Sie sind auf Seite 1von 0

Exploringfastfoodconsumptionbehaviours

andsocialinfluence

SubmittedinfullrequirementforthedegreeofDoctorofPhilosophy

EmilyBrindal
B.Psych(Honours)

FacultyofHealthSciences,theUniversityofAdelaide,SouthAustralia
SchoolofPsychology;SchoolofMedicine

NOBLEResearchGroup;CSIROHumanNutrition,Adelaide,SouthAustralia

April2010


i


Contents

ListofTables_______________________________________________________________________________________v
ListofFigures______________________________________________________________________________________vi
OverviewofChapters____________________________________________________________________________vii
Summary_________________________________________________________________________________________viii
Declaration__________________________________________________________________________________________x
Acknowledgements_______________________________________________________________________________xi

1 ChapterOne:Obesity,changingeatingpatternsandfastfoods_____________________________ 1
1.1 Obesity____________________________________________________________________________________________1
1.1.1 Globalincreaseinobesityrates_______________________________________________________________2
1.1.2 Directandindirectconsequencesofobesity_________________________________________________4
1.1.3 Energyequilibrium:Anexplanationforobesity_____________________________________________5
1.2 Eatingpatternsandchangingdiets_____________________________________________________________7
1.2.1 Theincreaseintheconsumptionofmealspreparedawayfromthehomeand
nutritionalimplications ______________________________________________________________________________7
1.3 Understandingfastfoodconsumption_________________________________________________________9
1.3.1 Definingfastfoods _____________________________________________________________________________9
1.3.2 FastfoodconsumptioninAustralia_________________________________________________________13
1.3.3 Thefastfoodconsumer______________________________________________________________________16
1.3.4 Whydopeopleconsumefastfood?_________________________________________________________17
1.3.5 Underlyingreasonsfortheshifttowardconveniencefoods ____________________________19
1.4 Fastfoodandobesity__________________________________________________________________________21
1.4.1 Fastfoodconsumption,weightgainandobesity:Areviewofexistingstudies_________22
1.4.2 Energydensity:Howfastfoodmaypromoteoverconsumption_________________________28
1.4.3 Mealdeals:Howfastfoodenvironmentsmayencourageconsumption_________________30
1.5 Summaryofliterature_________________________________________________________________________31
1.6 Objectivesofthecurrentresearch____________________________________________________________32
2 ChapterTwo:AstudyofthecurrentmacronutrientprofileoffastfoodinAustralia __34
2.1 Methods ________________________________________________________________________________________35
2.1.1 Theconstructionoftraditionalfastfoodmeals __________________________________________36
2.1.2 Healthieralternatives________________________________________________________________________37
2.1.3 Nutritionalcalculations______________________________________________________________________37
2.2 Results__________________________________________________________________________________________38
2.3 Discussion______________________________________________________________________________________42
2.3.1 Summaryofanalysis_________________________________________________________________________45
3 ChapterThree:SocialinfluenceandfoodintakeAreviewoftheliterature ___________46
3.1 Modellingandsocialnorms___________________________________________________________________47
3.1.1 SocialModelling______________________________________________________________________________47
3.1.2 SocialNorms__________________________________________________________________________________51
3.1.3 Normativeovereating________________________________________________________________________56
3.2 SocialFacilitation______________________________________________________________________________57
3.2.1 Socialfacilitationandeatingbehaviours___________________________________________________58
3.2.2 Thetimeextensionhypothesis _____________________________________________________________60
3.2.3 Beyondduration _____________________________________________________________________________64
3.3 Relationships:thepotentialmoderatorofsocialinfluence_________________________________67
3.3.1 Sourcesofinfluence__________________________________________________________________________68


ii


3.3.2 Researchonrelationshipsandsocialinfluence____________________________________________70
3.4 SocialInfluenceandfastfood_________________________________________________________________73
3.4.1 EnvironmentalPressure_____________________________________________________________________73
3.4.2 Normativeinfluenceandsocialfacilitation________________________________________________74
3.4.3 Thewhoofsocialinfluence ________________________________________________________________75
3.4.4 Fastfoodbehaviours_________________________________________________________________________76
3.5 Aimsofsubsequentstudies___________________________________________________________________77
4 ChapterFour:DevelopmentoftheFastFoodSurveyandPreliminaryExaminationof
SocialInfluencesonFastFoodConsumption_____________________________________________________78
4.1 DevelopmentoftheFastFoodSurvey(FFS)_________________________________________________79
4.1.1 TheFastFoodChoicesProgram(FFCP)____________________________________________________81
4.1.2 MeasuringCorrelatesofConsumption_____________________________________________________83
4.2 SurveyingattheRoyalAdelaideShow_______________________________________________________86
4.2.1 Procedure_____________________________________________________________________________________86
4.2.2 Assessmentofoutliers_______________________________________________________________________88
4.2.3 Adjustmentstoorders_______________________________________________________________________88
4.2.4 Recoding______________________________________________________________________________________89
4.3 Results__________________________________________________________________________________________90
4.3.1 Usability_______________________________________________________________________________________90
4.3.2 Finalsample__________________________________________________________________________________90
4.3.3 Therelationshipbetweendemographicfactorsandfastfoodconsumption___________ 92
4.3.4 Fastfoodbehaviours_________________________________________________________________________94
4.3.5 Fastfoodseatenandthepotentialforoverconsumption_________________________________99
4.3.6 Theinfluenceofmicroenvironmentalpressuresonfastfoodconsumption __________102
4.3.7 Socialinfluencesandfastfoodconsumption _____________________________________________104
4.4 Discussion_____________________________________________________________________________________108
4.4.1 TheFastFoodSurvey:Strengthsandlimitations_________________________________________109
4.4.2 Demographiccharacteristicsandfastfoodconsumption________________________________110
4.4.3 Behaviours,motivatorsandcontextsurroundingfastfoodconsumption______________112
4.4.4 Fastfood:Potentialforoverconsumption_________________________________________________113
4.4.5 Socialinfluencesandfastfoodconsumption _____________________________________________114
4.4.6 ConclusionsfromtheinitialFFS___________________________________________________________117
5 ChapterFive:SocialinfluencesonamountoffastfoodconsumedinonevisitTestinga
structuralmodel ____________________________________________________________________________________119
5.1 Testingexistingsocialinfluencemodelsinfastfood_______________________________________119
5.1.1 ExtendingFeunekesetal.smodelofsocialfacilitationtofastfoodconsumption_____121
5.1.2 Thegoalofthecurrentstudy_______________________________________________________________123
5.1.3 Usingsurveymethodstoexploresocialinfluence________________________________________124
5.2 Procedure _____________________________________________________________________________________126
5.2.1 AbriefintroductiontoInternetmethods _________________________________________________126
5.2.2 PilottestingoftheFastFoodSurvey(FFS)deliveredontheInternet__________________127
5.2.3 Recruitmentmethodsandoutcomes______________________________________________________128
5.3 Approachtoanalysis _________________________________________________________________________129
5.3.1 Descriptivedata_____________________________________________________________________________129
5.3.2 Methodsformodelling______________________________________________________________________130
5.3.3 Limitingthereverserelationshipbetweentimespenteatingandtheamounteaten_131
5.4 Dataadjustmentsandrecoding______________________________________________________________133
5.4.1 Assessmentofoutliers______________________________________________________________________133
5.4.2 Adjustmentstoreportedfastfoodordersandvariablerecoding_______________________133


iii


5.4.3 Codingfreeresponseitems________________________________________________________________135
5.5 Resultsfromdescriptivedata________________________________________________________________137
5.5.1 Demographiccharacteristicsofthefinalsample_________________________________________137
5.5.2 Reportedfastfoodbehaviours_____________________________________________________________138
5.5.3 Relationshipsbetweensocialvariables___________________________________________________139
5.5.4 Conclusionsfromdescriptiveanalyses____________________________________________________140
5.6 Modellingincreasedenergyintakefromfastfooditems__________________________________142
5.6.1 Confirmingthemodelofsocialfacilitationinfastfoodintake __________________________142
5.6.2 Conclusionsfrompreliminarymodelling_________________________________________________146
5.6.3 Theeffectofwhoispresentinthemodelofsocialfacilitation_________________________149
5.6.4 Personalrelationshipsandsocialfacilitationconclusions_____________________________152
5.6.5 Expandingthemodelofsocialfacilitationforfastfoodintake__________________________154
5.6.6 Conclusionsfrommodellingfastfoodintakeinmalesandfemales_____________________159
5.6.7 Alteredintakeinthosepeopleeatinginthepresenceofothers_________________________160
5.6.8 Conclusionsfromtheanalysisofthosewhoatewithothers____________________________162
5.7 ConclusionsfromtheFFSadministeredthroughtheInternet____________________________163
5.7.1 Themodelofsocialfacilitationforfastfoodconsumption ______________________________163
5.7.2 Theindirecteffectsofthepresenceofothersonintakeatasinglefastfoodeating
occasion:Mealdurationandeatingatmosphere________________________________________________164
5.7.3 Howwhoispresentinfluencesintakewhenconsumingfastfood______________________166
5.7.4 Microenvironmentalinfluencesonfastfoodconsumption_____________________________167
5.7.5 Sexdifferences,womenandsocialinfluenceduringfastfoodconsumption___________168
5.7.6 Limitationsofthecurrentstudy___________________________________________________________169
5.7.7 Directionsforfutureresearch______________________________________________________________171
6 ChapterSix:Fastfoodconsumption,minimaleatingandimpressionmanagement _174
6.1 Womenandeatinglightly:Areviewoftheliterature______________________________________175
6.1.1 Impressionmanagementandsexroles:Mechanismsbehindminimaleatingnorms _180
6.1.2 Fastfoodandimpressionmanagement___________________________________________________182
6.1.3 Aimsofthecurrentstudy___________________________________________________________________ 183
6.2 Method_________________________________________________________________________________________186
6.2.1 Development ________________________________________________________________________________186
6.2.2 Piloting_______________________________________________________________________________________187
6.2.3 Procedure____________________________________________________________________________________189
6.2.4 Calculationofenergyintake________________________________________________________________191
6.2.5 Datascreening_______________________________________________________________________________192
6.3 Preliminarystatistics_________________________________________________________________________193
6.3.1 Descriptivedata_____________________________________________________________________________193
6.3.2 Covariates____________________________________________________________________________________194
6.4 Hypothesistesting____________________________________________________________________________196
6.4.1 Analysisoftheeffectsofthepresenceofothers__________________________________________196
6.4.2 Lonedinersandreading____________________________________________________________________197
6.4.3 Intragroupmatchingofintake ____________________________________________________________198
6.4.4 Minimaleatingnorms ______________________________________________________________________200
6.5 Discussion_____________________________________________________________________________________202
6.5.1 Timeextensioninafastfoodrestaurant__________________________________________________203
6.5.2 Matchingnormsandtheirinfluenceonfastfoodconsumption_________________________203
6.5.3 Evidenceofminimaleatinginfastfoodconsumption ___________________________________205
6.5.4 Lonedinersandreading____________________________________________________________________207
6.5.5 Conclusions__________________________________________________________________________________208


iv


6.6 ACrossCulturalComparisonofFastFoodBehavioursinaNaturalisticEnvironment_211
6.6.1 AssessingeatingbehavioursinaNorwegiansample ____________________________________214
6.6.2 Results _______________________________________________________________________________________215
6.6.3 DiningatMcDonaldsinNorwaycomparedtoAustralia_________________________________216
6.6.4 Discussionandconclusions ________________________________________________________________217
7 ChapterSeven:Asummaryofthefindingsandimplicationsofthecurrentresearch 221
7.1 Overviewofcurrentresearch________________________________________________________________222
7.2 Socialinfluenceandfastfoodconsumptionbehaviours___________________________________224
7.2.1 Socialfacilitationandthetimeextensionhypothesis____________________________________224
7.2.2 Minimaleatingnormsandgender_________________________________________________________226
7.2.3 Wheredonormsstartandfinish? _________________________________________________________229
7.2.4 Thewhofactorofsocialinfluence________________________________________________________230
7.2.5 Dosocialinfluencesaffectfastfoodconsumption?_______________________________________232
7.3 Unravellingfastfoodconsumption__________________________________________________________233
7.3.1 Theculturesurroundingfastfoodconsumption _________________________________________234
7.3.2 Fastfoodconsumptionbehavioursandthepotentialforweightgain__________________236
7.3.3 Integratingknowledgeoffastfoodbehavioursandsocialinfluenceforhealth
promotion __________________________________________________________________________________________239
7.4 Conclusion_____________________________________________________________________________________240
References ___________________________________________________________________________________________242

Appendix1:PaperonmacronutrientprofileoffastfoodsinAustralia________________________I
Appendix2:NutritionalinformationfollowingFastFoodSurvey___________________________ VIII
Appendix3:FastFoodSurveyQuestions+InformationSheet _________________________________ X
Appendix4:ExamplescreenshotsofFastFoodSurvey(FFS)_______________________________XVIII
Appendix5:Completedpilotingform __________________________________________________________XXIII
Appendix6:RequestforparticipationinFastFoodSurveytoNWAHSmembers _______XXVI
Appendix7:SummaryofExplorationofPortionSizeItems _______________________________XXVII
Appendix8:Datacollectioninstrumentfortheobservationalstudy_______________________XLV

ListofTables

Table1:Exampledefinitionsusedinpreviousstudiesaboutfastfoods._________________________11
Table2:Descriptionofthedemographiccharacteristicsofthesample

(n=116)._________________91
Table3:Frequencyoffastfoodconsumptioninpilotsample(n=116)

.___________________________92
Table4:Totalvarianceexplained(AdjustedR
2
),betavaluesandsignificancelevelsforamodel
predictingtheamountofenergyconsumedatafastfoodeatingoccasionusingdemographic
factors(n=116)._________________________________________________________________________________________93
Table5:Frequencyofmealtypes,locationandtimeofconsumptionfortheRASsample(n=116).
___________________________________________________________________________________________________________95
Table6:Codedcategoriesandassociatedfrequenciesforbothopenresponseitems.___________99
Table7:Energy(KJ)andfatcontentoffastfoodconsumedbypilotparticipants

._____________ 100
Table8:Meanresponsetoitemsmeasuringsocialinfluenceoneating._________________________ 104
Table9:Correlationsbetweenenergyconsumedandratingsoftheeatingenvironmentfor
wholesample(n=116)._______________________________________________________________________________ 105
Table10:Modelsforsocialvariablespredictingenergyintakeatafastfoodeatingoccasion.
Totalvariance(AdjustedR
2
),betasandalphasarepresentedforeachpredictor(n=91). _____ 108
Table11:DescriptionofthedemographiccharacteristicsoftheInternetsample(n=407). ___ 137
Table12:FrequencyoffastfoodbehavioursreportedbytheInternetsample(n=407)._______ 138
Table13:Correlationvalues(Pearsonsr)forthesocialfacilitationmeasures

(n=407)._______ 140
Table14:Correlationvalues(Pearsonsr)forthesocialfacilitationmeasures.Valuesforfemales

(n=272)arepresentedabovethediagonal,valuesformales

(n=135)arebelowthediagonal. 143
Table15:Totalvarianceexplained(AdjustedR
2
)andbetavaluesforthedependentvariables
energyintake(Energy),timespenteating(Time)andpleasantnessofatmosphere(Atmos)for
bothmaleandfemalerespondentsintheInternetsample._______________________________________ 144
Table16:Totalvarianceexplained(AdjustedR
2
)andbetavaluesforthedependentvariables
energyintake(Energy),timespenteating(Time)andpleasantnessofatmosphere(Atmos)for
bothmaleandfemalerespondentsintheInternetsampleinthemodelincludingseparated
interpersonalrelationships._________________________________________________________________________ 150
Table17:Totalvarianceexplained(AdjustedR
2
)andbetavaluesforthedependentvariables
energyintake(Energy),timespenteating(Time)andpleasantnessofatmosphere(Atmos)for
bothmaleandfemalerespondentsintheInternetsamplefortheexpandedmodel.___________ 155
Table18:Predictorsofenergyintake(withassociatedBetavalues)infemaleparticipantswho
reportedeatingwithatleastoneotherperson(n=204).__________________________________________ 161
Table19:Descriptionofthesampleasfrequencydata(n=298)._________________________________ 193
Table20:Comparisonofenergyconsumed(KJ)andtimespenteatingbetweendifferentlevelsof
thevariablesmeasured
*
._____________________________________________________________________________ 194
Table21:Frequencyofmatchingordersforgroupsofdifferentsizesanddifferentsex
composition. __________________________________________________________________________________________ 198
Table22:MeansandstandarddeviationsforKJconsumptionfromfastfooditemsforbothtwo
way(sexbysexcomposition)andthreewayinteractioneffects(sexbysexcompositionby
groupsize)
*
.___________________________________________________________________________________________ 201
Table23:DemographiccharacteristicsofthesamplesofconsumersobservedinOslo(n=48)and
Adelaide(n=298). ____________________________________________________________________________________ 215


vi

ListofFigures

Figure1:PercentageofAustraliansmakingfastfoodpurchasesinlastfourweeks*. ____________13
Figure2:McDonalds,HungryJacks,PizzaHutandKFCrestaurantsintheAdelaidemetropolitan
Area.Redcirclesindicatemultiplerestaurantswithina1kilometreradius.______________________15
Figure3:Thepercentageofdailyguidelinesofmacronutrientsaccountedforbytraditionalfast
foodmeals.______________________________________________________________________________________________39
Figure4:Thepercentageofdailyguidelinesofmacronutrientsaccountedforbyhealthierfast
foodmeals.______________________________________________________________________________________________40
Figure5:Feunekesetal.s(1995)testedmodelofsocialfacilitation.NB.Numbersnexttoarrows
representpath/betavalues____________________________________________________________________________61
Figure6:Meantimespenteatingfordifferentgroupsizesatthereportedfastfoodeating
occasion(n=116)._____________________________________________________________________________________ 106
Figure7:BoxplotsfortimespenteatingfromtheInternetsample(InternetSs;inblue)andthe
fairsample(RASSs;ingreen)._______________________________________________________________________ 126
Figure8:Firsttimevisitorstowww.fastfoodstudy.com.auwithrecruitmentdatesandassociated
recruitmentmethodslabelled. ______________________________________________________________________ 129
Figure9:Activitiesprecedingconsumptionaccordingtotheresponsecategoriesfromthe
Internet(blue)andfair(green)samples.___________________________________________________________ 136
Figure10:ReasonsforfastfoodconsumptionfromtheInternetsample(blue)andthefair
sample(green)._______________________________________________________________________________________ 136
Figure11:Significantpathsfromgroupsizetoenergyintakefromfastfooditemsinmales
(n=131)._______________________________________________________________________________________________ 145
Figure12:Significantpathsfromgroupsizetoenergyintakefromfastfooditemsinfemales
(n=266)._______________________________________________________________________________________________ 145
Figure13:Significantpathsfrominterpersonalrelationshipstoenergyintakefromfastfood
itemsinmales(n=131).______________________________________________________________________________ 150
Figure14:Significantpathsfrominterpersonalrelationshipstoenergyintakefromfastfood
itemsinfemales(n=266).____________________________________________________________________________ 151
Figure15:Significantpathsforinterpersonalrelationships,timespenteatingandenergyintake
fromfastfooditemsinmales(n=131)._____________________________________________________________ 156
Figure16:Significantpathsforinterpersonalrelationships,timespenteatingandenergyintake
fromfastfooditemsinfemales(n=263).___________________________________________________________ 156
Figure17:Interactioneffectsforsubjectsex,groupcompositionandgroupsizeonKJintake
fromfastfoods. _______________________________________________________________________________________ 202



vii


OverviewofChapters

ChapterOne:
Reviewstheliteratureonobesity,shiftingeatingpatterns,fastfoodconsumptionand
anyassociationsbetweenthem.

ChapterTwo:
Presentsnutritionalprofilescomparingmacronutrientsofselectedfastfoodmeals
availableinAustralia.

ChapterThree:
Reviewstheexistingliteratureonthesocialinfluencesoneatingbehaviours.

ChapterFour:
DescribesthedevelopmentoftheFastFoodSurvey(FFS)andaninitialstudyexploring
demographic,environmentalandsocialcorrelatesoffastfoodconsumption.

ChapterFive:
Presentsamodelofsocialfacilitationthatistestedandrefinedinthecontextoffastfood
consumptioninasecondFFSsample.

ChapterSix:
Exploreshowminimaleatingnormscouldalterfastfoodintakeandpresentsan
observationalstudyfollowedbyasmallercrossculturalstudy.

ChapterSeven:
Summarisesthefindingsandimplicationsofthecurrentresearch.


viii


Summary

Therehasbeenanincreaseintheconsumptionofconveniencestylefoods.Thischange
hasoccurredconcurrentlywithaglobalriseinobesityrateswhichhasledtosome
researchersblamingtheincreasedconsumptionofbigbrandfastfoods(suchas
McDonalds)forexpandingwaistlines.Nutritionalprofilingintheinitialstudyshowed
thattheenergyprovidedinatypicalfastfoodmealseemedappropriateintermsofa
generaldailyintakebutthatincreasedmealsizes,poororderingdecisionsandchoiceof
fastfoodrestaurantcouldinfluenceenergybalanceandlongtermhealthoutcomes.

Eventhoughfastfoodsareoccupyingalargerpartofthediet,limitedpreviousresearch
hasexploredhowsocialinfluences(includingmodelling,socialnormsandsocial
facilitation)mayincreasetheintakeoffastfoods.Thereforetheaimofthisdissertation
wastoexploreenvironmental,socialanddemographicinfluencesontheamountoffast
foodconsumedatasingleeatingoccasion.

TheFastFoodSurvey(FFS)wasdevelopedandadministeredtotwosamplestocollect
informationincludingtheitem/seatenandanysocial,environmentalordemographic
influencesthatsurroundedtheconsumptiononparticipantsmostrecentvisittoone
ofthelargefastfoodchainsinAustralia.Resultsfromaninitialsample(n=116)revealed
boththeeffectivenessoftheprogramthatdeliveredtheFFSandsupportforthe
hypothesisthatenvironmentalandsocialfactorscouldinfluencetheamountoffastfood
consumed.

AsecondstudyusingtheFFSaimedtotestanddevelopanexistingmodelofsocial
facilitation(originallydevelopedusinggeneraleatingbehaviours)inthespecific
contextoffastfoodconsumption.Accordingly,alargersample(n=407)wasrecruited


ix


viatheInternet.Followingpathanalysis,therewassupportforthetimeextension
hypothesisinthecurrentdata;eatingwithotherpeoplepredictedthetimespenteating
whichsubsequentlypredictedenergyconsumptionfromfastfooditems.Beyondthe
simpleeffectsoftimeextension,furthermodellingshowedthatenvironmentalfactors,
includingreasons,forconsumptioncouldbeassociatedwithincreasedfastfoodintake.

AnalysisofdatafromthesecondFFSstudyshowedthatmenandwomenwere
influenceddifferentlybytheireatingenvironmenttherewasanegativedirecteffectof
otherpeopleonwomensenergyintakefromfastfooditems.Thetheoryofminimal
eatingsuggeststhatgenderrolesmayaltereatingbehavioursandofferssome
explanationforthisresult.AnobservationalstudywasconductedinMcDonaldsto
assesshowthepresenceofmaleandfemalecompanycouldinfluencefastfoodintakein
bothsexes.Acomparisonofenergyintakebyparticipantsex,groupsizeandthesex
compositionofthegrouprevealedsupportforminimaleatingnormsmeneatingwith
othermenatethemostfood,whilewomeneatingwithmenatetheleastfood.

Overall,thechapterspresentedinthisdissertationshowthat,inafastfoodconsumption
context,multiplesocialinfluencesoccur.Therefore,despitechangesinthetypesoffood
beingconsumed,themechanismsalteringeatingbehavioursmayberelativelystable.
Giventhepotentialassociationbetweenweightgainandtheconsumptionoffastfoods,
understandingtheseinfluencesisafirststeptowardfutureintervention.


x

Declaration

Thisworkcontainsnomaterialwhichhasbeenacceptedfortheawardofanyotherdegreeordiplomain
anyuniversityorothertertiaryinstitutiontoEmilyBrindaland,tothebestofmyknowledgeandbelief,
containsnomaterialpreviouslypublishedorwrittenbyanotherperson,exceptwhereduereferencehas
beenmadeinthetext.

Igiveconsenttothiscopyofmythesis,whendepositedintheUniversityLibrary,beingmadeavailablefor
loanandphotocopying,subjecttotheprovisionsoftheCopyrightAct1968.

Theauthoracknowledgesthatcopyrightofpublishedworkscontainedwithinthisthesis[Brindal,E.,
Mohr,P.,Wilson,C.,&Wittert,G.(2008).Obesityandtheaffectsofchoiceatafastfoodrestaurant.Obesity
ResearchandClinicalPractice,2(2),111117]resideswiththecopyrightholder(s)ofthoseworks.

IalsogivepermissionforthedigitalversionofmythesistobemadeavailableontheWeb,viathe
Universitysdigitalresearchrepository,theLibrarycatalogue,theAustralianDigitalThesesProgramand
alsothroughWebsearchengines,unlesspermissionhasbeengrantedbytheUniversitytorestrictaccess
foraperiodoftime.

EmilyBrindal
April2010


xi

Acknowledgements


Lareconnaissanceestlammoireducoeur
Gratitudeisthememoryoftheheart
JeanBaptisteMassieu

ThankstoCarlene,PhilandGaryforsupportingmymind;
ThankstoDavid,Grant,NadiaandGillyforgivingmestrength;
ThankstoallmembersoftheSharkTank,forgivingmeheart;
Thankstomyfamilyandfriendsforsupportingmysoul.

Iwouldliketoexpressmyappreciationtoallofthosepeoplewhohaveassisted
mewithcompletingmyresearch.Specifically,Iwouldliketoacknowledge:my
brotherforhelpingmewiththeprogramming,Nadia,Gilly,AslanandAdamfor
sittinginMcDonaldswithme,KeithConlonandDavidEllisforhelpingme
publicisemystudies,MannyNoakesforprovidingnutritionalfeedback,Tessafor
proofreadingandCSIROandtheNOBLEResearchGroupforprovidingsupport,
bothfinancialandemotional.


1

1 ChapterOne:Obesity,changingeatingpatternsandfastfoods

1.1 Obesity
Seidell(2000)emphasisedtheseriousnessoftherisingrateofobesitybyreferringtoit
asaworldwideepidemic.Subsequentpublicationshavealsonotedtheglobalscaleof
theproblem(Wadden,Brownwell,&Foster,2002).Obesityhasbeeninvestigatedfor
over40years(e.g.,Schachter,1968),andthereisanextensivebodyofrecentresearch
thathasattemptedtoidentifycausesfortheescalatingratesofincidenceindiverse
countries.

Generally,BodyMassIndex(BMI)isusedasapopulationlevelindicatorofobesity.BMI
iscalculatedbydividingweightbyheightsquared;thisprovidesanindicationofweight
relativetoheight.ABMIbetween25and29isgenerallyconsideredoverweightwhile30
andaboveisconsideredobeseinEuropids(WorldHealthOrganisation,1995).Thereis
slightvariationinthesecutoffsfordifferentethnicpopulations(WorldHealth
Organisation&InternationalAssociationfortheStudyofObesityInternationalObesity
TaskForce,2000).TherehasbeencriticismofBMIasameasureofobesityasitdoesnot
makeanyallowanceforbodyshapeormusclemass.Morefundamentally,BMIhas
limitedscopeasapredictorofobesityrelatedhealthproblemssuchasmetabolic
syndrome
1
.Inclinicalsettings,itisrecommendedthatobesitybemeasuredthrough
assessmentoffatdistribution,suchaswaistcircumferenceandhiptowaistratio,as
thesemethodsofmeasurementhavebeenshowntobebetterpredictorsofhealthrisks
associatedwithobesity(Haslam&Wittert,2009;PublicHealthServiceNational

1
Metabolicsyndromedescribesthepresenceagroupofcommonriskfactorsformyocardialinfarctionin
anindividual.Theseriskfactorsincludepoorglycemiccontrol,abdominalobesity,dyslipidemiaand
hypertension(Haslam&Wittert,2009).


2


InstitutesofHealthNationalHeartLungandBloodInstitute,1998;Seidell&Flegal,
1997).Inthepast,BMIhasbeenmostcommonlyusedtodefineobesity
2
.

1.1.1 Globalincreaseinobesityrates
Althoughvaryingmethodologiesanddefinitionsofobesityusedinnationalsurveyslimit
directcrossnationalcomparisons,multiplenationsreportsindicatesimilarincreasesin
theratesofobesity.OneanalysisoftheUSNationalHealthandNutritionExamination
Survey(NHANES)indicatedthatwhileratesofobesitywerestablethroughoutthe
1980s,inthepast20yearstheyhavedoubled(Baskin,Ard,Franklin,&Allison,2005).
ThenationalprevalenceofobesityamongAmericanadultsisnowreportedtobe30.4%
(Baskinetal.,2005).InNewZealand,theMinistryofHealth(2004)publishedareport
whichtrackedtheratesofobesitybetween1977and2003.Itshowedthegradualand
consistentriseinratesofobesitythroughoutthetimespan,withratesincreasingfrom
approximately10%tojustover20%.InsomenonWesterncountries,similarpatterns
havebeenobserved.Forexample,inKorea,inthesevenyearsfrom1995to2001,
obesityratesrosefrom13.9%to30.6%
3
ofthepopulation(Kim,Ahn,&Nam,2005).In
theUK,recentestimatesoftheprevalenceofobesitysuggestabout24%ofthe
populationisobese(Rennie&Jebb,2005).InotherpartsofEurope,theobesity
epidemichasbeenlessclearlyestablished.Between1993and2003,ratesofobesity
amongstPolishwomenrosefrom8.9%to15%.However,asimilartrendwasnot
observedformeninPoland(Milewiczetal.,2005).

Australiahasnotbeenuntouchedbytherisingratesofobesity.In1995,resultsofthe
NationalNutritionSurveyrevealedan18%prevalenceofobesityamongstAustralians

2
Ithasbeensuggestedthatestimationsofobesityrateswouldbehigherifwaistcircumferenceweremore
commonlyusedasapopulationestimateoftheproblem(Haslam&Wittert,2009).
3
Theauthorsusedamorestringentdefinitionofobesity(BMI=25)thanintheEuropeanandAustralasian
reportswhichusethestandardBMIof30.Thiswasjustifiedbytheresearchersgiventhedifferentshape
ofAsianpeople.


3


(Jackson,Ball,&Crawford,2001).In2000,theAusDiabstudyalsoprovidedasnapshot
ofobesityintheAustralianpopulationindicatingthat20.5%ofthe11,247participants
wereclassifiedasobese
4
(Dunstanetal.,2001).Subsequentfollowupassessmentsof
thesamesampleindicateda1.4kilogram(kg)weightincreaseoverfiveyears
5
(Barret
al.,2006).Followupassessmentsalsorevealedthattwiceasmanypeoplewhowere
overweightwhenfirstsurveyedbecameobesethanwentbacktonormalweight.The
AustralianInstituteofHealthandWelfarepublishedareporttrackingobesityfrom1980
to2001(Dixon&Waters,2003).Thisrevealedasimilarriseintheprevalenceofobesity
tothatreportedinothercountries.Thenumberofobeseadultsrosefrom9.3%in1980
to16.2%in2001
6
.

ThepopularmediahaverecentlyreportedthatAustraliasrateofobesityhassurpassed
NorthAmericas(McLean,2008a).ReportslabellingAustraliansastheworldsfattest
peopleappearedinmediainternationally(Larter,2008).Thesereportswerebasedona
reportpublishedbytheBakerInstitute(Stewart,Tikellis,Carrington,Walker,&O'Dea,
2008)whichindicatedthat26%ofAustraliansareobesearatehigherthanprevious
estimates.Itislikelythatthesampleuponwhichthisstudywasbasedwaspoorly
representativeofthegeneralpopulation.Theinstitutetooktheheightsandweightsof
10,000peoplewhowenttoafreebloodpressurecheckonanawarenessdayfor
CardiovascularDisease(CVD).Therefore,giventhenatureofthesurvey,itmayhave
attractedahighernumberofobesepeoplethanonewouldreasonablyexpectfromthe
population.TherehasbeenabodyofresearchabouttheassociationbetweenCVDand
obesity(AustralianInstituteofHealth&Welfare[AIHW]&NationalHeartFoundationof

4
AusDiabisaprospectivecohortstudyassessingdiabetesandassociateddisorders,includingoverweight
andobesity,usingavarietyofphysicalexaminations.Inthefirstphaseofthestudy,11,247people
completedphysicalassessments(Dunstanetal.,2001).
5
Onlythosewhowere65andolderlostweightduringthefiveyearperiod.
6
Resultswerereportedbysex.Thesepercentageswereaveragedtogetanapproximateestimateofthe
rateofobesityforthecombinedadultpopulation.


4


Australia,2004;Grundy,2004;Poirieretal.,2006).Althoughtheremaybesome
contentionregardingtheexactrateofobesityinAustralia,itisclearthatthecurrentrate
ofobesityishigherthanitwas20yearsago.ThisindicatesthatAustraliaisexperiencing
asimilartrendtotherestoftheworld,withrisingratesofobesity.

1.1.2 Directandindirectconsequencesofobesity
Asobesityratesrise,sodoesthenumberofpeoplelikelytosufferfromobesityrelated
disordersincludingdiabetes,heartdiseaseandcancer,allofwhichareassociatedwith
morbidityandmortality(Jamesetal.,2004;NSWCentreforPublicHealthNutrition&
NSWDepartmentofHealth,2003;OBrien&Webbie,2001).Someauthorshave
predictedthatobesitywillsoonpasssmokingasoneoftheleadingcausesof
preventabledeath(Sibbald,2002).Asindividualsgetheavierandtheprevalenceof
obesityrelatedhealthproblemsincreases,sodoesthedemandonthehealthcare
system.Thismakesincreasingratesofobesityaseriouspublichealthissueandamotive
forresearchtoidentifyitscauses.InAustraliaareportfromtheNationalPreventative
HealthTaskforce(2008)hasrecentlyoutlinedtheactualcostsofobesityindicatingthat
highbodymasswasresponsiblefor7.5%ofthetotalburdenofdiseaseandinjury
(p.5)in2003.

Furtherescalatingthecostofobesityaretheindirect,negativeeffectsofobesity
includingthestigmacurrentlyassociatedwithfatness.Despiteglobalincreasesinbody
weight,whatisconsideredtheidealbodyhasnotchangedtoresemblethisshift.
Childrenasyoungasfourhavedemonstratedadesireforanidealbody
7
(McLean,
2008b).Manyobesepeoplesufferthedamagingconsequencesoftheprejudicethey
experience(see,Harris&Walters,1991;Puhl&Brownell,2001,2003;Roehling,1999;

7
UnpublishedstudybasedonMcCabesworkatDeakinUniversity.


5


Wadden&Stunkard,1985).Researchcomparingchildrensstigmatisationofobese
childrenhasshownthatthisprejudiceisonethathasincreasedsincethe1960s(Latner
&Stunkard,2003).Theexistenceofsuchprejudicefurthersupportsthesocietal
pressuretoconformtoidealbodyshapes.Thiskindofpressureamongstapopulation
increasinglylesslikelytoachievesuchidealsmayincreasetheprevalenceofbodyimage
andeatingdisorders.Thereisalargebodyofliteratureandresearchonsuchissues(see
McCabe&Ricciardelli,2001a,2001b;McCabe&Ricciardelli,2003;Ricciardelli&
McCabe,2001).Klaczynski,GooldandMudry(2004)showedthatthemorepeople
valuedthebodyimageideal,thegreatertheirattempttocontroloverweight.Theyalso
reportedaperceptionthatobesitywascausedbypersonalshortcomings.

Finally,oneoftheotherconsequencesofobesityisthatitmaypromoteobesityinfuture
generations.Thereissomeevidencethatobesemothersaremorelikelytohaveobese
offspring(Shankar,Harrell,Gilchrist,Ronis,&Badger,2007).Furthermore,obese
parentsaremorelikelytoraiseobesechildren(Epstein,Wing,&Valoski,1985;Maffeis,
Talamini,&Tato,1998;Whitaker,Wright,Pepe,Seidel,&Dietz,1997).Somestudies
havealsosuggestedthattheoddsofbecomingobesecanincreasesimplythrough
knowingapersonwhobecomesobese(Christakis&Fowler,2007).Ifobesitypromotes
futureobesity,itisunsurprisingthatobesityratesarerising.Intheliteraturedebate
surroundswhetherornotobesityistrulyclassifiableasadisease(Heshka&Allison,
2001),nevertheless,thetransferrablenatureobesitycouldbeitsmostserious
consequence.

1.1.3 Energyequilibrium:Anexplanationforobesity
Aseriesofpublicationshaveexploredthefactorsthatmightcontributetoincreasing
ratesofobesity.Investigationsoftheepidemiologyofobesityhaveexploredthe


6


disruptiontoenergyequilibrium.Accordingtotheenergyequation,obesityoccurswhen
moreenergyisconsumedthanisexpended.Variableslikelytoincreaseenergyintake,
decreaseexpenditure,orboth,canthereforebeidentifiedasriskfactorsforobesity.The
equilibriumapproachunderliesmostoftheresearchexploringthefactorslikelyto
contributetoobesity(e.g.,Dodd,Welsman,&Armstrong,2008;Gerber&Corpet,1999;
Goris&Westerterp,2008;Rodriguez&Moreno,2008;Wells,Ashdown,Davies,Cowett,
&Yang,2007);itprovidesausefulstaringpointfromwhichhypothesescanbe
developed.

Thesearchforthepotentialcausalpathwaytoobesityhasfocusedondelineatingthe
contributionsofdietandfoodchoice,physicalactivityandpredilectionforsedentary
pastimes,geneticinfluences,environmentalandsystemvariablesthatimpactonfood
intakeandphysicalactivity,andapproachesthathavesynthesisedavarietyofpotential
contributors(Livingstone,2000;Swinburn,Caterson,Seidell,&James,2004;Weinsier,
Hunter,Heini,Goran,&Sell,1998).Otherclinicalresearchhasendeavouredtoestablish
causalitythroughexperimentalinterventions;onlylimitedsuccessisreported,with
weightmaintenancedifficulttomaintainaftertheinitialweightloss
8
(Scheen,2002).
Althoughsomeoftheseinterventionshavefocusedonphysicalactivity(Barretal.,
2006;Waddenetal.,1997;Wing,1999),themajorityhavelookedatmodifyingfoods
consumedincombinationwithlifestylefactors(Brinkworthetal.,2004;Shaietal.,
2008;Tay,Brinkworth,Noakes,Keogh,&Clifton,2008;YackobovitchGavan,Nagelberg,
Demol,Phillip,&Shalitin,2008).

Regardlessofdivergentviewsastowhetherenergyexpenditureorintakeismore
crucialformaintainingenergybalance,bothviewsarebasedonthepremisethatenergy

8
Theremaybephysiologicalreasonsforthisthatinvolvetheregulationanddefenceoffatmass(See,
Levin,2004,2007;Levin&DunnMeynell,2000;Mauer,Harris,&Bartness,2001).


7


equilibrium(orenergybalance)isanimportantaspecttoweightmaintenance,weight
lossandweightgain.Despitelimitedsuccessinweightmaintenance,clinicaltrialsand
interventionshaveshownthattargetingfactorsrelevanttoenergyequilibriumcanbea
successfulmethodforweightloss.

1.2 Eatingpatternsandchangingdiets
1.2.1 Theincreaseintheconsumptionofmealspreparedawayfromthehomeand
nutritionalimplications

Asavarietyoftechnologieshavebecomemoreaccessible,foodprocessinghasbecome
morecommoninindustryandinthehome(Earle,1997;Popkin,2001).Oneofthemost
attractivequalitiesoffoodprocessingisthatitcanmakefoodsmuchmoreconvenientto
obtainandconsumeconsequently,mealspreparedawayfromhomearenowwidely
availableandinexpensive.Theincreaseintheconsumptionofmealspreparedaway
fromhomehasbeenwitnessedinseveraldifferentpopulations.Forexample,Coxand
Foster(1985)notedashifttowardrestauranteatingandtheuseofconveniencestores
inthemid1980s.Morerecently,KantandGraubard(2004)reportedthatbetween1987
and2000thenumberofcommerciallypreparedmeals
9
eatenperweekbyindividuals
increasedby11%to2.77mealsperweek.Theyalsosupportedthecommonlyheld
beliefthatmealspreparedawayfromhomehavedifferentnutritionalqualitiesto
traditional,homecookedmealsandareenergydense,observingarelationshipbetween
thenumberofcommerciallypreparedmealseatenandoverallenergyintake.Howthis
impactedonoverweightandobesitywasnotclear,anditcannotbeassumedthat
greaterenergyintakeautomaticallyresultsinincreasedweight
10
.


9
Definedasthosepreparedinrestaurants,fastfoodplacesorcafeteriasincludingthoseeatenin,taken
awayordelivered.
10
Althoughtheaveragecommerciallypreparedmealisenergydense(thusincreasingenergyintake),
peopleconsumingthesemealscompensatebyincreasingtheirenergyexpenditureoreatlessthroughout
thedayandthereforekeeptheirenergyintakebalanced.Unfortunately,giventheconstraintsofthedata,
theauthorscouldnotreportspecificallyontheparticipantsamountofphysicalactivity.


8


DatafromtheUSDAsContinuingSurveyofFoodIntakesbyIndividuals(CSFII;WhatWe
EatinAmericaSurvey)providedamoredetailedpictureofchangingmealpatternsinthe
US.TheCSFIIisanationwidesurveythatwasundertakenacross50statesbetween
1994and1996.Itrequiredparticipantstorecalltheirintakeovera24hourperiodon
twononconsecutivedays,threetotendaysapart.Otherdetailsrecordedwerethetime
anddefinitionofeatingoccasions,waterintake,thesourceoffood,andinformation
aboutheightandweight(U.S.DepartmentofAgriculture,1997).Datawereobtainedfor
individualstwoyearsofageorolder(theintakeofchildrenbeingreportedbytheadult
interviewee).Thesamplecontained16,103casesincludingover9,000individualsabove
theageof20.Peoplewithlowincomeswereslightlyoversampled(J.Goldman&Borrud,
1997).

LinandFrazao(1999)comparednutritionaldatacollectedin1977bytheUSDAtothe
CSFIIdatafrom1995.Theyfoundthat,duringthistimeframe,thenutritionalqualityof
foodspreparedathomeincreasedmorethanthequalityoffoodspreparedawayfrom
thehome.Theauthorsnotedthatfoodspreparedawayfromhomeweretypicallyhigher
infats.Guthrie,LinandFrazao(2002)usedthesamedatasetstoestablishwhetherthere
hadbeenashifttowardsconsumingmealsandsnackspreparedawayfromhome.They
reportedthat1.78timesthepercentageoftotaldietarycaloriesconsumedcamefrom
itemspreparedawayfromhomebetween1994and1996comparedto1977.Guthrieet
al.sanalysesfurtherconfirmedthenotionthatthesefoodsaremoreenergydense,
higherinfatandlessnutritious(lowerinfibre,ironandcalcium).

Guthrieetal.(2002)assessedthesourcesofcaloriesfromfoodspreparedawayfrom
homeandreportedthatfastfoodswerethehighestsourceoffoodpreparedawayfrom
homeat12%,followedbyfoodseatenfromrestaurants,whichaccountedfor10%.Ina


9


slightlydifferentanalysisoftheCSFIIdata,Binkley,EalesandJekanowski(2000)
substantiatedGuthrieetal.sfindings,showingthatalargevolumeoffoodsprepared
awayfromhomearebeingconsumedandthatfastfoodaccountsforahighproportion
ofthesefoods.

