISTICA,version8.0. Stinco,C.M.,Fernámdez-Vázquez,R.,Heredia,F.J.,Meléndez-Martínez,A. J.,&Vicario,I.M.(2014).Spectroradiometryvs.digitalimageanalysis incolourmeasurementinjuicesfromdifferentorangeandmandarin varieties. OPA,47,139–144. Stinco,C.M.,Fernández-Vázquez,R.,Meléndez-Martínez,A.J.,Heredia,F. J.,Bejines-Mejías,E.,&Vicario,I.M.(2012).Influenceofdifferentbackgroundsontheinstrumentalcolorspecificationoforangejuices.InJ. Caivano&M.Buera(Eds.),Color in food. Technologigal and psychophysical aspects(pp.143–150).NewYork:CRCPress.Taylor&FrancisGroup. Vigneau,E.,Qannari,E.M.,Punter,P.H.,&Knoops,S.(2007).Segmentation ofapanelofconsumersusingclusteringofvariablesaroundlatentdirectionsofpreference.Food Quality and Preference,12,359–363. Zulueta,A.,Esteve,M.J.,&Frígola,A.(2007).Carotenoidsandcolorof fruit juice and milk beverage mixtures. Journal of Food Science, 72, C457–C463. How to cite this article:Fernández-VázquezR,StincoCM, HernanzVilaD,HerediaFJ,ChayaC,VicarioIM.Internal preferencemappingofmilk–fruitbeverages:Influenceof colorandappearanceonitsacceptability.Food Sci Nutr. 2018;6:27–35. https://doi.org/10.1002/fsn3.494