1.3 Understandingfastfoodconsumption
ItisclearthatmealspreparedawayfromhomecomprisealargerpartofWesternised
diets.DatareportedfromtheCFSIIalsosuggestsfoodspreparedawayfromhomeare
energydense.PapersfromGuthrieetal.(2002),Binkleyetal.(2000)andKantand
Graubard(2004)implicatefastfoodsintheshifttowardtheconsumptionoffood
preparedawayfromthehome.Thusitappearsthatfastfoodsareimportant
contributorstocurrentWesterndietsandpotentiallyimplicatedintheassociation
betweendietarychangeandrisingratesofobesity.

1.3.1 Definingfastfoods
Fastfoodisaspecificvarietyofconveniencefoodwhichiscommonlyassociatedwith
highenergydensity.Yet,therehasbeenambiguityinthedefinitionoffastfoodinthe
existingempiricalresearch.Somestudieshavepreferredtoleavedefinitionofthe
phraseopentotheirparticipants.FastfoodintheCSFIIquestionnairereferstofood
purchasedfromafastfoodplace,pizzaplacewithoutdefiningfastfoodplacesfor
participants(U.S.DepartmentofAgriculture,1997).Otherstudieshaveusedthestyleof
servicetodifferentiatefastfoodsfromotherconveniencefoods(Driskelletal.,2006).
Evenwithinanalysesofthesamestudy,definitionsusedtoanalysethedatahave
differed.Pereiraetal.(2005)combinedcafeteria,restaurantandfastfoodsintheir
definitionoffastfood.Whereasadifferentanalysisofthesamedataseparatedthe
definitionintorestaurantandfastfoodcategories(Duffey,GordonLarsen,Jacobs,


10


Williams,&Popkin,2007).SeeTable1(onthefollowingpage)forexamplesofpapers
whichhaveusedthetermfastfoodandhowthistermhasbeendefined.

Inanempiricalcontextitisimportanttohaveacleardefinitionofwhatconstitutesfast
foodtoimproveboththevalidityandreliabilityofresearchinthearea.AstudybyDunn,
Mohr,WilsonandWittert(2007)showedthatthedefinitionoffastfoodcanbe
inconsistentevenwithinasinglesample.Theyallowedparticipantstodefinefastfoods
fromalistofdifferentfooditems.Theyfoundthatalthoughpeopleindicatedavery
broaddefinitionoffastfoodwhenaskedwhatstyleoffoodsclassifiedasfastfood
(includingburgers,fishandchips,meatpiesandpasties,andboughtsandwiches),when
discussingfastfoodtheygaveexampleswhichfocussedonlyontraditionalfastfood
itemssuchasburgersandchipsfromlargefranchises.Thesedataaresupportedby
researchshowingthatmostfastfoodpurchasesareburgers,friesandpizza(Driskellet
al.,2006).

Onecrucialstepinarrivingatacleardefinitionoffastfoodsisthedifferentiationoffast
foodsfromconveniencefoodsasacategory.Thisdistinctionneedstooccurbecause
thereisawidearrayoffoodsthatcanbelabelledconveniencefoods(fromfrozenmeals
tomeatpiestoprepackagedsalads).Fastfoodfitscomfortablywithintheconvenience
foodgroup,butitremainsadistinctsubcategoryofthisgroup.Usingthestyleofservice
typicaltofastfoodrestaurantscandistinguishfastfoodfromotherconveniencefoods.
Evenwiththisfurtherrefinement,thedefinitionisstillinclusiveofanextensiverangeof
foods.Forexample,Asiancuisineplacesinlocalfoodcourtscouldmatchthis
descriptionbutmaynotbeassociatedwithfastfoodbytheaverageconsumer.



11


Table1:Exampledefinitionsusedinpreviousstudiesaboutfastfoods.

Author/s Definitionoffastfoodused Specificexamplesoffered
CARDIAstudy:
Duffey,GordonLarsen,
Jacobs,Williams,and
Popkin(2007)

Bigbrandfastfood
companies
BurgerKing,Wendys,Arbys,
PizzaHutandKentuckyFried
Chicken
Pereiraetal.(2005) Foodsfromcafeterias,
restaurantsandfastfood
companies
CSFII:U.S.
Departmentof
Agriculture(1997)

Foodfromafastfoodplace,
pizzaplace
None
Driskell,Mecknaand
Scales(2006)
Foodfromafastfood
restaurantsomewherefood
canbeordered,purchasedand
receivedwithinroughlyten
minutes
None
French,Harnackand
Jeffery(2000)
Mealsfromfastfood
restaurants
None
French,Story,Neumark
Sztainer,Fulkersonand
Hannan(2001)
Foodfromfastfood
restaurants
McDonalds,BurgerKingetc.
Guthrie,LinandFrazao
(2002)
Thosepurchasedfromafast
foodplacethathadquick
servicebutdidnothave
waitersortableservice
None
Jeffery,Baxter,McGuire
andLinde(2006)
"Fastfood"restaurants,quick
serviceburger,quickservice
roastbeef,andquickservice
pizzaparlour
McDonalds,LongJohn
Silvers,TacoBell
Popkin,SiegaRiz,
HainesandJahns(2001)
Fastfood Mexicanfood,Chinesefood,
frenchfries,hamburgersand
cheeseburgers
Reidpath,Burns,Garrad,
MahoneyandTownsend
(2002)
Fivelargestquickservice
restaurantsinAustralia
McDonalds,HungryJacks,
KFC,PizzaHutorRedRooster
RoyMorgan(2006) Fastfood Initiallynone,thenasked
aboutpurchasesfrom
McDonalds,KFC,Subway,
HungryJacksandDominos
(Pizza)
Ungeretal.(2004) Somethingtoeatfromfast
foodrestaurants
MacDonalds(sic),Burger
King,Dominoes(sic),Pizza
Hut,TacoBell,orotherfast
foodrestaurants



12

Itappearsthattheclearestandmostconsistentdefinitionoffastfoodsisonethat
involvesbigbrandfastfoods(Duffeyetal.,2007).Reidpathetal.(2002)classifiedfast
foodasalltheitemsavailablefromthefivelargestquickservice
11
restaurantsin
Australia.Astheserestaurantsrepresentthemajorityofnationalfastfoodchains,they
shouldberepresentativeoftheculturalunderstandingoffastfoodwithinthegeneral
population.Definingfastfoodinthiswayalsorestrictsthenumberofpotentialfoods
andrestaurantsthatcanbeincludedwithinthecategory.Thislimitiscrucialforboth
researchersandparticipantstoensurehomogeneityininterpretation.

InReidpathetal.soriginaldefinitionoffastfood,thetopfivefastfoodchainsin
AustraliawereMcDonalds,HungryJacks(a.k.a.BurgerKing),KFC(KentuckyFried
Chicken),PizzaHut,andRedRooster(Reidpathetal.,2002).Sincethispublication,two
newdominantcompanieshaveemergedinthefastfoodmarketinAustralia:Dominos
PizzaandSubway.DominosPizzanowclaimstobeAustraliaslargestpizzamaker,
havinghad360storesopenacrossAustraliain2005(Domino'sPizzaEnterprisesLtd.,
2005).Subwayhasalsocementeditsplaceamongstthemajorfastfoodchainsin
Australia.McDonalds,KFCandSubwayaloneaccountforover65%ofmarketsharefor
takeawaymealsinAustralia(Burke,2007).Subwayisadifferentstyleoffastfoodwhere
itscustomerschoosebetweenapredeterminedsetofingredientstocreatetheirown
sixinchorfootlongrollwithfilling.Subwayisalsodifferentfromthetraditionalstyleof
fastfoodasitactivelypromotesahealthyimagetoconsumers.Forexample,Subway
spokespersonJaredFoglereportshavinglost245poundsontheSubwaydietandis
theirselfproclaimedweightlosshero(DoctorsAssociatesInc.,2002).RoyMorgan
(2006)datasupporttheshiftinmarketsharesincetheReidpathetal.paper.Many

11
Quick-service restaurant is a term used in the industry that is synonymous with fast food restaurant. As is the
case with the term fast food, there is no clear definition as to what exactly constitutes quick-service.


13


peopleinthestudyindicatedthattheyhadeatenfromSubwayand/orDominosPizzain
thepastfourweeks.Thepercentagesforpeoplereportingeatingfastfoodfromdifferent
fastfoodchainscanbeseeninFigure1.
Thebigbrandcharacterisationoffastfoodisusefulfordefiningfastfoods.Theactual
companiesthataretheinthetopfastfoodchainswillshiftrelativetomarkettrendsand
cohorts.Therefore,researcherswillneedtobewaryoftheseshiftswhenestablishing
whichfastfoodcompaniesrepresentalargeportionofthemarketshareinagiventime
andlocation.

1.3.2 FastfoodconsumptioninAustralia
AlthoughCFSIIdataindicatethatfastfoodisoccupyingalargepartofthedietintheUS,
thereislimiteddataaboutfastfoodhabitsinAustralia.Dataaboutfoodspendingfrom
theAustralianBureauofStatisticsHouseholdExpenditureSurvey(2000)supportdatain
theUSshowingthatfoodspreparedawayfromthehomearebecominganincreasing
63
31
23
16
14
10
0
15
30
45
60
75
AnyFast
Food
McDonal ds KFC Subway Hungry
Jack's
Domi no's
*DatareconstructedfromRoyMorganpressreleaseSeptember15,2006
Q:Howmanytimeshaveyouvisitedorpurchased(fastfoodor
from[companyname])inthelastfourweeks?
Figure1:PercentageofAustraliansmakingfastfoodpurchasesinlast
fourweeks*.

NOTE:
This figure is included on page 13
of the print copy of the thesis held in
the University of Adelaide Library.


14


partofthediet.Atthetimeofthesurvey,anaverageof$127perweekwasspenton
foodsandnonalcoholicbeverages,withmealseatenawayfromhomeoccupyingthe
largestpercentageofthisspending(26.77%).Furthermore,takeawayandfastfoods
accountedfor55.88%($19perweek)ofallmealseatenout.Therefore,itappearsthat,
asintheUS,conveniencefoodsareoccupyingalargepartofthedietinAustralia.

ItisundeniablethatbigbrandfastfoodsarereadilyavailableinAustralia.Majorfast
foodrestaurantsexistinallthemetropolitancentres.Forexample,inAdelaide
(Australia'sfifthlargestcity:AustralianBureauofStatistics,2007)therearemultiple
fastfoodlocationsinallmetropolitanregionswithmanysuburbshavingseveral
restaurants(seeFigure2onthefollowingpage).

Despitethehighnumberofquickservicerestaurants,itisdifficulttodeterminehow
manyAustraliansareactuallyconsumingfastfood.AsurveyconductedbyIndependent
GrocersofAustralia(IGA)suggestedthat57%ofpeopleeatfastfoodatleastoncea
week(FOODweek,2008).Alargereportonfastfoodconsumptionconductedbymarket
researchcompanyRoyMorgan(2006)aimedtogetasnapshotoffastfoodconsumption
habitsacrossAustralia.BetweenJuly2005andJune2006,thecompanyasked25,000
Australiansaged14yearsandoveriftheyhadpurchasedfastfoodinthefourweeks
precedingtheinterview.Sixtythreepercentofrespondents(over15,000people)
indicatedthattheyhadconsumedfastfoodinthepastmonth.Acomparisonofthese
datawithresultscollectedinternationallybyRoyMorganshowedthatproportionately
fewerAustraliansreportedhavingfastfoodthanpeopleintheUS(80%)andNew
Zealand(67%),butmorethanintheUK(56%).Mohr,Wilson,Dunn,BrindalandWittert
(2007)reportedthat32.7%of17,354respondentstoanationwideACNielsensurvey


15


indicatedthattheyconsumeddineinfastfoodseveraltimesperweek.Only15.6%of
thissampleindicatedthattheyneverconsumeddineinfastfoods.

Figure2:McDonalds,HungryJacks,PizzaHutandKFCrestaurantsinthe
AdelaidemetropolitanArea.Redcirclesindicatemultiplerestaurantswithina1
kilometreradius.


16


1.3.3 Thefastfoodconsumer
Dunn,Mohr,WilsonandWittert(2008)reportedthatpeoplehadnegativestereotypes
aboutthetypicalfastfoodconsumer.Theynotedthat,generally,participants
descriptionofsomeonewhoeatsfastfoodincludedalesserindividual,lackingin
knowledgeorselfcontrol,andpossiblydrivenbyhedonisticoreconomicfactors
(p.333).Thesearesimilarconceptionstothoseassociatedwithobesepeople(Cowan,
Cowan,Hiler,Smalley,&Sehnert,1992;Lobera,Fernandez,Gonzalez,&Millan,2008).
Althoughthisperceptionofafastfoodconsumerisnegativeandjudgementaland,
despitethefactthatahighportionofAustraliansreporthavingconsumedfastfoods
withinamonth,fastfoodsmaybemoreattractivetosomeconsumersthanothers.Inthe
US,datafromtheCSFIIindicatethatclosetoathirdofthepeoplesurveyedatefastfood
duringatypicalday,witholdermalesofhigherincomereportinghigherintakes
(Bowman&Vinyard,2004).AdifferentanalysisofthecompleteCSFIIdataindicated
thatahighproportionoffastfoodconsumersarechildren(Paeratakul,Ferdinand,
Champagne,Ryan,&Bray,2003).Intheirsampleofadults,Jefferyetal.(2006)reported
thattherewasarelationshipbetweenhavingchildrenandeatingatfastfood
restaurants.Thismayindicatethatchildrenareattractedtofastfoodwhichmay,inturn,
attracttheirparents.Asidefromfamilies,thereisalsoevidencethatyoungadultsare
regularconsumersoffastfood.Forexample,Schroderetal.(2007)reportedthatfast
foodconsumersweremorelikelytobeyoungerandmoreeducatedthannonfastfood
consumers.TheCARDIAstudyofyoungadultsreportedthatwhitewomenwerethe
leastfrequentfastfoodconsumers(Pereiraetal.,2005).However,thisanalysisofthe
datawasrestrictedtoadultsandexcludedthoseundertheageof20.IntheUS,some
researchershaveassertedthatfastfoodintakeoccursfrequentlyinadolescenceage
groupsstatingthatupto75%ofadolescentseatfastfoodatleastonceaweek(French


17


etal.,2001).Othershavereportedthatadolescentseatfastfoodonaveragebetween
twoandthreetimesaweek(Ungeretal.,2004).

Thetypicaldemographiccharacteristicsoffrequentfastfoodconsumersappearsimilar
inAustraliansamples.Forexample,ithasbeenestimatedthat22%ofadolescentsin
Australiaeatfastfooditemseveryday(Savige,Ball,Worsley,&Crawford,2007).Mohret
al.(2007)analysedpredictorsoffrequencyoffastfoodconsumptionintheirlarge
nationalsample.Theyreportedthatbeingyounger(under45),havingahigherincome
andgreaterfondnessfortechnologywereallvariablesassociatedwithmorefrequent
fastfoodconsumption.Beingmalewasalsoasignificantpredictoroffrequencyof
takeawayconsumption.Occupationalstatusandeducationlevelweretheonly
demographicfactorsthatdidnotsignificantlypredictpatronage.

1.3.4 Whydopeopleconsumefastfood?
Manystudiesinvestigatingthereasonswhypeopleconsumefastfoodshavepointedto
convenience.TheIGAsurveyreportedthatpeoplegenerallyeatfastfoodbecauseof
convenientlocationsandtimeconstraints(FOODweek,2008).SchroderandMcEachern
(2005)surveyed100undergraduatestudentsandreportedthatfastfoodpurchasesin
theUKweremostlyimpulsive(57%)withasmallsubsetofpeopleroutinelyeatingfast
food(26%).Theyreportedthatbrandvalue,nutritionalvalue,ethicalvalueandfood
qualitytogetheraccountedforover50%ofthevarianceinfastfoodpurchasing
behaviour.Recently,BryantandDundes(2008)surveyedNorthernAmericanand
Spanishpeopleabouttheirperceptionsoffastfoods.Theyfoundthatmostpeoplerated
tasteandflavourasimportantfactorsforencouragingfastfoodconsumption.Cultural
differenceswereapparent,withUSstudentsratingvalueformoneyasmoreimportant
thanSpanishstudents.InaseparateUSstudyexploringthereasonsforfastfood


18


consumptionincollegestudents,similarfactorswerereported;mostpeoplerated
convenienceasafactorinfluencingtheirconsumption.Costandmenuchoicesalso
relatedtothenumberoffastfoodmealspurchased(Driskelletal.,2006).

Asidefromtheobviousbenefitsoffastfood(itisquick,easyandgenerallycheap),some
researchershavealsoreportedonhowexperientialaspectsoffastfoodareattractiveto
consumers.AsurveyofKoreanfastfoodusersdiscoveredthatfastfoodconsumption
wasnotpurelyautilitarianbehaviour.Park(2004)reportedthathedonicaspectsoffast
foodwerelikelytodeterminewhichrestaurantconsumersvisited.Suchfactorsrelated
totheexperienceofeatingfastfoodincludingthetasteandtheeatingenvironment.
Ootherstudieshavesuggestedthattheremaybefactorsinafastfoodenvironment
beyondthefoodandconveniencethatattractpeopletofastfood.Forexample,asmall
percentage(14%)ofpeopleinBryantandDundes(2008)surveyindicatedthatthey
likedtheabilitytosocialiseatfastfoodrestaurants.Driskelletal.(2006)reportedthata
thirdofthewomenintheirstudycitedsocialreasonstoexplaintheirfastfood
consumption.

Theexperientialdriversofconsumptionmayaffectcertaingroupsmorethanothers.
Fastfoodrestaurantsmaybedistinctfromothereatingoutplacesastheyarein
convenientlocations,areopenlonghoursandprovidelowcostfoods.Thesequalities,
amongstothers,mayattractcertaingroupsofpeopletofastfoodrestaurants.For
example,olderpeoplemaybeattractedtofastfoodlocationsastheyprovidea
convenientandaccessibleplaceforsocialising(Cheang,2002).

Ultimately,therearemanyaspectsreportedthatattractpeopletofastfood.Factorsthat
guideconsumerchoiceinmultiplecontextshavebeendocumentedandexplored.These


19


factorshaveincludedfamiliarity,priceandtaste(Prescott,Young,O'Neill,Yau,&
Stevens,2002;Steptoe,Pollard,&Wardle,1995),butconsumerchoiceisacomplex
behaviourwhichcannotbedefinedbyratingtheimportanceofaseriesoffactors
(Scheibehenne,Miesler,&Todd,2007).Inrealworldsituations,fooddecisionsaremade
withinthecontextoftimepressures,specificenvironments,individualpreferencesand
socialvariables.Socialcognitionresearchandtheorysuggeststhatdecisionswill
generallybemadeinwayswhichrequiretheleastcognitiveload,oftencalledheuristics
(Petty&Cacioppo,1986;Shah&Oppenheimer,2008).Itisonlyinidealsituationswhere
choicescanbemadebyweighingallimportant,relevantfactors.Themodeloffood
cognitionsuggeststhatimportantcriteriaareweighedinordertomakefooddecisions.
Theamountoftimeavailabletomakethesedecisionswillshiftchoicesfromthebest
optiontothemostidealoneatthetime(Scheibehenneetal.,2007).Priceisusually
amongstimportantpredictorsoffoodchoicebutnotconsistentlybecausecontextwill
changethefactorsthatareimportant.Ultimately,choicewillbedrivenbyfactorsthat
willfulfilthemostimportantneedatthetime(Scheibehenneetal.,2007).

1.3.5 Underlyingreasonsfortheshifttowardconveniencefoods
Itisclearthattherehasbeenatrendforpeopletoeatfewerhomepreparedmeals,
preferringtheconvenienceoffastfoodsandotherprepreparedmeals.Althoughthere
havebeenmanysurveysreportingontheshiftineatingpatternsandexploringthe
reasonsforfastfoodconsumption,therearefewpapersexploringwhyconvenience
seemstobeacurrentdriveroffoodchoices.Thereissomeindicationthatconvenience
mayevenoutweighpersonalpreference.IntheIGAsurvey,anoverwhelmingmajorityof
respondents(91.5%)indicatedthattheywouldprefertocooktheirownmealsrather
thaneatfastfood(FOODweek,2008).Possibleexplanationsforwhythereisan
increaseddemandforconveniencefoodsarediscussedbelow.


20

Timescarcity.JabsandDevine(2006)introducedtheconceptoftimescarcitywhenthey
wereexploringfoodchoices.Timescarcityistheperceivedabsenceoftimetogetdaily
tasksdone.Withincreasingdemandsontime,freetimebecomesmorevaluableand
increasestheneedtomultitask.Indeed,eatingnowoccurssimultaneouslywithother
activities.Forexample,oneinfiveAmericanpeoplereporteatinginthecar(Glanz,
2002).Suchshiftsmayhavenegativedietaryoutcomes.Ingeneral,eatingwhile
distractedhasbeenrelatedtoincreasedintake(Bellisle,Dalix,&Slama,2004).
Televisionviewingisaspecificexampleofmultitaskingwhileeating.Thenegative
dietaryoutcomesofeatingwhilewatchingtelevisionhavebeenwidelyinvestigated
(Moray,Fu,Brill,&Mayoral,2007;Motl,McAuley,Birnbaum,&Lytle,2004;Snoek,van
Strien,Janssens,&Engels,2006;Stroebele&deCastro,2003;VandenBulck&Van
Mierlo,2004).Familieswhofrequentlyeattogetherinfrontofthetelevisiongenerally
demonstratelesshealthyeatingpatterns(Coon,Goldberg,Rogers,&Tucker,2001).As
feelingsoftimescarcitygrow,foodchoicesarelikelytochange(Jabs&Devine,2006).

Changingfoodculture.Slowfoodmovementsupporterssuggestthatmanyfoodsaway
fromthehomeareconsumedbecausefoodhasbecomealesscentralpartofdaily
interaction(seePetrini,2003).Supportersofthemovementbelievethatfoodneedsto
reestablishitselfaspartofsocialinteractionsratherthanbeinganemptysustenanceor
vice.Oneoftheunderlyingsentimentsassociatedwiththesevaluesistheideathatas
lesstraditional,homecookedfoodisconsumed,thenutritionalandsocialvalueoffood
inaculturefalls(SlowFood,2008).Indeed,someadolescentsreporthavingnofamily
mealsduringtheweek(NeumarkSztainer,Hannan,Story,Croll,&Perry,2003).Ifpre
preparedfoodsarepurchased,familieslosetheopportunitytospendtimetogether
learningaboutthepreparationandqualitiesofdifferentfoods.White(2007)


21


hypothesisedthatfoodpreparationskillscouldpredictthehealthinessofadiet.
Therefore,theculturethatsurroundsfoodandfoodpreparationcouldbeanimportant
factorforhealthyfoodchoices.

1.4 Fastfoodandobesity
Itisclearthattherehasbeenbothariseintheprevalenceofobesityandashiftineating
patterns.Thefactthatthesechangeshaveoccurredacrossasimilartimeframehasled
manypeopletosuggestthatchangesineatingpatternsmayhavenegativehealth
outcomesforthepopulation.InKorea,forexample,therewasadramaticincreaseinthe
prevalenceofobesitythroughoutthe1990s(Kimetal.,2005),whileduringthesame
timeframethepopularityoftakeawayfoodsalsogrew(Park,2004).Thesechangesmay
bepurelycoincidentalandoffernoindicationofcausality.Nevertheless,thesearethe
verystyleofassociationsthathaveledtotheexplorationasfastfoodsasapotential
contributortotheobesityepidemic.

Inthepublic,theblameforobesityhasregularlybeenshiftedawayfromchangesin
eatingpatternstolargefastfoodcorporationssuchasMcDonalds.Inrecentyears,two
featurefilmshavebeenproducedthatspecifically,andnegativelytargetMcDonalds
(SupersizeMeandFastFoodNation).Thesefilmsalonehavereachedaudiencesof
millionsandinsomecaseshavebeenshowntoalterconsequentdietarybehaviours
(Cottone&ByrdBredbenner,2007).InAustralia,fastfoodhasnotbeenimmunetosuch
negativeattention.Recently,researchershavepubliclycriticisedorganisationssuchas
CricketAustraliaforthesponsorshiptheyacceptfromfastfoodcompanies(Colagiuri&
Caterson,2008).



22


ThehighvisibilityandadvertisingexpenditureofcompanieslikeMcDonaldsmay
contributetothelargevolumeofnegativepublicitythattheyreceive.McDonaldsclaim
thattheysell,morethan75hamburgerspersecond,ofeveryminute,ofeveryhour,of
everydayoftheyear(Spencer,Frank,&McIntosh,2005:p.379),demonstratingthe
sheerscopeofthecorporation.Furthermore,thefastfoodindustryhasbeenestimated
tobeworthUSD99.6billionglobally(Datamonitor,2005).

1.4.1 Fastfoodconsumption,weightgainandobesity:Areviewofexistingstudies
Thepotentialreachoftheeffectsoffastfoodconsumptionmaybethereasonwhy,to
date,multipleresearchreportshavebeeninitiatedtoexploretherelationshipbetween
obesityandfastfood.Iftheindustryisindeedaddingtotheobesityepidemic,itcouldbe
ahighlyinfluentialcontributorasitaccountsforasubstantialportionofcurrentfood
intake.

Someauthorshaveassociatedthelocalityoffastfoodrestaurantswithobesity.For
example,Reidpathetal.(2002)suggestthatthehigherdensityoffastfoodlocationsin
poorerpostaldistrictsinAustralia(accordingtosocioeconomicstatus)createsan
environmentthatpromotesobesity.Theauthorsdidfindahigherdensityoffastfood
outletsinthepoorerof267postaldistrictsinMelbourne,Australia.Thisstudy
highlightsoneoftheprimarylimitationsofecologicaldataitdoesnotindicatehow
manypeopleareactuallyoverweightintheseenvironmentsorifpeopleinareaswitha
highdensityoffastfoodrestaurantsactuallyconsumemorefastfood.Furthermore,
studiesthathavesurveyedresidentsintheUShavefailedtosupportthesuggestionthat
closeproximitytofastfoodrestaurants(athomeandatwork)resultsinhigherintakeof
fastfoodortotheBMIstatusoftheresidents(Jeffery,Baxter,McGuire&Linde,2006).



23


TheCSFIIdatahasbeenusedtoassesstheeffectoffastfoodonotherfoodchoicesand
weightstatus(Bowman,Gortmaker,Ebbeling,Pereira,&Ludwig,2004;Bowman&
Vinyard,2004;Paeratakuletal.,2003).Paeratakuletal.(2003)demonstratedthatfast
foodusewasrelatedtolowerintakeofhealthierfoods,suchasvegetables
12
.Thistrend
appearstobeconsistentwhenappliedtothedietsofchildrenaswellasadultsinthe
CSFIIdata(Bowmanetal.,2004;Paeratakuletal.,2003).BowmanandVinyardnoted
thatfastfoodconsumershadslightlyhigheroddsofbeingoverweight.Binkley,Eales
andJekanowski(2000)reportedthatthosewhoreportedeatingfastfoodintheCSFII
wereonaveragefrom.8to1kgheavierthanthosewhodidnot.However,these
comparisonsarelimitedbythefactthatthosesubjectsclassifiedaseatingfoodprepared
awayfromhomeandfastfoodweredeterminedonlyonthebasisofthefoodsconsumed
inoneortwo24hourperiodsofdietaryrecallandnotgeneraldietaryhabits.

French,HarnackandJeffery(2000)comparedtheuseoffastfoodrestaurantsovera
threeyearperiodamongNorthAmericanwomen.Allparticipantswereenrolledina
weightlossintervention.Theirpreinterventiondataindicatedapositiverelationship
betweentheregularityoffastfoodconsumptionandbeingyounger,unmarriedand
heavier,andhavinghigherBMIandlowerincome.Theauthorsalsoconfirmedprevious
assertionsthatenergyintakeincreaseswiththeintakeoffastfood.Furthermore,the
resultsconfirmPaeratakuletal.s(2003)finding(takenfromtheCSFIIdata)showing
thattheconsumptionofvegetablesdecreaseswithincreasingfrequencyofvisitstofast
foodrestaurants.Moretelevisionviewingwasalsoassociatedwithgreaterfastfood
consumption.Ratesofparticipationinphysicalactivitywerenotlinkedtofrequencyof
fastfoodconsumption.Whenanalysingdataattheendoftheinterventionperiod,

12
J effery et al. (2006) supported these associations in a separate sample. After surveying over 1000 US
residents, they reported that eating fast food was negatively related to vegetable consumption and positively
related to BMI.


24


Frenchetal.(2000)foundthatanincreaseofonlyonefastfoodmealaweekwas
associatedwithadailyenergyintakeincreaseof234.4kilojoules(kJ)andaweightgain,
overandabovetheaverageweightgain,of.72kg.Thefindingsofthisstudyofferwhat
appearstobethemostcomprehensivepictureofboththefastfooduserandtheeffects
thatfrequencyoffastfoodconsumptioncanhaveondietandweightstatus.

French,Story,NeumarkSztainer,FulkersonandHannan(2001)reportedsimilarresults
basedupontheexaminationoffastfooduseamongadolescentsintheUS.Studentswho
reportedeatingfastfooddemonstratedpoorerfoodchoicesandanincreasedenergy
intake.Thestudyrevealedsomeotherinterestingfindingswhichpaintthepictureofthe
fastfoodconsumer.Apositiverelationshipwasrevealedbetweeneatingfastfoodand
theamountofunhealthyfoodavailableathomeaswellastheamountofsimilar
conveniencefoodseaten(foodfromrestaurants,snacksandprepreparedmeals).As
wasreportedfortheadults,itappearedthatthemorefastfoodeatenbyconsumers,the
moretimetheconsumerspentwatchingtelevision.

AstudyofthefastfoodhabitsofAfricanAmericanshasconfirmedthenegativeeffects
thateatingfastfoodcanhaveonthediet.Aftersurveyingroughly650collegestudents,
Satia,GalankoandSiegaRiz(2004)reportedthathigherfrequencyoffastfood
consumptionwassignificantlyassociatedwithlowervegetableintakeandhigherfat
intake.SimilarfindingshavealsobeenreportedamongstSpanishpopulations(H.
Schroder,Fito,Covas,&Investigators,2007).

Evidenceshowingrelationshipsbetweenunhealthybehaviours(eatingenergydense
foodandsomesedentaryactivities)appearstoberobustandissupportedbytheresults
ofUtter,NeumarkSztainer,JefferyandStory(2003).Whilenotspecificallyinvestigating


25


theintakeoffastfood,theysuggestedthatbaddietarybehaviours(includinghigher
consumptionoffriedfoods)andhoursspentinsedentaryactivitieswereassociated
withhigherBMI.Fromthesedataitappearedthatitwasonlycertainsedentary
behaviours(specifically,frequenttelevisionviewing)thatrelatedtounhealthydietary
patternssomesedentaryactivities(doinghomeworkandreading)wereactually
relatedtobetterdietaryhabits.

Ifeatingunhealthy,energydensefoodisrelatedtospendingmoretimedoingactivities
lowinenergyexpenditure,accordingtotheequilibriumapproach,thisislikelytoresult
inobesity.Itisexactlythesetypesofassociationsthatamplifythepotentialoffastfood
asariskfactorforobesity.Ofparticularconcernistherelationshipbetweentelevision
viewingandfastfoodconsumption.Manyauthorshavenotedthenegativeeffectsthat
televisionviewingcanhaveonbothphysicalactivityandhealth.Forexample,ithas
beensuggestedthattherestingmetabolicrateinchildrenisactuallylowerwhen
watchingtelevisionthanwhenresting(Klesges,Shelton,&Klesges,1993).However,this
findinghasnotbeenconsistentlysupportedbysubsequentresearch(Coopera,Klesges,
DeBonc,Klesges,&Shelton,2006;Dietz,Bandini,Morelli,Peers,&Ching,1994).When
exploringtheincreasingrateofobesityinchildreninAustralia,OldsandHarten(2001)
pointedtochanginglifestylesasareasonfortheincrease.Theyarguethatincreased
screentime(watchingtelevision)andhighermotorvehicleusagearepotentiallymore
importantforaccountingforrisesinchildhoodobesitythanasinglefactorsuchasfast
foodconsumption.JefferyandFrench(1998)reportedapositiverelationshipbetween
increasedtelevisionviewingandfastfoodconsumptionandBMI.Thiseffectwasfound
inwomenonly.Thestudyalsofoundthattelevisionviewingpredictedweightgainin
womenwithhighincomes.Theauthorsconcludedthatfastfoodsandtelevisionviewing
maybecontributingtoobesityratesintheUS.


26

TwoanalysesofdatafromtheCoronaryArteryRiskDevelopmentinYoungAdults
(CARDIA)studyhavefurtherexploredfastfoodconsumptionandweightgain
associations.TheCARDIAstudyisaprospectiveepidemiologicstudyofthe
determinantsandevolutionofcardiovasculardiseaseriskfactorsamongyoungadults
(Duffeyetal.,p.202).OneaspectoftheCARDIAstudyexaminedtheeffectsoffastfood
intakeoveralongerperiodoftime(15years)thanpreviouslongitudinalresearch.
SimilartoFrenchetal.s(2000)results,aninitialanalysisofCARDIAdatarevealedthat
increasesinthenumberoffastfood(and/orrestaurant)mealseatenresultedinrisesin
bodyweight(Pereiraetal.,2005).SubsequentanalysisofthedatabyDuffeyetal.
(2007)separatedfastfoodsfromrestaurantfoods,comparingthefrequencyofbig
brandfastfoodandrestaurantfoodconsumptionbetweenyearssevenandtenofthe
study.Theydiscoveredthatincreasesinfastfoodconsumptionoverthethreeyear
timeframewereassociatedwithincreasesinBMIwhileincreasesinrestaurantfood
consumptionwerenot.Theyalsonotethatthosepeopleeatingfastfoodmore
frequentlyatbothtimeperiodshadahigherBMI(Duffeyetal.,2007).

Notallstudieshavefoundassociationsbetweenfastfoodandweightstatus.Simmonset
al.(2005)exploredBMIandfastfoodconsumptioninpeoplelivinginruralAustralia.
TheyreportedthatBMIwasgenerallyhigherinpeopleinruralareascomparedto
metropolitanareas.TheyfoundthattherewasnorelationshipbetweenBMIandfast
foodconsumption,furtherconcludingthattheavailabilityoffastfoodisunrelatedto
obesity.Thestudyshowedthatthosepeoplewhoreportednoteatingtakeawayfoodhad
alowerwaistcircumferencethanthoseeatingtakeawayfoods.Itthereforedoesnot
completelydismissarelationshipbetweenfastfoodconsumptionandweight.


27


Nevertheless,theseresultsindicatethatcautionneedstobemadewhenextrapolating
resultstodifferentpopulations.

Rosenheck(2008)recentlycompletedasystematicreviewoftheliteratureexploringthe
effectsoffastfoodconsumption.Intotal,thereviewincluded16studies.Theauthor
concludedthattherewassomeevidenceofalinkbetweenfastfoodconsumptionand
theriskofweightgain,butemphasisedthatthislinkcannotbeextrapolatedtoacausal
relationship.Despitelackofcausationinthisrelationship,theauthorpointedtothe
associationbetweenfastfoodintakeandincreasedenergyintake.Theauthoralsonoted
thattherelationshipbetweentheconsumptionoffastfoodandincreasedenergyintake
ismorestronglyevidentintheliteraturethanadirectassociationbetweenfastfood
intakeandobesity.

Thereseemstobeevidenceofanassociationbetweenfastfoodintakeandweightgain
intheliterature.Itisimportanttounderstandthewayinwhichfastfoodcouldassociate
withpositiveenergybalanceandevenobesity.Despitealackofevidenceforadirect
linkbetweenfastfoodintakeandobesity,thepositiverelationshipbetweeneatingfast
food,certainsedentarybehaviours,andconsumptionofotherfattyorenergydense
foods,andthenegativeassociationwithhealthyeating,allpointtoapatternofoverall
obesogenic
13
livingwhichmayresultinweightgainand,inthelongterm,obesity.
Nutritionalandenvironmentalcharacteristicsoffastfoodshavebeenmorecommonly
citedasreasonsfortheweightgainthathasbeenseeninpeoplewhoeatfastfoods.The
SouthAustraliangovernmentenquiryintofastfoodconsumptionandobesitypointsto

13
Swinburnhasfocusedhisresearchoninvestigatingobesogenicenvironments(e.g.,Swinburn,Egger,&
Raza,1999).Anobesogenicenvironmentisonewhichpromotesobesitybyincreasingfoodconsumption
(particularlyofenergydensefoods)and/ordecreasingphysicalactivity.Beyondobesogenic
environments,theremaybecertainbehavioursthatclustertogetherthatacttopromoteobesity.


28


thecompositionoffastfoodasapotentialdriverfortherelationshipbetweenweight
gainandfastfoodconsumption(SocialDevelopmentCommittee,2007).

1.4.2 Energydensity:Howfastfoodmaypromoteoverconsumption
MultiplestudiesanalysingtheCSFIIdatacommentonthenutritionalcompositionoffast
foodandhowthismayhavedeleteriouseffectsondietaryintake(Bowmanetal.,2004;
Bowman&Vinyard,2004;Paeratakuletal.,2003).Allattempttoillustratethatfastfood
isenergydenseand,whenincludedaspartofaneverydaydiet,increasesenergy
intake.BowmanandVinyard(2004)didthisbystatingthat,whenconsumed,amealof
fastfoodprovidedmorethan33%oftypicaldailyenergyneedsandfatintakethey
alsoshowedthatfastfoodconsumptionwasrelatedtofailuretomeetseveral
nutritionalrecommendations.Bowmanetal.(2004)showedthatenergyintakewas
higherinfastfoodconsumerscomparedtononconsumersin9to19yearolds.Results
fromeachofthestudiesshowedthatthefoodisenergydenseandthatenergyintake
washigherondayswhenfastfoodwaseatencomparedtodayswhenitwasnot.

PrenticeandJebb(2003)havetriedtoverifytherelationshipbetweenfastfoodand
obesity.Theysuggestamechanisticlinkbetweenthetwo.Firstly,theynotethepositive
correlationbetweentherisingaccessibilityoffastfoodandratesofobesity.Secondly,
usingnutritionalinformationprovidedbyfastfoodrestaurants,theyshowthatfastfood
ishighinfatandthereforeenergydense.Lastly,theypositthatenergydensefoods
createpassiveoverconsumption.Thistermreferstotheideathatthehumanbiological
systemisflawedatrecognisingthepresenceofenergydensefoodsanddoesnot
compensatefortheincreasedenergyintake(whichresultsfromthefactthatfats
containalmosttwicetheenergycontentofothermacronutrients)insubsequentfood


29


intake
14
.Followingthislineofreasoning,PrenticeandJebbsuggestthatfastfood
promotesoverconsumptionwhichinthelongtermresultsinobesity.Theirassertions
areplausible,butdonotproveacausallinkage.

Otherauthorshavealsoanalysedfastfoodstoestablishhowtheirnutritionalcontent
couldbeconsideredenergydenseandthereforepromoteweightgain.Maloufand
Colagiuri(1995)investigatedthenutritionalcontentofmealsfromseveralmajorfast
foodchainstoestablishtheeffectthatdifferentamountsoffastfoodconsumptionwould
haveonrecommendeddailyintakes.Theyconcludedthatevenonemealoffastfooda
weekunfavourablyaffectsthedietbyincreasingenergyandfatintakeandlowering
fibreconsumption.Theauthorsofferednojustificationforthemealstheyconstructed
beyondthefacttheywereconstructedfromthemenusofthefastfoodchainsofinterest.
TheMcDonaldsmeal,forexampleconsistedofaBigMac,largefriesandamediumthick
shake.Thisisanunusualcombinationofitemswhichwouldhavetobepurchased
separatelyasthereisnocurrentmealthatconsistsoftheseitemsinthesesizes.A
standardBigMacwithlargefrieswouldcomewithalargeCocaCola.ACocaColawould
lessenthetotalfatintake.Furthermore,sincethispaperwaspublished,thenutritional
contentofmanyfastfooditemshaschanged.IncontrasttoMaloufandColagiurispaper,
Rice,McAllisterandDhurandhar(2007)recentlydemonstratedthataweeklydiet
composedentirelyoffastfoodcouldactuallybehealthierthanthetypicaldietintheUS.
ThisperspectivewasquicklyretortedwithStender,DyerbergandAstrup(2007)citing
dangersinherenttohighlyenergydensefoodsandlargeportionsizes.Theyalso
disagreewiththefatcontentoffastfooditemsusedinRiceetal.spaper.Stenderetal.
arguethattheconstituentsoffastfoodarenotconstantworldwideandthatinfood

14
Passive overconsumption is a term used frequently when discussing the consumption of fats. See Blundell and
MacDiarmid (1997) for a full description of the physiological drivers behind passive overconsumption.


30


itemsavailableincountriesoutsidetheUS,fastfoodremainshighfat(especiallyin
transfattyacids).

1.4.3 Mealdeals:Howfastfoodenvironmentsmayencourageconsumption
Asidefromthenutritionalcompositionoffastfoods,thewayinwhichtheyaresoldmay
alsoencourageoverconsumption.Mealdealsofferavarietyofitemspackagedtogether
forasingleprice.Typically,afastfoodmealconsistsofaburger,asideorder(usually
chipsorfries)andadrink.Giventhatpriceandconvenienceareamongstthestrongest
predictorsoffoodchoice(Prescottetal.,2002;Steptoeetal.,1995),itislikelythatmeal
dealscontributetothepopularityoffastfood.Theyhavetwocomponentsthatmay
resultinexcessiveconsumption:increasingportionsizethroughupsizingandbythe
inclusionofasugarsweetenedbeverage.

Upsizing.Someauthorshaveexploredwhethermealdealsandsupersizingareareas
wherethefoodindustryhashelpedpromoteobesity(Edwards,Engstrm,&Hartwell,
2005).Upsizingissomethingthatcharacterisesbigbrandfastfoodenvironments.This
generallydenotesincreasingthesizeofthesideordersthataccompanyaburgeror
chickenitem.ItwasthisveryaspectofMcDonaldsmealdealsthatreceivednegative
publicityinthefilmSupersizeMe
15
.Bypaying12%more,onaverage,anAustralian
consumercanhave23%moreenergyand25%morefat(CameronSmith,Bilsborough,
&Crowe,2002).Thisaloneisnotaproblemunlessconsideredinrelationtoresearch
showingthatwhenpeopleareservedmorefood,theyeatmore(Jefferyetal.,2007;
Wansink,Painter,&North,2005;Wansink&Park,2001).Theseenvironmental
pressuresonfoodconsumptionpotentiallylinkfastfoodtoexcessiveconsumption.


15
Soonafterthefilmdebuted,McDonaldsdiscontinuedtheirsupersizeoptionintheUS.Theoptionto
upsizefromsmalltomedium,ormediumtolargestillexistsgloballyinfastfoodrestaurants.


31


Softdrinks.Fastfoodmealsareusuallyaccompaniedwithasugarladenbeverage,with
manyfastfoodcompaniesdealinguniquelywithonesoftdrinkproducer
16
(forexample,
CocaColaorPepsi).Ithasbeensuggestedthatfastfoodsandassociatedbeverage
consumptionmayberesponsibleforalargeproportion(40%)oftheincreaseincaloric
sweetenerconsumedperdayintheUSoverthepast30years(Popkin&Nielsen,2003).
PapersanalysingtheCFSIIdataindicatethatrespondentswhoatefastfooddranktwice
theamountofsugarsweetenedbeveragesthanthosewhoreportednotconsumingfast
food(Bowman&Vinyard,2004).Bothstudiesindicatedthatfastfoodconsumptionwas
associatedwithincreasedsoftdrinkintake.Thisrelationshipiscauseforconcernwhen
consideredinlightoftheresearchrelatingsoftdrinkconsumptiontobodyfat(Gillis&
BarOr,2003),theincreasedriskofobesity(peradditionaldailyservingofsoftdrink;
Ludwig,Peterson,&Gortmaker,2001)andweightgain(Raben,Vasilaras,Moller,&
Astrup,2002).Ifsoftdrinkspromoteweightgainandfastfoodconsumptionencourages
softdrinkconsumption,thenthisrelationshipcouldbepotentiallyharmful.

1.5 Summaryofliterature
Thecurrentreviewoftheliteratureoneatingpatterns,fastfoodconsumptionand
obesityhasshownthattherehasbeenashiftinWesternisedeatingpatternswith
increasesinthepurchaseandconsumptionofconveniencestylefoods,predominantly
fastfoods.Thisshiftisimportantbecausetherehasbeensomesuggestionthatfastfood
consumptionrelatestoobesityaseriousandrisingglobalissue.Thepublished
literatureonfastfoodreviewedsuggestedthatconveniencewasamajordriverfor
consumptionandthatincreasingtimepressuresandchangingculturalvaluesoffood
mayunderliethisdrive.Althoughmanyresearchershaveexploredfastfoodandobesity,
evidencesuggestsfastfoodconsumptionismoredirectlyrelatedtoweightgainthan

16
Thiskindofpracticeisalsoreferredtoasmultibrandingeachfastfoodcompanyhasanagreementas
towhichcompaniesprovidetheirbeverages.


32


obesity.Severalauthorshaveproposedthatthisassociationresultsfromthetypical
natureofbigbrandfastfoods;theyarehighfatandenergydense.

1.6 Objectivesofthecurrentresearch
Thefirstobjectiveofthisresearchwillbetodeterminethenutritionalcompositionof
differentstylesoffastfoodcurrentlyavailableinAustralia.Ifmostfastfoodsareenergy
dense,thenitisindeedpossiblethatincreasedconsumptionresultsinenergyimbalance
andweightgain.Thereisalargevarietyoffastfoodsavailableinthecurrentfastfood
marketwhichmaymeanthathighenergydensityisnotapredeterminedqualityoffast
food.Itisimportanttodevelopamorecomprehensiveandrelativeunderstandingof
whetherandhoweatingfastfoodcouldresultinweightgain.Anutritionalanalysisof
typical,popularfastfoodmealsinAustraliawillattempttoachievethisobjective.

Thekeyobjectiveofthisresearchwillbetoexplorehowsocialinfluencesrelatetothe
intakeoffastfoods.Theliteratureonsocialinfluences(includingmodelling,social
normsandsocialfacilitation)oneatingbehaviourreviewedinChapterThreesuggests
thatavarietyofsocialfactorsareassociatedwithincreasedintakeatmealandsnacking
occasions.Therefore,subsequentstudiespresentedinthisdissertationwillaimto
assesshowsocialinfluencescanincreasetheintakeofbigbrandfastfoods.Thereare
limiteddataavailableaboutfastfoodconsumptionbehavioursandconsequentlythese
behaviourswillbeassessedusinganexplorativeapproach.Thisinitialapproachwillbe
usedwiththeintentionofdiscoveringwhethersocial,environmentalanddemographic
variablesareassociatedwithfastfoodconsumption.Theaimofsubsequentstudieswill
betorefinetheunderstandingofanyrelationshipbetweensocialinfluencesandfast
foodconsumption.Thiswillinclude:(1)testingwhetherexistingsocialfacilitation
models(thathavebeenprimarilydevelopedfromdataonathomeeatingoccasions)


33


applytofastfoodconsumption,(2)tailoringthemodelofsocialfacilitationtoinclude
factorspotentiallyrelevanttofastfoodconsumption,(3)assessingtheeffectsofother
socialinfluenceswithinthemodelofsocialfacilitationand(4)investigatinghowsocial
normscanalterintakeatafastfoodeatingoccasion.

Theproposedobjectivesandresearchwillcontributetopreviousresearchinseveral
ways.Firstly,despitethecommontendencyforfastfoodconsumptiontobeassociated
withtheobesityepidemic,thereislimitedresearchthatdescribesfastfood
consumptionbehaviours.Secondly,althoughthereisliteraturediscussingsocial
influencesoneatingbehaviours,thishasnotbeenappliedtofastfoodconsumption.Fast
foodconsumptionrepresentsanimportantfocusforresearchoneatingbehavioursasit
occupiesanincreasingpartofWesterniseddiets,butmoreimportantly,increasesinthe
consumptionoffastfoodsmayassociatewithweightgain.Finally,previousstudieshave
tendedtofocusonlyononeformofsocialinfluenceateatingoccasions.Thisresearch
willattempttoexploretheaffectsofmultiplesocialinfluences.


34

2 ChapterTwo:Astudyofthecurrentmacronutrientprofileoffastfoodin
Australia

Aseatingpatternschange,morepeoplethaneverbeforeareeatingfoodspreparedaway
fromhome.Fastfoodpurchasesaccountforthehighestproportionofspendingonthese
foods(AustralianBureauofStatistics,2000).Withmoreconsumerslookingforaquick
bite,thenumberoffastfoodchainshasincreasedandexistingchainshaveupdated
theirmenustomeetincreasingdemand.McDonalds,forexample,havemademany
healthychangestotheirmenu.Theseincludethecreationofthehealthychoicemenu;
shrinkingthesizeofmanyoftheiritems;changingthecompositionoftheirmeat
productstocontainmoremeatandlessfat;swappingcookingoils;andoneMcDonalds
haseventrialledsellingSpriteZerotolessenthelocalcommunitysintakeofadded
sugars(M.J.Johnson,2005).Ontheotherhand,KFCpubliclyrejectedmakingchangesto
theirmenuincludingchangingtheiroils(Burke,2007).

Inthepast,fastfoodconsumptionhasbeenassociatedwithweightgainonthebasisthat
fastfoodsareenergydenseandpromoteotherunhealthybehaviours.However,new
menusandfastfoodchainsinthecurrenteatingenvironmentofferconsumerscontinual
choicewhichmeansthatthedecisionforapersontoconsumefastfoodisnot
accompaniedwithanypredeterminedoutcomesastheycanchoosebetweenavariety
ofchainsanddifferentmenuoptions.Yet,manyresearchershavepreviouslyassumed
thatfastfoodisenergydenseandfewhaveactuallyexploredhowenergydense
differentvarietiesoffastfoodare.Giventhecurrentvarietyinfastfoods,thereislikely
tobevarietyinthenutritionalcompositionofmanyfastfoods.Forexample,Stender,
DyerbergandAstrup(2007)showedthatlevelsoftotalfatinseeminglyidenticalmeals


35


coulddifferdramaticallybetweencountries.Itisthereforeahastyassumptionthatall
fastfoodsareenergydenseanditisimportantthatresearchersknowthenatureofthe
foodtheyarediscussing.

Intheirnutritionalanalysisoffastfoods,MaloufandColagiuri(1995)showedthatan
Australianfastfoodmealcouldraiseoverallfatintakeandhavegenerallyadverse
effectsoneatingpatterns.Themarkethaschangedsodramaticallysincethepublication
ofthispaperthatitisunclearhowunhealthyfastfoodinthecurrentmarketis.
Undoubtedly,asRice,McAllisterandDhurandhar(2007)attemptedtodemonstrate,
choicescanbemadethataltertheimpactthateatingfastfoodhasonaconsumers
health.Whetherthistranslatestocurrent,Australianfastfoodsisunknown.Exploring
thenutritionalcontentofAustralianfastfoodswillhelpclarifythis.Furthermore,
identifyinghealthierfastfoodchoicesmayfacilitatethedevelopmentofuseful
nutritionaladviceforconsumers.

Theaimofthefollowinganalysiswastoexaminetheimpactoftraditionalascompared
tohealthierchoicemealsonenergyconsumptionandmacronutrientcompositionboth
betweenandwithinthemostpopularfastfoodrestaurantsinAustralia
17
.

2.1 Methods
SixofthemajorfoodchainsinAustraliawerechosenforcomparison:McDonalds,
HungryJacks,KFC,DominosPizza,RedRoosterandSubway.Thesechainsaccountfora
largeportionofthefastfoodmarketshareinAustralia.Thetopthreeofthese
companies(McDonalds,KFC&Subway)aloneaccountforover65%ofmarketsharefor
takeawaymeals(Burke,2007).TheAustralianbasedwebsitesofthecorporationswere

17
SectionsofthischapterhavebeenpublishedinObesityResearchandClinicalPractice.SeeAppendix1
forthecompleteversionofthispaper.


36


accessedandthemanufacturersnutritionalinformationwascollectedinOctoberof
2005.Theaccuracyofthedatawasnotassessed.Attheendof2005,afollowupvisitto
restaurantsintheAdelaideCBDandsurroundingsuburbswasusedtocheckwhether
menuinformationwascurrent.Wheninconsistencieswerefound,theywereresolvedby
followupphonecallstothecompanyinformationlines
18
.Twotypesofitemswere
chosenfromeachchain:amealconstructedfromtraditionalitemsandahealthier
alternative.

2.1.1 Theconstructionoftraditionalfastfoodmeals
Traditionalfastfoodmealswereconstructedtoresembleapackagedmeal,includinga
burger,friesorchipsandasugarsweetenedbeverage.Theburgerswerechosenifthey
couldbeconsideredasignatureburgersynonymouswiththefastfoodchain(e.g.,aBig
MacfromMcDonalds).Burgerswerealsoselectedonthebasisthattheywerehighly
comparablebetweenchains.DominosPizzaandSubwaydonotsellburgersandthese
weresubstitutedwiththemostpopularitemsoldfromthemenuofeachcompany;the
portionsizeswerechosentobesimilartotheweightofburgersfromtheotherchains.
Thismeantthata6ratherthan12rollfromSubwaywasincluded.TheDominosPizza
mealdealincludeditemsthatwerenotpackagedforindividuals(2pizzas,agarlicbread
and1.25litreCocaCola),sothreeslicesofpizzawerechosenasaportionbecausethis
wasroughlytheweightoftheburgeritems.Thebeveragesizewasbasedonasingle,
averagesizedserve(250millilitres)accordingtotheservinginformationonthelarger
bottle.



18
Someoftheoptionsthathadnutritioninformationavailableonthecompanywebsitesnolongersold
theseproducts.Forexample,KFCnolongerofferedMashies.Themostsignificantdifferencewasthat
RedRoosterdidnotlistanyoftheirhealthieroptionsontheirwebsiteasthesehadbeenrecently
introducedandthewebsitehadnotbeenupdated.Toobtainnutritionaldatafortheseproducts,aphone
callwasmadeandthedescriptionandaccompanyingnutritionalinformationfortheseproductswasused
todeterminethemealsincludedinthisanalysis.


37


2.1.2 Healthieralternatives
Healthieritemsweredrawnfromthosemarketedasahealthierchoice(e.g.,lowinfat)
andwereselectedtorepresentaclosealternativetotheburgeritems,thussaladswere
excluded.Theseitemsweregenerallybaguettes.Baguetteswereanalysedwithnoside
ordersbecauseatallbutonechain,theyweresoldasasingleitemmealorwitha
bottleofwater.AtMcDonaldsavarietyofchoiceswereavailablethatweremarketedas
healthyandwereconsideredreasonablereplacementsforthesignatureburger.
Althoughaleanbeefburgerrepresentstheclosestmatch,thebaguettewiththelowest
energycontentwasincludedsinceitwasthemostcomparabletothehealthieritems
selectedfromtheotherfastfoodlocations.DominosPizzaandKFChadnomarketed
healthieralternativesatthetimeofdatacollection.Thereforenohealthyalternatives
wereincludedforthesechains.

2.1.3 Nutritionalcalculations
Kilojoule(KJ)andmacronutrientinformationforbothtypesofmealfromthesixchains
wasconvertedintopercentageofdailyallowanceasdeterminedfromnutrient
guidelines.Thisallowedcomparisonbetweenchainswhilealsogivingfastfood
consumptionadailycontext.ItisworthnotingthatMcDonaldsusethisformat,which
theyrefertoasGuidelineDailyAmounts(GDAs)forgivingnutritionalinformationon
theirproducts.

ChoosingageneraldailyKJamountonwhichtobasedailymacronutrientrequirementis
difficultbecauseenergyrequirementsvaryaccordingtoindividualfactors.However,a
KJamountwaschosentoallowastandardforcomparison.NutrientReferenceValuesfor
AustraliaandNewZealand(NationalHealthandMedicalResearchCouncil,2006)
providesarangeofrecommendeddailyintakesseparatedbyagegroup,sex,sizeand


38


physicalactivitylevel(PAL).Forwomenbetweentheagesof19and39whoarefairly
sedentary(PAL1.4),themidpointdailyenergyrecommendationis8400KJ.The
midpointofenergyrequirementsforwomenintheAustralianGuidetoHealthyEating
(AGHE:Kellett,Smith,&Schmerlaib,1998)isslightlylessconservativeat9250KJ.
However,themoreconservativevaluewaschosenasitcorrespondswiththeAmerican
DailyValuesystemwhichuses2000kcal(8400KJ),anamountdeemedappropriatefor
females19to30yearsold.McDonaldsGDAforenergyisalsobasedonthisamount.

Tocalculatedailyguidelineamountsformacronutrients,percentagesofenergyfrom
macronutrientslistedintheAGHEwerereferredto.Midpointvaluesoftherangefor
adultswereused.Theresultingpercentageswere55%,18%and27%ofdailyenergy
fromcarbohydrates,proteinsandfatrespectively.TheAGHErecommendsthatnomore
thanonethirdofthetotalfatconsumedshouldbesaturated;thisvaluewasalso
incorporatedintothedailyintake.ConversionofgramamountsofmacronutrientstoKJs
usedthestandardsinCodexAlimentariuswhere1gramofcarbohydrateorproteinis
equalto17KJand1gramoffatisequalto37KJ(CodexAlimentariusCommission,2006).

2.2 Results
Figures3and4summarisemacronutrientintakesfromvariousitemchoicesateachof
thefastfoodchains.


39

38.95
60.98
46.02
53.42
38.11
33.60
51.25
38.93
45.06
38.60
31.01
30.70
37.90
29.60
32.50
31.00
21.80
48.62
93.48
69.66
68.84
50.41
47.80
71.94
100.82
61.67
76.35
63.13
64.11
35.47
52.10
36.72
51.37
49.12
37.53
47.37
0 25 50 75 100
Domino's
3 Slices of Meatosaurus (222g)
Piece of Garlic Bread (24g)
Glass of Coca-cola (250mL)
Hungry J ack's
Whopper (269g)
Regular French Fries (116g)
Regular Coca-cola (454mL)
McDonald's
Big Mac (201g)
MediumFrench Fries (105g)
MediumCoca-cola (407mL)
Red Rooster
Crispy Fillet Burger (213g)
Regular Chips (147g)
Bottle of Coca-cola (390mL)
KFC
Original Fillet Burger (211g)
Regular Seasoned Chips (133g)
Can of Pepsi (375mL)
Subway
Six Inch Chicken Fillet Sub (231g)
Mayonnaise (15mL)
M&M Cookie (37g)
Regular Coca-cola (407mL)
Percentage of estimated daily requirements
Carbohydrates
Saturated fat
Total fat
Protein
KJ without drink
Kilojoules (KJ )
Figure3:Thepercentageofdailyguidelinesofmacronutrientsaccountedforbytraditionalfastfoodmeals.


40


Figure4:Thepercentageofdailyguidelinesofmacronutrientsaccountedforbyhealthierfastfoodmeals.

19.33
17.86
24.67
16.92
27.43
24.96
30.36
26.87
13.21
11.91
6.53
8.81
18.11
7.34
4.89
7.34
18.03
18.03
31.28
17.40
0 25 50 75 100
Hungry J ack's
Spicy Chicken Baguette (233g)
McDonald's
Chicken Tandoori Deli-choices (243g)
Red Rooster
Chicken D'Lite Baguette (235g)
Subway
Six Inch Roasted Chicken Sub (223g)
Fat-free Honey Mustard (15mL)
Percentage of estimated daily requirements
Carbohydrates
Saturated fat
Total fat
Protein
Kilojoules (KJ )


41


TheaverageKJfromthetraditionalfastfoodmealwas47.5%ofthe8400KJdailyenergy
requirement.ThemealatHungryJacksprovidedthemostenergybutalsohadthe
largestportionsizes.Inthisregard,themealwhichwasconstructedforDominosPizza
mostlikelybenefitedfromthesmallersizedsideorderanddrinkandaccountedforthe
lowestKJamount.Withandwithoutabeverage,theSubwaymealrepresentedthe
lowestKJintake.

Thetraditionalfastfoodmealsaccountedforhighamountsofthedailyguidelineforfat.
TheOriginalFilletCombofromKFCwastheonlymealthatprovidedlessthan30%ofits
totalenergyfromfat(28.7%).Whopper(HungryJacks)andBigMac(McDonalds)
mealsprovidedmorethan40%oftotalenergythroughfat.Despitethefactthatthe
traditionalmealfromSubwayrepresentedthelowestpercentageofKJforameal,
analysisofitsenergyratiorevealedthatitprovided33.9%ofitsoverallenergythrough
fats.Allthetraditionalmealscontainedenoughsaturatedfattoaccountforclosetotwo
thirdsofthedailyguidelineamount.TheHungryJacksmealcontainedenough
saturatedfatforanentiredaysintake;equivalenttoalmosttwocompleteSubway
ChickenFilletSubmeals(Subway)insaturatedfat.

Replacementhealthychoiceitemsaccountedforconsiderablylessofthedailyguideline
forenergy,fatandcarbohydratesthantraditionalmealswhilegenerallyretainingthe
amountofprotein.Fatswerereducedmostnotablybythehealthychoices.Healthy
optionsaccountedfor39to80%ofthedailyamountforfatscomparedtothetraditional
meals.ThehealthyoptionatHungryJackswasagainthehighestintotalfat.However,in
comparisontotheWhopperMeal(HungryJacks),theSpicyChickenBaguette(Hungry
Jacks)alsorepresentedthelargestdecreaseintheamountofenergyandfatprovided
whencomparingtraditionalandhealthierfastfoodsmeals.Despiteofferingthelowest


42


amountsoffat,RedRoostersBaguettewasthehighestintotalKJ.Thisdifferencewas
mainlyaccountedforbytheenergycontributionofprotein.Subwayshealthieroption
representedthesmallestdecreaseofthepercentageofcarbohydrateandoverallenergy
andshowedaslightincreaseintermsofprotein.

2.3 Discussion
Theresultsshowedthatselectivityinfastfoodchainpatronagecanaffectenergyintake
by50%;allfastfoodisnotthesame.Thisamountcanbehalvedagainthroughthe
purchaseofahealthyalternativeatmostfastfoodchains.Removingasugarsweetened
beveragealsooffersaneasywaytoreduceenergyintakefromafastfoodmeal.This
findingreinforcesRiceetal.s(2007)argumentthatfastfoodcanbeeateninanon
detrimentalfashion.Awarenessoftheeffectofchoicecouldhelpconsumerswho
frequentfastfoodrestaurantsbyimprovingthehealthinessoftheitemsselected.Italso
isanindicationthatnotallfastfoodsaredirectlycomparableintermsofpotential
healthoutcomes.

Thisstudyislimitedpredominantlybythreefactors.First,itisrestrictedtothe
comparisonofonlyasmallnumberofitemsinaseaoffastfoodchoices.Theseitems
werechosentoberepresentativeofbothtraditionalandhealthierfastfoodchoices,and
aimedtocharacterisebothendsofthebroadspectrumoffastfoodsavailable.Including
furtheritemsforcomparisonwouldprimarilyincreasethedescriptivevalueofthe
comparison,whilenotchangingtheunderlyingvalueofthemessage.Second,the
nutritionalcomparisonsinthisstudywereundertakenusingtheinformationprovided
bythecompanies.Whilesomemayquestionthevalidityofsuchinformation,the
absenceofindependentlyobtainednutritionalinformationforalltheitemsmadethe
useofmanufacturersnutritionalinformationanecessity.Usingthepubliclyavailable


43


informationalsolimitedthecomparisontomacronutrientsbecauseitisdifficultto
obtaindetailssurroundingthefoodcompositionwithregardtomanykey
micronutrients.Third,dailyguidelineamountsofferonlyaroughguideformany
consumers.Theinformationisusefulforcomparingfoodchoiceandprovidinggeneral
informationonapublichealthlevel,however,directnutritionalinterventionshouldbe
tailoredtoindividuals.

ThereweresizeabledifferencesintheKJandmacronutrientcontentoffastfooditems
bothbetweenandwithinfastfoodrestaurants.Muchofthedifferencebetweenchains
wasadirectresultofportionsize.Unsurprisingly,thechainswiththebiggerburgers
providemoreofallthemacronutrients.Giventhatlargerportionsizeshavebeen
consistentlyshowntopromotefoodintake(Wansink,Painter&North,2005;Wansink&
Park,2001),thisisanimportantelementtoconsiderwhenconsideringhowfastfood
mealsmayinfluenceintake.Thesizeofthemealpurchasedcanbeeasilymodified
withintherestaurantbyorderingasmallermeal.However,thereareenvironmental
factorsinafastfoodrestaurantthatmayincreasethepressuretohavelargermeals
whichalsoneedtobeconsidered.Forexample,valuemealspromoteincreasedintake
byappealingtoindividualdesiresforagooddeal(Vermeer,Steenhuis,&Seidell,2009).

Theoccasionaltraditionalfastfoodmealasamealreplacementisnotinitselfgoingto
upsetthedailyenergybalance,especiallyifitiseateninplaceofthemainmealofthe
day.Thereisnoagreeddefinitionaboutwhatconstitutesamealbutifitissupposedthat
amealprovides30%(upto50%,ifitisamainmeal)ofdailyenergy,afastfoodmeal
couldbeincorporatedintoadaysintakewithoutcompromisingenergybalance.
Althoughfastfoodconsumptionhasbeenassociatedwithweightgain(Frenchetal.,
2000),notallpeopleconsumingfastfoodareoverweight.Giventheotherobesogenic


44


behavioursthattendtobeassociatedwithfastfoodconsumptionitmaybethateating
fastfoodisunhealthyincombinationwithotherbehaviourssuchconsumingless
vegetables(Paeratakul,Ferdinand,Champagne,Ryan&Bray,2003)andwatchingmore
television(French,Harnack&Jeffery,2000;Mohr,Wilson,Dunn,Brindal&Wittert,
2007).Thus,asOldsandHarten(2001)suggested,focussingonfastfoodaloneisa
narrowapproachtocurbingthemuchpublicisedobesityepidemic.

Ontheotherhand,itisclearfromtheresultsthatthefatcontentofthetraditionalfast
foodmealishigh.Infact,oneofthetraditionalfastfoodmealsprovidesenoughfatto
accountforanentiredaysintake(theHungryJacksWhoppermeal).Althoughnot
associateddirectlywithobesity,dietshighinbadfats,suchassaturatedandtrans
fasts,areassociatedwithothernegativehealthconsequences,includingtheincreased
riskofcardiovasculardisease(Huetal.,1997;Lichtensteinetal.,2003)andeven
dementia(Kalmijnetal.,1997).Thehealthierfastfoodchoicesarefocussedonreducing
thefatcontentofthemealsanddonotablylessenthefatprovided.Choosingahealthy
choiceisthereforeawayforconsumerstoreduceoverallfatintakefromfastfood,
furthermorereducingtheenergydensityoftheirmeals
19
.

Theconsumptionofatraditionalfastfoodmealdoesnotsolelyexplainweightgain,
althoughthereisvalidityintheargumentthattraditionalfastfoodsfrombigbrandfast
foodcompaniesareenergydense(Prentice&Jebb,2003;Stenderetal.,2007).The
cumulativeeffectoffastfoodconsumptionincombinationwithotherobesogenic
behaviours(watchingmoretelevisionandeatingotherenergydensefoods)islikelyto

19
Therearefurtherchangesthatcanbemadetotraditionalitemsthatcanreducetheamountoffatthey
contain.Forexample,assuggestedintheSouthAustraliangovernmentenquiry,changestocooking
methodshaveprovenbeneficialinreducingtheamountofsaturatedandtransfatsintraditionalmeals
(SocialDevelopmentCommittee,2007).Whilesomecompanieshavepursuedsuchchanges,includingthe
useofbetteroils(B.Smith,2006),othercompanieshavebeenmorereluctanttochangetheircooking
techniques(Burke,2007).


45


havenegativeoutcomesforhealthandweightstatus.Giventhenutrientprofileofmeals
commonlyconsumedfromquickservicerestaurants,itseemsverylikelythatincreasing
thefrequencywithwhichsuchfoodiseatencouldhavenegativehealthconsequences,
includingweightgain.

2.3.1 Summaryofanalysis
Thestudypresentedinthischaptercomparedavarietyofbigbrandfastfooditems
(bothtraditionalandhealthier)inordertogainabetterunderstandingofthenutrient
contentofthereadilyavailableandalsomostcommonlypurchaseditemsfromthemost
popularoutletsinthequickservicerestaurantsectorinAustralia.Itrevealedthatthere
werecleardifferencesbetweentheoverallenergyandfatcontentsofmealsbothwithin
andbetweendifferentfastfoodchains.However,traditionalfastfooditemswerehighin
fatandenergydense.Withinasupplyanddemandframework,thegreatervarietyof
theseitemsrelativetothelimitedhealthierchoiceswouldsuggestthatthetraditional
itemsthataremorecommonlypurchasedandconsumed.However,thereislittle
publisheddataonthis.Nevertheless,factorsthatincreasetheamountoffastfood
consumedmayleadtobothhigherenergyandfatintake.Thereforefurtherexploration
offastfoodconsumptionandfactorsthatmayberelatedtoitsconsumptionisjustified.
Thenextsectionwillreviewliteraturethatexploreshowavarietyofsocialinfluences
canencouragefoodintake.


46

3 ChapterThree:SocialinfluenceandfoodintakeAreviewoftheliterature

TheliteraturereviewedinChapterOnehighlightedtheextenttowhichtheobesity
epidemicmightbelinkedbacktotheincreasingenergydensityoftheWesterndiet.It
alsohighlightedthatfastfoodconsumptionmaybeacontributortothischange.In
ChapterTwo,nutritionalprofilesweredevelopedfordifferenttypesoffastfoodto
explorehowincreasesinfastfoodconsumptionorpoororderingdecisionscouldhave
deleterioushealthconsequencessuchasweightgain.Thisstudyshowedthatincreases
intheconsumptionoftraditionalfastfoods(i.e.,mealsconsistingofburgersandfries)
frombigbrandfastfoodcompanies(i.e.,McDonalds,BurgerKingetc.)couldrelateto
weightgainandpotentiallytoobesityinthelongterm,thusaddressingtheinitial
objectiveofthisresearchwhichwastoexplorethewayinwhichfastfoodscouldbe
contributingtotheobesityepidemic.Thesecondmajorobjectiveofthisresearchwasto
explorehowsocialinfluencesrelatetofastfoodconsumption.

Fooddecisions,suchaswhattoeat(includingfoodandvolume),howoften,whenand
withwhom,aremadewithincomplexeatingenvironments.Theseenvironmentsare
multifacetedanditisdifficulttodeterminehowdecisionsabouteatingwillbemadein
anygivencontext(Scheibehenne,Miesler&Todd,2007).TheBoundaryModelof
Overeating(Herman,Olmsted,&Polivy,1983;Herman&Polivy,1984)hasemphasised
theimportanceoftheenvironmenttoeatingdecisions,suggestingthatinthecontextof
modern,Westernfoodconsumption,eatingrarelyoccurswhenpeoplearehungry.
Accordingtothismodel,eatingcommonlyoccursinazoneofbiologicalindifference
thatmakespeoplemoresusceptibletoenvironmentalcuesandsocialnormswhich
dictatewhenandhowmuchintakeshouldoccur.Bythislogic,ifthenormsandcuesthat


47


asocialenvironmentprovidesencourageconsumptiontheseenvironmentsbecome
onesthatcouldalsopromoteoverconsumption.Therefore,socialfactorscouldbea
potentiallyimportantpredictorofexcessenergyintake.

Theresearchliteraturesuggeststhatsocialvariableshaveanimportantinfluenceon
eatingbehaviour.Whendiscussingsocialinfluences,Pfeffer(1985,p.400)writes,the
effectofothersintheindividualsenvironmentonbothattitudesandbehaviorisoneof
theoldestandmostprominentthemesintheliteratureofbothsociologyandsocial
psychology.Althoughcurrentdefinitionstendtoconsidersocialinfluencestobe
purposeful(see,Caildini,1993),inthecontextofeatingbehaviours,socialinfluence
referstobyproductsofthecomplexsocialinteractionsoccurringbetweenindividuals
inanaturalenvironment.Therehavebeenvarioustypesofsocialinfluenceidentifiedas
potentiallyalteringfoodconsumption.Theseperspectiveshavebothsharedfeatures
anduniqueones.Researchconcernedwithsocialinfluencecanbebroadlydividedinto
twomajortheoreticalperspectives:modellingandsocialnormativetheory,andsocial
facilitationtheory.Bothhavebeenlinkedtofoodintakeandprovideabasisfor
suggestingthatchangesinfastfoodintakeataneatingoccasionmightbelinkedtosocial
variables.

3.1 Modellingandsocialnorms
3.1.1 SocialModelling
Literatureusingthemodellingapproachdominatedearlyresearchintotheeffectsof
socialfactorsonfoodintake.Modellingtheoryisbasedonthepremisethatlearning
occursthroughobservingothers(Bandura,1977).Sociallearningisoperationalisedby
measuringtheextenttowhichatargetbehaviour(e.g.,havingasecondhelpingata
buffet)istakenupbyanindividualafterexposuretothedemonstrationofthat


48


behaviourinthemodeller.Becausemealsareconsumedfrequentlyinsocialsettings
(Marshall,1995)thereisstrongpotentialforsociallearningtooccur.Furthermore,
therearemanypotentialcredibleandinfluentialmodelsforeatingbehaviour:parents
(especiallyatayoungage),colleagues,friends,andevencelebrities(towhomexposure
occursthroughthemedia).

Socialmodellinghasbeenshowntobeanimportantinfluenceonarangeofhealth
behaviours,mostnotablysmoking(Leatherdale,Brown,Cameron,&McDonald,2005)
andtheconsumptionofalcohol(M.D.Wood,Read,Mitchell,&Brand,2004).Social
modellingoffoodchoicesandintakehasalsobeenexamined.Thecorefeaturesof
modellingstudiesincludetheuseofahighlycontrolledlaboratorysettingandthe
manipulationoftheindependentvariableconditions(i.e.,thesocialmodel)throughthe
useofaconfederate.Thisisusuallyaperson(ideallyblindtothehypothesesoraimsof
thestudy)whoisdirectedtobehaveinawaydesignedtopotentiallyinfluenceanothers
behaviour.Often,whenexaminingfoodconsumption,modellingstudieshaveinvolved
augmentationandinhibitionconditions.Intheformer,itishypothesisedthatthe
confederateactionswillsubsequentlyincreasethetargetbehaviour(e.g.,amountoffood
consumed)whereasinthelatter,theconfederateactionswilldecreasethebehaviourof
theparticipant.

Inanearlyexampleofthemodellingstudiesonfoodconsumption,Rosenthaland
McSweeny(1979)testednormalandoverweightstudentsreceptivitytoexternalfood
consumptioncues.TheirstudywasunderpinnedbySchachters(1971)externality
theoryofobesity.Thistheorypositsthatobesityresultsfromindividualdifferencesin
receptivenesstoexternalcuesintheenvironment.Morespecifically,itissuggestedthat
obesepeoplearelessattunedtotheirinternalsatietycuesandrelyonexternalcuesto


49


determinetheirhungerlevelsandmakefoodchoices.Itisbasedonthehypothesisthat
anobesepersonhasdifficultyregulatingtheirenergyintakeeffectivelyandismore
likelytobeguidedbyexternalvariables,includingotherpeoplesbehaviour.Usingthis
theoreticalbasis,RosenthalandMcSweeneyproposedthattheextenttowhichtherate
andvolumeofconsumptionwouldbealteredbyaconfederatesmodellingwouldvary
betweenobeseandnonobeseparticipants.Theytestedthesehypothesesintwo
experiments.Inthefirst,participantsweretoldthatthestudyinvolvedgettingtoknow
someoneoverameal.Theywerethenplacedwithaconfederatewhoateeitherquickly
orslowly.Duringthelunch,theconfederatewhoconsumedfoodatafastratewasable
toincreasesignificantlythespeedwithwhichtheparticipantate.Contraryto
predictions,thisinfluencewasmorepronouncedforthenormalweightparticipants.

InthesecondstudyRosenthalandMcSweeney(1979)investigatedtheamountoffood
beingconsumed,ratherthanrateofconsumption.Theyarguedthatthismeasurewas
likelytobeabettercorrelateofobesity.Theconfederatesateeitheralargeorsmall
numberofcrackersinfrontoftheparticipant,whoseintakeamountwasalsomonitored.
Participantintakeamongtheobesewasnotpredictedbyexposuretothemodel
althoughexposuretoamaleconfederatedidleadtoincreasedintakeacrossweight
groups.AlthoughneitherofRosenthalandMcSweeneysexperimentsprovidedsupport
forSchachters(1971)theory,theydosuggestthatfoodconsumptionrateandamount
canbeinfluencedbysocialcontextualvariables.Otherstudiesduringthistimeperiod
(e.g.,J.C.Conger,Conger,Costanzo,Wright,&Matter,1980;Miller&Ginter,1979;
Nisbett&Storms,1974)alsosupportedthefindingthatwhileconfederatescanbeused
successfullytoalterfoodintake,theeffectstheycreatearenodifferentforobeseand
nonobesesubjects.



50


deLucaandSpigelman(1979)alsousedSchachters(1971)externalitytheoryto
investigatetheeffectsofamodelonsnackintake.Theypairedobeseandnonobese
subjectswithobeseandnonobeseconfederatestoobservedifferencesinconsumption
ofdifferentcandies(M&Msandjubelollies).Inallconditionstheconfederateateten
lollies.Theresultsindicatedthatobeseparticipantsatemorelollieswhentheywere
placedinthepresenceofanobesemodelthanwheninthecompanyofanonobese
model.Thesamewastruefornonobeseparticipants(i.e.,theyatemorelollieswitha
nonobesemodelthanwithanobeseone).Althoughdifferencesinconsumptionwere
largerforobeseparticipants,theresultsofferlittlesupportforSchachterstheory,
becauseallsubjectsweresensitivetotheeffectoftheconfederateandrespondedin
equivalentwaystodifferentmodels.Thestudycanbeinterpretedasshowingthe
influenceofsocialstigmatisation;obesepeoplemayhavefeltmorecomfortableeating
lollieswithanotherobeseperson(believingtheywouldbejudgedlessseverely)and
thereforeexperienceddisinhibition.TakenwithRosenthalandMcSweeneys(1979)
results,thisfindingwouldsuggestthatsocialmodellinghasapronouncedimpacton
intake.Theresultsalsosuggestthatthetypeofmodel(obeseversusnonobeseinthis
case)presentcanaltertheeffect.

Goldman,HermanandPolivy(1991)hypothesisedthatthelevelofhungeranindividual
experiencedwouldchangetheirreceptivenesstosocialmodelling.Intheirsecond
experiment,subjectswereaskedtodeprivethemselvesoffoodforeitheralong(24
hour)orashorterperiodoftime(4hour).Thelattergroupwerealsogivenapreload
(i.e.,amilkshakeonarrival)tominimisetheirhunger.Participantsweretoldtheywere
takingpartinatastetest,andplacedinaroomwithaconfederatewhoateeitherlarge
orsmallamountsofthetargetfood.Contrarytoexpectations,bothgroupswereequally
susceptibletosocialinfluence.Furthermore,althoughthefastedparticipantsreported


51


beingsignificantlymorehungrythanthenonfastedparticipates,theydidnotavail
themselvesoftheallyoucaneatopportunity.Interpretationoftheresultwas
somewhatcompromisedbyinabilitytoconfirm24hourabstinenceinthefastedgroup
andthefailuretomeasureperceivedpalatabilityofthefoodprovidedinthe
experimentalmanipulation.

Thereissufficientempiricaldatatosuggestthattheeatingpatternsofothers(complete
strangers,inmostcases)changesintakeandthereforethateatingismalleabletosocial
influences.Themodellingliteratureislargelyexperimental,meaningthattheecological
validityisquestionable.ThemajorgapinmodellingresearchidentifiedbyHerman,Roth
andPolivy(2003)isthefailureofmostresearcherstotietheirresultsbacktotheory;
modellingstudiesappeartobemoreconcernedwithshowingtheeffectsanddifferences
insusceptibilitybetweengroupsofsubjectsthanexplainingtheirobservationofa
generalmodellingeffect.Thiscouldbedue,inpart,tothefactthatalthoughthestudies
aredescribedasmodellingstudies,whethersociallearningasdefinedbyBandura
(1977)isachievedremainsdebatable.Theoutcomesoftheabovemodellingstudiesdo
notdemonstratealeveloflearningsomuchasconformity.Theyresembletheoriginal
conformitystudiesbyAsch(1955)thatshowedthatsocialpressurescouldchange
subjectsratingoflinelength.Withinagivencontextorsituation,conformitycanbe
likenedtocompliancewherepeoplefollowothersbehaviourorexpectationstoget
awardoravoidpunishment(beingjudgednegativelyinthecurrentcase:Wiggins,
Wiggins,&VanderZanden,1994).

3.1.2 SocialNorms
Theconceptofsocialnormsandtheextenttowhichthisinfluencesanindividuals
behaviouristhefundamentalpremiseunderlyingsociologyandsocialpsychology.


52


CialdiniandTrost(1998,p.152)definesocialnormsasrulesandstandardsthatare
understoodbymembersofagroup,andthatguideand/orconstrainsocialbehaviour
withouttheforceoflaws.Theconceptofamealisonenormthatinfluencesfoodintake
behaviourandexistswithineveryculture.Thepeoplepartakingofthemealhavea
sharedappreciationofwhatfoodsareeatentogetheraswellaswhenandwhereitis
appropriatetoeatcertainfoods(Aikman&Crites,2005).

Roth,Herman,PolivyandPliner(2001)describedtwosocialnormsthatgoverneating:a
matchingnormandaminimaleatingnorm.Theformernorm,basedontheprincipleof
conformity,involvesmatchinglevelsofintaketothatofothers;thisistheresult
commonlyobservedinthemodellingliterature.Theminimaleatingnormdescribeshow
theeaterengagesinimpressionmanagementbydemonstratingcontrolovertheir
intake.Thisnormexistsinsocietieswheretheabilitytolimitconsumptionisviewedas
desirable.PlinerandChaiken(1990)havehelpedvalidatetheminimaleatingnorm.
Theyshowedthatthenumberofcrackerseaten,intheirstudy,relatednegativelyto
ratingsoffemininityofafemaleparticipant.Martin,PlinerandLee(2004)further
exploredimpressionmanagementandfoundthattheweightstatusofthepersonbeing
observedcouldchangeperceptionssurroundingeating.

ThenotionofminimalandmatchingnormsoriginatedfromastudybyRothetal.
(2001).Inamodifiedmodellingstylestudy,theysubstitutedthenormalconfederate
withanabsentconfederate;detailsofapreviousparticipantsbehaviour(including
listsofresponses)wereprovidedtotheexperimentalparticipants.Thelist(often
accidentallyleftbehind)providesparticipantswiththedetailslikethenumberof
crackerseatenbyvariousparticipants.Anadditionalconditionwasalsointroducedinto
therevisedexperimentalparadigm;theexperimentersmanipulatedwhethersubjects


53


wereobservedornot.Thismanipulationwasdesignedtotestwhetherparticipants
eatingbehaviourwasdrivenmorebythenecessitytoconformtothenormsprovided
(asrepresentedbythelist)ortheneedforimpressionmanagement(asmanipulatedby
otherpersonobservations).Theresultsindicatedthatparticipantsconformedtothe
standardssetbytheabsentconfederatesintheaugmentation(eatingmore)and
inhibition(eatingless)conditionscomparedtothenonorm(control)condition,when
eatingalone.However,whenbeingobservedbyapersonwhowasnoteating,
participantsdidnotconformtotheconstructednorms.Thisindicatedthat(a)absent
confederatescouldbeeffectivemodelsforintake;(b)normscouldbeusedto
manipulateintake;and(c)thatsocialnormscouldberejectediftheyconflictedwith
impressionmanagementneeds.

Leone,PlinerandHerman(2007)laterattemptedtoassesswhethermatchingor
minimaleatingnormswoulddominateinasituationinwhichthenormativebehaviour
wasambiguous.Usingabsentconfederates,theypresentedparticipantswithboth
minimaleatingandmatchingnormssimultaneously.Noclearadherencetoeithernorm
waswitnessed.Theysuggestedthatwithoutcleardirectioninthewayofnorms,
confusionresultedandparticipantsateasmuchastheywanted.Intheirsecondstudy,
Leoneetal.madetheconfederateinformationmoreclearandfoundthatpeoplemore
closelymatchedtheirbehaviourtotheabsentconfederate.Theauthorsconcludedby
suggestingthatinsituationsthataredesignedtoencourageintake,eatingwillbe
inhibitedbynorms.

Leoneetal.s(2007)studycanbeviewedassupportiveofthesocialnormative
interpretationofmodellinginfluencesoneating.Inaseriesofpublications,Herman,
RothandPolivy(2003)havedevelopedacomprehensiveoverviewofnormativetheory.


54


Theyproposethateatingnorms(e.g.,matchingnorms)existasaresultoftheinherent
ambiguitysurroundingwhatisanappropriateamounttoconsume.Inthishazeof
uncertainty,thewayinwhichothersbehaveisusedtoguideanddeterminefoodintake.
NormativetheoryisadevelopmentofaspectsoftheBoundaryModelofOvereating,
whichsuggeststhatintodayssocietyeatingoccursinazoneofbiologicalindifference
(Herman&Polivy,1984).

Hermanetal.(2003)describethesocialnormativemodelofeatingasessentially
inhibitorymeaningthatcessationcuesarethosewhichdictateconsumptionwhen
eatinginthepresenceofotherpeople.Hermanetal.suggestthatthereisanabundance
ofcuesintheenvironmentthatencourageintake(e.g.,availabilityofpalatablefood,
greatvarietyandeasyavailability)anditisonlywhendecidingwhentoceaseeating
thatpeoplebecomeattentivetosocialcuesandusethemtoguidebehaviour.Thereis
evidencetosupportthenotionthatcuessuchaspalatabilitycandriveandincrease
consumption.deCastro,Bellisle,DalixandPearcey(2000)reportthatpalatability
accountsforonlyasmallamountofvarianceinmealsize(<5%)infreelivinghumans
20
.
AlthoughthisfindinginitiallyappearstodisagreewithHermanetal.s(2003)
suggestion,deCastroetal.(2000)furtherwritethatalthoughpalatabilityaccountsfor
littlevariance,itremainssignificantevenwiththeinclusionofvariousfactorsincluding
thetimeofdayandthenumberofpeoplepresent.Theyalsopositthatthelowvariance
accountedforcouldbeanartefactofthefactthatgenerallyfoodsconsumedarepleasant
aspeopledonotaspiretoeatunpleasantfoods(especiallyinafreeliving
environment).Theauthorsusethefindingthatonly9.3%ofmealsrecordedindiet
diarieswereratedasunpleasanttosupportthisidea.Thus,palatabilitymightbeenough
toinitiateintakeasissuggestedbysocialnormativetheory.

20
Thisdescribespeoplewhowerenotinaclinicalsetting.Freelivingmeansparticipantsweresimply
askedtokeepadiaryoftheirintakeastheycarriedontheirlifeasnormalforthem.


55

Thesocialnormativemodelofeating(andovereating)hasattractedsomecriticism.One
involvestheverynatureofnorms;theyexistatthecommunity,group,family,situational
andevenpersonallevel.Eachsetofnormsmayacttoaffectamounteatenatanyone
time.Forexample,HermanandPolivy(2005)describesituationalandpersonal
normsthatmightguidebehaviourinspecificeatingsituations(e.g.,abuffetataparty
versuscocktailsatadinner).Theconstraintsofthesituationmaysetlimitson
judgementsofappropriateportionsizesjustastheindividualspersonalbeliefsabout
foodmightdefinenormsforappropriatebehaviour.

StudiessuchasthatbyRothetal.(2001)haveempiricallydemonstratedtheexistenceof
multiplenormsandfurtherindicatedthattheycanbeofvaryingstrengthswhilealso
beingconflicting(encouragingandinhibitingintake).Discoveringhownormsmay
mediateandmoderateeachotherpresentsachallengeforfutureresearchonnorms.For
example,theremaybesituationalfactorsthatmoderatetheeffectofsocialnorms
observedinexistingstudies.PlinerandMann(2004)showedthatfoodpreferences
couldmoderatetheeffectsofsocialnorms.Usingabsentconfederates,theirstudy
revealedthatalthoughsubjectsdidconformtotheimpliedsocialnorm(i.e.,increased
thenumberofcookieseatenintheaugmentationnorm)forpalatablecookies,theeffect
oftheimpliednormontheconsumptionofunpalatablecookieswasnotsignificant.
Usingtheseresults,PlinerandMannquestionedtheinfluenceofnormswhenexplaining
foodpreferenceandarguedthatalthoughtheamounteatencanbemanipulatedthrough
theuseofnorms,preferencesremainrelativelystable.Althoughthisappearsto
challengetheeffectivenessofutilisingsocialinfluenceasamechanismforachieving
improvementstothequalityoffoodconsumed,otherstudieshaveshownthe
importanceofmodellinginthedevelopmentofhealthyfoodpreferencesandfood


56


choices.Afteranalysingtheresultsoftheirfocusgroups,Kubik,LytleandFulkerson
(2005)reportedontheimportanceofrolemodelsinmakinghealthierfoodchoices(and
increasingthelikelihoodofbeingactive).

Regardlessofwhetheractualfoodchoicesaredrivenbysocialnorms,thereiscertainly
strongempiricalevidencetoconcludethattheamounteatenis,atleastpartially,
influencedbytheperceivedactionsofothers,suggestingaroleforsocialinfluencein
overeating.Socialnormsmayhaveastronginfluenceoneatingbehaviourandarean
importantconsiderationinresearchofeatingbehaviours.

3.1.3 Normativeovereating
Herman,PolivyandLeone(2006)haveextendedtheirsocialnormresearchtopropose
thenormativemodelofovereating.Thenormativemodelofovereatingisdesignedto
accountfortheincreasingprevalenceofobesityandincorporatesmultipleformsof
socialinfluenceaswellassomeothercrucialenvironmentalfactors.Itpositsthatcertain
normscanbeheldresponsibleforpermittingandencouragingovereating.Accordingto
thistheoreticalperspective,TypeIovereatingarisesinresponsetoenvironmental
pressure.Theauthorscitethetoxicenvironmentinwhichweexistandhowthis
providescuesforovereating.Theymentiontwofactorsthatcreatethisdangerous
environment.Thefirstfactorisportionsize.Thereisalotofevidenceshowingthat
thereisatendencyforindividualstoeatwhatisputinfrontthem.Evenifitisnotall
eaten,simplyhavingalargerportionsizereliablyincreasesintake(e.g.,Wansink&Park,
2001).Thefactthatincreasingportionsizesisaconstituentofchangingeatingpatterns
makesthetoxicityofportionsizeevenmoreimportantwithincontemporaryeating
environments.AnotherfactorwhichtheauthorsbelievehelpcreatesTypeIovereating
issocialinfluence.Theynotetheresultsfromsocialfacilitationtheories(discussed


57


below)whichhaveshownthateatingwithotherspromotesgreaterfoodintake.The
modelalsoconsidersthefindingsfrommodellingandnormativeapproaches
21
.
Thereforethemodelconsiderssocialinfluencesonintakewhilealsoaccountingfor
environmentalfactors.

EssentiallyTypeIIovereating(disinhibtion)isseenwhenotherwiserestrainedeaters
losetheirrestraint.Ratherthanrevertingtoconsumingnormalamountsfollowingthe
lossofrestraint,theseeatersloseallrestraintandovereat.TypeIIovereatingisnotas
widelywitnessedasTypeI.ItissuggestedthatTypeIIovereatingaffectsonlydieters.
Howandwhenthisstyleofovereatingoccursisunclear.Hermanetal.(2006)suggest
thatdistress,cravingsandeatinganyforbiddenitemsmayruinthediet,intheeyesof
thedieter,resultingindisinhibitedeating.Thiseffectmayalsobetheresultofego
depletion(Kahan,Polivy,&Herman,2003).Accordingtothisnotion,dietersuseahigh
leveloftheiregostrengthtosticktotheirdietsanditiseasyfortheiregotobecome
depletedandforthemtofail.Thiscouldhavelongtermconsequences;failureasaself
perceiveddietercouldhavenegativeconsequencesforselfperceptionandpromote
furtherfailure.Weightcycling(dieting/bingeing/dieting)mayhavenegative
implicationsforbothmentalandphysicalhealth(U.S.DepartmentofHealthandHuman
Services,2008).

3.2 SocialFacilitation
Uziel(2007)recentlydescribedsocialfacilitationasthemostbasiceffectofoursocial
environmentonbehavior:theeffectofthemerepresenceofothersnearus(p.593).
Allport(1920)wasthefirsttodefinesocialfacilitation,describingitasthewayinwhich

21
Althoughintheirtheoryofnormativeovereating,Hermanetal.(2003)separatesmodellingandsocial
normtheoriesintodistinctapproaches,whethertheyareclearlydivergentisquestionable.Socialnorm
theorycouldbeviewedsimplyasaniterationoftheearlymodellingstudies.Itrepresentsanevolutionin
thetheoreticalgrounding,terminologyandstudydesignofthesocialmodellingresearch.


58


thepresenceofotherscanalterperformanceonnonsocialtasks.Zajonc(1965)later
theorisedthatthiseffectwastheresultofincreaseddrivewithinanindividualwhen
doingtasksinthepresenceofothers.Hefurthersuggestedthatthealterationindrive
wouldacttoimproveperformanceonsimpletasksandinhibititduringmore
demandingones.Areviewofthesocialfacilitationliteraturesuggestedthatthespeedat
whichasimpletaskcanbecompletedcanimproveminimallyinthepresenceofothers
(Bond&Titus,1983).PublicationsfollowingZajoncshavesuggestedavarietyofways
inwhichsocialfacilitationmaychangeperformance:itmayinfluencelevelsofself
awareness,selfmonitoringandattention/distraction.Socialfacilitationresearchhas
exploredhowthepresenceofotherscanaltercognitive(Klauer,Herfordt,&Voss,2008)
andsportingperformance(Carron,Burke,&Prapavessis,2004)andbehaviourssuchas
shopping(Sommer,Wynes,&Brinkley,1992).Althoughsocialfacilitationresearchhas
accountedforonlylowlevelsofvarianceinsomebehaviours(Bond&Titus,1983),the
effecthasbeenshownclearlyoneatingbehaviours(J.M.deCastro,1990).

3.2.1 Socialfacilitationandeatingbehaviours
Socialfacilitationresearchutilisesadistinctresearchparadigmthatdiffersfromthat
usedinmodelling/socialnormsresearch.Thelatterischaracterisedbyalargely
artificialeatingsituation:foodtypesareconstrained,peopleeatwithstrangersandtheir
behaviourismonitored.Intherealworld,thetypicaleatingsituationisvastlydifferent
and,asScheinebenneetal.s(2007)modeloffoodcognitionhasestablished,realworld
situationscanchangethepreferencesfor,andthedriversof,foodchoice.Mieselman
(1992)hascriticisedstudiesofsocialinfluenceforbeingtooartificialandhaspointedto
theneedtobalancethiskindofexperimentationwithnaturalisticobservations.Itisthis
approachthatsocialfacilitationresearchofeatingbehaviourattemptstoutilise.



59


Socialfacilitationresearchoneatingbehavioursprimarilyutilisesdietdiarytechniques.
Thedietdiaryrequiresparticipantstorecordeatingbehaviours,includingwhatwas
eaten,theenvironmentwhereitwaseatenandhowlongwasspenteating(andany
otherfactorsrelevanttoindividualstudies,suchasmood)overafixednumberofdays.
Thismethodisdesignedtoallowformorenaturalmeasurementoffoodintakewhile
alsoobtainingapictureofgeneraldietarybehavioursandinfluencesonit
22
.

deCastrohasutilisedthedietdiarytechnique,arguingthatthismethodofdata
collectionallowsexperimenterstomakeunobtrusiveandaccurateobservationsoffree
livingpeople(J.M.deCastro,2000).Howevernoformofdietarymeasureiswithout
criticism.Dietdiaries,foodfrequencyquestionnairesanddifferentformsoffoodrecall
allhaveadvantagesanddisadvantages(see,Binghametal.,1988;Bonifacj,Gerber,Scali,
&Daures,1997;R.K.Johnson,2002).Almostallmethodssufferfromunderreportingof
foodintake;peopletendtoforgetitemssuchascondimentswhendetailingthefoods
theyhaveeatenandmayrespondinsociallydesirableways,underestimatingbadfoods
andoverestimatinggoodfoods.Amorefundamentalflawwiththedietdiarymethodis
thatitmayalsoalterpeoplesbehaviourastheyaremorecloselymonitoringtheirfood
intake(J.M.deCastro,2000).Threatstothereliabilityandvalidityoffooddiarieshave
beentackledwithvaryingdegreesofsuccess.Forexample,thecommontendencyto
underreportintakehasbeenaddressedbyrequiringparticipantstogettheirintakes
substantiatedbypeoplewhowerepresentduringtheeatingoccasion.Morerecently,
participantshavealsobeenrequiredtotakephotosofthefoodtheyhaveconsumed
(see,J.M.deCastro,2000).Althoughthedietdiarytechniqueremainsimperfect,its
applicationhasincreasedresearchintotheeffectofsocialinfluenceoneatingin
naturalisticsettings.

22
Oncethedataarecollected,itcanbeusedtolookatspecificfoodintakesorspecificmeals.


60

Inoneoftheearlysocialfacilitationstudies,deCastroanddeCastro(1989)usedthe
dietaryinformationinthediariestocalculatetheamountofdietaryenergy
23
consumed
onvariousoccasionsthroughoutasevendayperiod.Thisinformationwasthen
separatedintomealsusingtheamountofdietaryenergyconsumedandlengthofa
reportedeatingoccasion.Theyfoundthatthenumberofotherpeoplepresentatan
eatingoccasionsignificantlypredictedtheamounteatenatmeals.Infact,includingthe
numberofpeoplepresentdoubledthetotalamountofvarianceoftheamounteaten
explainedintheirmodel.Later,deCastro(1990)reportedthateatinginthepresenceof
otherpeoplewasassociatedwiththeamounteatenandthetimetheyspenteating.
Usingdietdiariesandcorrelationalanalysesinaseparatesample,deCastroandBrewer
(1992)suggestedthatapositivelinearrelationshipexistedbetweenamountconsumed
duringamealandthenumberofpeoplepresent.Theyreportedthatthepresenceofjust
asingleotherpersoncouldincreasetheamounteatenby28%andthatmealswereover
75%largerwheneatingwithotherscomparedtoeatingalone.

3.2.2 Thetimeextensionhypothesis
Howsocialfacilitationworksinataskassimpleaseatingisstillamatterofdebate.de
Castros(1990;1994)initialsuggestionwasthatthepresenceofothersincreasesthe
timespenteatingandthereforealsotheamountconsumedamodelhenamedthe
timeextensionhypothesis.Socialfacilitationresearchinotherfieldshasshownthat,for
simpletasks,therateatwhichtheyareperformedincreasesinasocialsetting.deCastro
(1990)arguedthattherateofeatingmaydrivetheeffectsofsocialfacilitationinhis
earlyfindingsalthoughhesubsequentlyconcludedthatmealdurationwasmore
important.

23
Asmeasuredinunitsofkilocalories(kcal)orkilojoules(KJ).Thecolloquialuseofthetermcalories
referstokcal(1kcal=4.184KJ).


61

Feunekes,deGraafandvanStaveren(1995)presentedamoredetailedversionofthe
timeextensionhypothesisinanefforttounderstandwhatotherfactorscouldincrease
durationand/orintake.InlinewithdeCastroshypothesis,theysuggestedthatthe
durationofamealmediatessocialfacilitation.Usingpathanalysistheyattemptedto
developamoreexplanatorymodelofsocialfacilitation.Althoughdirectrelationships
betweenthenumberofpeoplepresentandtheamounteatenwerelargelynon
significant,theyfoundamediatingeffectoftheamountoftimespenteating;the
presenceofothersdirectlyincreasedthedurationofamealwhichthenincreasedintake.
Theirmodelalsoindicatedthatthepresenceofotherscouldimproveatmospherewhich,
inturn,increasedthemealdurationandhencethemealsize(seeFigure5).

Feunekesetal.(1995)proposalismoreexplanatorythandeCastrosoriginaltime
extensionhypothesis.Itisonlyafterconsideringanddismissingavarietyofother
variables(e.g.,hungerandavailability),throughstatisticalanalyses,thatFeunekeset
al.smodelofsocialfacilitationwascreated.Otherresearchershaveprovidedfurther
supportfortheideathatmealsizeisaffectedbymealduration.Weber,Kingand
Atmosphere
No. Others
Time Eating
Food Intake
.45
.32
.18
.31
Figure5:Feunekesetal.s(1995)testedmodelofsocialfacilitation.
NB.Numbersnexttoarrowsrepresentpath/betavalues


62


Meiselman(2004)citedastudybyPliner,Bell,KinchlaandHirsch(2003)showingthat
extendingthedurationofamealincreasedintakeregardlessofthenumberofpeople
present(inthiscase1,2or4people).Thisresearchnotonlycorroboratedthetime
extensionhypothesisbutalsoshowedthatitcanbewitnessedandmanipulatedina
laboratorysetting.

Whileinvestigatingtheeffectsofsocialfacilitationintheelderly,Mathey,Zandstra,de
GraafandvanStaveren(2000)foundthatthetimeavailabletoeatamealcouldbe
extendedwithoutincreasingintake;anobservationcontrarytothetimeextension
notion.Theymanipulatedsocialsettingbyputtingsubjectsineitheracozysetting
(foodeateninagroup,talkingallowed,atmospheredesignedtobedesirable)oranon
cozyone(foodeateningroupsbutindividualsisolatedandnotalkingpermitted).As
wellasalteringthesocialsetting,theresearchersusedpreloads
24
toassesstheroleof
physiologicalreactionstoconsumption.Theyfoundthatthepreloadwasmore
influentialthansocialsettingonlevelsofconsumption.Participantsatelessaftera
carbohydraterichpreload,butnodifferenceinoverallintake,ineachsetting,wasfound.
Thus,withtheadditionofthiscarbohydraterichpreloadinthegroupsetting,themeal
waslongerbutnotlarger.Giventhedesignofthestudy,itisdifficulttodeterminewhich
factorscouldhavebeeninstrumentalincreatingthisexceptiontotimeextensionlike
theories.Forexample,itcouldbetheelderlysample,thelaboratoryenvironment,the
preloadoracombinationofthesefactorsthatstoppedtheincreaseinconsumption
despiteprolongingmealtime.

TheFrenchParadoxalsohighlightspotentialproblemswiththetimeextension
hypothesis.ThisparadoxisbasedontheobservationthatFrenchpeopledemonstrate

24
Apreloadisfoodordrinkthatisgiventotheparticipantbeforetheexperiment,proper,inorderto
decreasehunger.Often,preloadsaregivenintheformofaflavouredmilkshake.


63


lowerbodyweightandriskofheartdiseasethanthepeoplefromtheUSeventhough
theyhaveadietthatispurportedlyhighinfat.Rozin,Kabnick,Pete,FischlerandSheilds
(2003)havesuggestedthatsmallerportionsizesexplainthesedifferences.They
observedthatportionsizesinMcDonaldsinFranceweresmallerthanintheUSandyet
Frenchpeoplespentlongereating.Inastudyofdifferencesbetweendietarypatternsfor
NorthAmerican,FrenchandDutchpeople,fewdifferenceswereobservedintheamount
eaten,timespenteatingandtherateofeatingbetweentheFrenchandtheNorth
Americanparticipants(J.M.deCastro,Bellisle,Feunekes,Dalix,&deGraaf,1997).
Interestingly,theDutchparticipantsatesmaller,morefrequentmealsbutatemore
slowlyandspentlongereatingthem.Theseresultshighlightthepotentialrelevanceof
cultureasamediatorormoderatorofsocialfacilitationandeating.Theideathatsome
culturesmayeatmoreslowlyandconsumelessfoodisconsistentwithMatheyetal.s
(2000)findingsthattimecanbeextendedwithoutincreasingtheamounteaten.

BellisleandDalix(2001)examinedhowlevelsofdistractionwhileeatinginfluencedthe
amountconsumed.Threeformsofdistractionwerecomparedintheirstudy:anaudio
tapepromptingpeopletothinkaboutfood,atelevisionplaying,andprovisionofagroup
environmentforeating.Thefoodserveddidnotvary.Theresultsindicatedthat
participantsdescribedthefoodasbettertastingwhenconsumedwithothers.The
amounteatendidnotchangesignificantlyfrombaselineunderanyoftheconditions.
Therearetwopossibleexplanationsforwhysocialfacilitationwasnotdemonstratedin
thisresearch.ThestudywasconductedinFranceandmaythereforebeconfoundedby
differenteatingstyles.Furthermore,inallconditions,peoplereportedeatingfortime
periodsthatwereatleast30minutes.Thisisconsiderablylongerthantheaverage12
minutesspenteatingamealpreviouslyreportedinNorthAmericanstudies(Pliner,Bell,
Hirsch,&Kinchla,2006).Giventhatdurationhasrepeatedlybeenshowntoaffectsocial


64


facilitation,ceilingeffectsattachedtotheoverlylongmealdurationmayhaveswamped
socialfacilitationeffects.WithoutknowingatypicaleatingdurationinFrance,itis
difficulttodeterminethis.

3.2.3 Beyondduration
Althoughtherehasbeensomedebateabouttheeffectofdurationoneating,thereare
manypapersthatconfirmdeCastrostimeextensionhypothesis.Thesepapers,
however,offerlittleinsightintotheunderlyingmechanismsbehindthetimeextension
hypothesis.StroebleanddeCastro(2004)suggestedthatthepresenceofothersmay
increasearousal,whichinturnincreasesintake.ThishypothesisconcurswithZajoncs
drivetheory(1965).StroebleanddeCastroincorporateddietdiarieswithdatafrom
heartratemonitorsintheirstudy.Unlikethepreviousresearchusingdietdiaries,this
studyalsoincorporatedeatingoccasionsatrestaurants.Theresearchersfoundthat
arousal(measuredthroughheartrate)didnotsignificantlychangeindifferentsocial
companyandsettings.Participantsdidreportdifferentlevelsofemotionsanddifferent
intakesaccordingtotheeatingenvironment.Again,themorepeoplepresent,thelonger
andthelargerthesizeofthemeal.Theauthorsextendedthisfindingbyshowingthat,
whenwithothers,peoplehadhappiermeals.Furthermore,whenpeoplerated
themselvesasmoreexcited,theyatemorefoodandhigheramountsoffat.Finally,the
diariesshowedthateatingamealinarestaurantincreasedintake.Stroebleandde
Castrousedtheirresultstoofferfurthersupportforthesocialfacilitationtheoryof
eatingbutaddedthatthelocationandatmosphereoftheeatingoccasionmayalsoalter
intake.ThisisconsistentwithFeunekesetal.s(1995)modelofthetimeextension
hypothesis,whichincorporatesatmosphereasanimportantdriveroftimeextension.



65


LumengandHillmen(2007)alsousedarousaltheoryasanexplanationfortheeffectsof
socialfacilitationwheneating.Theycovertlyrecordedchildreneatinginalaboratory
environmentdesignedtoresembleanaturaleatingoccasion.Theauthorsfoundthatas
groupsbecamelarger,thetimespenteatingdidnotincreasebutthatthechildrensrate
ofeatingincreased,whichthenactedtoincreasetheiroverallintake.Theyalsoreported
that,asthenumberofmembersinthegroupincreased,thenumberofsocial
interactionsdecreased.Theseresultsdidnotcompletelydispelsocialfacilitationtheory,
asthosechildrenwhoateforlongerdideatmorefood;itwassimplythatthepresence
ofothersdidnotincreasethedurationoftheeatingoccasion.Itispossiblethatchildren
haveadifferentapproachtoeatinganddosowithgreaterlevelsofrivalry(especiallyin
thepresenceoftastysnackfoods).Feelingsofcompetitionwouldexplainwhy,asthe
numberofothersincreased,therateofeatingalsoincreased.

ReddanddeCastro(1992)attemptedtomanipulatethesocialconditionsinwhich
peopleatetogainfurtherinsightintohowsocialfacilitationmightimpactinthereal
world.Overthreeseparatefivedayperiods,participantswereaskedtoeatalltheir
mealsalone,withothers,ornormally.Theyrecordedtheirfoodintakes,thelengthof
theireatingoccasionsandthecompanypresentindietdiariesovereachofthese
periods.Therewerefewdifferencesobservedbetweenthenormaleatingandeating
exclusivelywithothersconditions.Theauthorsnotedthatthiswaslikelytobethe
resultofthefactthatnormaleatingwasofteninthepresenceofothers.Thestudy
confirmedprevioussocialfacilitationresults;inthenormalcondition,wheneatingin
groups,peopleatemoreandateforlongerthanwhentheywerealone.Whenpeople
wereforcedtoeatalone,thecompositionofwhattheyatechanged;participants
reportedeatingfewercaloriesfromfatthanconsumedwheneatingnormally.Results
fromStroebleanddeCastros(2004)studymayexplainwhyparticipantsatelessfatty


66


foodswheneatingalonecomparedtoeatingnormally.StroebleanddeCastroobserved
thathigherfeelingsofexcitementwereassociatedwithhighfatintake.InReddandde
Castrosstudy,theoppositeeffectmayhavebeenoccurring;eatingalonewasunexciting
andresultedinlowerfatintake.ThisspeculationissupportedbyReddanddeCastros
owndatawhichindicatedthatwhensubjectsatealonetheyreportedgreaterlevelsof
depression.

Fewstudieshaveattemptedtountanglewhetheritisthetimespenteatingorgroupsize
thatisthemajordriverofsocialfacilitation.Pliner,Bell,HirschandKinchla(2006)
conductedanexperimentalstudythataimedtotestthetimeextensionhypothesisby
separatingtheeffectsofgroupsizeversusmealduration.Plineretal.assignedgroupsof
menorwomentooneofsixexperimentalconditionsaccordingtodifferenteating
lengths(normal12minutes;long36minutes)andgroupsizes(alone,groupoftwo
orgroupoffour).Byattemptingtodisruptnaturalgroupformationandmealdurations,
theyaimedtoassesswhichhadagreatereffectonintake.Duringandafterthe
experimentparticipantswereofferedaplateofpizzaandabowlofcookies.Results
fromthestudyindicatedthatpeopleatemorewhengivenlongerperiodsoftimetoeat
(36minutes)butnotwheninlargergroups
25
.Theauthorssuggestthattheseresults
supportdeCastros(1990)suggestionthatsocialfacilitationeffectsaremediatedby
mealdurationastherewaslittleevidenceofadirecteffectofgroupsizeandintake.
Theyalsosuggestthateatingforlongermaynotbedeterminedbythepeoplepresent
butbysimplyhavingprolongedaccesstofood(whilebeingpromotedtoeatby
environmentalcues).


25
Inoneoftheanalyses,lonedinersatethemostcookieswhichiscontrarytoothersocialfacilitation
findingsandmaybereminiscentofadisinhibitioneffect.


67


Plineretal.(2006)alsonotedsomeevidenceofmatchingintheirresultspeoplein
pairsateamountsthatwerethemostsimilar.Plineretal.suggestedthatthesizeofthe
groupofeatersmayaffectintakedifferentlybecausethenumberofpeoplepresentwill
altertheinfluencethateachindividualhasindictatingthenorm.Inotherwords,ina
freeeatingenvironment,withonlytwopeople,matchingnormativebehaviouriseasy
becausethebehaviourofothersisclearlydiscernible.Asmorepeopleareaddedtothe
group,itbecomesmoredifficulttodeterminewhatothersareeating,andtherefore
moredifficulttodefineappropriateintake.

Authorsusingvaryingmethodshaveattemptedtoexplorealternativefactorsthatmay
underliesocialfacilitation.Regardlessofthesedifferentapproaches,mealdurationhas
repeatedlybeenshowntobeanimportantmediatoroftheeffect.Therefore,itislikely
that,asdeCastro(1990)suggestedinhisoriginaltimeextensionhypothesis,eating
withotherscanacttoprolongthetimespenteatingandconsequentlytheamountof
foodconsumed.

3.3 Relationships:thepotentialmoderatorofsocialinfluence
Eagly(1983)arguedthatrolesdictatehowtobehaveandareaccompaniedbyaseriesof
expectations.Forexample,beingaprofessionalinaresearchenvironmentcomeswith
differentexpectationstobeingamusicianinapopularband.Earlyresearchfocused
primarilyonthenumbersofpeoplepresentinattemptingtounderstandsocial
facilitation.Itislikelythatwhothecoeateristotheeaterwillalsomoderateor
mediatetheeffectonconsumption.Iftherolethatisbeingsatisfiedinasocialsituation
canchange,thenitwouldbeexpectedthatdifferentrolesmaychangehowsocial
influenceoccurs;relationshipswithcertainpeoplemaypromoteovereatingmorethan
others.Avarietyofstudieshaveshownthatthesocialrelationshipsharedwiththe


68


personwhoproduces,orfacilitates,socialinfluencecanalterhowthisinfluenceis
witnessed(e.g.,Clendenen,Herman,&Polivy,1994;J.M.deCastro,1994).All
relationshipsinvolvenorms,andthesenormsmaypromotehealthiereatinginsome
casesandunhealthiereatinginothers.Thus,inordertounderstandthemechanisms
thatdrivesocialinfluencesonconsumptionitisimportanttodeterminewhetherthe
rolesofpeopleinthesocialsituationimpactdifferentiallyoneatingbehaviour.

3.3.1 Sourcesofinfluence
VerleghandCandell(1999)proposedthattherearebothprimaryandsecondary
referencegroupsthatinfluencefoodchoices.Theprimaryreferencegroupincludes
peoplewithwhomemotionaltiesareshared.Thesearepeoplewhoarecloseandhave
frequentcontactwiththetargetperson.Researchhasshownthatfamiliesinfluenceboth
fatintake(Hertzler&Frary,1996)andbreakfastbehaviours(Billonetal.,2002).
ZimmermanandConnor(1989)conductedahealthpromotioncampaignandnotedthat
fatconsumptionwasinfluencedbyfamily,friendsandcoworkers.Theyalsosuggest
thatfamilymembersarecrucialfordetermininghowsuccessfullypeoplemaintain
dietarybehaviourchange.AccordingtoVerleghandCandell,secondaryreference
groupsincludefriends,institutionsandthemassmedia.VerleghandCandellhave
shownthatdespitetheclosenessoftheprimarygroup,situationalvariablesmeanthat
thesecondaryreferencegroupcanhaveastronginfluenceoneatingbehaviour.

Inastudyusingamixtureofqualitativeandquantitativeresearch,Feunekes,deGraaf,
MeyboomandvanStaveren(1998)surveyed361socialnetworks(consistingofparents,
child,andbestfriendofparentorchild)andinterviewedtenfamilies.Theyconcluded
thatfamilyrulestendtoinfluenceoverallfoodintakeandthatconversationsaboutfood
occurbetweenfamilymembersmorefrequentlythanamongstfriends,whichsupports


69


theideaofthepervasivenessoftheprimaryreferencegroup.Althoughitappearsthat
friendshavelittleinfluenceoffoodchoice,Feunekesetal.foundthattheconsumptionof
snackfoodswasassociatedwiththefoodintakeofbestfriends.Thus,whensnacking,
friendsmaybeconsideredanimportantreferencegroupandpossesstheabilityto
changeconsumptionoratleastdefinethenormforsnackfoodintake.Furthermore,it
couldsuggestthatcertainfoodsareassociatedmorecloselywithcertaintypesof
interpersonalrelationships.

McIntosh,Fletcher,KubenaandLandmann(1995)investigatedthetypeofpeoplewho
influencefoodchoicesintheelderly.AfreeresponseformatallowedMcIntoshetal.to
identifythesocialsourcesconsideredimportantinthedecisiontochangeaneating
behaviour(e.g.,reduceredmeatintake,inthiscase).Theparticipantsindicatedthat
bothprimaryreferencegroupmembersandsecondarygroups(includingthemass
media)influencedtheirdecisions.Amajorityoftheelderlysubjectsreportedbeing
influencedmoststronglybythemassmedia
26
.

Althoughtheexactnatureoftheinfluencethatthemassmediahasoneatingbehaviours
isdebatable,itisoftencitedasanimportantsourceofinfluence.Forexample,theNSW
ReportonObesity(2003)concludedthatadvertisingmaybepartlyresponsibleforthe
promotionofunhealthybehaviours.Manyresearchershavereportedthatthe
advertisingofunhealthyfoodsdominatesviewingtimes(Batada,Seitz,Wootan,&Story,
2008;Chapman,Nicholas,&Supramaniam,2006;Galcheva,Iotova,&Stratev,2008)and
thenegativeeffectsthattheseadvertisementscanhaveonchildren(Kelly,King,
Bauman,Smith,&Flood,2007;Marshall,O'Donohoe,&Kline,2007;Morleyetal.,2008).

26
Thisstudyhighlightsoneofthedifficultieswithdeterminingwhoisanimportantsocialinfluence.Itis
achallengingtaskforpeopletoidentifywhoinfluencestheirbehaviour.Thepeopleclosesttothem(the
primaryreferencegroup)arelikelytoaffectbehaviourthemost,butmayalsodosoinawaywhichisless
visiblethaninputfromexternalinfluences(suchasthemassmedia).Thismaymeanthatpeopleareless
likelytospontaneouslycitefamilyandpartnersasinfluencingtheirbehaviour.


70


Themechanismbywhichthisoccursmaybeimplicitlylinkedtosocialnormative
theoriesinwhichthemassmediacreatesaconsumptionnormthatinfluencesfood
choicesfortheworse.McIntoshetal.s(1995)studywiththeelderlyranksthemass
mediaasoneofthemostpowerfulformsofsocialinfluence.Exactlywherethemediafits
inasocialnetworkandwhenitisusedasareferenceforfoodintake,islessclear.
Furthermore,giventhewidereachingnatureofthemassmedia,itisdifficulttoquantify
anyimpactitmayhaveoneatingbehaviour.

3.3.2 Researchonrelationshipsandsocialinfluence
deCastro(1994)observedthatdifferentsociallydefinedrelationshipscanaffectsocial
facilitation.Onthebasisofdietdiarydatafromalargesample(over500people),he
reportedthatallmealseatenwithothers,regardlessoftheirrelationshiptothe
participant,tendedtobelongerandlarger.Healsofoundthatwhentheotherwasa
spouseorfamilymember,theamountwasstilllargerbuttherateofconsumptionwas
faster,whereaswhentheotherwasafriendorcolleague,largermealswereeatenmore
slowly.deCastroarguesthattheresultssupporthistimeextensionhypothesisand,in
addition,socialdisinhibitionassociatedwithsharedeatingwithfamiliarpeopleresults
inincreasestotheamountconsumed.Thisisaresultwhichisconsistentwiththeresults
ofDubeetal.(2007),whoarguedthatcertaintypesofsocialinteractioncanincrease
intake.Moregenerally,thedifferencesobservedinintakeindeCastrosstudycould
reflectthedifferentpurposesthataccompanyeacheatingoccasion.

Clendenen,HermanandPolivy(1994)usedanenvironmentalsituationdesignedto
imitateacasualfilmnighttofurtherexaminesocialinfluence.Thedesigninvolvedthe
manipulationofgroupsize(alone,pairsorfours),relationshipofmembersinthegroup
(selfnominatedfriendsorstrangers),andthefoodoffered(coldmeatsorcookies).


71


Althoughtheyfoundthatparticipantswhoatethemostdidspendlongereating,theydid
notfindalinearrelationshipbetweenthenumberofpeoplepresentandamountoffood
consumed;individualsingroupsoffourdidnoteatsignificantlymorethanindividuals
inpairs.Participantsingroupsoffriendsatemorethanthoseinagroupofstrangers,
andgroupsoffriendsconsumedsignificantlymorecookieswheneatingtogether.

Thefamiliarityhypothesis,basedonthetheoryofimpressionmanagement,positsthat
thefamiliaritybetweenpeoplemoderatesthedrivetopresentpositivelyandtherefore
theextenttowhichmealtimebehaviourisinfluenced.Inotherwords,whenpeopleeat
withfamiliarothersthereislessdesiretoimpressorprojectacertainimage(Salvy,
Jarrin,Paluch,Irfan,&Pliner,2007).Supportforthishypothesiscanbefoundina
numberofstudies.Tice,Butler,MuravenandStillwell(1995)showedthattheir
participantshaddifferentstylesofselfpresentationwithfriendsthanwithstrangers
andthatasfamiliarityincreased,behaviourchanged.Similarly,Learyetal.(1994)found
lessneedtoimpressinsamesexgroupsoffamiliarpeople.

Hetherington,Anderson,NortonandNewson(2006)exploredhowbothfamiliarityand
distractioncouldinfluencetheeffectofsocialfacilitation.Inalaboratorystudy,they
manipulatedthelevelofdistractionandthefamiliarityoftheparticipants.AsinBellisle
andDalixs(2001)study(describedearlierinSection3.2.2,page68),therewereseveral
conditionsmanipulatingthelevelofdistraction,includingthepresenceofothersand/or
atelevision.Theresearchersmeasuredtheactualtimespenteating,definedas
discernibleeatingbehaviourratherthanthetimespentatthemeal(aswasrecordedby
Clendenenetal.,1994).Thiswassomethingthatcouldnotbedeterminedindietdiary
studieswherethetimespenteatingisbasedpurelyonthetimethatparticipantsbegan
andceasedeating.InHetheringtonetal.sstudy,whenpeopleatealone,theyatethe


72


least,regardlessofthepresenceofthetelevision.Eatingwithothersincreasedenergy
intake18%abovebaselineandextendedthetimespenteating.Theresearchersnoted
thatupto40%ofthetimespentatamealwithotherswasactuallyspenttalkingwhile
theremainingtimewasspentconsumingfoodordrinkitems.Eatingwithstrangersalso
resultedinahighlevelofdistractionfromthefoodbutdidnotsignificantlyincrease
intake,suggestingthatsocialinteractiondoesnotincreaseintakesimplybytaking
peoplesmindsoffwhatandhowmuchtheyareeating.Finally,participantsate50%
morecakewhenfriendswerepresent.Theauthorssuggesthighlypalatable,high
energysnacksmaybemorepopularwithincertainsocialinteractions(i.e.,groupsof
friendscomparedtostrangers).

Salvyetal.(2007)alsoexperimentallyexploredtheeffectsoffamiliarityonself
presentationofeatingbehaviours.Participantssatinaroomwithastranger,afriendor
apartnerandweregiventenminutestotalkandeatbiscuits,bothsweetandsavoury.
LikeLearyetal.(1994),Salvyetal.alsomanipulatedthesexcompositionofthegroup
(mixedversussamesex).Theyfoundthatbothfemaleandmaleparticipantsatemore
foodwhentheywerewithfamiliarpeople.Thisconfirmsthatthepeoplewithwhomone
eatswithisanimportantinfluenceonbehaviour.

Multiplestudiessuggestthatinterpersonalrelationshipsmaydeterminethenatureofan
eatingoccasionandextentoftheinfluencethatonepersonhasonanother.Inorderto
determinetheinfluencethatdifferentpeoplemayhaveoneatingbehaviour,itiscrucial
toassesswhoispresent.



73


3.4 SocialInfluenceandfastfood
Despitethecommonassertionthatfastfoodconsumptionisobesogenicandthatthe
increasednormativepressuretoeatmealspreparedawayfromhomeisoneofthe
driversoftheobesityepidemic,researchidentifyingtheantecedentsandconsequences
offastfoodconsumptionissparse.Publishedresearchhastypicallyfocussedondefining
thefastfoodconsumeraswellasexploringthefactorsimportantinfoodchoice.
Traditionalbigbrandfastfoodproductsareenergydenseand,forthisreasonalone,itis
importanttounderstandtheirconsumption.Itishighlylikelythatindividual,
environmentalandsocialvariableswillallacttopromoteorinhibitintake.Thesocial
influenceliteratureindicatesseveralpossiblewaysthatsocialfactorsmayaffectfast
foodintake.Manyconsumersreportthattheirfastfoodconsumptionisdrivenbytaste,
costandconvenience(Bryant&Dundes,2008;Driskelletal.,2006;M.J.A.Schroder&
McEachern,2005).Althoughsocialfactorsarenotfrequentlycitedbyconsumersas
influencingtheirfastfoodeatingbehaviours,itispossiblethatthesefactorsdoaffect
fastfoodconsumption.Vartanian,HermanandWansink(2008)monitoredsocial
influencesataneatingoccasionandthenaskedparticipantstoexplaintheirfoodintake.
Theyfoundthattasteandhungerwereusedbyparticipantstodescribetheirintake
despiteclearevidenceofsocialinfluences.Therefore,althoughfastfoodconsumersdo
notacknowledgesocialinfluencestheymayexistatafastfoodeatingoccasion.

3.4.1 EnvironmentalPressure
Themodelofnormativeovereatingsuggeststhattheremaybeenvironmental
antecedentstooverconsumption.Inafastfoodeatingenvironment,thesecuesmaybe
moreexplicittoconsumersthaninahomeeatingenvironment.Upsizingandmeal
dealsarecommonelementstomanyfastfoodfranchises.Theyaredesignedtosell
morefoodforlessmoney(CameronSmith,Bilsborough&Crowe,2002).Beyondthese


74


moreovertinfluences,fastfoodeatingenvironmentsmayhavesimilarinfluencesas
generaleatingenvironments.Atmosphericmeasuressuchasapleasantambience
(Feunekesetal.,1995),thelocation(atrestaurantsversusathome:Stroebele&de
Castro,2004)andtastyfood(J.M.deCastroetal.,2000)havebeenshown
independentlytoincreaseintakeateatingoccasionsforavarietyoffoods.Howthese
aspectsoftheenvironment(ormicroenvironmentalinfluences)interactaswellas
whethertheycanincreasetheintakeoffastfoodsisunknowninthecurrentliterature.

3.4.2 Normativeinfluenceandsocialfacilitation
Althoughthepresenceofothersisdirectlyrelatedtoimprovedatmosphere,thereare
manyotherwaysinwhichsocialvariablesmayinfluenceintake.Matchingnorms,
wherebypeoplesintakesarematchedtothosearoundthem,andsocialfacilitation,
wherebytheintakeamountsincreasewiththenumberofotherpeoplepresentatthe
eatingoccasion,havebeenconsistentlyshowntoincreaseintakeofmealsandsnacks.It
isunclearwhetherthesetwomechanismsactuallyrepresentthesameoradifferent
socialinfluencebut,together,theyhighlighthowintakecanbeinfluencedbysocial
variables.Theextenttowhichthisfindingisreplicatedinfastfoodeatingenvironments
isyettobeaddressed.

Itispossiblethatsocialfacilitationeffectsonconsumptioninfastfoodrestaurantswill
differfromthoseobservedinothereatingenvironments.Theverynatureoffastfooda
mealthatisconvenientandquickhighlightspossibledifferencesfromgeneraleating
behaviours.Thespeedyserviceandconsumptionoffastfoodmaymakeitimmunefrom
theeffectsoftimeextensionwhicharehypothesisedtounderliesocialfacilitation
effects.Yet,BellandPliner(2003)reportedthatwhenfoodwaseateningroupsof
peopleatafastfoodlocation,thetimespenteatingwassignificantlylongerthanthe


75


durationofmealsforthoseeatingalone.Admittedly,theeffectwassmallerthanthat
reportedinarestaurantandaworksitecafeteria,butitwassignificant,nonetheless.Bell
andPlinersstudywasobservationalanddidnotincludedetailsofthefoodsconsumed
anditwasthereforenotpossibletodeterminewhethertheincreasedmealdurationled
toincreasedintake.Consideringtheclearlyportionedvaluemealsthatarenormally
providedatfastfoodrestaurants,itispossiblethatwhilethedurationofameal
increased,theamounteatendidnot.Yet,StroebleanddeCastro(1992)reportedthat
socialfacilitationeffectscouldaltertheamountofconsumedandthetimespenteating
atmealsinrestaurants.Nevertheless,therestrictivetimingandspeedynatureoffast
foodsislikelytolessenvariabilityintheamounteatenandtimespenteating.Research
isneededtoclarifytheroleofsocialfacilitationinfastfoodconsumption.

Incontrasttothelimitationsthatafastfoodenvironmentmayhaveontheabilityto
observetheeffectsofsocialfacilitation,theclearlyportioneditemsmaymeanthatfast
foodeatingnormsaremoreclearlydetermined.Havingmealslabelledassmall,medium
andlargemaymakeiteasierforpeopletoadheretomatchingnorms(orderingsimilar
amountstothosepeoplearoundthem).Asidefrompotentiallyhavingestablished
environmentalnorms,thereisnoevidenceofhowandindeedwhethereatingnorms
operateduringafastfoodeatingoccasion.

3.4.3 Thewhoofsocialinfluence
Therehavebeenstudiesthatsuggestthattheinterpersonalrelationshipsharedwiththe
peoplepresentataneatingoccasionmayaltertheeffectofthepresenceofotherson
foodintake(Clendenenetal.,1994).Furthermore,therehasbeensuggestionthat
certaineatingbehavioursmaybemorecloselyassociatedwithcertaintypesofpeople
(e.g.,snackingbehavioursandfriends:Feunekesetal.,1998).Comparedtoregular


76


eating,fastfoodbehavioursmaybesubjecttodifferentsourcesofinfluence.Fastfood
mayrepresentadistincttypeofmealwithcertaintypesofpeopleinfluencingbehaviour
morethanothers.Whetherandhowdifferentpeoplealterfastfoodintakemayprovide
usefulinsightintoanysocialinfluencespresentduringfastfoodconsumption.

3.4.4 Fastfoodbehaviours
Manyofthequestionspertainingtohowsocialinfluencemayalterfastfood
consumptionexistbecauseofpaucityinresearchthatattemptstoexploresocialfactors
andfastfoodintake,but,morefundamentally,becauseofalackofavailableresearchon
fastfoodbehaviours.Withoutpreexistingknowledgeofthesebehaviours,itisdifficult
toascertainwhatinfluencesareimportantfordeterminingtheamountoffastfood
consumed.Menuavailabilitysuggeststhattraditionalfastfooditemsaremorepopular
thanhealthieralternatives.Incontrast,McDonaldshavecitedsalesfromtheirHealthy
Choicemenuasareasonforsalesgrowth(Gumbel,2004).Thereissomeevidencefrom
consumersurveysthatthepurchaseofhealthyfastfooditemsisincreasing(DiPietro,
Roseman,&Ashley,2004).Withoutanydataonactualorderingchoices,itisdifficultto
knowwhetherhealthieroptionssimplyexpandMcDonaldspotentialmarket(i.e.,
consumerswhowouldnotusuallyfrequentMcDonaldsareattractedthereforthe
healthyoptions)orwhethertheseitemsrepresentalargeproportionoftheirsales(i.e.,
existingconsumershavechangedtheirfoodchoices).

Thereislimiteddatathatdescribesthewayinwhichconsumerseatbigbrandfastfood
items.Forexample,itisunclearwhetherconsumerspurchasevaluemealstoeatas
mealsorsnacks.Asanoccasionalmealreplacement,fastfoodmaynotdramatically
swayenergyequilibrium(seeChapterTwo).Ontheotherhand,iffastfoodintakeis
additionaltothreeregulardailymeals,energyexcessislikelytoresult.Other


77


informationaboutthewaythatbigbrandfastfoodsareeateninAustraliaisneeded,
includingthemotivationsandcontextofconsumption.Giventhepotentialforsocial
factorstoincreasegeneralfoodintake,understandingthesocialdynamicassociated
withfastfoodintakecouldrevealimportantinsightsaboutoverconsumptionandweight
gain.

3.5 Aimsofsubsequentstudies
Thesocialinfluenceliteraturereviewedshowedthatsocialfactorscanaltereating
behaviour.Studiesoneatingbehaviourshaveshownsignificanteffectsofsocialnorms
andsocialfacilitation.Noneofthesestudieshaveexploredhowsocialfactorsmay
influencefastfoodconsumption.

Theaimofsubsequentstudieswasthereforetoexplorehowsocialinfluenceoperatesat
afastfoodeatingoccasion.Limitedexistingresearchonsocialinfluenceandfastfood
consumptionbehavioursprecludedtheformulationofspecifichypotheses.Therefore,
severalresearchquestionswereusedtoguidethedevelopmentofthefollowingstudies
presented.
Theseresearchquestionswere:
1. Domicroenvironmentalpressuresinafastfoodrestaurantinfluencethe
amounteaten?
2. Cansocialfacilitationand/orthetimeextensionhypothesisincreasethe
amountoffastfoodeaten?
3. Whatsocialnormativeinfluencesaffectfastfoodconsumptionbehaviour?
4. Ifsocialinfluencefactorsareinfluentialinfastfoodconsumption,are
thesemoderatedbydifferentinterpersonalrelationships?


78

4 ChapterFour:DevelopmentoftheFastFoodSurveyandPreliminary
ExaminationofSocialInfluencesonFastFoodConsumption

Theliteratureonfastfoodconsumption(reviewedinChapterOne)andsocialinfluence
(reviewedinChapterThree)suggeststhattherearemanyfactorspotentiallyassociated
withincreasedfastfoodintake.Itisimportanttoclarifytherelationshipbetweenfast
foodconsumptionandtheoccurrenceofoverweightorobesitythathaspreviouslybeen
reported.Thepossibilitythatindividualcharacteristics,includingsex,age,socio
economicstatus(SES)andbodymassindex(BMI),mayassociatewithfrequencyoffast
foodconsumptionhasbeensuggestedinanumberofpreviousstudies(Bowman&
Vinyard,2004;Frenchetal.,2000;Mohretal.,2007;H.Schroderetal.,2007).Whether
thesesamedemographicsrelatetoquantityconsumedinonemealislessclear.For
example,teenagersmayberegularconsumersoffastfood(Ungeretal.,2004),butmay
consumerelativelylittleoneachvisittotherestaurant.

Therehavebeenseveralreportsindicatingthatfastfoodconsumptionismotivatedby
convenience(FOODweek,2008;M.J.A.Schroder&McEachern,2005).Althoughthe
motivatorsforfoodchoice(e.g.,fastfoodsversusotherfoods)areimportant,
understandingwhatstylesoffastfoodspeoplechoose,whethertheyconsumetheseas
mealsorsnackswhether,andhowtheybalancetheirenergyintakethroughoutthe
remainderofafastfoodconsumptionday,willhelptoexplainhowfastfoodcouldbe
overconsumedandresultinweightgain;thiswassuggestedinthenutritionalanalysis
reportedinChapterTwo.



79


Despitethefactthatsocialinfluence,inmanyforms(i.e.,environmentalpressure,social
facilitation,socialnorms),hasbeenshowntoincreaseintakeatavarietyofeating
occasionswitharangeofdifferentfoods,thisresearchhasnotcurrentlybeenextended
tofastfoodeatingbehaviours.Todate,researchhasfocussedonlyontheconsumption
ofmultiplestylesofsnackfoodsincludingcrackers(Rosenthal&McSweeney,1979),
cookies(Pliner&Mann,2004),lollies(deLuca&Spigelman,1979),savourysnacks
(Clendenenetal.,1994),cake(Hetheringtonetal.,2006),orgeneraldietaryintake(J.M.
deCastro,1990,1994).

Withminimalinformationavailableonfastfoodbehaviours,asurveywasdevelopedin
ordertocollectdataon:(a)whatfastfooditemsarebeingconsumed,and(b)thesocial
andenvironmentalcircumstancessurroundingthisconsumption.Thegoalsofthis
surveyweretwofold:toexplorehowactualfastfoodchoicescouldincreasetheamount
eatenandtoassesswhetherdemographiccharacteristicsandsocialfactorsarerelated
tofastfoodintakeatasingleeatingoccasion.Thischapterwilldiscussthedevelopment
andtestingoftheFastFoodSurveyandinitialanalysesexploringthetypesoffastfoods
beingeaten,andhowsocialcontext,socialinfluencesandindividualcharacteristicsmay
alterenergyintakeatafastfoodeatingoccasion.

4.1 DevelopmentoftheFastFoodSurvey(FFS)
TheFastFoodSurvey(FFS)wasdevelopedtoassessfastfoodhabitsandenable
explorationofvariablesthatmayincreasefastfoodconsumption.Itwasintendedto
measureconsumptionandpotentialpredictorsinaquick,butreliablemannerthat
wouldmaintaintheenthusiasmoftheparticipantsandprovidepotentiallyuseful
individualfeedbackaboutconsumptionbehaviour.Withthecurrentliteratureoffering


80


littledirectionregardingfastfoodconsumption,theFFSwasdesignedtocapturemany
aspectsofbehaviourandcontextinordertoassesstheimpactofpotentialmicro
environmentalpressures,andsocialanddemographicvariablesonconsumption.

TheFFScapturedpatternsoffastfoodintakefrombigbrandfastfoodfranchisesin
Australia:McDonalds,HungryJacks,KFC,DominosPizzaandRedRooster.Thereason
forlimitingtheexplorationofconsumptiontothesecompanieswastominimise
ambiguityintheoperationalisationoffastfood(seeTable1,page11).AlthoughSubway
hasadominantpartofthefastfoodmarketinAustralia,itwasexcludedfromthe
definitionofbigbrandfastfoodsusedinthisstudy.Thereasonforthiswasbecause
Subwaydoesnotprovidetraditionalfastfoods(burgers,friedchicken,pizzaetc.)
27
.The
focusontraditionalfastfoodsintheFFSwaschosenasincreasedintakeofthesestyles
ofitemsismorelikelytobeassociatedwithnegativehealthoutcomes.

IntheFFS,participantswereaskedtorecalltheirmostrecentfastfoodeatingoccasion
includingwhen,whatandwithwhomtheyate.Toaidintherecalloftheitems
consumed,theFastFoodChoicesProgram(FFCP)wasdeveloped(see,
www.fastfoodstudy.com.au).Throughacomputerinterface,theFFCPprovidedamenu
ofpotentialfoodchoicesavailableateachrestaurantandaskedtheparticipantto
identifytheirselections,includingtheamountofeachitemconsumed.Analgorithmwas
developedandutilisedthatallowedfeedbacktorespondentsregardingthenutrients
theyhadconsumed
28
.Followingtheirmenuselections,participantswerepresented
withaseriesofsurveyquestionsattemptingtoassesstheenvironmental,socialand

27
AlthoughthereareitemsontheSubwaymenuthataremoreenergydense(e.g.,chickenfilletor
meatballsubs),theitemstheysellareallsandwichesanddifferdramaticallyfromtraditionalfastfood
items.
28
Participantswereonlymadeawareofthenutritionalfeedbackcomponentofthesurveyaftertheyhad
completedallthequestions.Thiswasusedsoparticipantswouldnotbeprimedtobeconcernedaboutthe
nutritionalconsequencesoftheirconsumption.


81


demographicfactorsthatmightcorrelatewithfastfoodconsumption.Finally,
participantswereofferedthechancetoenteradrawforadoublemoviepass.Thefinal
versionoftheFFSwasprogrammedentirelyinMacromedia(nowAdobe)Flash.Each
componentofthesurveywillbediscussedindetailbelow.

4.1.1 TheFastFoodChoicesProgram(FFCP)
ThefirsttasktobecompletedwithintheFFSwastheFFCP.Thisaskedparticipantsto
recalltheitemstheyhadeatenwhentheylastvisitedoneofthenamedfastfood
restaurants.TheFFCPwasdevelopedtoassistparticipantsrecall.TheFFCPincluded
interactive,illustratedmenusthatparticipantsnavigatedtoselecttheitemsthey
consumed.Whentheitemswereselected,thecorrespondingnutritionalinformation
wasstoredwithintheprogramtoallownutritionalfeedbackuponcompletionofthe
survey.Giventhecommercialnatureofthefoodslisted,theprogramincludeddetailed
Trademarkinformation.Finally,theprogramcontainedananimated,stepbystep
tutorialthatshowedparticipantshowtoidentifyandselecttheitemstheyconsumed.

TheFFCPwascreatedinanumberofstages.Theseinvolvedtheillustrationand
programmingofover400productsfromthefivemajorchains.Illustrationswereall
doneusingAbodePhotoshopandMacromediaFlash
29
.NavigationintheFFCPwas
designedtoreflectmenulayoutswithintherestaurants;itemsweregroupedasburgers,
sides,drinks,dessertsandsoon
30
.Itwasanticipatedthiswouldallowforintuitive,
simplenavigation.


29
Carewastakentoensurethattheitemsweredrawntoaccuratelyresemblethefoodstheywerebased
on.Forexample,theBigMacwasdrawnwiththreebuns,twobeefpatties,lettuceandcheese.Thislevelof
accuracywasdonetoensurethatthoseparticipantswhowerenotfrequentfastfoodconsumerscould
recognisetheitemstheyhadeatenwithoutrelyingonfamiliaritywithproductnames(e.g.,Whopper).
30
Aseriesofpromotionalfooditems,notlistedonthecompanywebsites,wereavailableinsome
restaurants.TheseitemswereexcludedfromtheprimaryFFCPbecausetheyareavailableonlyfora
limitedtime.


82


Thecompanywebsiteswereusedtosourcealistoftheitemsavailablefromeachchain,
andtheaccompanyingnutritionalinformationfortheseitems.Toverifythattheonline
menuwasaccurate,followupvisitsweremadetoonerestaurantofeachofthechains
31
.
Thefinalsectionoftheprogramhadafreetextentryareaallowingparticipantstodetail
anynonstandardfoodconsumed(e.g.,specialpromotionaloffers).

Afterparticipantshadmadetheirmenuselections,theprogramdirectedthemtoa
checklistscreen.Thisallowedparticipantstocheckthattheyhadaccuratelyenteredthe
itemstheyate.Hereparticipantscouldalsoadjustthequantitiesofitemstheyate
and/orremoveitemsfromtheirorder.Toaccountforanyalterationstotheordersthat
werenotdetailedintheprogram,thechecklistwasfollowedbythefreetextentryarea
thataskedparticipantsiftherewereanyalterationstothesizeorcompositionofthe
itemstheyhadselectedortodetailanyfoodwaste.Thisalsoincludedthequestions
allowingentryofanyitemsconsumedthatwerenotlistedinthesurvey.

Thefinalcomponentdeliveredbytheprogramwasthenutritionalfeedbacksystem.
Uponcompletionofthesurvey,participantsweregiventheoptiontoviewthe
nutritionalinformationoftheitemstheyhadconsumed.Iftheyselectedthisoption,they
weredirectedtoanutritionalpanelincludingalloftheinformationfortheitemsthey
hadselected.Theinitialnutritioninformationpanelwasfollowedbymoredetailed
nutritionfeedback.Thefeedbackrelatingtoenergyandnutrientconsumptionwas
developedinassociationwitharegistereddieticianusinginformationfromthe
AustralianGuidetoHealthyEating(Kellettetal.,1998).Informationincludedcommon
sourcesofeachmacronutrientandsodium;itsuggestedsomehealthimplicationsfor

31
Atthetimethatthesurveywasbeingdeveloped,RedRoosterreleasedanewhealthylineoffoodsthat
wasnotlistedonthewebsite.ContactwasmadewithRedRoosterandthenutritionalinformation
accompanyingthenewitemswasemailedbythecompanyforincorporationintotheprogram.


83


excessiveintakeofeachandgavetherecommendedrangesofdailyintake.Thedietician
gaveadviceregardinghealthoutcomesofexcessiveconsumptionaswellassuggesting
somealternative,moreinterpretableguidelinesfortheintakeofprotein,basedonbody
weight.Thisinformation,asitappearedintheFFCPcanbeseeninAppendix2.

4.1.2 MeasuringCorrelatesofConsumption
Questionsaboutthebehaviourssurroundingthefastfoodeatingoccasion,potential
socialinfluences,perceptionsofportionsize
32
andindividualdemographicswere
measuredinthesectionfollowingtheFFCPandpriortonutritionalfeedback.Afulllist
ofthequestionsandtheaccompanyinginformationsheetcanbeviewedinAppendix3.

Intotal,therewere38questionsaccompanyingthesurvey.Theseweredesignedwitha
mixtureofdropdownmenus,radiobuttons,spinnersandfreetextentryareasand
incorporatedintotheinitialsectionofthesurvey.Toavoidmissingdata,eachpage
includedverificationthatallnecessaryquestionshadbeenanswered.Ifitemswere
overlooked,amessageinformingparticipantsthattheyhadmissedaquestionappeared.
ScreenshotsofthefinalsurveycanbeviewedinAppendix4.

Twosetsofquestionsattemptedtogaugetheroleofsocialinfluenceonfastfood
consumption.Fourquestionsweredesignedtomeasurenormativeinfluences.Allwere
ratedon7point,anchoredLikertstylescales.Thesequestionsaskedparticipantsto
giveratingsoftheawarenessofotherswhileeating,theamountothersate,howaware
otherswereofwhattheparticipantconsumedandconcernforwhattheotherpeoplein
theeatingenvironmentthought.Inalltheseitems,lowerscoresindicatedless

32
Discussionoftheseitemsisbeyondthescopeofthisdissertationandhasbeenwrittenforpublicationin
aseparateforum.ItcanbeviewedinAppendix7.


84


normativeinfluence.Questionsonawarenesswereratedfrom1(notatallaware)to7
(veryaware).Thetwoitemsaboutconcernwererankedfrom1(alittle)to7(alot).

Toassesseffectsofsocialfacilitation,fouritemswereincludedinthesurvey.These
includedaquestionontheamountoftimespenteating,ratingsofatmosphere(two
items)andthenumberofpeoplepresentattheeatingoccasion.Itemsmeasuring
atmospherewithinthefastfoodrestaurantincludedquestionsabouthowsocialand
howpleasantparticipantsthoughttheatmospherewas.Theatmosphereoftheeating
environmentwasratedona7point,Likertstylescalewhere1representednotatall
social/pleasantand7representedextremelysocial/pleasant.Responsesindicating
thenumberofpeoplepresentweredividedaccordingtoavarietyofinterpersonal
relationships.Forexample,participantscouldindicatetheywerewithfourpeopleand
thatthegroupconsistedoftwofriends,onechildandtheirpartner.Thisenables
relationshiptobeexaminedasamoderatorormediatoroftheeffectofnumberpresent.
Ninedifferentcategoriesofrelationshipwereutilised.Thesewere:partner,children,
parent,sibling,otherrelative,colleague,friend,membersofastructuredsocialgroup
andanothergroup(withafreetextentryarea).

Thenextseriesofquestionswereincludedtoinvestigatethecontextormicro
environmentalaspectsoffastfoodconsumptionbothonthespecificoccasionrecalled,
andingeneral.Theeightquestionsdescribingthecontextofthespecificeatingoccasion
wereansweredthroughamultiplechoiceformat
33
.Questionsrelatingtothespecific
eatingoccasionincluded:theformatofthepurchase(takeaway,drivethru,dinein);
locationthefoodwaseaten(restaurant,car,home,other);thedayofpurchase;themeal

33
Theresponseformatwaspilotedtoensurethattheoptionsprovidedwereappropriate.Thiswasthe
extentofinitialpilotingasnoscaleswerebeingdeveloped.Allvariablesmeasuredwereobserved
variablesthatdescribeddiscretebehaviours,perceptionsorparticipantcharacteristics.


85


occasion(breakfast,lunch,dinner,inbetweenmealssnack);thepartofdayduring
whichtheitemswereeaten(05001159morning,12001359midday,141659
afternoon,17001829evening,163011.59night,124.59latenight);whethertheitems
wereboughtinamealdealorseparately;andthereasonforfastfoodconsumptionand
anyactivitiesthatprecededtheeatingevent(twoopenendedquestions).Asingleitem
assessedhowmuchrespondentsenjoyedthefooditemstheyconsumedontheoccasion
reportedfrom1(notatall)to7(extremely).Thethreeitemspertainingtogeneralfast
foodbehavioursincludedfrequencyofpatronage,tendencytoorderthesameitemsand
theappealofupsizing.

Threeitemsaddressedhowthefastfooditemsconsumedimpactedondailyeating
patterns.Thisincludedratingsofthesizeoftheeatingoccasionrelativetoanormal
one.Thatis,ifthefastfoodwaseatenaslunch,participantswereaskedifthiswasa
smaller,typicallyorlargersizedlunchthanwhattheywouldusuallyeat.Asecond
questionaskedtheextenttowhichthetotalamountoffoodeatenduringthedaywas
differenttousual.Afreeresponseitemaskedparticipantstodetailanychangestotheir
eatingpatternforthedayofconsumption(priortoandfollowingfastfoodintake).The
formertwoquestionswereratedona7point,Likertstylescalefrom1(much
less/smaller)to4(typical)to7(muchmore/bigger).

Finally,therewerethirteendemographicquestions.Amongstthesewereestimationsof
heightandweighttoallowthecalculationofBodyMassIndex.Participantswereasked
toindicatetheirconfidencewiththeheightsandweightstheyreportedusinga4point
scale(from1notatallconfident,to4veryconfident).Otherdemographic
informationincluded:age(inyears);sex;annualhouseholdincome(20,000andless,
20,00140,000,40,00160,000,morethan60,000);postcode(toallowcalculationof


86


socioeconomicdisadvantage);countryofbirth(Australia,UKorIreland,Other);levelof
education(BelowSecondary,Secondary,Trade/Apprenticeship,Certificate/Diploma,
Bachelordegreeorhigher);currenttypeofemployment(Fulltimeemployed,Part
time/Casualemployed,Unemployed,Homeduties/Retired,Student,Other);sizeof
immediatefamily;maritalstatus(Marriedorlivingwithpartner,Separated/Divorced,
Widowed,NeverMarried);whethertheparticipantwasapensioner;andwhetherthe
participantwasofAustralianAboriginalheritage.

4.2 SurveyingattheRoyalAdelaideShow
TheFFSwasinitiallyadministeredtoaconveniencesampletoallowpreliminary
explorationofthelinksbetweensocialandenvironmentalvariablesandconsumption
behaviourwhilealsopilotingtheinterfaceundermonitoredconditions.Ethicsapproval
forthisstudywasreceivedfromtheUniversityofAdelaideandtheethicscommitteeat
CSIRO,HumanNutrition.

4.2.1 Procedure
TheFFSwassetupatasurveystation(orkiosk)whereparticipantscouldsitand
completethequestionnaireononeoffourlaptops
34
.ThisstationwaslocatedinaCSIRO
displayintheJubileePavilionattheannualRoyalAdelaideShow.TheRoyalAdelaide
Show(RAS)isanagriculturalfairandoneofthelargestannualcommunityeventsin
SouthAustralia.TheRAShasmanynonagriculturalactivitiesincludingshowrides,
displaysandstallssellingmerchandiseandfood.Itrunsforninedaysfromthe

34
AlthoughthesurveywasdesignedusingInternetspecificsoftware,thefacilitiesprovidedattheRoyal
AdelaideShowdidnotincludeanInternetconnection.Thereforetheoriginalprogramminghadtobe
modifiedslightly.Aserverlaptopwassetupthatlinkedtothesurveyandstoredthedata.Theonly
changesthatneededtobemadetothesurveyrelatedtothebackend(thedatabase)andthereforedid
notchangethesurveyinterface.Thesurveystilloperatedthroughawebbrowserandthereforeappeared
nodifferenttotheInternetversionofthesurvey.Themaindifferencewasthataresearchprojectofficer
waslocatedwithinthekiosktoprovideanyassistanceandreceivefeedbackabouttheFFSanditsmethod
ofadministration.


87


beginningofSeptembereachyearandisreportedtoattract35%oftheSouthAustralian
population(RoyalAgricultural&HorticulturalSocietyofSA,2006).

OverthefirstsevendaysoftheRAS,peoplewalkingpastthedisplaywereaskedifthey
hadrecentlyconsumedfastfoodandwhethertheywereinterestedindoingan
interactivefastfoodsurvey.Theywereofferedthechancetobeenteredintoadaily
drawforadietbookandaseparatedrawforafreedoublepasstoamovie.Interested
peopleweredirectedtoaboothandthenloggedintothesurveybyeithertheresearch
projectofficerinattendanceorbyanadministrativeassistant.Participantshadthe
opportunitytoasktheprojectofficerforassistanceiftheyranintodifficultiesorwere
confusedbyanypartofthesurvey.Administrativeassistantswereinstructedtoprovide
assistanceonlywiththeproceduralaspectsofthesurvey.Assistantsandparticipants
werealltoldthatthesurveywasdesignedtounraveleverydayfastfoodconsumption
behaviours.

Recruitmentwasaimedatthosepeoplewhohadconsumedanyitemfromoneofthe
fivechainswithinthe12monthsprecedingdatacollection.Givenpreviousdata
confirmingthepervasivenessoffastfoodconsumption,itwasexpectedthatthiswould
encompassalargeproportionofthepopulation.Aswellashavingconsumedfastfood
withinthistimeframe,participantshadtobeconfidentthattheycouldrememberthe
specificdetailsofthisconsumption(i.e.,time,place,otherpeoplepresent).Potential
participantsalsohadtohaveconsumedanitemfromoneofthefastfoodrestaurants
withinAustraliaasthenutritionalinformationwasbasedonAustralianstatistics.People
undertheageof16wereexcludedfromthestudy.Thisagelimitwasimposedforethical
reasonsandtorestricttheinvestigationtothosepeopleoldenoughtobeassumedtobe
independentintheirfoodchoices.


88

4.2.2 Assessmentofoutliers
Afterrecruitment,datawerecompiled.InsevendaysofrecruitingattheRAS,thefour
computerstationsobtained126responses.Thesedatawerethenscreenedforoutliers
andcodedbeforeanalysis.Dependentvariableswerescreenedforoutliersusingz
scoresandstemandleafplots.Fivecaseswereremovedfollowingthisscreening
35
.
MahalanobisdistancesindicatednocaseswithextremeChiSquarevalues(p<.01).After
thisscreening,twocaseswereremovedbecausetheywereincomplete.Foreaseof
interpretation,threecaseswereremovedbecausetheyrelatedtoconsumptionofa
drinkonly
36
.

4.2.3 Adjustmentstoorders
AnumberofparticipantsusedthefreetextentryareasfollowingtheFFCPtoindicate
thattheordertheyhaddetailedneededadjusting.Intotal,24orderadjustmentswere
indicatedbytherespondents.Ofthese,12ordersneededthestandardportionsizeof
theitemsadjustedduetofoodwaste.Inmostcases,thereportedestimatesofvariations
wereexact,forexample,IhadhalfaBigMac.Whereitwasindicatedthatanitemwas
shared,theportionwasdividedbythenumberofpeoplepresent.Whentheadjustments
weremoreambiguous,forexample,Ihadabiteofhotcakes,theiteminquestionwas
purchased,theapproximatesizeandweightofthechangeassessed,andthevalue
calculatedfromthat.Sevenadjustmentsindicatedmenuitemsthattheparticipantcould
notfind.OnlyoneoftheseitemswasnotactuallylistedintheFFCP(aCrunchiesundae).

35
Stemandleafplotsindicatedoneextremecaseforenergyconsumed;thispersonconsumedover
9200KJ.However,thestandardisedzscoreforthiscasedidnotindicatethatthiscasewasover3.29
standarddeviationsawayfromthemean.Asthedistributionwaswithinnormallimitsandzscoresdid
notidentifythiscaseasanoutlier,itwasretainedinthesample.Peoplewhoreportedspendinglonger
than40minuteswereidentifiedasoutliersusingzscorecriteria;stemandleafplotsconfirmedthis.
Therewerefivepeoplewhoreportedeatingfor41ormoreminutesthatwereremovedfromthesample.
36
Recruitmentwastargetedatpeoplewhopurchasedanyitems.Onlyasmallnumberofpeopleonlyhada
drinkandthismayskewtheresultingcalculationofenergyintakefromfastfooditems,therefore,these
peoplewereexcludedfromtheanalysis.


89


Contactwasmadewiththefastfoodcompanytodiscoverthenutritionalcontentofthis
item.Thefinalfiveorderadjustmentspertainedtomenuitemcustomisation.Inall
instancesthesewereremovalofavegetableitem(pickleortomato)fromaburger.No
energyandfatadjustmentsweremadeforthese,astheseitemsareverysmallandhave
littleimpactonoverallenergy.Theadjustedenergyamountswereusedinall
subsequentanalyses.

4.2.4 Recoding
Severalitemswererecodedbeforetheanalyses.Therewassubstantialvariationinthe
frequencyoffastfoodpatronage.Answersrangedfromonceeveryfewmonthsto
almostdaily.Sixcategorieswerecreatedtoreduceskewandimproveinterpretabilityof
thedatatheseincluded:onceeveryfewmonthsbutlessthanonceamonth,monthly,
morethanonceamonthbutlessthanfortnightly,fortnightly,onceaweekand
twiceaweekormore.

Giventheskewpresentinthedataforthenumberofpeoplepresent,allthosewho
reportedeatingingroupswithfiveormorepeoplewerecombinedintoacategory
labelledlargegroup
37
.Thismeantthatgroupsizehadfivelevels:lonediner,pair,
groupofthree,groupoffourandlargegroup(i.e.,fiveormore).

Thepostcodevariablewasusedtocalculatetherelativesocioeconomicdisadvantageof
theareawheretheparticipantlived.Todothis,dataabouttheIndexofRelative
SocioeconomicDisadvantage(IRSD)wastakenfromtheABSwebsite(AustralianBureau
ofStatistics,2008a).Thisindexgivesanindicationofsocioeconomicdisadvantage
accordingtopostalzonesthroughoutAustralia.Scoresrangefrom200to1200with

37
Thedistributionofthegroupsizevariablewasskewed.Uponcloserexaminationofthedata,itwasclear
thatthisskewwaslikelytheresultofafewrespondentswhohadeateninlargegroupsofpeople(20).


90


lowerscoresindicatingmoresocioeconomicdisadvantage.Scoresarecalculatedusing
Censusinformationaboutthenumberofhighandlowincomehousesinapostalarea
andothervariablessuchasaccesstoresources(AustralianBureauofStatistics,2008b).

4.3 Results
4.3.1 Usability
Commentsfromparticipantsaddressedtotheresearchstaffindicatedthattheyfound
theFFSandtheFFCPengagingandeasytocomplete.Mostparticipantscompletedthe
surveywithoutanyassistance.Questionsthatrelatedtotheproceduralaspectsofthe
surveyoftenwereaccountedforbyinstructionsalreadyonthescreen.Itwassuspected
thatthesequestionsresultedfromparticipantsdesiretoskipaheadwithoutgetting
boggeddownintheinstructions.Becausethestudywasdesignedtobecompleted
withoutguidance,thesequerieswereresolvedbydirectingparticipantstowardthe
instructionsonthescreen.Thereweregenerallynootherquestionsfollowingthis
direction.Onepersonindicatedthathewasunabletoreadthequestionsduetolearning
difficultiesbutstillwishedtocompletethesurvey.Hisanswerswererecordedbyan
assistantwhoreadthequestionsandpossibleresponseoptionsastheyappearedonthe
screen.

4.3.2 Finalsample
Therewere116peopleinthesampleaftertheoutliershadbeenremoved.Respondents
rangedinagefrom16to61yearsofage,withameanageof30(SD=12.16).
Participantsreportedhavingbetween0and8immediatefamilymembers.Frequencies
forthedemographicfactorsaredescribedinTable2.



91

Table2:Descriptionofthedemographiccharacteristicsofthesample

(n=116).
Variable
1
% Variable %
Sex(male) 38.8 OfAboriginalorigin 4.3
Highestlevelofeducation
completed
Employmentstatus
Bachelor 34.5 Fulltime 38.8
Trade/certificate 23.3 Student 28.4
SecondarySchool 36.2 Parttime/casual 20.7
BelowSecondarySchool 6.0 Homeduties/retired 4.3
Householdincome Unemployed 3.4
20,000andbelow 14.7 Maritalstatus
20,001to40,000 20.7 Married/livingwith
partner
45.7
40,001to60,000 21.6 Nevermarried 50.9
60,000andabove 43.1 Other

3.4
Receivepension 11.2 Stateofresidence
Countryofbirth SouthAustralia 95.7
Australia 82.8 Victoria 2.6
UK 5.2 Queensland 1.7
Asia 6.0
Other 6.0
1
Someofthecategoriesforsomeofthevariableswerecombinedbecauseoflowcell
frequencies;

Separated/divorced/widowed

Therewasadisproportionnumberoffemalesrepresentedinthesample.Therewasalso
ahighernumberofpeopleinthehighincomerangeandpeoplepossessingaBachelors
DegreethanwouldbeexpectedinthegeneralpopulationaccordingtotheSouth
AustralianCensusdata(AustralianBureauofStatistics,2007).ThemeanfortheIRSD
was1005.09(SD=60.06).Therewerepeoplefromboththelowestandhighestdeciles
ofdisadvantage(AustralianBureauofStatistics,2008c).Thesurveyattractedahigh
numberofAustralianbornrespondents,andtherewasmorethantwicethenumberof
peoplereportingtobeofAustralianAboriginalheritagethanwouldbeexpectedinthe
SouthAustralianpopulation(AustralianBureauofStatistics,2007).



92
FrequencyoffastfoodpatronageisreportedinTable3.Mostpeoplereported
consumingitemsfromoneofthefastfoodchainsbetweenonceafortnightandoncea
week.
Table3:Frequencyoffastfoodconsumptioninpilot
sample(n=116)

.
Variable %
Frequencyofpatronage
Morethanonceevery6months 15.5
Onceamonth 13.8
Onceafortnight 25.0
Onceaweek 22.4
Twiceaweek 15.5
Morethantwiceaweek 7.8

onlypeoplewhohadconsumedfastfoodwereaskedto
participate.

4.3.3 Therelationshipbetweendemographicfactorsandfastfoodconsumption
Giventheexistingthemeinpreviousresearchofassessingdemographiccharacteristics
andfastfoodconsumption,therelationshipbetweenseveraldemographicsandfastfood
intakeaswellasfrequencyofconsumptionwasinitiallyanalysed.

Malesfastfoodchoicesindicatedthattheyatesignificantlymorekilojoules(M=4057,
SD=1872)thanfemales(M=3253,SD=1714),t(114)=2.34,p<.05.Peoplewho
reportedworkingfulltimeatemore(M=4003,SD=2040)thanthosewithanyother
occupationalstatus(M=3286,SD=1605),t(114)=2.11,p<.05.Therewasno
differenceinthefrequencyofvisitstoanyofthefastfoodchainscorrespondingwith
employmentstatus.

Bivariaterelationshipsbetweenenergyintakeorfrequencyofconsumptionand
demographicfactorswerecalculated.Therewasaweak,positiverelationshipbetween
ageandenergyintake,r(116)=.20,p<.05.Therelationshipsbetweenenergyintakeand


93
levelofeducation,andenergyintakeandhouseholdincomewerenotsignificant(p>.05).
Therewasaweak,negativerelationshipbetweenenergyintakeandsocioeconomic
disadvantage,r(114)=.21,p<.05.Frequencyofpatronagewasnotrelatedto
demographics,includingage,income,levelofeducationorsocialdisadvantage.Itwas,
however,positivelyassociatedwithenergyintakeattheeatingoccasion,
r(116)=.24,p<.05.

Tofurtherexploretherelationshipbetweendemographiccharacteristicsandintake,a
multipleregressionwasperformed.Sex,age,themeasureofsocioeconomic
disadvantage(IRSD),householdincome,highestlevelofeducationcompleted,being
married(versusnot),beingfulltimeemployed(versusnot)andfrequencyof
consumptionwereincludedinthemodel.ResultscanbeseeninTable4.

Table4:Totalvarianceexplained(AdjustedR
2
),betavaluesand
significancelevelsforamodelpredictingtheamountofenergyconsumed
atafastfoodeatingoccasionusingdemographicfactors(n=116).
AdjustedR
2
=.120
Predictors
Beta
Alpha
Sex .14 .14
Age .16 .11
IRSD .24 .02
Householdincome .06 .61
Levelofeducation .08 .42
Married .04 .76
Fulltimeemployed .12 .25
Frequencyofconsumption .17 .06

Thefinalmodelwassignificant,F(8,105)=2.93,p<.01andaccountedfor12%ofthe
totalvariancefortheamountofenergyconsumedatasinglefastfoodeatingoccasion.
Theonlysignificantpredictoroftheamounteatenwasthelevelofsocioeconomic
disadvantageoftherespondentssuburb.Thiswasnegativelyrelatedtotheamount


94
consumed,indicatinggreaterconsumptionwasassociatedwithhighersocioeconomic
disadvantageorviceversa.Selfreportedfrequencyofconsumptionwaspositively
relatedtoenergy,butthisonlytrendedtowardsignificance.

Thesamecombinationofdemographiccharacteristicswasincludedtopredict
frequencyofconsumption.Thismodelfailedtoreachsignificance.

BMIandfastfoodintake.Thirtythreemalesand49femaleswereconfidentinboththe
heightsandweightstheyprovided
38
.BodyMassIndex
39
(BMI)wascalculatedforthe
participantsconfidentintheirheightandweights.ThisgroupreportedameanBMIof
25.87(SD=5.22)withvaluesrangingfrom17.63to43.83.BMIwasnotsignificantly
relatedtotheenergyintakeonthefastfoodeatingoccasion.Therewasaweak,positive
relationshipbetweenBMIandfrequencyofpatronage,r(82)=.25,p<.05.

4.3.4 Fastfoodbehaviours
Asthereislimited,publishedresearchonfastfoodbehaviours,thefrequenciesoffast
foodbehavioursandmotivatorsandcontextofconsumptionwereexplored.Aspects
suchasmealdealsandfoodchoiceswereanalysedtoassesswhethermicro
environmentalpressuresinfluencetheamountoffastfoodeaten(therebyaddressing
thefirstresearchquestionposedinSection3.5).

Descriptionofgeneralfastfoodbehaviours.Frequencyofdifferentpurchasingbehaviours
canbeseeninTable5.

38
Peoplewithconfidenceintheirheightandweightmeasureswereofaslightlydifferentdemographic
descriptiontothoselackinginconfidence.Theyhadhigherincome,t(114)=2.57,p<.05,andagreater
levelofeducation,t(114)=2.89,p<.01,thanthosenotconfidentintheheightsandweightsprovided.
Therewerenodifferences(atthep<.05level)forenergyintake,ageorsocialdisadvantagebetweenthe
groups.
39
Weight[kgs]dividedbyheight[m]

squared.


95
Table5:Frequencyofmealtypes,locationandtimeofconsumptionfortheRASsample(n=116).
Variable % Variable %
Formofpurchase Timeofday
Takeaway 43.1 Morning(05001159) 10.3
Drivethru 27.6 Midday(12001359) 25.0
Dinein 29.3 Afternoon(14001659) 16.4
Locationeaten Evening(17002029) 35.3
Home 29.3 Night(20302359) 9.5
Restaurant 31.0 Latenight(00000459) 3.4
Car 16.4 Dayofweek
Other 23.3 Monday 5.2
Ordereditemsasamealdeal 58.6 Tuesday 8.6
MealOccasion Wednesday 13.8
Breakfast 6.0 Thursday 15.5
Lunch 36.2 Friday 13.8
Dinner 49.1 Saturday 25.9
Inbetweenmealsnack 8.6 Sunday 17.2

Therewerefewclearpatternsofbehaviourssurroundingfastfoodconsumption.Most
respondentspurchasedtheiritemsinatakeawayordrivethruformat;lessthanathird
ofthesampleconsumedtheiritemsintherestaurant.Mostpeoplereportedconsuming
theirfastfooditemsfordinnerorlunch.Atthetimeofthesurvey,onlyMcDonalds
offeredabreakfastmenuanditisthereforeunsurprisingthattherewasonlyasmall
groupofbreakfasteaters.Frequenciesofdifferentmealoccasionscorrespondedto
frequenciesofthepartofdaywhentheitemswereeaten.Finally,justover40%ofthe
participantsreportedeatingfastfoodontheweekend.

Motivatorsandcontextoffastfoodconsumption.Questionsaboutthereasonsforfast
foodconsumptionandtheactivitiesbeforeeatingthefastfooditemsneededtobecoded
toallowanalysisofthefrequencyofdifferentresponses.Theopenendedresponsesfor
thequestionWhydidyoueatfastfoodonthisoccasion?werecodedintocategories


96
accordingtorecurringthemes.Asmostresponseswereonlyafewwordslong,the
themeswerefairlyapparent.Incaseswherethereweretworeasonsgiven,forexample,
wantedit,itwaseasy,thefirstoftheseoptionswaschosenasthiswasassumedtobea
primaryreason
40
.

Anumberofrespondentsindicatedsimplythattheyatebecausetheywerehungry
(labelledBasicNeeds).Oneinitialthemethatarosefromthisopenresponseitemthat
relatedtofastfoodconsumption,wasanattractiontotraditionalfastfoods.Subthemes
ofthisincludewantingfastfoodasaresultofthetypeoffood.Sometimesthesecravings
werespecificallystated,Ihadacraving,inotherinstancestheywerelessdirectly
stated,wasdrunkandwantedit.Therefore,responsesthemedaroundcravingsand
needswerecombinedintoacategorylabelledAttractiontotheFood.

Thesecondthemewhicharosefromtheopenendedresponsespertainedtoanother
typicalcharacteristicoffastfoods:convenience.Manypeoplestatedsimplyitwaseasy
asareasonforeatingfastfood.Lessobviousexamplesofreasonsencompassedbythe
GeneralConveniencemotivatorincludednothavingtocook.Thereseemedtobea
distinctsetofitemsthatrepresentedareasontypifyingadifferenttypeofconvenience.
Theseitemsincluded:convenientlocation,accesstotoilets,bankingfacilitiesandthe
convenientpricingofthefood.Theseallrelatetothefastfoodenvironmentwhichthe
restaurantcreatesandwerethereforelabelledIncidentalConvenience.

Thefinalreasoncitedintheopenresponseitemsfocusedonsocialdemands;
participantsindicatedthattheyateitbecauseotherpeoplewanteditorboughtitfor

40
TheonlytimethiswasnotthecasewaswhenthefirstreasonwasIwashungry.Thisreplywas
generallyuninformativeastothereasonsforeatingfastfoodinsteadofanyotherfood.Wherethisanswer
wastheonlyresponselistedbyaparticipantitwascodedassuch.


97
them,ortheyhadtoeatittobepartofasocialoccasion.Assuch,theseresponseswere
categorisedasbeingmotivatedbyOtherPeople.

Forthesecondopenresponseitemactivitiesprecedingfastfoodconsumptionthe
sameprocesswasusedforcoding.Again,descriptionsoftheactivitiesbeforethefast
foodconsumptionweregenerallyonlyafewwordslong;Iwasshopping,atthepub
andsoon.Overhalfofthesampleindicatedthattheywereoutandaboutbefore
consumingthefastfooditems.In35oftheseresponses,participantswrotethatthey
wereactuallyonthewaysomewhereelse
41
accordinglyitemsincludingdescriptionof
thisactivitywerecodedasBetweenActivities.

Nineteenpeopleinthesamplewrotethattheywereoutparticipatinginsomekindof
socialactivity,includingspendingtimeatthepub,walkingaroundthefair(RAS)andso
on.Eighteenparticipantsindicatedthattheywereshoppingbeforetheirfastfood
consumption.Theseinstancesweregenerallystatedsimplyasshopping.Considering
thenumberofpeoplewhorespondedthisway,Shoppingwasdefinedasadistinct
category.Elevenpeoplewrotethattheywereparticipatinginamoderatelevelof
physicalactivity(suchasanorganisedsport).Althoughthisactivitywasnominatedby
lessthantenpercentofthesample,itwasdistinctand,forthepurposesofcoding,called
ModeratePhysicalActivity.Forotherinstancesthatincludedsocialising,thelabelSocial
Activitieswasused
42
.


41
Twentyoneofthosepeoplewhoindicatedthattheyweretravellingwrotethattheyweregoingtoor
fromspecificlocations.Workwasthemostcommondestinationspecified(10people).Elevenparticipants
indicatedthattheyweredrivingwithnospecifieddestination.
42
Althoughthepresenceofotherpeoplewasnotexplicitlystatedinmanyofthese,forexample,atthe
pub,theywerestillassumedtobesocialactivities.Thedataonthenumberofotherpeoplepresentwhen
eatingthefastfoodindicatedthatnoneofthepeopleengaginginsocialactivitiesbeforeeatingdined
alone.


98
Asidefromthoseoutandabout,20peopleindicatedthattheyweredoingactivities
typicallycompletedathome.Theseincludedhomeworkorstudy,sleeping,watching
television,readingandthinking.Alloftheseactivitiesrequiredlowphysicalactivity
levels.Accordingly,thecategoryNotMuchwasusedtodescribetheseactivities.

Thefinalgroupofparticipantsindicatedthattheyweredoingactivitiesoutsideoftheir
homes.Twentythreepeoplewrotethattheyhadbeenatworkoratuniversity.These
responseswerecodedintoadistinctgroupWork.Fourpeoplehadmissingdatafor
theactivitiesprecedingtheirfastfoodconsumption.

Followingtheidentificationofthesethemes,motivatorsforfastfoodconsumptionitems
werecodedintothefollowingreasons:BasicNeeds,AttractiontotheFood,General
Convenience,IncidentalConvenienceandOtherPeople.Theactivitiespreceding
consumptionwerecategorisedas:BetweenActivities,SocialActivities,ModeratePhysical
Activity,Shopping,Work(oruniversity)andNotMuch.Thefrequenciesofresponsesin
eachofthecodedcategoriesforbothopenresponseitemscanbeseeninTable6(over
thepage).

Tovalidatethescoringofthesethemes,therawresponsesweregiventoasecond
researcher.Shewasinformedofthecategoriesandthebasicprocessforcodingand
askedtocodealltheresponses.Agreementbetweenthetwocodingsetswas100%.

Analysisofthefreeresponsedataaccordingtothesecategoryassignmentsindicated
thatenergyconsumptionfromthefastfoodeatingoccasionreporteddidnotvary
accordingtostatedmotivationtoconsumeoractivitiesprecedingconsumption(p>.05).



99
Table6:Codedcategoriesandassociatedfrequenciesforbothopenresponseitems.
ActivitiesPrecedingConsumption(n=111)
% Examples oftypicaldescriptions
BetweenActivities 25.9 Drivingintothecity;Headinghomefromwork
SocialActivities 17.2 Spendingtimewithfriends;Atacarnival
Shopping 15.5 Shopping;Buyinggroceries
NotMuch 14.7 Watchingtelevision;Sleeping
Working 12.9 Atwork
ModeratePhysicalActivity 9.5 Playingnetball;Housecleaning
ReasonsforConsumption(n=116)
% Exampleoftypicaldescriptions
GeneralConvenience 32.8 Itwaseasy;Didnthavetocook
Basicneed 24.1 Iwashungry;Itwasmealtime
Incidentalconvenience 17.2 Itwasthere;Itwaslate,nootheroptions
Attractiontofood 12.9 Cravingfattyfood;Hadbeendrinking
Otherpeople 12.9 Someoneboughtitforme:Everyonewasdoingit

4.3.5 Fastfoodseatenandthepotentialforoverconsumption
ChapterTwoshowedthatorderandbrandchoicecouldpotentiallychangetheamount
ofdietaryenergyconsumed.Thefollowingsectiondescribesanalysesdesignedtoassess
whetherthesenutritionaldifferenceswouldbereflectedintheactualordering
behavioursreported.

Differencesinenergyandfatintakebetweenchains.Theaverageenergyconsumedfrom
thefastfooditemswas3564KJ(SD=1812).Giventheindicationinthenutritional
profiling(ChapterTwo)thatsomefastfoodcompaniesoptionsweremoreenergy
densethanothers,energyconsumedwasassessedbetweeneachfastfoodbrand.The
highestproportionsofpeopleinthesamplesreportedconsumingitemsfromHungry
JacksfollowedbycloselybyMcDonalds.Theamountofenergyconsumedbythebrand
offastfoodeatencanbeviewedinTable7.



100
Table7:Energy(KJ)andfatcontentoffastfoodconsumedbypilotparticipants

.
Variable
KJ

%KJfromfat
n M SD M SD
Wholesample 116 3564 1812 39.36 8.99
McDonalds 38 2933
ad
1328 36.25
bc
9.84
HungryJacks 40 4338
c
1796 43.29
a
8.99
KFC 23 3885
cd
2028 39.34
ac
5.66
DominosPizza 12 2184
ab
1614 34.71
c
5.66
RedRooster 3 4320 869 45.15 8.87

Sharedsuffixesindicatedifferencesthatarenotsignificantatthe.05level

Analysisofvariance(ANOVA)comparingenergyintakebetweenthedifferentbrands
43

revealedsignificantdifferencesexisted,F(3,109)=7.53,p<.001.Posthoccomparisons
usingBonferronisadjustmentsrevealedthatthosewhoateatHungryJacksate
significantlymoreenergythanthoseatMcDonaldsorDominosPizza.Thosewhoateat
KFCalsoatesignificantlymorethanthoseeatingDominosPizza.

Tofurtherassessenergydensityoftheitemsconsumed,thepercentageofenergyfrom
fatwascalculated;everygramoftotalfatconsumedfromthefastfooditemswas
convertedinto37KJofenergy(CodexAlimentariusCommission,2006).Thetotal
kilojoulesconsumedfromfatwerethendividedbytheoverallenergyintake.Theextent
towhichenergyfromfastfoodswasdominatedbyfatvariedsignificantlybetweenfast
foods(excludingRedRooster),F(3,109)=5.78,p=.001.DunnettsT3
44
posthoc
comparisonswereusedtoexplorewherethisdifferencelay.Peopleconsumedahigher
percentageoffatinitemsconsumedfromHungryJackscomparedtothosefrom
DominosPizzaorMcDonalds.Meansandstandarddeviationsforthepercentageoffat
fromthefastfooditemsconsumedcanalsobeenseeninTable7 (above).

43
GiventhesmallnumberofpeoplewhohadeatenRedRooster,theseparticipantswereexcludedfrom
thecomparison.
44
Levenestestrevealedaviolationoftheassumptionofvariance.


101

Makinghealthychoices.ThenutritionanalysisinChapterTworevealedthatsomeofthe
majorfastfoodchainsofferedhealthyalternativesthatcoulddramaticallyreduce
energyintakefromfastfooditems.Toexplorewhethertheseoptionswerebeing
selectedbyconsumers,orderinformationwasanalysedforitemseatenfrom
McDonaldsandHungryJacks
45
.Itemsthatwereadvertisedaslessthan10gramsoffat
perservewerecodedashealthychoiceoptions;thiscriterionwasusedbyMcDonalds
andHungryJackstoadvertisehealthyalternativesatthetimeofthesurvey.Ofthe13
primaryburgerandbaguetteoptionsavailableonthemenuatHungryJacks,onewas
lowfat(7.7%).McDonaldsoffered16burgerorbaguetteoptionsandfiveofthese
containedlessthantengramsoffat(31.3%).Orderdescriptionsindicatedthatsix
ordersintotalincludedalowfatburger/baguetteitematMcDonalds(20.0%ofall
orderscontainingaburgerorbaguette).Oneofthebaguetteswasaccompaniedbyfries.
OnlyoneoftheordersatHungryJacksincludedthelowfatbaguettefromHungryJacks
(2.6%ofallorderswithaburgerorbaguette).Theparticipantswhohadorderedlowfat
itemswereofvaryingdemographiccharacteristics;therewerefourfemalesandthree
males,ranginginagefrom19to61,residinginareasrangingfromthehighestdecilesof
disadvantagetothelowest.Therewerenodifferencesinanyofthedemographic
characteristicofthissamplecomparedtothosewhoorderedtraditionalfastfood
items.

Dailyimbalance.Toassesswhetherparticipantsatefastfooditemsasmainmeal
replacements,reportedmealoccasionswereassessed.Mostofthefastfoodeating
occasionswereclassifiedasmeals.Therewasnodifferenceintheamountofenergy

45
Orderdescriptionswerecodedaccordingtowhethertheycontainedaburgerorbaguette,chips,a
chickenitem(nuggetsorchickenportions)and/orabeveragetogiveanindicationofwhatstyleofitems
werepopularateachchain.AsKFCandDominosPizzadidnotofferanyhealthyalternativesatthetime
ofthesurveyandtherewereonlylimitedordersfromRedRooster,ordersfromthesechainswere
excludedfromtheanalysis.


102
consumedatmealoccasions(M=3652,SD=1837;breakfast,lunchordinner)as
comparedtosnackingoccasions(M=2636,SD=1244),t(114)=1.71,p>.05.

Inordertoexplorewhethertheitemseatenwerereplacementsforotheritems,and
whethertherewereanycompensatoryeatingbehavioursinthedaywhenthefastfood
intakeoccurred,theopenendedquestionaboutalterationstodietarypatternswere
analysed.Overhalfoftherespondents(n=67)chosetoleavethisitemaboutalterations
todietarypatternsblank.Correspondingtothelackofdetailineatingchange,themodal
responsefortheamountoffoodeatenduringthedaywastypical(aratingof4).
Relativetothespecificeatingoccasion,themodalresponsewasalsotypical.Overhalf
ofthosewhorespondedtothequestionaboutchangesinintakethroughouttheday
indicatedthattheyhadskippedormissedamealonthedayoftheirfastfood
consumption(20.7%ofthewholesample).Thetotalenergyconsumedbypeoplewho
reportedskippingameal(n=24,M=3371,SD=1804)didnotdiffersignificantlytothe
amounteatenbythosewhodidnot(n=92,M=3615,SD=1820),t(114)=.59,p>.05.
Otherrespondentstothisquestionindicatedthattheyhadeatenmorefoodthanusual
(n=6),thatthemealoccasionoccurredinadifferenttimeofdaythanusual(n=6),and
thatonthedayofconsumption,theyatemoreconveniencefoodsthantheyusually
would(includingthefastfoodconsumptiontheydetailed;n=9).

4.3.6 Theinfluenceofmicroenvironmentalpressuresonfastfoodconsumption
Measuresincludingappealofupsizing,tendencytoorderthesameitems,purchasinga
mealdealandenjoymentoffoodwereanalysedtoexploreanypotentialmicro
environmentalinfluencesonfastfoodconsumption.



103
Manyparticipantsdescribedtheconceptofupsizingasunappealingwiththemodal
responsebeing1(veryunappealing).Themeanscoreforthisitemwasalsobelowthe
midpoint(M=2.78,SD=1.97).Incontrast,themodalresponseforthetendencytoorder
thesameitemswas7(verytrueofme).Themeanwasabovethemidpoint(M=4.51,
SD=2.03).Therewasnorelationshipbetweenthetendencytoorderthesameitemsand
frequencyofpatronage,r(116)=.03,p>.05ortheamounteatenonthereported
occasionr(116)=.12,p>.05.However,amoderatepositiverelationshipexistedbetween
theappealofupsizingandfrequencyofpatronage,r(116)=.26,p<.05aswellasthe
amounteaten,r(116)=.32,p<.001.Therelationshipbetweentheenjoymentofitems
eatenandthetotalenergyconsumedwasnotsignificantforeithertheamounteaten,
r(116)=.11,p>.05orfrequencyofconsumption,r(116)=.13,p>.05.

Overhalfoftheparticipantsorderedtheiritemsaspartofamealdeal.Havingordereda
mealdealincreasedtheamountofkilojoulesfrom2657(SD=1580)to4205
(SD=1697).Thisdifferencewassignificant,t(114)=4.98,p<.001.Crosstabswereused
tocomparethenumberofpeopleeatingmealdealsacrossthedifferentbrandsoffast
food(excludingRedRooster).Thenumberofpeoplehavingmealdealswas52.6%,
75.0%,56.5%and25.0%forMcDonalds,HungryJacks,KFCandDominosPizza
respectively.ChiSquarerevealedasignificantdifferencebetweenthegroups,

2
(3,110)=10.60,p<.05.Ahigherpercentageofpeoplethanexpectedorderedmeal
dealswheneatingHungryJacks;lesspeoplethanexpectedreportedorderingameal
dealwhenhavingDominosPizza.



104
Thesizeofthemealdeals
46
wasalsoexamined.Acrossallchains,amajority(61.8%)of
the68peoplehavingamealdealhadaregularormediumsizedmeal.Thiswasfollowed
bylargemeals(22.1%)andsmallmeals(16.2%)respectively.Therewasamoderate
relationshipbetweenthesizeofthemealpurchasedandtheappealofupsizing,
r(68)=.27,p<.05.Fiftythreepointfourpercentofpeoplewhopurchasedalargemeal
ratedtheappealofupsizingbelowthemedian.

4.3.7 Socialinfluencesandfastfoodconsumption
Themeansforthemeasuresofsocialinfluence,theratingsofatmosphereandthetime
spenteatingcanbeseeninTable8.
Table8:Meanresponsetoitemsmeasuringsocialinfluenceoneating.
Question M SD
Howawarewereyouofthepeoplearoundyouwhenyou
wereeating?
2

4.81 2.01
Howmuchfastfooddoyouthinkthosearoundyouate?
1
4.40 1.59
Howawareoftheamountyouatedoyouthinkthose
peoplearoundyouwere?
2

3.35 1.90
Howconcernedwereyouaboutwhatthosepeoplearound
youthoughtofyou?
1

2.18 1.71
Approximatelyhowlongdidyouspendeating?

(inminutes)
14.18 8.34
Pleaseratehowsociabletheatmospherewaswhenyouate
thisfastfood
3

3.84 1.84
Pleaseratehowpleasanttheatmospherewaswhenyou
atethisfastfood
4

4.53 1.57

n=116;

n=91meansonlyforthosewhoatewithothers
Ratingscaleresponseformat:
1
1=alittle,7=alot;
2
1=notatallaware;7=veryaware;
3
1=notat
sociable;7=verysociable;
4
1=notatpleasant;7=verypleasant

Mostpeoplereportedeatingwithoneotherperson(27.6%).Diningalonewasthenext
mostfrequentresponse(21.6%)followedbygroupsofthree(19.8%),four(17.2%)and

46
Mealdealsizewasdefinedaccordingtothesizeofthefriesordrinkwithprioritygiventofrieswhen
thetwovaried.


105
largegroups(13.8%).Therelationshipbetweengroupsizeandtheamountofenergy
consumedwasnotsignificant,r(116)=.18,p>.05.Somepreviousauthorshavenot
supportedalinearrelationshipbetweenthenumberofpeoplepresentandtheamount
eatenandsuggestedthatasimpledichotomyofaloneversuswithothersismore
discriminating(Pliner,Bell,Hirsch&Kinchla,2006).Analysescomparingthosewhoate
incompanytothosewhoatealonewasstillnonsignificant,t(114)=.97,p>.05.Those
whoatewithothersdidspendlongereating(M=11.00,SD=6.51)thanthosewhoate
alone(M=15.05,SD=8.61),t(114)=2.19,p=.03.Table9showsthecorrelations
betweenthesocialfacilitationvariables.

Table9:Correlationsbetweenenergyconsumedandratingsoftheeating
environmentforwholesample(n=116).
Variable 1
KJ
2
Size
3
Time
4
AtmosS
1.Energy(KJ) 1.00
2.Groupsize(Size) 0.11 1.00
3.Timespenteating(Time) 0.14 0.10 1.00
4.AtmosphereSociability(AtmosS) 0.04 0.54* 0.16 1.00
5.AtmospherePleasantness 0.13 0.24* 0.14 0.67*
*significantatp<.01level

Theatmosphericmeasuresweresignificantlyandpositivelyrelated.Groupsizewasalso
positivelyrelatedtoimprovedratingsofatmosphere.Theseweretheonlysignificant
relationshipsbetweenthesocialfacilitationvariables.Althoughtherewasadifferencein
thetimespenteatingbetweenparticipantswhoatewithothersversusthosewhoate
alone,therewasnoevidenceofalinearrelationshipbetweenthesevariables.The
relationshipbetweentimeeatingandgroupsizeisplottedinFigure6.


106

Itisclearthatalthoughthereappearedtobeapositive,linearrelationshipbetweenthe
peoplepresentandtimeeatingasgroupsizeincreasedfromalonedinertoagroupof
threediners,thisrelationshipreversedasthegroupsizeincreasedfromthreemembers
tolargergroups.

Assessmentoftheperceptionsofthenormsthatsurroundedtheeatingoccasionwas
restrictedtorespondentswhoatewithatleastoneother
47
.Theonlynormative
influencethatrelatedtotheamountconsumedwastheperceptionofhowmuchfood
thosepeoplearoundwereeating,r(91)=.26,p<.05.Normativeperceptionswere
associatedwitheachother;ratingsofhowmuchpeoplethoughtothersaterelated
moderatelyandpositivelytogeneralawarenessofthosearound,r(91)=.33,p<.01,and
moderatelywiththelevelofconcernforwhatthosearoundthoughtofthem,
r(91)=.45,p<.01.Concernforwhatothersthoughtoftheparticipantwasalsoweakly

47
Chisquarecomparisonsandttestsindicatedthatwerenostatisticaldifferencesbetweenthosewhoate
withothersforanyofthedemographiccharacteristics.
Figure6:Meantimespenteatingfordifferentgroupsizesat
thereportedfastfoodeatingoccasion(n=116).


107
andpositivelyrelatedtohowmuchtheparticipantsthoughtthosearoundthemate,
r(91)=.23,p<.05.

Theninecategoriesofinterpersonalrelationshipdescribedearlierwerereducedto
moresimplegroups.Parents,siblings,otherrelativesandchildrenwerecombinedintoa
categoryofFamily.Partnerwasleftasadistinctcategory.Friends,membersofasocial
groupandcolleagueswerecombinedinaFriendscategory.Allthoseparticipantswho
hadlistedanothercategorywereputintothemostappropriateofthesegroups.Of
thoseeatingwithothers,mostateinthepresenceofamember/softheirfamily
(n=65);thiswasfollowedbypeopleeatingwiththeirpartner(n=34),andfriends
(n=32)
48
.

Asimplemultipleregressionmodelwasconstructedtoassesstheeffectsofthesocial
factorsonthosewhoatewithothers.Allthenormativeitemsandthesocialfacilitation
items(groupsize,timeeatingandpleasantnessofatmosphere
49
)wereincluded.Limited
demographicswerealsocontrolledfor(sex,ageandsocioeconomicdisadvantage).The
onlysocialfactorsignificantlyinfluencingtheamounteatenwastheperceptionofhow
muchtheotherspresentwereeating.Thiswaspositivelyassociatedwithactualintake.

Giventhepotentialinfluencethatthetypeofpersonpresentmayinfluenceeating
behaviours,asecondmodelwasconstructed.Thismodelincludedthesamesocial
factorsbutthegroupsizevariablewasreplacedwiththenumberoffamilymembers
present,thenumberoffriendspresentandwhethertherewasapartnerpresent.
Includingmoredetailedinformationaboutthepeoplepresentimprovedtheamount

48
Thesenumbersdonotsumtothetotalsample(91)asmanyparticipantsateinthepresenceofmultiple
peoplewhohaddifferentrelationshipswiththeparticipant,forexample,oneparent,onefriend.
49
Sociablenessofatmospherewasnotincludedasitwasmoderatelyrelatedtoboththeratingsofthe
pleasantnessoftheatmosphereandgroupsize.


108
variancethatthemodelaccountedforbyroughly5%.Italsorevealedthatwhowas
presentcouldhaveeffectsontheamounteaten;thenumberoffamilypresenthada
significant,negativerelationshipwiththetotalamounteaten.Althoughthepresenceofa
partnerhadapositiveeffect,itwasnotsignificant.Theperceptionofhowmuchthe
otherspresentateremainedasignificantpredictoroftheenergyintakefromfastfood
itemsatthesingleeatingoccasion.Descriptionsofbothmodelsandbetavaluescanbe
seeninTable10.

Table10:Modelsforsocialvariablespredictingenergyintakeatafastfood
eatingoccasion.Totalvariance(AdjustedR
2
),betasandalphasarepresented
foreachpredictor(n=91).
Predictors
Model1
AdjustedR
2
=.119
Model2
AdjustedR
2
=.167
Beta Alpha Beta Alpha
Timespenteating .00 .98 .02 .85
Atmospherepleasantness .08 .46 .13 .25
Generalawarenessofothers .02 .84 .07 .57
Amountotherspresentate .29 .01 .32 .01
Perceptionofhowmuchothers
thoughtyouate
.05 .69 .08 .57
Concernofjudgement .03 .81 .02 .87
Sex .19 .09 .22 .06
Age .17 .12 .21 .06
IRSD .22 .04 .21 .07
GroupSize .15 .13
Numberfriendspresent .02 .88
Numberfamilypresent .29 .04
Partnerpresent .13 .27

4.4 Discussion
Thepurposeofthisstudywastoundertakeapreliminaryexplorationoffastfoodeating
behaviourswiththegoalofgainingabetterunderstandingofthevarietyoffactorsthat
couldincreasetheamountoffastfoodeatenonasingleoccasionandhowthesemight
bestbeoperationalised.Afurtheraimwastotesttheacceptabilityandusabilityofthe


109
FFSandtheeasewithwhichitcouldbecompletedwhenonline.Todothis,anInternet
surveywasdevelopedandadministeredatalocalagriculturalfairtoasmall
conveniencesampleofpeoplewhohadconsumedfastfood.

4.4.1 TheFastFoodSurvey:Strengthsandlimitations
Itwasclearthatthesurveyrepresentedanacceptableinstrumenttomeasurevarying
aspectsoffastfoodconsumption.DatacollectionviatheFFSwasefficient;participants
appearedtohavelittledifficultyengagingwiththeprogramandmanywereinterested
inthenutritionalfeedbackthatwasprovided(50%ofrespondentsviewedthe
nutritionalinformation).Demographicvariablesandmeasuresofsocialinfluenceinthe
FFScouldbeusedtosuccessfullypredictenergyintakeatasingleeatingoccasion.
Althoughasmallpercentageofthetotalvariancewasaccountedforineachofthe
regressionmodels,thefactthatthesefactorscouldsignificantlypredictintakeindicates
thatfurtherinvestigationiswarranted.

Thesurveyattemptedtoaccountforalargevarietyofcorrelatesforfastfood
consumption.Althoughmostmeasuresworkedsuccessfully,resultsfromotherswere
lessinformative.Specifically,theitemsaboutvariationingeneraldietaryintakeonthe
dayoffastfoodconsumptionwereleftlargelyunansweredwhichmeantthatresponses
tothisitemcouldonlybeusedinadescriptivesense.Thismeantthattheimpactoffast
foodintakeondailyeatingpatternscouldnotbeestablished.Nevertheless,skipping
mealswascommonlyreportedasachangetousualeatingpatterns.Totruly
understanddailyenergybalancing,foodintakewillneedtobemeasuredbymore
sophisticateddietarymeasuressuchasfoodfrequencyquestionnairesordietdiaries,or
throughethnographicassessmentswhichhavehigherparticipantburden.



110
Althoughmuchpreviousresearchhasexploredfrequencyoffastfoodpatronage,itis
importanttoidentifythedriversofenergyintakeatanyonemealatthemicrolevelas
aninitialsteptounderstandingfastfoodbehaviours.Exploringtheamountoffastfood
eatensimultaneouslywiththereportedfrequencyoffastfoodconsumptionisanovel
aspectoftheFFS.Analysingbothmeasuresrevealedthattheinfluencesthatpredict
energyintakeatasinglefastfoodvisitdifferfromthosethatpredictdifferencesin
frequencyofattendanceatafastfoodrestaurant.

Thesurveywascompletedbyasmall,conveniencesamplewhichwasnotrepresentative
ofthewiderpopulation.Thislimitsextrapolationoftheseresultstothegeneral
population.Furthermore,therewassomeevidencethatthefastfoodconsumption
reportedbytheseparticipantsmayhavebeeninfluencedbythecontextofthesurvey.
ManyreportedconsumingHungryJacks(theonlyofthefastfoodchainoperatingatthe
RAS)andseveralindicatedthatconsumptionoccurredwhileatthefairwhenanswering
theopenendedquestions.Thismeansthattheconsumptioninthissamplemaynotbe
representativeoftypicalfastfoodeatingoccasions.Therefore,althoughtheseresults
revealedsomeinterestingfindings,theyneedtobeinterpretedcautiouslyuntil
corroboratedbylarger,morerepresentativesamples.

4.4.2 Demographiccharacteristicsandfastfoodconsumption
Previousauthorshavenotedrelationshipsbetweenindividualdemographicsandthe
regularityoffastfoodconsumption(Mohretal.,2007;H.Schroderetal.,2007).The
currentsurveyfailedtosupportthesefindingsnoneofthedemographicfactors
measuredsignificantlypredictedfrequencyoffastfoodpatronage.Thesampleinthe
currentstudywassmallwhencomparedtothatofpreviousstudies(thesurveyreported
byMohretal.includedalmost20,000).Thismayhavelimitedthepredictivepowerof


111
themodelforfrequencyoffastfoodconsumption.Preliminarycorrelationcomparisons
suggestedarelationshipbetweenwithhigherBMIandhigherfrequencyofpatronage
(orviceversa)whichsupportspreviousstudiesthathaveshownthatfastfood
consumptionmaybeassociatedwithBMI(Duffey,GordonLarsen,Jacobs,Williams&
Popkin,2007).

Thedemographicfactorsaccountedformoreofthetotalvarianceoftheenergy
consumedatasingleeatingoccasionthanforfrequencyoffastfoodpatronage.There
wasaneffectofgenderontheamounteatenonthesingleoccasionassessed.Males
consumedmorethanfemalesonthesingleeatingoccasion,afindingconsistentwith
theirhigherdailyrequirements(Kellett,Smith&Schmerlaib,1998).Althoughsex,age
andoccupationalstatuswereinitiallysignificant,whenconsideredincombinationwith
otherdemographicvariables,thesevariablesfailedtocontributesignificantlytothe
predictionofenergyintake.Theonlysignificantdemographicpredictorofenergyintake
atafastfoodeatingoccasionwasthelevelofsocioeconomicdisadvantageofthe
participantsresidentlivinginmoredisadvantagedareasassociatedwithhigherfast
foodconsumptionatasingleeatingoccasionandviceversa.Thereisnoexisting
literatureonhowdemographicfactorsinfluencetheamounteatenatasinglefastfood
eatingoccasion.Perhaps,peoplefrommoredisadvantagedareashaveahigher
appreciationofthefinancialbenefitsofeatinggreateramountsoffoodatfastfood
restaurants.AsCameronSmith,BilsboroughandCrowe(2002)noted,foronlyasmall
increaseinprice,energyintakecanbeincreasedsubstantially.

Finally,althoughonlytrendingtowardsignificance,apositiverelationshipbetweenthe
frequencyoffastfoodconsumptionandenergyintakeatasingleeatingoccasionwas
found.Understandinganypotentialrelationshipbetweenfrequencyoffastfood


112
consumptionandtheamountconsumediscrucial;ifpeoplewhovisitfastfood
restaurantsmorefrequentlyalsoorderandconsumemorefood,itislikelythatthis
associationwillresultinoverconsumptionfromfastfooditems.

4.4.3 Behaviours,motivatorsandcontextsurroundingfastfoodconsumption
Mostpeopleinthecurrentsamplereportedeatingfastfoodonceaweekoroncea
fortnight;ratescomparabletothosereportedinlargersurveys(FOODweek,2008).
Multiplestudieshavereportedthatconvenienceandcostdrivefastfoodconsumption
(Scheibehenne,Miesler&Todd,2007;M.J.ASchroder&McEachern,2005).Thecurrent
findingssuggestthatconveniencecanbebothgeneralandincidental.Theresultsalso
suggestedthatthefooditselfmaybeastrongmotivatorforasubsetofconsumers.
Interestingly,costwasspontaneouslycitedasareasonforconsumptionbyonlythreeof
therespondents.

ThecontextoffastfoodconsumptionreportedbyparticipantsintheFFSsupported
previousindicationsfromsurveyswithundergraduatestudentsthatfastfoodpurchases
aremostlyimpulsive(M.J.A.Schroder&McEachern,2005).Itseemedthatfew
respondentsplannedtoeatfastfoodandthosewhodidusuallyindicatedthatitwas
partoftheirroutine(i.e.Fridayisjunkfoodnight).Approximately16%ofpeopleate
theirfastfooditemswhileinthecarandcloseto30%purchasedthefoodwhileinthe
car.Boththesefrequenciesindicatethewayinwhichpeoplehavesatisfiedtheirneed
forconveniencewithfastfoods.Thereisnogreatmysterytofastfoodconsumption,itis
easyandsatisfiesitslabel;itisfast.Inaworldwithincreasingtimescarcity,wherefood
andsharingmealsmaybelessintegraltotheculture,itisnotsurprisingthatastyleof
eatingwhichisnotfocusedontheentiretyoftheexperience,butcanbeslippedin
betweenothercommitmentsandactivitiesmaybebecomingincreasinglypopular.


113

4.4.4 Fastfood:Potentialforoverconsumption
Thereweresignificantdifferencesintheamountofenergyconsumedbetweenthefast
foodchains.ThefactthatthemealsatHungryJackscomeinlargerportions(discussed
inChapterTwo)was,mostlikely,oneofthereasonsforthehigherenergyconsumption
atthischaincomparedtomealsfromMcDonaldsandDominosPizza(whichsellless
energydensemealsandsmallerportions).TheitemsconsumedfromHungryJackshad
higherlevelsoffatandweremorecommonlypurchasedasmealdealscomparedto
otherchainswhichmayalsoaccountforthehigherenergyconsumedatHungryJacks
comparedtosomeotherchains.

Orderscontaininghealthychoiceitemswerenotcommonplace.InMcDonalds,there
weremorelowfatoptionsavailableandmorepeopleorderedthesetypesofitemsthan
atHungryJacks.Thisagainsuggeststhatthepurchasingenvironmenthasacrucial
impactonbehaviourwhichisconsistentwiththeproposal,inthemodelofnormative
overeating,thatenvironmentalpressurescanpromoteoverconsumption(Herman,
Rother&Polivy,2003).Suchasmallnumberofpeopleorderedthesestylesofitemsthat
furtheranalysesontheorderscontaininghealthychoiceswererestricted.

Microenvironmentalpressures,suchasmealdealsandmenuoptions,wereassociated
withintake.Thewaythatmealsarepromotedandconstructedislikelytobean
importantfactorpromotingincreasedintake.

Althoughthepurchaseofmealdealswascommonplace,overall,respondentsfoundthe
conceptofupsizingunappealing
50
.Halfofthepeoplewhoreportedeatingalargemeal

50
Onlymealdealscanbeupsized.


114
(accordingtotheirorderdescriptions)alsoratedtheappealofupsizingbelowthe
median.Thissuggestssomeincongruitybetweenattitudesandbehaviour,asthe
defaultsizedmealdealatmostfastfoodrestaurantsismediumorregular.Therefore,
peoplewhoatelargemealsdid,ineffect,upsizethemduringtheorderingprocess.An
aspectofsocialdesirabilitymayhavebeeninfluentialinresponsestothisquestion,
especiallyinthecontextofthenegativepublicitythatupsizinghasreceived(especially
throughthefilmSupersizeMe
51
).

4.4.5 Socialinfluencesandfastfoodconsumption
Fastfoodconsumptionwasassociatedwithsocialfactors.Agroupofrespondents
directlyidentifiedotherpeopleasthereasonfortheirfastfoodconsumptioninthe
openendedresponseformat.Citingotherpeopleasthereasonforconsumptionwas
usuallyaccompaniedwithsuggestionsofafeelingofexternallocusofcontrol.For
example,everyoneelsewasdoingitorsomeoneelseboughtitsoIhadtoeatit.The
groupofpeopledirectlyidentifyingothersasthereasonfortheirconsumptionwasa
small,inlinewiththefindingsofVartanianetal.(2008)whichsuggestedthatoften
participantsdonotcitesocialfactorsasinfluencingtheirbehaviour.Nevertheless,the
fastfoodconsumptionreportedoftenappearedtobesocialinnature(otherpeoplewere
presentinmanyoccasions).ReddanddeCastro(2006)reportedthatinfreeliving
environmentspeoplenaturallytendedtoeatwithothers.Consistentwiththis
observation,intheFFS,amajorityofrespondentsreportedeatingwhileotherswere
present.Furthermore,thepresenceofotherswasassociatedwithmoresociableand
pleasanteatingenvironmentsforrespondentswhenaskedtoratetheatmosphereatthe
meal.


51
Themovieisactuallyaboutsupersizingwhichisupsizingtofriesandadrinkonesizelargerthanlarge.


115
Overall,socialfactorssuccessfullycontributedtothemodelpredictingintakeatasingle
fastfoodeatingoccasion.Howthesefactorsrelatetodifferentsocialinfluencetheories
willbediscussedinthefollowingsections.

Normativeinfluence.Oneofthestrongestsocialpredictorsofenergyintakeataneating
occasionwastheperceptionoftheamounttheothersaroundate.Thisnormwas
designedtoassessthelevelofmatchingreducingorincreasingintaketomatchthe
otherpeoplesintakeataneatingoccasion(Roth,Herman,Polivy&Pliner,2001).
Previousfindingssupportingtheexistenceofmatchingnormshaveshownthatwhen
eatinginthepresenceofpeoplewhoeatmorefood,peoplewillmodifytheirown
behaviourtomatchit.Theactualamountthatthecoeatersconsumedattheeating
occasionrecalledintheFFScannotbedeterminedfromthesedata.Itmaybethatpeople
inthesamplewhoatemorefoodjustifiedtheirincreasedconsumptionbyadjustingthe
perceivednormoftheenvironment.Although,itisdifficulttomeasuretheperceptionof
normsinanenvironmentthroughasurvey,itisnotimpossible.Forexample,theTheory
ofPlannedBehaviourincludesassessmentofsocialnormsthroughsurveymethods
(Ajzen,1985).

Socialfacilitation.Therewasnorelationshipbetweenthenumberofpeoplepresentand
theamounteaten,anobservationcontrarytothatreportedbydeCastroandBrewer
(1992).Therewasalsonorelationshipbetweenthenumberofpeoplepresentandthe
timespenteating,whichisalsocontrarytosocialfacilitationfindings(J.M.deCastro,
1990,1994;J.M.deCastro&Brewer,1992).Therelationshipbetweenthenumberof
peoplepresentandtimespenteatingappearedtobepositiveonlyuptothepresenceof
threeotherpeople.Itmaybethatasgroupsgetlarger,theeffectofotherpeoplebecame
morediffuseandreducedtheeffectoftimeextensionthathasbeenobservedin


116
previousstudies.Onlyasmallproportionofrespondentsateinlargegroupsinthis
preliminarystudyand,accordingly,datafromalargersampleisrequiredtounderstand
therelationshipbetweenthetimeeatingandthenumberofpeoplepresentthatwas
observedinthisstudy.Astherewasadifferenceinthetimespenteatingbetweenthose
eatinginagroupversusthoseeatingalone,thesedataareatleastpartiallyconsistent
withthetimeextensionhypothesis.

If,aswassuggestedbyFeunekes,deGraafandvanStaverens(1995)model,increased
energyintakeisindirectlyinfluencedbythepeoplepresentthroughthetimespent
eating,evidenceofarelationshipbetweentheamounteatenandnumberofpeople
presentmayhavebeenmissedinanalysesmeasuringonlydirecteffects.Itisalso
possiblethattherestrictedvarianceintimetakentoeatafastmealensuredminimal
effectsfromsocialvariables.Again,furtherresearchwillbeneededtoclarifyifthisis
indeedthecase.Moredetailedexaminationofthepatternoftheseassociationsandhow
theyimpactonfastfoodconsumptioniswarranted.

Interpersonalrelationshipsandfastfoodconsumption.Aswassuggestedintheliterature
(J.M.,deCastro,1994;Feunekes,deGraaf,Meyboom&vanStaveren,1998;Verlegh&
Candel,1999),itappearsthatinterpersonalrelationshipsmaybeanimportant
moderatoroftheeffectsthatotherscanhaveontheamounteaten.Analysinggroupsize
alonerevealedlittledirecteffectontheamounteatenforthoseeatingwithothers.When
interpersonalrelationshipswereaddedtothismodel,itnotonlyimprovedthetotal
varianceexplainedforenergyintake,butalsorevealedthatsomepeopleincreased
intake(i.e.,partners)whileothersdecreasedit(i.e.,familymembers).Althoughthe
effectofpartnerswasnotasignificantone,theeffectoffamilywas.Thus,evenwhen
accountingforseveralothersocialfactorsandsomedemographics,thenumberoffamily


117
memberspresentwasrelatedtodecreasedintakeatthefastfoodeatingoccasion.
Previousresearchhasshownthatfamilyhasaprominentinfluenceoneatingbehaviours
(e.g.,Zimmerman&Connor,1989).Yet,socialmodellingstudieshavetendedtofocuson
theeffectsofstrangersversusfriendsontheamounteaten(e.g.,Rosenthal&
McSweeney,1979;Rothetal.,2001).Eatingwithmorefamilymembersmaymakeafast
foodeatingoccasionmorepracticallyorientatedorlessofaspecialevent.Thismayin
turn,decreasethemotivationtooverindulge(deCastro,1994).Thisfindingwillneed
confirmationandfurtherexplorationinlargersamplesbeforeitcanbebetter
understood.

4.4.6 ConclusionsfromtheinitialFFS
TheFFSisanacceptableandfeasibleapproachforsurveycollectionoffastfoodintake
frompeopleacrossarangeofages,educationallevels,socioeconomicstatus,and
culturalbackgrounds.Thecurrentstudyshowedthattheclimateofthefastfood
environment(asdefinedbyvariablessuchasthewaytheseitemsarepackagedtogether
andthesizeoftheseitems)isapotentialsourceofvariationintheamountoffastfood
consumed.Therefore,thefastfoodenvironmenthasuniqueenvironmentalpressures
thatmaypromoteconsumption.Thatisnottoinferthatinfluencesongeneraleating
behaviourarenotalsoapplicabletofastfoodconsumption,asthecurrentresultsalso
providedsomesupportforthepotentialinfluenceofsocialfactors,suchasthepresence
ofotherpeopleandtheperceivedenvironmentalnorms,ontheamountofenergy
consumedfromthesinglefastfoodeatingoccasion.Agreaterunderstandingofany
relationshipsbetweenintake,andsocialfacilitationvariables(includingthenumberof
peoplepresent,andthetimespenteating)canonlybedevelopedthroughadditional
datacollection.Corroboratingthesefindingsinanew,largersamplewillalsoallow
moredetailedanalysisofintakewhichmayhelptorevealsomeoftheindirecteffectson


118
consumption.Assuch,theaimofthenextstudywastoadministertheFFStoalarger
samplewiththegoalofmodellingmultiplefactorsthatmaybeimportantforincreased
intakeinafastfoodenvironment.

119

5 ChapterFive:Socialinfluencesonamountoffastfoodconsumedinonevisit
Testingastructuralmodel

InitialtestingoftheFastFoodSurvey(FFS)attheRoyalAdelaideShow(RAS;alocal
fair)indicatedthatavarietyofthesocialfactorsproposedtoinfluencefastfoodeating
behaviourwarrantedcloserexamination.Forexample,therewasnotstrongsupportfor
themodelofsocialfacilitationandthetimeextensionhypothesis,asneitherthenumber
ofpeoplepresentnorthetimespenteatingwererelatedtotheamountoffastfood
consumedatasingleeatingoccasion.Itwasunclearwhetherthiswasduetothelimited
samplesizeoraresultofthenatureoffastfoods.Theaimofthesubsequentstudywas
to:1)exploretheobservedrelationshipsusingmoresophisticatedstatisticalanalyses,
2)assesswhetheranexistingmodelofsocialinfluencecouldbereplicatedwhen
measuringfastfoodintakebehaviours,and3)toextendthismodeltoprovideamore
detaileddescriptionofanysocialfacilitationeffectsonfastfoodconsumption.Thisstudy
willthereforeaddressthesecondresearchquestionposedinSection3.5:Cansocial
facilitationand/orthetimeextensionhypothesisincreasetheamountoffastfood
eaten?

5.1 Testingexistingsocialinfluencemodelsinfastfood
Feunekes,deGraafandvanStaverens(1995)modelistheonlypublishedattemptto
demonstratehowsocialfacilitationmayoperateduringaneatingoccasion.Itshowsthat
thepresenceofothersincreasesthetimespenteating,whichinturnincreasesthe
amounteatenormealsize(SeeFigure5,Section3.2.2page61).Thismodelrepresentsa
soundbasisforexploringtheeffectsofsocialfacilitationonfastfoodconsumptionasit
providesamodelforpreliminarytestingandcanbedevelopedtoincludeavarietyof
otherfactorsthatmayaddpredictivepowertothemodel.
120

InordertotestdeCastros(1990;1992)originalproposalthateatingwithothers
increasestimespenteatingandconsequentlymealsize,Feunekesetal.(1995)
administereddietdiariestoasmallsampleofDutchvolunteers(n=50)
52
.Thesediaries
collectedinformationon:theitemseaten(amountandtypeoffood),thenumberof
peoplepresent,thetimeatthebeginningandendofthemeal,activitieswhileeating,
sociablenessoftheeatingatmosphere,howpleasantthefoodwas,howdifficultitwasto
getadditionalfoodandhowmuchtheparticipantintendedtoeat.Theresearchersthen
assessedwhichoftheabovefactorswererelatedtointakeatavarietyofeating
occasions(breakfast,lunch,dinnerandsnacks).Theirpreliminaryanalysesrevealeda
positiverelationshipbetweenmealdurationandenergyintakeatallmealoccasions
(exceptlunch).Theiranalysesalsoshowedapositiverelationshipbetweenthenumber
ofotherpeoplepresentandhunger,ratingsofatmosphereandthepleasantnessoffood.
Therewaslittleevidenceofadirecteffectofthenumberofpeoplepresentonthe
amounteaten
53
.

Followingpathanalysis,theauthorsconcludedthattimespenteatingmediatedthe
effectthatthepresenceofothershadonintake.Theyfurtherconcludedthatthis
mediationispartlyexplainedbytheeffectofpeopleonthepleasantnessofthe
experience;ratingsofatmospherewereimprovedwithmorepeoplepresent.
Nonetheless,theassociationbetweenatmosphereandmealdurationwasonlyof
borderlinesignificance.Therewasnodirecteffectofthepresenceofothersonthe
amounteaten.Theresearchersalsoreportthattherewerenomediatingeffectsof

52
Thiswasoriginallytwosamplesthatwerecombinedforstructuralequationmodellingintheirfinal
sample.
53
Intheirfirststudy,onlytheconsumptionofbreakfastwassignificantlyassociatedwiththenumberof
peoplepresent.Inthesecondstudy,onlytheamounteatenatsnackingoccasionswasassociatedwiththe
numberofpeoplepresent.Whenthesampleswerecombined,therewasonlyevidenceofasignificant
relationshipforbreakfastmeals.Feunekesetal.(1995)concludethatoverallnodirecteffectbetweenthe
numberofpeoplepresentandtheamountoffoodconsumedwaswitnessed.
121
hungerandthetasteandavailabilityoffoodbetweenthenumberofpeoplepresentand
mealsizes.

Feunekesetal.(1995)claimsocialfacilitationoffoodintakeismediatedbymeal
duration.Thebivariaterelationship(alsoreferredtoasatotaleffect)betweenthe
numberofpeoplepresentandtheamounteatenwasweak,therefore,whetherthe
relationshiptheypresentrepresentsoneofmediationisdebatable(see,Baron&Kenny,
1986).Ithasbeenacknowledgedthatmediationcanexistevenwhentherelationship
betweenthemediatedvariablesisnotsignificant(see,Judd&Kenny,1981,p.207).
Furthermore,ithasbeensuggestedthatanindirecteffectcanbesignificantinthe
absenceofasignificanttotaleffect(Preacher&Hayes,2004).Themodelpresentedby
Feunekesetal.showedanindirecteffectofthenumberofotherspresentontheamount
eatenthroughmealdurationwhichmayalsobereferredtoasamediatingeffect.The
timespenteatingappearstohaveanimportantrelationshipwiththenumberofpeople
present.Thisassociationhasbeenshowninstudiesusingdietdiaryrecords(deCastro,
1990,1991,1994)andinobservationalsettings(Bell&Pliner,2003;Sommer&Steele,
1997).Furthermore,experimentalstudieshaveshownthatprolongedaccesstofoodcan
increaseintakeregardlessofthenumberofpeoplepresent(Pliner,Bell,Hirsch&
Kinchla2006).Therefore,itispossiblethatthepresenceofothersindirectlyincreases
intakebyincreasingthetimespenteating.

5.1.1 ExtendingFeunekesetal.smodelofsocialfacilitationtofastfoodconsumption
Socialfacilitationresearchsuggeststhateatinginthepresenceofotherscanincrease
theamountoffoodeaten(Clendenen,Herman&Polivy,1994;J.M.deCastro,1990;J.M.
deCastro&Brewer,1992).Giventhemoderatelystrongevidenceforconsistenteffects
ofsocialfacilitationonfoodintake,andtheacknowledgedenergydensityoffastfoods,it
isimportanttoexploretheimpactofsocialinfluencesonconsumptionoffastfoodin
122
ordertounderstandhowthisbehaviourcouldimpacthealth.deCastrohasexplored
socialfacilitationacrossmanydifferentcontextsbuthasnotextendedthesestudiesto
fastfoodconsumption(J.M.deCastro,1990,1991,1994;J.M.deCastroetal.,2000;J.M.
deCastroetal.,1997;J.M.deCastro&Brewer,1992).Thereisalsogoodevidencethat
otherscanindirectlyincreasetheamountoftimespenteating(Bell&Pliner,2003;
Feunekesetal.,1995;Sommer&Steele,1997).Ifsocialfacilitationisdependentupon
themediatingeffectoftime,asFeunekesetal.(1995)suggested,thefastnatureoffast
foodmaylimittheextenttowhichthepresenceofothersinfluencestheamount
consumed.

Onelimitationofattemptingtoapplythemodelofsocialfacilitationtofastfood
consumptionisthenatureoffastfoodpurchases.Plineretal.(2006)suggestedthat
socialfacilitationmaybeexplainedbyprolongedaccesstofoodinanenvironmentthat
promotesfoodintake.ThepreviousFFSstudyshowedthatmost(70.7%)purchases
weretakentobeconsumedawayfromtherestaurant.Feunekesetal.s(1995)study
wasconductedinahomeenvironmentwhereaccesstofoodisminimallyconstrained.
However,someoftheirresultssuggestedthatincreasesinmealsizeinthepresenceof
othersmaynotoccurspontaneouslyduringtheeatingoccasionbutmaybeanticipated
beforeeating.Forexample,theynotedasmalleffectofthepresenceofothersonthe
volumeoffoodaparticipantintendedtoeatbeforetheeatingoccasion.When
participantsatewithotherstheamounttheyintendedtoeatpriortothemealincreased.
Inaneatingenvironmentsuchasarestaurant,theamountisdeterminedatthetimeof
ordering.Thus,socialfacilitationofconsumptionvolumeandmealenergydensityinthe
restaurantsettingareevidencedatthetimeofordering(c.f.,Stroeble&deCastro,2004).
Theinfluenceofsocialfactorsonconsumptionintentionhasalsobeenobservedin
studiesofpurchasingbehaviour(Sommer,Wynes&Brinkley,1992).Theabove
observationssuggestthatitmaybepossibletoobservesocialfacilitationeffectsinthe
123
fastfoodconsumption,regardlessofwhetherfoodisconsumedintherestaurantor
elsewhere.

5.1.2 Thegoalofthecurrentstudy
Thegoalofthecurrentstudywastoexploretheindirectormediatingeffectsofthe
otherpeoplepresentataneatingoccasionbyconfirmingandextendingFeunekesetal.s
(1995)modelofsocialfacilitationinthecontextoffastfoodconsumption.

InFeunekesetal.s(1995)model,therewasevidencethatthenumberofpeoplepresent
attheeatingoccasionindirectlyaffectedenergyintakethroughmealdurationand
eatingatmosphere.Acommonwayfordeterminingthemagnitudeofindirecteffectsis
bymultiplyingthestandardisedpathcoefficientsthatareassociatedwithanindirect
effect(Keith,2006;MacKinnon,Krull,&Lockwood,2000).Usingthismethod,the
strengthsoftheindirectpathsinthemodelreportedbyFeunekesetal.are0.04
(.45*.18*.51)throughatmosphereand0.16(.32*.51)throughmealduration
54
.Thefirst
aimofthecurrentstudywastoassesstheutilityofFeunekesetal.smodelinthecontext
offastfoodconsumption.Specifically,thisaimwastodescribewhethertheindirect
effectsofthenumberofpeoplepresentthroughmealdurationandatmosphere
presentedinFeunekesetal.smodelwouldalsobepresentinamodelbasedonasingle
fastfoodeatingoccasion.

Thesecondaimofthismodellingwastorefinethemodelofsocialfacilitationforfast
foodconsumption.TherewassomeevidenceintheinitialFFSstudythatpeoplewith
differentinterpersonalrelationshipstotherespondenthadvaryingeffectsonthe
amounteaten.Specifically,thenumberoffamilymemberspresentnegativelypredicted
energyintakefromfastfooditemswhilethenumberoffriendspresentandthepresence

54
Feunekesetal.donotactuallyreportthesevaluesintheirpaper.
124
ofapartnershowedminimaltotaleffects.Therefore,thesecondaimofthismodelling
wasinitiallyaddressedbyexploringwhetherinterpersonalrelationshipssharedwith
thepeoplepresentaltersanyoftheeffectsobserved.

Feunekesetal.(1995)consideredtheeffectoffullnessratingsanddiscoveredthatthey
didnotsignificantlycontributetotheiroriginalmodel.Therearelikelytobemanyother
environmental,contextualandbehaviouralmeasuresthataltertherelationships
betweentheamounteaten,thetimespenteatingandtheeatingatmosphere.
Consequently,thethirdaimofmodellingwastoexpandthemodelfurtherbythe
inclusionofotherfactorsassociatedwiththesevariables.

Thefinalanalysisaimedtoexplorehowsocialfactorsandnormsmayacttoincreasefast
foodintakeatasingleeatingoccasion.Inthepreviousstudy,asignificantrelationship
wasfoundbetweentheparticipantsperceptionofhowmuchthosearoundhimorher
hadeatenandtheamountactuallyeatenbytheparticipant(seeSection4.3.7,page105).
Itwashypothesisedthatthesenormsmayaddextraexplanatorypowertothemodelfor
thosepeoplewhoateinthepresenceofothers.

Thisresearchwillthereforeextendpreviousresearchbybothtestingthemodelofsocial
facilitationforanincreasinglycommoneatingbehaviourthathasnotbeenpreviously
explored(i.e.,fastfoodconsumption),andbyextendingthismodeltoprovidea
thoroughdescriptionoftheeffectsofmanyvariablesatasinglefastfoodeating
occasion.

5.1.3 Usingsurveymethodstoexploresocialinfluence
Previoussocialfacilitationstudiesofeatingbehaviourhavealmostexclusivelyuseddiet
diaries(e.g.,J.M.deCastro,1990,1994;J.M.deCastro&Brewer,1992;Feunekesetal.,
125
1995;Stroebele&deCastro,2004)withonlytwousingalternativeapproaches
includingexperimentalandobservationalmethods(Clendenenetal.,1994;Sommeret
al.,1992).Dietdiariescollectdetaileddietaryinformationaboutdailyintakes.Theyare
limitedbythehighparticipantburdentheyimposeandareoftenexpensiveto
administerbecausetheyinvolveparticipanttrainingandmayinvolvefinancial
reimbursementfortheparticipantstime.Thisiswhydietdiarystudiesofteninvolve
smallsamples(althoughthereareexceptions,deCastro,1994,hadover500
participants).Surveymethodscanobtaindatafromawidecrosssectionofthe
populationbutrequireparticipantrecallofintake,whichcanlimittheaccuracyand
levelofdetailthatcanbeobtained.

Becausethepurposeofthecurrentstudywastoexaminetheinfluenceofthepresence
ofothersonenergyintakeatafastfoodeatingoccasion,amealrecallapproachwas
usedinwhichrespondentsdescribed,indetail,thecircumstancesoftheirmostrecent
fastfoodmeal.Initialworkatanagriculturalfair,describedintheChapterFour,verified
theutilityofthisapproach.

Askingpeopletorecallthetimetheyspenteatingatthelastoccasiontheyatefastfood
usingsurveymethodsmayhaveproveddifficultforconsumers,therebyquestioningthe
reliabilityofrecall.Inpreviousresearch,timespenteatinghasbeenrecordedbyan
observerorbytheparticipantduringthemeal.Inasurvey,estimatesoftimespentare
dependentonaccuracyofrecall.Attemptsweremadetoassistrecallbyproviding
primingcuesdesignedtoassisttherespondenttorecreatetheexperience,butitis
difficulttojudgetheeffectivenessofthisapproach.Tocheckconsistencyinestimatesof
timespenteating,thetimesreportedbythefairsample(fromtheinitialFFSstudy)were
comparedtothoseoftheInternetsample.Themeaninthefirstsamplewas14.18
minutes(SD=7.98)whileintheInternetsampleitwas14.34minutes(SD=8.39).As
126
canbeseeninFigure7,thedistributionswereclosetoidentical
55
.Thisindicatesgood
consistencyintherecalloftime
56
.

5.2 Procedure
5.2.1 AbriefintroductiontoInternetmethods
MarketresearchersbeganusingInternetsurveysovertenyearsago(Comley,1996).
Internetmethodscanbebothcheaperandmoreefficient(Granello&Wheaton,2004;
Schmidt,1997)thantraditionalformsofsurveying,improvingresponseratesby10to
15%comparedtomailoutsurveys
57
(Cooketal.,2000).Asidefrombetterresponse
rates,thespeedofresponsemaybefasterforInternetsurveys,savingaresearcherupto
aweekwaitingforpostedsurveystobereturned(Deutskens,deRuyter,Wetzels,&
Oosterveld,2004).Moreover,thereissomeevidencethatsomeparticipantspreferthis
formofsurveying(B.Smith,Smith,Gray,Ryan,&Team,2007;Zhang,1999),perhaps

55
TheoutliersindicatedintheInternetsampledistributionwerenotoutliersintheinitialscreeningand
wereleftinthesampleastheydidnotexceedtheoriginal40minutebenchmarkfromthefairsample.
56
Itdoesnotindicatewithinsubjectreliabilityintherecalloftime.Inordertodothis,alongitudinal
approachwouldberequired.
57
ThehigherresponseratesforInternetsurveysmaybeartificiallyinflatedbecauseundeliverableemails
aregenerallyremovedfromtheinitialresponsepool,whereasundeliverablelettersaremuchharderto
trace(Cook,Heath,&Thompson,2000).
RAS Ss Internet Ss
T
i
m
e

S
p
e
n
t

E
a
t
i
n
g
40
35
30
25
20
15
10
5
0
Figure7:Boxplotsfortimespenteatingfromthe
Internetsample(InternetSs;inblue)andthefairsample
(RASSs;ingreen).
127
becausetheycanbecompletedandsubmittedmoreconvenientlythanphoneorpostal
surveys(Gruber,Szmigin,Reppel,&Voss,2008).

OneoftheconcernswithInternetsurveyingisbiasarisingfromsubpopulation
variabilityinaccesstotheWorldWideWeb.Althoughthisissueisalsotrueforphone
surveying(noteveryonehasaccesstoaphone),ithasbeennotedasofparticular
concerninmanydiscussionsofInternetmethods(Duffy,Smith,Terhanian,&Bremer,
2005;Mathy,Kerr,&Haydin,2008;Sheehan&Hoy,1999).Accessibilityisbecomingless
ofaconcernasthenumberofInternetconnectionscontinuestoexpand.Sixtysix
percentofAustralianhouseholdsreportedsomeformofInternetconnectioninthe
recentAustralianCensus(AustralianBureauofStatistics,2007).Itisdifficulttomeasure
howmanypeoplehaveaccesstotheInternetoutsideoftheirhomesaswork,libraries,
cafesandevensomepetrolstationsarenowbecomingaccesspoints.Schillewaertand
Meulemeester(2005)reportedthatparticipantswithoutanInternetconnectionat
homecompletedtheonlineversionoftheirsurveyatwork.Furthermore,therehave
beenseveralpapersindicatingthatInternetsamplingcanreachrepresentative
populations(Best,Krueger,Hubbard,&Smith,2001;Gosling,Vazire,Srivastava,&John,
2004;Mathy,Schillace,Coleman,&Berquist,2002).

5.2.2 PilottestingoftheFastFoodSurvey(FFS)deliveredontheInternet
PilottestingofanyInternetsurveyisacriticalfirststepbeforerolloutontheWorld
WideWeb(Bucher,2002).TheFFSprogramwaspilotedtoassessitsdownloadlength
andaccessibilityviatheInternet.Aformwassentwiththelinktotheprogram,and
participantswereaskedtocompletetheformandreturnittotheresearcher.Thelinkto
program,alongwiththerequiredpassword(uniquetothepilotversion),wassenttoa
conveniencesampleofacquaintances.Thepilotversionwasavailablefortendays.An
exampleofonecompleteresponseandasummaryofthefeedbackreceivedcanbeseen
128
inAppendix5.Intotal,eightpeoplecompletedthepilotversionoftheprogram.All
feedbacktheyprovidedwasincorporatedintotherevisedversionoftheprogram;
typographicalmistakeswerecorrectedandnavigabilitywasimproved.Information
regardingexpecteddownloadandcompletiontimeswasaddedtotheinstructions.This
includedthecreationoftwoprogressbars
58
.Thesefunctionsweredesignedtokeep
potentialparticipantsawareofthelengthofthesurvey.

5.2.3 Recruitmentmethodsandoutcomes
Toaidparticipantrecall,asimpledomainnamewasregistered
(www.fastfoodstudy.com.au)
59
.Recruitmentoccurredovera90day
60
period.Theweb
addresswaspromotedandmarketedwidelyinthecommunityandpeoplewereasked
tologinandcompletethesurvey.Promotionalstoriesandinterviewsinthelocalmedia
encouragedhitstotheFastFoodStudyhomepage.Snowballsamplingwasalsoused,
withanemailsenttoacquaintancesoftheresearchteamaskingthemtocompletethe
surveyandtoforwardtheemailontotheirfriends.Finally,atargetedmailout
supplementedtherecruitmentinanattempttoattractahigherproportionofmale
respondents.ThreehundredmaleparticipantsregisteredfortheNorthwestAdelaide
HealthSurvey(See,Grantetal.,2008fordetaileddescriptionofthissample)weresenta
letterpromotingthewebsiteandaskingthemtotakepartinthesurvey(seeAppendix
6).Recruitmentdrivestopromotethehomepageprovedsuccessful,withpeaksinfirst

58
Thefirstshowedtheprogressofthedownloadasapercentageofthecompletedownload.Thesecond
progressbarwasinsertedintothesurveyquestionsandindicatedhowmanymorepagesofthesurvey
questionsremained.
59
Inordertoensureaccessbypeoplewhomisrememberedthesiteaddress,anattemptwasmadeto
ensurethatthesitewasidentifiableviaasearchengine.Utilisationofasimpleandexplanatoryweb
addressalsoensuredthatpeoplewhowereinterestedinthetopicandsurfingtheweb,wouldalsofind
thewebsite.Carewastakentomakethehtmlcodeofthehomepageascleanaspossible.Thismeantthat
thecodewaswrittentocontainonlytheelementsnecessaryforthepagetodisplaycorrectly.Tagswere
writtenintotheprogrammingthatallowedmetasearchenginessuchasGoogletosearchforrelevant
contentsofthepage.Thesitewasthenregisteredwiththemajorsearchengines(Google,Yahooand
MSN).
60
RecruitmentwasbetweenNovember2006andMarch2007butceasedovertheholidayseason;from
endofDecembertoendofJanuary.
129
timevisitorsfollowingeachattempt.Figure8showsfirsttimevisitors
61
tothe
www.fastfoodstudy.com.auhomepage.Thepeaksfollowingpublicityarelabelledwith
theassociatedrecruitmentattempts
62
.




5.3 Approachtoanalysis
5.3.1 Descriptivedata
Priortomodelling,descriptivedatawereanalysedtoaddresstwogoals.Thefirstwasto
provideapointofcomparisoninthefastfoodeatingbehavioursreportedinthecurrent,
largersamplecomparedtotheRASsample(hereinreferredtoasthefairsample).The
secondgoalofdescriptiveanalyseswastoassessthetotaleffects(bivariate
relationships)ofthesocialfacilitationvariablespresentedbyFeunekesetal(1995).


61
FirsttimevisitorinformationiscollectedfromtheIPaddressofthepersonaccessingthesurvey.A
personvisitingthewebsitefromthreedifferentcomputerswillregisterthreefirsttimevisitorhits.
62
Forethicalreasons,thecounterofthehomepagewasnotattachedtothesurveypages,thereforeitis
impossibletodeterminehowmanyofthesehitsresultedincompletionofthesurvey.Therefore,thisdata
issimplyaguideforinterestinthesite.
0
10
20
30
40
50
60
70
80
90
1
9
/
1
1
/
0
6
2
6
/
1
1
/
0
6
3
/
1
2
/
0
6
1
0
/
1
2
/
0
6
1
7
/
1
2
/
0
6
2
4
/
1
2
/
0
6
3
1
/
1
2
/
0
6
7
/
0
1
/
0
7
1
4
/
0
1
/
0
7
2
1
/
0
1
/
0
7
2
8
/
0
1
/
0
7
4
/
0
2
/
0
7
1
1
/
0
2
/
0
7
1
8
/
0
2
/
0
7
2
5
/
0
2
/
0
7
4
/
0
3
/
0
7
1
1
/
0
3
/
0
7
1
8
/
0
3
/
0
7
11/02:Channel7News
8/12:1
st
Email
20/11:5AARadio
8/03:NWAHSMailout
24/01:2
nd
Email
N
u
m
b
e
r

o
f

v
i
s
i
t
o
r
s

t
o

w
e
b
s
i
t
e

Figure8:Firsttimevisitorstowww.fastfoodstudy.com.au withrecruitmentdatesandassociated
recruitmentmethodslabelled.
130
5.3.2 Methodsformodelling
AswasdonebyFeunekesetal.(1995),modellinginthecurrentstudywasperformed
usingpathanalysis.Pathanalysisallowstestingofcausalmodelswhilealsoassessing
directandindirecteffectstobemeasured.Itsabilitytoestimateindirecteffects
differentiatesitfromotherlinearequationtechniques(Pedhazur&Schmelkin,1991).
Preliminarymodellingwasdevelopedusingtherecursive
63
modelpresentedin
Feunekesetal.spaper.Giventherecursivenatureofthemodelandtheabsenceof
latentvariables
64
,pathanalysiswasperformedusingsequentialregressions(also
referredtoasanordinaryleastsquaresregressionapproach).Thismethodallowsthe
pathcoefficientsandindirecteffectstobecalculated,but,asisthecasewithmost
modellingtechniques,doesnotdeterminecausation(Keith,2006).

Allindirecteffectsofthepeoplepresentattheeatingoccasionwerecalculatedby
multiplyingstandardisedpathcoefficients(Keith,2006).Todeterminethesignificance
oftheseindirecteffects,Sobels(1982)equationswereused.Theseequationsarebased
onasymptoticdistributiontheoryandtestwhetherthemagnitudeoftheindirecteffect
issignificantlydifferentfromzero.Toassistinthesecalculations,anonlinecalculator
wasused(Preacher&Leonardelli,2001).

InaccordancewiththerecommendationsofTabachnickandFidell(1989),datawere
screenedtoensurelinearity,normality,homoscedasticityandavoidmulticollinearity

63
Recursivemeansthatthereisnofeedbackloopbetweenvariablesandallofthearrowsinthemodel
pointinonlyonedirection.
64
Latentvariablesarenotdirectlyobservableandarecalculatedthroughindicatorvariablesthatare
observableordirectlymeasured.Therefore,latentvariablesuseindicatorsofanunderlyingconceptand
typicallyrequireconfirmatoryfactoranalysistoconfirmthatthesemeasurestapintothesameconstruct
(Kline,2005).Byrne(2001)identifiedselfconceptasacommonexampleofalatentvariable.
131
andsingularity
65
beforethemodelsweredeveloped.Followingthisscreening,the
measuresenergyconsumedandtimespenteatingweretransformed
66
.

5.3.3 Limitingthereverserelationshipbetweentimespenteatingandtheamounteaten
Measuringthetimespentataneatingoccasionhasbeencommoninpreviousstudies
butmaybeconfusedwiththeeffectofamounteatenontimetaken(bigmealstake
longertoeat).Thefocusofthisstudyiswhetherextratimespentintheestablishment
thanisneededtoconsumeamealleadstoanincreaseintheamounteaten.Therefore,in
anefforttoclarifyinterpretation,theamountoftimerequiredtoeatfastfooditemswas
subtractedfromtheparticipantsestimateofthetimespenteating.

Assessingtheamountoftimeneededtoeatafastfoodmeal.Fortyonepeopleinafood
courtwereobservedinordertodeterminehowlongtheytooktoconsumeavarietyof
takeawaymeals(includingbakedpotato,preparedsalads,meatpiesandtraditionalfast
food)
67
.Eachofthedinersobservedwaseatingwithoutdistractionsincludingother
peopleand/orreadingmaterials,computersandsoon.Time(inminutes)wasrecorded
fromthetimeofthefirstbiteuntilthelastbite(anddidnotincludetimetocompletea
drink).Theaveragetimespenteatingwas7.61minuteswiththemodaltimebeing

65
Normality,homoscedasticityandlinearitywereassessedthroughexaminationofthescatterplotandthe
normalprobabilityplotforeachregression(includingthepreviouslytransformedvariables).Inspection
ofeachofthesegraphsdidnotrevealanyobviousabnormalities.Toavoidmulticollinearityand
singularity,preliminarybivariate correlationswereperformedonallthevariablestobeincludedinthe
model.Nonewereveryhighlyrelatedtoeachother(r>.80).Foreachregression,tolerancesandthe
varianceinflationfactors(VIF)werecalculatedtoassessthelevelmulticollinearityforeverypredictor.No
toleranceswereclosetozero;allhadvaluesabove.5.Afterthisscreening,allmeasureswerejudged
suitableforinclusioninaregressionanalyses.
66
Toassessnormalityinthedependentvariables,skewnessandkurtosisvaluesweredividedbytheir
standarderrorstodeterminehowmanyzscoresthedistributionwasawayfromnormal.Foratmosphere,
thedistributionwaswithinthreestandarddeviationsfromnormalforskewness.Forkilojoulesconsumed,
theskewwasoverfivestandarddeviationsfromnormalitywhileitwasnineawayfortheextratimespent
eating.Histogramswereusedtoexplorethedirectionoftheskew.Bothshowedapositiveskew.Toadjust
thisskew,thesevariablesweretransformedusingsquareroottransformationsasissuggestedby
TabachnickandFidell(1989,p.85).
67
ObservationalmethodswereusedratherthandatafromlonedinersintheinitialFFS.Asdatafromthe
FFSdidnotincludedescriptionofactivitiesduringconsumption,itwasunclearhowmanylonediners
mayhavebeendoingactivitiessuchasreadingwhichmayhaveprolongedthetimetheyspenteating
wheneatingalone.
132
8minutes.Datafor15peoplewhoateeitherMcDonaldsorHungryJackswere
separatedfromthedataforothernonbigbrandfastfoodmeals.Themeantimewas
7.73minutesforthefastfoodconsumersandthemodaltimespenteatingwas8
minutes.

Thirteenofthe15peopleobservedeatingbigbrandfastfoodsinthefoodcourtwere
consumingmealsthatconsistedofaburger,friesandadrink.DatafromtheinitialFFS
supportedtheideathatorderscontainedfairlyconsistentcomponents.Foodanddrink
itemsselectedintheFFSwerecodedaccordingtoitemdescription(i.e.,ifapersonhad
sixchickennuggets,thiswascodedasoneitem;ifapersonhadacheeseburgerandaBig
Mac,thisconstitutedtwoitems).Ofthe116respondents,93.1%hadorderscontaining
threeitemsorless(includingthedrink)
68
.

Givenconsistencyinthetypesoffastfoodmealseatenandtheconsistencyof8minutes
neededtoeatthesetypesofmeals,thisvaluewasconsideredenoughtimetoconsumea
fastfoodmeal(andmorethanenoughtimetoeatafastfoodsnack).Subsequently,8
minuteswassubtractedfromthetimespenteatingreportedbyparticipants.All
recalculatedtimespenteatingscoresthatwerebelowzerowereconvertedtozero.
Althoughthislimitedthevariabilityoftimeinthesample(21.9%ofpeoplewerecoded
aseatingforzeroextraminutes),thiswasdonetoeaseinterpretationofthevariable.
Thenewvariablethereforerepresentedextratimespenteatingatthefastfoodeating
occasion.


68
Inmostcases,thosewhoconsumedmoreitemsatefromKFCwheremealstendedtohavemore,smaller
components.Theonlyexceptionweretwopeople(onewhoateatMcDonaldsandonewhoateHungry
Jacks)whoconsumedasoftserveconeinadditiontoaburger,friesanddrink.
133
5.4 Dataadjustmentsandrecoding
5.4.1 Assessmentofoutliers
Followingallrecruitmentdrives,therewere434responsestothesurvey.Elevencases
wereremovedfromthedatasetbecausetheyhadincompletenutritioninformation
69
.
Foreaseofinterpretation,threecaseswereremovedbecausetheyhadonlyconsumeda
drink.Afurthertwocaseswereremovedbecausetheparticipanthadboughttheiritems
(fortheonemealoccasion)frommorethanoneofthefastfoodchains.

Aftertheinitialscreening,datawerescreenedforunivariateoutliersonthedependent
variablesenergyconsumedandtimespenteating.Zscorecriteriaandstemandleaf
plotsindicatedsevenoutliersfortheamountofenergyconsumed
70
.Thedatawerealso
screenedformultivariateoutliers.SixcaseswithextremeMahalanobisscores(p<.01)
wereidentifiedasmultivariateoutliersandremoved(Tabachnick&Fidell,1989).

5.4.2 Adjustmentstoreportedfastfoodordersandvariablerecoding
ParticipantsindicatedthattheordertheyhadenteredintotheFastFoodChoices
Program(FFCP)neededadjustingin74instances.Theseweredoneinthesamewayas
presentedinSection4.2.3.Ofthe74adjustments,20ordersneededthestandard
portionsizeoftheitemsadjustedand54ofthetotaladjustmentsrelatedtotheactual

69
Althoughthesurveyincludedprogrammingthatpreventedincompletesurveyanswers,somepeople
failedtoselectfastfooditemsandcompletedthesurvey.
70
Therewassomediscrepancybetweenoutliersidentifiedviathedifferentprocedures.Thestemandleaf
plotsindicatedtwelveoutliers;peoplewhoconsumedover8300KJfromtheirlastfastfoodeating
occasion.Zscorecriteriarevealedonlythreeoutliers(9500KJorgreater).Examinationofthemeal
contentsforthosereturningenergytotalsbetween8300KJand9500KJsuggestedthattheupperfigure
wasalarge,butlikelymealdescription.Thelargestorderbelow9500KJwasanorderof9474KJ,which
wascomprisedoftwoburgers,alargefries,asmalldrinkandamediumdrink.Consequently,9500KJwas
usedasthecutofffordeterminationofoutliers.Fouroutliersweredetectedbybothstemandleafplots
andzscoresfortimespenteating.Thesewereallpeoplewhohadeatenforover40minutesandwere
deleted.
134
itemsselectedfortheorder
71
.Twocaseshadorderalterationsthatcouldnotbe
calculatedandwereconsequentlydeleted.

Datawererecodedforthevariables,frequencyofconsumption,groupsizeandtheIndex
ofRelativeSocioeconomicDisadvantage(IRSD)accordingtothemethoddescribedin
Section4.2.4.Manyofthevariablesinthesurveywerecategorical(e.g.,brandoffast
food,employmentstatus,reasonsforconsumption)andwererecodedforinclusionin
regressionmodelsviautilisationofdummycoding
72
(Tabachnick&Fidell,1989).
Variablestreatedinthisfashionwerereasonsforfastfoodconsumption,activities
precedingfastfoodconsumption,andbrandoffastfood.Toavoidsingularity
73
,one
referencegroupwasexcludedfromtheanalysisineachcase(Tabachnick&Fidell,
1989).Therearenoestablishedstatisticalrecommendationsforchoosingareference
grouptobeexcludedfromcoding.Choosingwhichresponsecategorytoexcludefrom
themodelisimportant.Thesignificanceofthedummycodedvariablesinthemodelis
determinedrelativetothereferencecategoryduringtheregression.Themostcommon
response(thereforethemodalresponse)forbrandoffastfoodeaten(McDonalds),and
themostcommonactivitybeforeconsumption(betweenactivities)werechosenasthe
referent.Asignificanteffectarisingwiththesevariablesthereforeindicatedasignificant
differencefromthemode.Thereferentforthevariablereasonsforconsumptionwasthe
responsebasicneeds.Thisresponseindicatedthatthepersonsimplyatethefood

71
Thirtyninepeoplereportedthatitemsneededtobeaddedtotheorder.In61%ofthesecases,the
menuitemwasavailabletoselectintheFFCPbutnotfoundbytheparticipant(usuallyacondiment).The
nutritionalinformationforitemsthatweremissingfromtheFFCPwasobtainedviaphonecallsandemail
totherelevantcompanies.Theother15adjustmentstoorderspertainedtocustomisationofthe
compositionofstandardmenuitems,forexample,Hadcheeseburgerwithnomeat.Minorchangestothe
burgerinvolvingsuchingredientsasonionsandpickleswerenotadjustedforinthecalculationofenergy
content.Whenextrasaucewasordered,adjustmentsweremadetotheorder.
72
Dummycodinginvolvescreatingdichotomous(i.e.,yesversusnovariables)fromcategoricallabels.
73
Singularitymeansthatthepredictionofonevariablecanbepredictedperfectlybyanother.Thisoccurs
indummycodingbecauseiftherethreecategoriesofavariableandthesearecodedintothreegroups,
informationonmembershipoftwogroupscanpredictmembershipofthethird.Ifaresponseisnotin
categoryoneortwo,thenitwouldbelongtocategorythree.
135
becausetheywerehungry(adrivercommontoanyeating)withsignificantdifferences
fromthisseenaspotentiallyinformativeforfastfoodconsumption.

Tolimitthenumberofvariablesincludedinmodelling,othercategoricalmeasureswere
recodedintodichotomousvariablesrepresentingthepresenceorabsenceofthemost
frequentresponsecategory.Employmentstatuswascodedintofulltimeemployment
(fulltimeemployedversusnotfulltimeemployed),maritalstatuswascodedintomarried
(versusnotmarried),eatingoccasionwascodedintodinner(orother),locationwhere
thefoodwaseatenwascodedintoathome(ornotathome)anddayofweekwhenthe
foodwasconsumedwascodedintoeatenontheweekend(versusduringaweekday).

Inthemodelsincludingtheinterpersonalrelationshipswiththepeoplepresentwhile
eating,thedataontheotherspresenthadtoberecoded.Twovariableswerecreated
thatassessedthenumberoffriendsandfamilymemberspresentbycollapsingsome
socialrelationshipcategories.Thiswasdoneinawayidenticaltothatreportedin
Section4.3.7inthepreviousChapter
74
.Athirdvariablewasalsocreatedthatindicated
thepresenceofapartnerattheeatingoccasion.Apersoncouldhaveeatenwithpeople
representingmultipleinterpersonalrelationships.Forexample,theymayhaveeaten
withapartner,twofriendsandafamilymember.

5.4.3 Codingfreeresponseitems
ThecategoriesusedtocodetheopenresponseitemsdevelopedfromtheinitialFFS
studywereappliedtotheInternetsample.Thesecategoriesprovedsuitableforthenew
dataset;distributionsofresponsesweresimilaracrossbothsamples.ChiSquaretests

74
Thedescriptiveinformationforfriendsandfamilypresentindicatedthatthereweresomeoutliersin
bothofthesemeasures.Thenumberoffriendspresentrangedfrom0to20whileforfamilytherangewas
from0to13.Aswasdoneforthegroupsizevariable,theseitemswererecodedtoreducethedistribution
oftheseinthemodels.Thusbothvariableswererecodedfrom0to5ormorefriendsorfamilypresent.
136
revealedthatthefrequencyofresponseswasnotsignificantlydifferentineither
question(seeFigure9andFigure10).

Activities before fast food consumption


Social
acitivites
Not much Acitivity Shopping At work Going
somewhere
P
e
r
c
e
n
t
40.0%
30.0%
20.0%
10.0%
0.0%
RAS Ss
Internet Ss
Figure9:Activitiesprecedingconsumptionaccordingtotheresponse
categoriesfromtheInternet(blue)andfair(green)samples.
Between
Activities
Working
Moderate
PA Shopping
Social
Activities
Not
Much
Activitiesprecedingconsumption
Reasons for fast food consumption
Incidenta conveniece
OthersAttraction to food
Others
General Conv
Hunger
P
e
r
c
e
n
t
40.0%
30.0%
20.0%
10.0%
0.0%
RAS Ss
Internet Ss
Basic
Needs
General
Convenience
Attraction
toFood
Other
People
Incidental
Convenience
Reasonsforconsumption
Figure10: ReasonsforfastfoodconsumptionfromtheInternetsample
(blue)andthefairsample(green).
137
5.5 Resultsfromdescriptivedata
5.5.1 Demographiccharacteristicsofthefinalsample
FrequenciesfordemographicdetailsofthesamplecanbeseeninTable11.

Table11:DescriptionofthedemographiccharacteristicsoftheInternetsample(n=407).
Variable % Variable %
Sex(male) 33.2 Martialstatus
Highestleveleducationcompleted Married/livingwithpartner 49.9
Bachelor 56.5 Nevermarried 44.5
Trade/certificate 23.4 Other

5.7
SecondarySchool 19.2 Stateofresidence
BelowSecondarySchool 1.0 SouthAustralia 67.3
Householdincome NewSouthWales 8.4
20,000andbelow 8.1 Victoria 7.6
20,001to40,000 10.1 WesternAustralia 7.9
40,001to60,000 26.5 Queensland 5.9
60,000andabove 55.3 Tasmania 1.0
Countryborn Undefined 2.0
Australia 80.8 Frequencyofpatronage
UK 5.2 Once/year 4.4
Asia 4.2 Morethanonce/halfyear 16.5
Other 9.8 Once/month 21.4
Employmentstatus Once/fortnight 22.9
Fulltime 57.5 Once/week 19.7
Student 20.9 Twice/week 9.1
Parttime/casual 13.8 Morethantwice/week 6.1
Homeduties/retired 4.7 Receivepension 5.7
Unemployed 1.7 OfAboriginalheritage 2.9
Other 1.5

separated/widowed

Despiteattemptstorecruitmaleparticipants,thesurveyattractedahigherproportion
offemalerespondents,consistentwiththepreviousFFSstudy.Theaverageageofthe
Internetsamplewas31.45(SD=11.82)witharangefrom16to74yearsofage.The
numberintheparticipantsimmediatefamiliesrangedfrom0to8members.Themean
138
IRSDvalueforthesamplewas1010.70(SD=65.98).Therewererespondentsfromall
decilesofsocioeconomicdisadvantage.Themodalresponseindicatedthatmost
respondentsresidedinareaswiththesecondlowestlevelofsocioeconomic
disadvantage(AustralianBureauofStatistics,2008c).

5.5.2 Reportedfastfoodbehaviours
Thefrequencydatasupportedtrendsseeninthedatacollectedfromthefairsample(see
Table12).Theonlyexceptiontothiswasthatthereweresubstantiallyfewerpeople
whoatemealdealsintheInternetsamplecomparedtothefairsample(approximately
25and50%respectively)
75
.

Table12:FrequencyoffastfoodbehavioursreportedbytheInternetsample(n=407).
Variable % Variable %
Fastfoodbrand Dayofpurchase
McDonalds 40.8 Monday 7.4
HungryJacks 23.1 Tuesday 8.4
KFC 21.1 Wednesday 10.6
DominosPizza 8.1 Thursday 14.0
RedRooster 6.9 Friday 18.2
Locationeaten Saturday 25.1
Home 37.8 Sunday 16.5
Intherestaurant 31.0 Mealoccasion
Inthecar 16.5 Breakfast 5.7
Other 14.7 Lunch 38.6
Formofpurchase Dinner 44.2
Takeaway/drivethru 69.5 Inbetweenmealssnack 11.5
Dinein 30.5 Orderedamealdeal 24.6

Thepercentageofpeopleeatingfromeachofthefastfoodcompanieswasconsistent
withtheirreportedmarketshare(BISShrapnel,2003).

75
Whenreviewingthedescriptionsofthefastfooditemsordered,someappearedtobemealsandyet
participantsindicatedthattheywerenotpurchasedaspartofamealdeal.Someparticipantsindicated
theyhadeatenaburgerwithmediumfriesandamediumdrinkandalsoansweredthattheyhadnoteaten
amealdeal.Itisvirtuallyimpossibletoordertheseitemswithouthavingtheminamealdeal.
139
Thegroupsizeofpeoplepresentduringthefastfoodconsumptionwasmostcommonly
oneotherperson(i.e.,twopeopleinagroup;38.8%ofthesample).Indescendingorder,
themostfrequentcategoriesoftheotherpeoplepresentduringtheeatingoccasionwas
alone(26.3%),twootherpeople(12.8%),inagroupoffourormoreothers(11.3%)and
withthreeothers(10.8%).Again,asintheinitialFFSstudy,wheneatingwithothers,
peoplemostcommonlyatewiththeirpartner(45.7%).

Eightyfivepercentofthesamplewasconfidentintheirselfreportedheightsand
weights
76
.BMIrangedfrom16.9to48.2withanaverageof25.5.Therewasaweak
relationshipbetweenBMIandtheamountofenergyconsumedfromfastfooditems,
r(346)=.24,p<.001.

5.5.3 Relationshipsbetweensocialvariables
Correlationswerecomputedbetweentheamountofenergyconsumedfromfastfood
items,theextratimespenteating(hereinreferredtoastimespenteating),ratingsof
atmosphere,enjoymentoftheitemseatenandgroupsize.Theresultscanbeseenin
Table13(onthefollowingpage).

Thenumberofpeoplepresentwassignificantlyassociatedwithincreasedtimespent
eating.Therewasasmall,negativeassociationbetweenthenumberofotherspresent
andenjoymentoffooditems.Timespenteatingwascorrelatedwithimprovedratingsof
theeatingatmosphere,butnottotheenjoymentoftheitemsconsumed.Sociablenessof
theatmospherewasmorestronglyrelatedtogroupsizethanthepleasantnessofthe
atmosphere.

76
Thoseconfidentintheirselfreportedheightsandweightswereslightlydifferenttothosewhowere
not.Thereweremorepeopleinthe20,000to40,001incomebracketconfidentintheirheightand
weightthanexpected,
2
(3,404)=14.25,p<.01.Otherwise,therewerenostatisticaldifferencesbetween
thoseconfidentintheirheightsandweightsandthosewhowerenotforage,sex,employmentstatus,
maritalstatus(allp>.05).
140
Table13:Correlationvalues(Pearsonsr)forthesocialfacilitationmeasures

(n=407).

1
KJs
2
Size
3
Time
4
AtmosS
5
AtmosP
1.Energyconsumed(KJs) 1.00

2.Groupsize(Size) 0.04 1.00

3.Timespenteating(Time) 0.17
***
0.23
***
1.00

4.Sociablenessofatmosphere
(AtmosS)
0.01 0.48
***
0.21
***
1.00

5.Pleasantnessofatmosphere
(AtmosP)
0.05 0.13
**
0.14
**
0.50
***
1.00
6.Enjoymentofitems

0.12
*
0.10
*
0.04 0.18
***
0.46
***
***
p<.001;
**
p<.01;
*
p<.05

5.5.4 Conclusionsfromdescriptiveanalyses
ResultsfromthefullInternetsamplelargelyreplicatedresultsobtainedatthefair
despitesignificantdemographicdifferencesinthesamples.Thefrequenciesof
behaviourreportedwereverysimilarbetweensamples.Furthermore,thereasonsfor
andactivitiesprecedingconsumption,timespenteatingandamounteatenwerevery
similarinbothsampleswhichprovidesgoodvalidationoftheFFS.

TheFFSsamplerecruitedfromtheInternethadahighpercentageofwomen,peoplein
thehighesteducationandincomebracketsandahighproportionoffulltimeemployed.
Therefore,thissamplewasnomorerepresentativethantheinitialFFS.Thesexbiasin
bothsamplesislikelytoresultfromthehealthrelatednatureofthesurvey
77
.The
Internetnatureofthesurveymayhaveattractedabiasedsample(Schillewaert&
Meulemeester,2005).Alternatively,fastfoodconsumersmayalsorepresentaunique
groupofthepopulationasithasbeenreportedthatdemographiccharacteristicsare
associatedwithfrequencyofconsumption(Schroderetal.2007;Mohretal.,2007).


77
OfallthewomenstudyinganonschoolqualificationinAustralia,13.9%areinthefieldofhealth.In
contrast,only3.5%ofallthemenstudyingatthislevelarestudyinghealth(AustralianBureauof
Statistics,2007).
141
Thebivariaterelationshipsanalysedlargelysupportedthetotaleffectsreportedby
Feunekesetal.(1995).Peoplepresentrelatedtothetimespenteating,ratingsof
atmosphereandenjoymentoffooditems.Furthermore,thedirectrelationshipbetween
thenumberofpeoplepresentandtheamounteatenfailedtoreachsignificance(aswas
thecaseintheinitialFFSstudy).OnerelationshipthatdidnotreplicateFeunekesetal.s
wasthattheenjoymentofthefooditemsrelatednegativelyratherthanpositivelytothe
numberofotherpeoplepresent.Thenegativerelationshipbetweenenjoymentofthe
foodandthenumberofpeoplepresentcontradictsresearchwhichhassuggestedthat,
wheneatingwithothers,peopleperceivefoodastastingbetter(Bellisle&Dalix,2001;
Stroebele&deCastro,2004).Thecontradictoryrelationshipmayresultfromthenature
offastfood;itisverypredictableandlikingofthefoodmayberelativelystable.

Finally,thesecorrelationssupportthoseintheinitialFFSstudywhichshowedthatboth
ratingsofthesociablenessandpleasantnessofatmosphereareassociatedwithgroup
size.Inthecurrentresults,bothmeasuresalsorelatedtothetimespenteating.
Sociablenessoftheatmospherewasmorehighlycorrelatedwithgroupsizeandthetime
spenteatingthanthepleasantnessofatmosphere
78
.Giventhefactthatpleasantnessof
atmosphereislessconceptuallysimilartothenumberofpeoplepresentthan
sociablenessofatmosphere,andthatbothmeasuresrelatedtovariablesinsimilarways,
itrepresentsamoreappropriatemeasureofatmosphere.Therefore,pleasantnessof
atmosphereratherthansociableness(usedinFeunekesetal.s,1995originalmodel)
willbeusedinfutureanalyses.


78
Intheoriginalmodelofsocialfacilitation(Feunekes,deGraaf&vanStaveren,1995),themeasureof
atmosphereaskedparticipantstoratethesociablenessoftheiratmosphere.Thiscanbecriticisedasit
maybeviewedasaproxymeasureforthenumberofpeoplepresent.Thefactthatthenumberofpeople
presentwasrelatedtoratingsofatmospherecould,therefore,beseenasanexpectedresult.Inorderto
overcomethisissue,twomeasuresofatmospherewereincludedintheFFS.Thefirstwasidenticaltothe
measureseeninFeunekesetal.smodel;thesecondaskedparticipantssimplytoratehowpleasantthe
atmospherewas.
142
Inconclusion,therelationshipsassessedheresupportedthosereportedattheinitial
stagesofmodellingundertakenbyFeunekesetal.(1995)forgeneraleatingbehaviours.
GiventhesimilaritiesbetweenthecurrentresultsandthosereportedbyFeunekesetal.,
itispossiblethatanindirecteffectofthenumberofotherpeoplepresentthroughthe
timespenteatingexists.Bivariaterelationshipsoffernoindicationofindirecteffectsand
thereforemodellingwasperformedtofurtherexplorewhetherandhowsocial
facilitationoperatesinthecontextoffastfoodconsumption.

5.6 Modellingincreasedenergyintakefromfastfooditems
AswasevidentintheinitialFFSstudy,therewasasignificantdifferenceinenergyintake
forparticipantsexinthelargerFFS,withmales(M=4264,SD=1830)eating
significantlymorethanfemales(M=3307,SD=1598),t(405)=5.41,p<.001.Allmodels
weresubsequentlycreatedseparatelyforeachsex.

5.6.1 Confirmingthemodelofsocialfacilitationinfastfoodintake
Toaddresstheinitialaimsofthecurrentstudy,initialconfirmatorymodelswererun
usingthevariablesthatFeunekesetal.(1995)foundtohaveasignificantinfluenceon
foodintakewhilealsocontrollingfordemographicfactors.Bivariaterelationships(also
referredtoastotaleffects)werecalculatedforeachofthepredictorsforfemalesand
malesseparately.ThecorrelationcoefficientsforthetotaleffectsaredisplayedinTable
14(overthepage).Althoughtotaleffectsweresmallfortherelationshipbetweengroup
sizeandtheamounteaten,theywereinreversedirectionsformalesandfemales.
Atmospherewaspositivelyrelatedtoboththetimespenteatingandgroupsizein
femalesbutnotinmales.
143

Table 14: Correlationvalues(Pearsonsr)forthesocialfacilitationmeasures.Valuesforfemales

(n=272)arepresentedabovethediagonal,values
formales

(n=135)arebelowthediagonal.

1.
KJs
2.
Size
3.
Time
4.
AtmosP
5.
Age
6.
Married
7.
Income
8.
Ed
9.
FTE
10.
IRSD
1.Energyconsumed(KJs) .09 .22
***
.03 .21
***
.04 .02 .05 .10 .04
2.Groupsize(Size) .05 .27
***
.16
*
.02 .06 .04 .06 .15
*
.09
3.Timespenteating(Time) .26
**
.22
**
.19
**
.09 .03 .02 .10 .10 .05
4.Pleasantnessofatmosphere
(AtmosP)
.00 .09 .05 .03 .11 .10 .11 .01 .14
*
5.Age 22
**
.13 .14 .14 .34
***
.12 .22
***
.05 .06
6.Married .05 .04 .07 .02 .49
**
.29
**
.01 .15
*
.03
7.Income .09 .04 .03 .02 .21
*
.45
***
.16
**
.35 .14
*
8.Levelofeducation(Ed) .02 .10 .13 .12 .16 .08 .21
*
.14
*
.11
9.Fulltimeemployed(FTE) .09 .07 .09 .03 .14 .39
***
.43
***
.15 .08
10Socioeconomic
disadvantage(IRSD)
.04 .14 .06 .11 .19
*
.12 .05 .29
**
.06

***
p<.001;
**
p<.01;
*
p<.05
144
Theinitialregressionanalysedenergyintakeusingtimespenteating,groupsize,
pleasantnessofatmosphereandthedemographicvariablesaspredictors.Thisregression
wasfollowedbyregressionspredictingthetimespenteatingandratingsofatmosphere.
Foreverystepofthemodel,sixdemographicvariables(age,beingmarried,working
fulltime,householdincome,IRSD[socioeconomicdisadvantage],andlevelofeducation)
wereenteredtocontrolforanypotentialdifferencesthesemaycreate.Allvariables
wereenteredintoeveryregressionsimultaneously.Theresultingbetavaluesforboth
malesandfemalescanbeseeninTable15.

Table15:Totalvarianceexplained(AdjustedR
2
)andbetavaluesforthedependentvariables
energyintake(Energy),timespenteating(Time)andpleasantnessofatmosphere(Atmos)for
bothmaleandfemalerespondentsintheInternetsample.
Step1:Energy Step2:Time Step3:Atmos

Males

Females

Males

Females

Males

Females

AdjustedR
2
.098 .103 .073 .104 .017
ns
.050
Predictors
Timespenteating .32
*
.28
**

Atmosphere .05 .01 .04 .17
**

Groupsize .03 .15
*
.24
**
.26
**
.14 .18
*
Age .34
*
.24
**
.20 .07 .20 .01
Married .11 .05 .09 .02 .19 .10
Householdincome .03 .05 .09 .02 .06 .07
Levelofeducation .02 .05 .19
*
.10 .11 .08
Fulltimeemployed .04 .08 .15 .12 .06 .09
Socioeconomic
disadvantage(IRSD)
.10 .03 .02 .03 .08 .14
*

n=131;

n=266;
**
p<.01,
*
p<.05;
ns
Modelwasnotsignificant

Significantpredictorswereusedtoconstructafigureoftherelevantpathwaysbetween
outcomemeasuresformales(SeeFigure11)andforfemalesinthesample(seeFigure
12).

145

Figure12:Significantpathsfromgroupsizetoenergyintakefromfastfooditemsinfemales
(n=266).
Atmosphere
GroupSize
TimeEating
Energy
IRSD
Age
.18
-.24
.26 .28
-.14
.17
-.15
Figure11:Significantpathsfromgroupsizetoenergyintakefromfastfooditemsinmales
(n=131).
GroupSize
TimeEating
Energy
Age
.24 .32
-.34
Levelof
education
.19
146
Males.Inthepredictionofenergyintakeinmales,onlytwoofthevariablesentered
contributedsignificantlytothevarianceexplainedbythemodel.Thestrongestpredictor
wasage,whichhadanegativerelationshipwithenergyintake.Theothersignificant
predictorwasthetimespenteating,whichpositivelypredictedenergyintake.Groupsize
wasasignificantpredictoroftimespenteating.Theindirecteffectofthenumberof
otherspresentwas.08(.24*.32)throughtimespenteating(Sobelstatistic=2.16,p<.05).
Theanalysisforatmospherefailedtoreachsignificance.

Females.Infemales,therewasevidenceofanindirecteffectoftimespenteatingon
energyintake.Thevalueofthiseffectinfemaleswascomparabletothatseeninmales:
.07(.26*.28;Sobelstatistic=3.13,p<.01).Groupsizealsohadadirectrelationshipwith
energyintakeinthefemaleswithinthesample.Thedirectionoftheinfluenceofothers
wasweakbutnegative,contrarytothepositiverelationshipsproposedintheliterature.
Atmospherewasasignificantinfluenceinfemales;groupsizepredictedratingsof
atmosphere.Theindirecteffectofothersthroughratingsofatmosphereandthenthe
timespenteatingwasquitesmall:009(.18*.17*.28)
79
.

5.6.2 Conclusionsfrompreliminarymodelling
Thepathanalysissupportedthemediatingeffectofmealdurationsuggestedby
Feunekesetal.(1995).Thesupportfortheeffectsofbothtimeandatmosphereas
influencesontotalenergywasclearerinthemodelforwomenascomparedtomen.
Atmospherewasnotrelatedtothenumberpeoplepresentortothetimespenteatingin
men.Womenandmenmaythereforebeinfluenceddifferentlybyatmosphere.

79
Sobeltestcannotbecalculatedforindirecteffectsthroughmultiplevariables(i.e.,mealdurationand
eatingatmosphere).Ifallthepathsaresignificant,theindirecteffectcanalsobetreatedassignificant
(Kline,2005).
147
Feunekesetal.(1995)reportedthattheassociationbetweenratingsofatmosphereand
timespenteatingwasofborderlinesignificance.Iftherewasaneffectforwomen,but
notmen,collapsingthesamplesmayhaveobscuredtheresultintheirstudy.

Thedifferencesseeninthemodelsformenandwomenareworthnoting.Separatingthe
modelintomenandwomenindicatedthatthetotaleffectofthenumberofotherpeople
presentonintakewaspositiveinmenandnegativeinwomen.Contrarytopredictions
arisingfromtheoriesaboutsocialfacilitation,increasingnumbershadaninhibitory
effectonintakeinwomenwhichiscontrarytothetheoryofsocialfacilitation.Evenafter
accountingforthepositiveeffectthattimeextensionhadontheamounteaten,larger
groupsizesnegativelyinfluencedintake.Conversely,inmentherewasweakevidenceof
theeffectofsocialfacilitation.Modellingrevealedapositive(butnonsignificant)total
effectofthenumberofotherpeopleontheamounteatenandamediatingeffectofthe
timespenteating.

Thefactthatthetotaleffectoftheotherspresentattheeatingoccasionontheamount
eatenwasnegativeinwomenwhiletheindirecteffectwaspositivewarrantsfurther
discussion.Whenmediatedbytimespenteating,groupsizeincreasedintake.In
contrast,thenumberofpeoplepresentnegativelypredictedtheamountconsumed.
Furthermore,includingtimespenteatingintheregressionmodelstrengthenedthe
relationshipbetweenthenumberofotherspresentandenergyintakefromthefastfood
items,therebyincreasingthemagnitudeofthiseffectby.06andmakingitsignificant
80
.


80
Themodelwasrunwithonlytimespenteatingandgroupsizetoassesswhetheritwastimespenteating
thatcreatedthiseffectinfemalesandnotanyofthedemographicvariables.Inthisregression,bothpaths
weresignificant.Thepathbetweentimeandenergywasstrongerandintheoppositedirectionto(=.26,
p<.001)totheeffectofgroupsize(=.16,p=.01).
148
TheeffectobservedinthecurrentmodelisconsistentwithMacKinnon,Krulland
Lockwoods(2000)descriptionofinconsistentmediationorsuppressioneffects.The
authorswrite,...situationsinwhichdirectandindirecteffectsoffairlysimilar
magnitudesandoppositesignsresultinanonzerobutnonsignificantoverall
relationshiparecertainlypossible(MacKinnonetal.,p.175).

Therefore,itappearsthatpositiveindirecteffectsandnegativedirecteffectscanoccur
simultaneouslyandmayacttoreducethetotaleffectbetweentwovariables.Thiswas
clearwhenconsideringthetotalindirectanddirecteffectsofthenumberofpeople
presentinthemodelforwomen.Thesumoftheindirectanddirectpathsequatestothe
totaleffectoftheotherpeoplepresentontheamounteaten(indirecteffects=.08
(.07+.009);directeffect=.16,totaleffect=.09).Thereforethesuppressingeffectof
eatingdurationmayacttoincreasetheoverallinhibitoryeffectthatthepresentof
othershadontheamounteaten.Thereisdebateabouttheexactdefinitionofsuppressor
effects
81
.Nevertheless,accordingtothedefinitiongivenbyMacKinnonetal.(2000),the
observedrelationshipsinthecurrentmodelrepresenttheinconsistentmediationofthe
numberofpeoplepresentontheenergyconsumedbythetimespenteating.More
generally,thismeansthattheremaybesharedvariancebetweentimeandthenumber
ofotherpeoplepresent,notrelevanttotimespenteating,whichimprovedthe
predictivevalidityofthetwomeasures.Thisfindingalsorevealedimportantdetail
abouttherelationshipbetweenthepresenceofothersandtheamounteatenthatcould
notbediscoveredthroughpurelybivariateanalyses.

81
Horst(1941)firstdescribedsuppressioneffects.Oneofthecriteriaformeetingsuppressionwasthat
thesuppressorvariable(timespenteating)isnotdirectlyrelatedtotheoutcome(energyintake).Inthe
aboveexampleitis.However,Conger(1974)gaveamuchmoregeneraldescriptionofasuppressoreffect,
definingitasonevariableincreasingtheeffectofanother.Foradescriptionofmultipleformsof
suppressioneffects,seeSmith,AgerandWilliams(2006).
149
Feunekesetal.(1995)proposedthatfutureresearchfocusonwhodemonstratessocial
facilitationandwhodoesnot.Giventhelimitedvariabilityinafastfoodenvironment,it
wassurprisingthattheindirecteffectofthenumberofpeoplepresentthroughthetime
spenteatingcouldbesignificantlydifferentfromzerointhemodelforfastfood.Itwas
alsonoteworthythatsocialfacilitationeffectswereclearerinmalefastfoodconsumers
thanfemaleconsumerswhodemonstratedeffectsoftimeextensionbutanoverall
inhibitoryeffectofthepresenceofothers.

5.6.3 Theeffectofwhoispresentinthemodelofsocialfacilitation
TheinitialFFSstudyshowedthatthenumberofcertaintypesofpeoplepresent(i.e.,
family)negativelypredictedintakeinfastfoodeaters,whereaseatingwithfriendsand
familieshadlittleeffect.Itwasunclearhowreplacingthegeneralgroupsizemeasure
withoutcomesseparatingthenumberofpeopleofdifferentinterpersonalrelationships
tothenodalpersonwouldaffectthemodel.Pathanalyseswereperformedidenticallyto
theproceduredescribedabove.Theonlydifferencewasthatgroupsizewasreplaced
withthenumberoffamilypresent,thenumberoffriendspresentandthepresenceor
absenceofapartner.Thiswasdoneinordertogainamoredetailedunderstandingof
howthetypeofpeoplepresentmaydirectlyorindirectlyalterintakeandtoaddressthe
finalresearchquestionposedinSection3.5.ResultingdataispresentedinTable16(on
thefollowingpage).

Again,significantpathswereusedtodrawafigureofthemodelofsocialfacilitation.The
figuresformenandwomencanbeseeninFigure13andFigure14(onpages150and
151respectively).
150
Table16:Totalvarianceexplained(AdjustedR
2
)andbetavaluesforthedependentvariables
energyintake(Energy),timespenteating(Time)andpleasantnessofatmosphere(Atmos)for
bothmaleandfemalerespondentsintheInternetsampleinthemodelincludingseparated
interpersonalrelationships.
Step1:Energy Step2:Time Step3:Atmos

Males

Females

Males

Females

Males

Females

AdjustedR
2
.101 .100 .110 .108 .001
ns
.043
Predictors
Timespenteating .35
***
.28
***

Atmosphere .03 .02 .04 .17
**

No.Familypresent .08 .08 .23
**
.20
**
.12 .16
*
Nofriendspresent .05 .15
*
.07 .18
**
.09 .09
Partnerpresent .13 .09 .28
**
.19
**
.05 .10
Age .32 .25
***
.16 .08 .20 .01
Married .19 .06 .04 .06 .19 .13
Householdincome .03 .05 .09 .01 .05 .07
Levelofeducation .03 .03 .19
*
.08 .11 .09
Fulltimeemployed .02 .09 .19 .11 .06 .10
Socioeconomic
disadvantage(IRSD)
.08 .02 .01 .04 .08 .13
*

n=131;

n=266;
***
p<.01,
**
p<.01,
*
p<.05;
ns
Modelwasnotsignificant









Figure13:Significantpathsfrominterpersonalrelationshipstoenergy
intakefromfastfooditemsinmales(n=131).
Partner
TimeEating
Energy
.35
Levelof
education
.19
No.family
.23
.28
151

Males.Aswasseeninthepreviousmodelofsocialinfluence,thedirecteffectsofthe
presenceofotherpeoplewerenotsignificant.Thenumberoffriendspresentdidnot
indirectlyinfluenceintakeatasinglefastfoodeatingoccasioninmales.Theindirect
effectofhavingapartnerpresentwas.10(.28*.35;Sobelstatistic=2.25,p<.05)through
timespenteatingand.08(.23*.35;Sobelstatistic=2.17,p<.05)forthenumberoffamily
memberspresent.Thetotaleffect(.00)ofthenumberoffamilypresentwascancelled
outbyindirect(.08)anddirecteffects(.08).Overall,thetotaleffectofthepresenceofa
partnerwasslightlynegative(.03)withapositiveindirecteffect(.10)andanegative,
directeffect(.13,notsignificant).

Females.Inthemodelforfemales,thepeoplepresentfromalltypesofrelationships
indirectlyinfluencedtheamounteatenthroughthetimespenteating.Familyhadthe
greatestindirecteffectonenergyconsumed.Theeffectofthenumberoffamilypresent
Figure14:Significantpathsfrominterpersonalrelationshipstoenergyintakefromfastfood
itemsinfemales(n=266).
Atmosphere
No.friends
TimeEating
Energy
IRSD
Age
.19
-.25
.18
.28
-.13
.17
-.15
No.family
Partner
.20
.16
152
throughtimeeatingwas.06(.20*.28;Sobelstatistic=2.39,p<.05)and.008
(.16*.17*.28)
82
throughthepathfromeatingatmospheretotimespenteatingtothe
amounteaten.Theeffectofpartnersandthenumberoffriendspresentwasindirectly
influencedbytimespenteating,butnotbytheeatingatmosphere.Thiseffectwas.05for
bothpartners(.19*.28;Sobelstatistic=2.36,p<.05)andthenumberoffriendspresent
(.18*.28;Sobelstatistic=2.61,p<.01).Thedirectandindirecteffectsofthepresenceofa
partner(indirecteffect=.05,directeffect=.08)andthenumberoffamilymember
present(indirecteffect=.07,directeffect=.09)wereverycloseandresultedinsmall
totaleffects,.03forpartnerand.02forthenumberoffamilymemberspresent.The
numberoffriendspresentwastheonlyrevisedgroupsizepredictorthatdirectly
predictedenergyintake.Thecombineddirectandindirecteffectscreatedanegative
totaleffectofthenumberoffriendspresent(.15+.05=.10).

5.6.4 Personalrelationshipsandsocialfacilitationconclusions
Separatingthenumberofpeoplerepresentingdifferentsocialrelationshipsdidnot
changethetotalvarianceexplainedbythemodels.Itdid,however,provideinformation
astothetypesofpeoplewhomaybedrivingtheeffectswitnessedintheconfirmatory
modelofsocialfacilitationforfastfoodconsumption.

Inmen,themoredescriptiveinterpersonalpredictorsrevealedthattimeextensiondid
notrelatetointakeforallsocialgroups;thenumberoffriendspresenthadnoeffecton
thetimespenteating.Althoughitwasnotsignificant,itwasinterestingthatthe
presenceofapartnerhadanegativedirecteffectandapositiveindirecteffectonenergy

82
Sobeltestcannotbecalculatedforindirecteffectsthroughmultiplevariables(i.e.,mealdurationand
eatingatmosphere).However,ifallthepathsaresignificant,theindirecteffectcanalsobetreatedas
significant(Kline,2005).
153
intakeinmales.Thetotaleffectofthepresenceofapartnerwassmallbutnegative(
.03).Nevertheless,itquerieswhethersocialfacilitationoffastfoodconsumptionoccurs
inmenwheneatingfastfoodwiththeirpartners;havingpartnerspresentmaychange
menseatingbehaviourwheneatingfastfood.Thisrelationshipfailedtoreach
significanceinthecurrentsampleandwillneedtobeexploredfurtherinalarger
sample.

Inwomen,theindirecteffectsofthenumberoffamilypresentseeninthemodelwere
similartothosepresentedbyFeunekesetal.(1995).Thenumberoffamilymembers
presentpredictedimprovedeatingatmospherewhichthenpredictedthetimespent
eating.Itispossiblethatmanyoftheeatingoccasionsanalysedinthedietdiaries
occurredwithfamilypresent.ThisdatawasnotpresentedinFeunekeetal.spaper.

Splittingthegroupsizepredictorintodifferentinterpersonalrelationshipsforwomen
alsorevealedthatthenumberoffriendspresenthadthestrongestdirecteffecton
energyintake;itdecreasedconsumption.Thenumberoffriendspresentalsohada
positiveindirecteffectonconsumptionbyincreasingtimespenteating.Overall,the
modelsincludinginterpersonalrelationshipsratherthanageneralgroupsizepredictor
haveextendedtheresultsoftheconfirmatorymodelsbyofferingamoredetailed
descriptionoftheinfluenceofrelationshiponconsumption.Intheexistingliterature,
thereisgoodevidencethatinterpersonalrelationshipsmaymoderatetheeffectsof
socialinfluence(e.g.,deCastro,1994;Clendenen,Herman&Polivy,1994).Thenatureof
apersonsrelationshiptothepeoplewithwhomsheorheeatsislikelytochangethe
natureoftheeatingoccasionandtheamountconsumed.

154
5.6.5 Expandingthemodelofsocialfacilitationforfastfoodintake
Thelowtotalvarianceforthedependentvariablesexplainedininitialmodellingmeans
thattheremaybeotherpredictorsthatcouldcontributetothemodel.Resultsofthe
initialFFS(ChapterFour)indicatedthatthebrandoffastfoodconsumedwasassociated
withsignificantdifferencesinenergyconsumption.Previousresearchhasalso
suggestedthatpalatabilityoffoodandlocationofconsumptioncanalteratmosphere,
andpossiblyindirectlyaffectatmosphere(J.M.deCastroetal.,2000;Stroebele&de
Castro,2004).Activitiesbeforeconsumptionmayinfluencethereasonsforconsumption
andultimatelythelevelofintake.IntheinitialFFSstudy,therewassomeindicationthat
frequencyofconsumptionrelatedtoenergyintakeintherecalledfastfoodmeal(this
relationshiponlytrendedtowardsignificance).Giventhepotentialimplicationofthis
relationshipforweightgain,itisimportanttoassesswhetherthisrelationshipisone
thatremainsaftercontrollingforavarietyofotherfactors.Therefore,thegoalofthe
followinganalyseswastodetermineiftheadditionofotherenvironmentaland
behaviouralvariablesalteredtherelationshipsbetweentheotherpeoplepresent,the
eatingatmosphere,thetimespenteatingandtheamounteaten.Subsequently,
predictorsmeasuringfastfoodbrand,activitiesprecedingconsumption,reasonsfor
consumption,enjoymentoffooditems,location,dayeatenandtheparticipants
frequencyofconsumptionwereaddedtothemodel.

Thestandardisedbetasthatresultedfromeachstepofthepathanalysisformalesand
femalesseparatelycanbeseeninTable17(onthefollowingpage).Again,modelswere
drawnusingpathsfromthesignificantpredictors.Theresultingmodelscanbeseenin
Figure16andFigure15(seepage156).
155
Table17:Totalvarianceexplained(AdjustedR
2
)andbetavaluesforthedependentvariables
energyintake(Energy),timespenteating(Time)andpleasantnessofatmosphere(Atmos)for
bothmaleandfemalerespondentsintheInternetsamplefortheexpandedmodel.
Step1:Energy Step2:Time Step3:Atmos
Males

Females

Males

Females

Males

Females

AdjustedR
2
.241 .242 .151 .090 .249 .373
Predictors
Timespenteating .39
***
.29
**

Atmosphere .01 .13 .06 .16
*

No.familypresent .09 .01 .24
*
.21
**
.04 .15**
No.friendspresent .09 .03 .05 .15
*
.11 .14*
Presenceofpartner .11 .13
*
.32
**
.20
**
.01 .05
Age .17 .14
*
.16 .06 .27
*
.07
Married .05 .06 .07 .07 .14 .07
Income .02 .05 .09 .01 .01 .01
Levelofeducation .07 .03 .23
*
.09 .01 .01
Fulltimeemployed .03 .07 .27
**
.11 .03 .02
Socioeconomic
disadvantage(IRSD)
.04 .01 .08 .03 .06 .11
*
Atefastfoodathome .18 .23
**
.13 .12 .28
**
.29
***
Atefastfoodonweekend .15 .05 .07 .06 .07 .04
Frequencyofconsumption .14 .04 .08 .07 .11 .05
AteHungryJacks .23
*
.22
**
.05 .07 .06 .08
AteKFC .31
**
.13 .15 .09 .05 .10
AteDominosPizza .18 .20
**
.16 .15
*
.17 .05
AteRedRooster .05 .02 .02 .06 .10 .04
Enjoymentofitems .04 .06 .18 .02 .37
***
.46
***
Ateforgeneralconvenience .12 .18
*
.09 .06 .05 .07
Atebecauseofothers .00 .03 .26
*
.07 .07 .05
Atebecauseofattractionof
food
.01 .16
*
.06 .09 .10 .13
*
Ateforincidental
convenience
.02 .10 .17 .09 .08 .02
Shoppingbefore .01 .04 .12 .05 .09 .01
Notdoingmuchbefore .10 .02 .11 .01 .10 .04
Doingmoderateactivity
before
.06 .04 .01 .07 .02 .03
Workingbefore .02 .03 .11 .01 .09 .01
Socialisingbefore .02 .10 .06 .01 .12 .01

n=131;

n=263;
***
p<.001,
**
p<.01,
*
p<.05
156
Figure15:Significantpathsforinterpersonalrelationships,timespenteatingandenergy
intakefromfastfooditemsinmales(n=131).
Atmosphere
No.Family
TimeEating
Energy
Enjoyment
ofFood
Others
MadeMe
Working
Fulltime
AteKFC
AteItems
atHome
Ate
HungryJacks
.31
.24 .39
.28 .37
.26
.27
.23
Levelof
Education
.23
Partner
.32
Age
.27
Figure16:Significantpathsforinterpersonalrelationships,timespenteatingandenergyintake
fromfastfooditemsinfemales(n=263).
Atmosphere
No.Friends
TimeEating
Energy
IRSD
Enjoyment
ofFood
General
Convenience
Attraction
toFood
Age
AteItems
atHome
Ate
Dominos
Ate
HungryJacks
-.13
-.14
.21
.23
.29
.29
.46 -.11
.18
.16
.15
.22
.16
.13
No.Family
Partner
.16
.20
.15
.14
-.20
.21
.15
157
Themodelpredictingtotalenergyconsumedwassignificantlyimprovedbythe
additionalpredictorsformen(F(17,103)=2.30,p<.01)andwomen(F(17,235)=3.79,
p<.001).Thepredictionofpleasantnessoftheatmospherewasalsoimprovedinboth
men(F(17,105)=3.37,p<.001),andwomen(F(17,237)=8.86,p<.001).Timespent
eatingwasnotsignificantlyimprovedbytheinclusionofadditionalpredictorsineither
men(F(17,104)=1.33,p>.05),orwomen(F(17,236)=0.71,p>.05).

Thedifferencesinsignificantpathsformalesandfemaleswereagainapparentinthe
morecomplexmodels.Theresultsarepresentedformalesandfemalesseparately.

Males.BothHungryJacksandKFCdirectlyandpositivelypredictedenergyintake
relativetothereferencecategory(McDonalds).Twodemographicvariables(ahigher
levelofeducationandnotworkingfulltime)andeatinginresponsetosocialpressure
predictedtimespenteating.Inclusionoftheadditionalpredictorshelpedtheregression
foratmospheretoreachsignificance.Enjoyingtheitemseatenandeatingathome
predictedbetterratingsofatmosphereinmales.Aswaswitnessedintheothermodels,
noneofthepredictorsforthenumberofotherspresentrelatedtoatmosphere.Boththe
numberoffamilymemberspresentandeatingwithapartnercontinuedtohavean
indirectrelationshipwiththeamounteatenthroughthetimespenteating.Theindirect
effectswerecomparabletothosewitnessedinthepreviousmodelling:.09(.24*.39;
Sobelstatistic=2.22,p<.05)fornumberoffamilypresentand.12(.32*.39;Sobel
statistic=2.35,p=.01)forpresenceofapartner.

Females.Thenegativedirectrelationshipbetweenthenumberoffriendspresentandthe
amounteatenwitnessedinthepreviousmodelfailedtoreachsignificancewiththe
158
inclusionoftheadditionalpredictors.Interestingly,eatinginthepresenceofapartner
demonstratedasignificantnegativerelationshipdirectlywithenergyintake.The
numberoffriendspresentandthenumberoffamilypresentbothassociatedwith
improvedratingsofatmosphereratingsalthoughthepresenceofapartnerdidnot.

Thepresenceofpeoplefromeachoftheinterpersonalrelationshipspositivelypredicted
thetimespenteatingaswasseeninthepreviousmodel.Theindirecteffectofthe
presenceofothersthroughtimewas.06forthepresenceofapartner(.20*.29;Sobel
statistic=2.37,p<.05)andthenumberoffamilypresent(.21*.29;Sobelstatistic=2.59,
p<.001)and.04forthenumberoffriendsmemberspresent(.15*.29;Sobel
statistic=2.01,p<.05).

Therewereseveralfactorsdirectlypredictingenergyintakeinwomen.Asseeninthe
modelformen,brandoffastfoodeatenwasadirectpredictorofenergyintake.Having
itemsfromHungryJacksresultedinincreasedenergy,whereaseatingDominosPizza
resultedinalowerenergyintakerelativetoeatingMcDonalds.Eatingathomeand
eatingforgeneralconvenienceorbecauseofattractiontothefastfoodalsopositively
predictedenergyintakefromthefastfooditems.Eatingathomeandbecauseof
attractiontothefoodalsopositivelyrelatedtobetterratingsofatmosphere.

Aswaswitnessedinthemodelformen,greaterenjoymentofthefooditemsstrongly
predictedbetterratingsoftheeatingatmosphere.Thenegativerelationshipbetween
socioeconomicdisadvantageandatmospherealsoremainedsignificantafterthe
inclusionoftheadditionalpredictors.

159
DominosPizzahadanegativeassociationwithenergyintakebutapositiverelationship
withtimespenteating.ThissuggeststhatwheneatingDominosPizza,femalesateless,
yetateforlonger.Sensitivityanalyseswereperformedandrevealedthatexclusionof
thisgroupdidnotchangetheoverallmodelandtherefore,itwasretainedinthefinal
model
83
.

5.6.6 Conclusionsfrommodellingfastfoodintakeinmalesandfemales
Itwasinterestingthat,inthepreviousmodel,itwasthenumberoffriendspresentthat
hadasignificanteffectonenergyintakeinfemales,butwhenothercontextualvariables
wereconsideredinthemodel,itwasthepresenceofapartnerthathadthegreatest
directinfluenceonenergyintakeinfemales.Thissuggeststhatsomeofthecontextual
variableshadsharedvariancewiththenumberoffriendspresentandmayhavereduced
themagnitudeofthiseffect.Itisalsonoteworthythatthepresenceofapartnerwasthe
onlygroupmeasurethatdidnotsignificantlyrelatetoratingsofatmosphereinfemales
ineithermodel.Partnersrepresentauniqueinterpersonalrelationshipandaprimary
sourceofinfluenceoneatinganditispossiblethat,ascomparedtofriendsandfamily
members,thepresenceofpartnershasadistinctinfluenceoneatingbehaviour.

Theadditionoftheextrapredictorsimprovedthepredictivepowerofthemodelof
socialinfluenceinfastfoodconsumption,withthenotedexceptionofthemodelfitwhen
predictingtimespenteating.Thismodelwasnotimprovedbytheinclusionofadditional
predictorssuggestingthatothervariables,notcapturedinthestudy,mayinfluencehow
longittakestoconsumeafastfoodmeal.Furthermore,timespenteatingwasnotwell

83
ThemodelswerecalculatedwithandwithoutthosefemaleswhohadeatenDominosPizza.Betavalues
fordirectinfluencesofenergyintakewerealteredminimally(byapproximately.01)butthesignificant
pathsremainedthesame.
160
predictedbythevariablesinthemodel.Thismaybebecauseofthenatureoffastfood
consumptionorthefactthatonlyalimitednumberofvariablesdesignedtoexplaintime
spenteatingspecificallywereincludedinthemodel.Finally,fewfactorsmaybeableto
increasethetimespenteatingfastfoodasthefoodisdesignedtobeconsumedquickly.

Inthemoredetailedmodelofsocialinfluencepresentedhere,additionalpredictors
wereincludedtobetterdescribeinfluencesspecifictofastfoodconsumption.Although
theextrapredictorsimprovedthemodeloverall,theydidnotfundamentallychangethe
indirectrelationshipswitnessedintheearliersocialfacilitationmodels.Thatis,the
presenceofotherpeoplepredictedthetimespenteating,whichthenpredictedthe
amounteaten.Inwomen,thepresenceofotherpeoplealsocontinuedtohaveanegative
indirecteffectontheamounteaten.Thissuggeststhatthemediatingorsuppressing
effectofmealdurationisimportantwithinthemodelofsocialinfluence.Italsosuggests
that,particularlyinwomen,thereissomeaspectofeatinginthepresenceofother
peoplethathasaninhibitoryeffectonfastfoodintake.

5.6.7 Alteredintakeinthosepeopleeatinginthepresenceofothers
Additionalanalyseswereundertakentofurtherexploreinfluencesonfemale
consumptioninthepresenceofothers
84
.Fourmeasuresofperceivednormswere
includedinthepathanalysisforenergyeatenfromfastfooditemsinthisgroup.


84
Originally,theintentionofthenextanalyseswastodevelopamodelforthosepeoplewhoateinthe
presenceofothers.Giventhedifferencesbetweensexesinthepreviousmodels,itdidnotmakesenseto
combinethesamplefortheseanalyses.Furthermore,therewerenotenoughmalesandfemaleseating
withotherstopermitsplittingthemodelintomalesandfemaleswhoatesocially(208femalesand92
malesatewithothers).Giventhehighernumberoffemaleswhoatewithothers,theanalysiswas
restrictedtofemalefastfoodconsumerswhoateinthepresenceofatleastoneotherperson.
161
Thepreliminaryanalysesshowedthatmodelfortimespenteatingwasnotsignificant
(p>.05)inthisgroup.Asthiswasthesecondstepofthepathanalysis,furtheranalysis
didnotoccur.Therefore,inplaceofatestofthemodelforthosewomenwhoateinthe
presenceofotherpeople,asinglemultipleregressionwasperformedthatanalysedthe
directeffectofthe32predictorsonenergyintake.Thebetavaluesresultingfromthis
analysiscanbeseeninTable18.

Table18:Predictorsofenergyintake(withassociatedBetavalues)infemaleparticipantswho
reportedeatingwithatleastoneotherperson(n=204).
Step1:Energyintake
AdjustedR
2
=.197

Predictor Beta Predictor Beta
Numberoffriendspresent .06 AteDominosPizza .19
**
Numberoffamilypresent .02 AteRedRooster .08
Atewithpartner .22
*
Ateforgeneralconvenience .19
*
Timeeating .24
**
Atebecauseofothers .03
Atmosphere .11 Atebecauseofattractionoffood .13
Age .18
*
Ateforincidentalconvenience .10
Married .12 Shoppingbefore .07
Income .06 Notdoingmuchbefore .02
Levelofeducation .00 Doingmoderateactivitybefore .05
Fulltimeemployed .11 Workingbefore .02
Relativesocialdisadvantage .01 Socialisingbefore .13
Atefastfoodathome .16 Enjoymentofitems .03
Atefastfoodonweekend .08 Awarenessofothers .01
Frequencyofconsumption .06 Amountyouthoughtothersate .03
AteHungryJacks .25
**
Concernforothersperceptions .06
AteKFC .16
*
Amountothersthoughtyouate .03

**
p<.01,
*
p<.05

Thetotalvarianceexplainedbythemodelwaslowerforthesamepredictorsinthe
modelofenergyintakeinwomeneatingsociallycomparedtoallwomeninthesample
(19.7%versus24.2%,respectively).Furthermore,theextrasubjectivenormpredictors
162
failedtoaddtovarianceexplainedbytheinitialmodeloffastfoodconsumptionfor
womeneatinginthepresenceofothers(F(4,172)=.36,p>.05).Mostofthepredictors
thathadasignificanteffectonenergyintakeonthesingleeatingoccasionwereidentical
tothoseseeninthemodelforallfemaleeaters.Eatingbecauseofanattractiontothe
styleoffoodandeatingathomedidnotremainsignificantinthefemaleswhoatewith
others.

Thestrengthoftheassociationbetweenthepresenceofapartnerandtheamounteaten
wasstrongerwhenconsideringonlythosefemaleswhoatewithotherpeople.Theeffect
wascomparable,butinanoppositedirection,tothatofthestrongestpredictors(i.e.,
eatingHungryJacksandthetimespenteating).

5.6.8 Conclusionsfromtheanalysisofthosewhoatewithothers
Theanalysisoffemalefastfoodconsumerswhoatewithothersrevealedthatthe
predictorswerebasicallyunchangedbutexplainedlessofthetotalvarianceforthe
amountconsumedatasinglefastfoodeatingoccasion.Thefactthatapproximately75%
ofthefemalefastfoodconsumersatewithothersislikelytoexplainthesimilarity
betweenthemodelforthecompletesampleandthatforthegroupwhichateinthe
presenceofothers.

Despitetheinclusionofthesubjectivenormvariablesdesignedtomeasurepotential
influencesinparticipantseatinginthepresenceofotherpeople,thesepredictorshad
littledirectinfluenceontheenergyconsumed.

163
5.7 ConclusionsfromtheFFSadministeredthroughtheInternet
5.7.1 Themodelofsocialfacilitationforfastfoodconsumption
Althoughthemodellingpresentedinthischapterwasbasedonthemodelofsocial
facilitationpresentedbyFeunekesetal.(1995),therewasminimalevidenceofsocial
facilitationinthemodelscreatedforfastfoodconsumption.Formen,thepresenceofa
partnerandthenumberoffamilymemberspresentattheeatingoccasionpredictedthe
timespenteatingwhichconsequentlypredictedtheamounteaten
85
.Consistentwith
Feunekesetal.smodel,therewasgoodevidenceforthesuggestionthatmealduration
mediatesand/orsuppressestheeffectofthepresenceofothersatafastfoodeating
occasion(discussedfurtherbelow).Finally,therewasalsosomeindicationthat
atmospherecouldpredictthetimespenteatingwhichthenpredictedtheamounteaten
infemales.

Incontrasttothemodelofsocialfacilitation,thedirecteffectofotherspresent
negativelyinfluencedtheamounteateninwomen.Therefore,thiseffectdoesnot
resemblesocialfacilitationsomuchassocialinhibition.Thetheorythatbest
accountsforthisfindinghasbeenlabelledminimaleatingnorms(Rothetal.,2001).
Thisdescribesthetendencytoeatlessinthepresenceofaneatingcompanioninorder
tomakeamorepositiveimpression.Thiseffecthasbeenfoundprimarilyinfemale
diners.Itmaybethecasethat,whenwomenatefastfood,theyreducedtheirintake
relativetothatconsumedinothercircumstances,inordertomeetthisnorm.


85
Therewassomeevidenceofaweak,negativedirecteffectofeffectofpartnersontheamounteaten,
questioningtheutilityofthismodelformaleswhoatewiththeirpartnerbutthiswasnotsignificantatthe
p=.05level.
164
Ifminimaleatingnormswereapparentinthefastfoodeatingoccasionsanalysedhere,it
isunclearwhetherthisissomethingthatisspecifictothenatureofthefastfoodeating
occasionsexploredorisageneraleffectseenwhenwomeneatinthecompanyofother
people.ThecurrentstudydeviatedfromFeunekesetal.s(1995)inseveralways.Itused
asurveymethodologywithalargesampleofAustralianconsumersandfocussed
specificallyonfastfoodconsumption.Nevertheless,deCastro(1994)reportedthatthe
effectsofsocialfacilitationongeneraleatingbehaviourswereclearerinwomendining
withmencomparedtomendiningwithothermen.Perhapsitiseatingfastfoodthat
createsthisinhibitorysocialinfluence.Furtherexaminationofminimaleatingnormsin
fastfoodenvironmentswillbeneededbeforethiscanbedetermined.

Eventhoughmostparticipantsinthestudypurchasedtheiritem/susingatakeaway
format,energyconsumptionwasassociatedwithanindirecteffectofmealduration.
ThisresultisconsistentwithFeunekesetal.s(1995)resultwhichindicatedthattheir
modelofsocialfacilitationcouldbeapplicableatmealpreparationstage,notjustthe
consumptionstage.Thefindingthateatingthefastfooditemsathomeincreasedintake
(inwomenonly)isalsoconsistent.Whenpeopleeatathomethereislessopportunityto
obtainmorefastfood.Thissupportstheideathattheamounttobeeatencanbe
influencedattheorderingstage(Sommeretal.,1992)andsuggeststhatintervention
couldbetargetedatastageearlierthanpurchasing.

5.7.2 Theindirecteffectsofthepresenceofothersonintakeatasinglefastfoodeating
occasion:Mealdurationandeatingatmosphere

ThemodelpresentedbyFeunekesetal.(1995),showedthatmealdurationmediated
theeffectofthepresenceofothersontheamounteaten,suggestingthatthetime
165
extensionhypothesis(deCastro,1990;1994)mightexplaintheeffectsofsocial
facilitation.Inthemodellingperformedinthecurrentstudy,theindirecteffectoftime
onenergyconsumedremainedinbothmalesandfemales,despiteusingfastfoodmeals
totestthetimeextensionhypothesisandaddingvariouspredictorstothemodel.In
females,thisindirecteffectexistedeveninthepresenceofanegativedirecteffectofthe
presenceofothersontheamounteaten.Thisindicatesthattheindirecteffectofthetime
spenteatingisanimportantvariableandrequiresconsiderationintheexplorationof
socialinfluence.

Despiteattemptstovalidatethereliabilityofthemealdurationmeasureusedinthis
study,itispossiblethatwhenparticipantsrecalledthetimetheyhadspenteating,they
wereinfluencedbythetimeextensionhypothesisusingimplicitknowledgeofthiseffect
tocalculateandrecallhowlongtheyspenteating.Inotherwords,itisnotunreasonable
tosuspectthatpeoplewhoatewithothersmayhavebelievedthattheyspentlonger
eating.Withouttestingthisinanobservationalsetting,itisdifficulttoconcludethe
magnitudeofthiseffectinthecontextoffastfoodconsumption.Nevertheless,the
resultsfromthisstudysuggestthatitdoeswarrantfurtherexaminationinafastfood
environment.

Thepresenceofotherpeopleassociatedwithboththetimespentataneatingoccasion
andtheperceivedpleasantnessofaneatingatmosphereinwomeninafastfoodsetting.
Thefactthatthepresenceofotherpeoplecanrelatetoimprovedeatingenvironment
(andviceversa)isnotsurprising.Thereisasocialstigmaassociatedwitheatingalone,
especiallywhendiningout(Stroebele&deCastro,2004).Thus,eatingwithothershas
twooutcomes:itallowsonetoeatforlongerperiodsoftimeandalsofindthe
166
experiencemorepleasant.Incontrasttowomen,formentheeatingatmospheredidnot
seemtohaveanyassociation(directorindirect)withtheamounteatenorthepeople
present.

5.7.3 Howwhoispresentinfluencesintakewhenconsumingfastfood
Previousresearchhasindicatedthatdifferentinterpersonalrelationshipsmaymoderate
socialinfluenceeffects(see,Section3.3,forfulldiscussionofthisresearch).Themodels
constructedfromthecurrentanalysessupportthesuggestionthatnotallpeoplehave
equivalentinfluencesonintake.Inmen,thenumberoffriendspresentatthefastfood
eatingoccasiondidnotsignificantlyinfluenceintake(directlyorindirectly).Incontrast,
eatingwithfamilyand/orapartnerhadanindirecteffectontheamounteaten(through
time).Howinterpersonalrelationshipsinfluencedintakeinwomenwaslessclear
becausetherelationshipsbetweenwhowaspresentandthetimespenteatingchanged
slightlywiththeinclusionofadditionalpredictorsaccountingforthecontextofthe
eatingoccasion.Nonetheless,itwasapparentthatdifferentinterpersonalrelationships
mayaltereatingoccasionsindifferentways.

Itislikelythatanumberoffactorsdeterminethewayotherdinersinfluenceintake.For
example,itmaybethatthepresenceofcertaintypesofpeoplecreatesacompletely
differenteatingenvironment.Indeed,Clendenenetal.(1994)foundthatthepresenceof
friendscomparedtostrangerscouldfacilitatedifferentfoodchoices.Thereisgood
evidencethatfamiliaritydeterminesdifferentlevelsofintake(see,Tice,Butler,Muraven
&Stillwell,1995).Itisalsolikelythattheleveloffamiliaritywiththepeoplepresent
underliesthewaythatpeoplewithdifferentinterpersonalrelationshipsinfluence
eating.
167
Describingthedifferentwaysthatwhoispresentataneatingoccasioncaninfluence
intakeisanimportantadditiontothecurrentunderstandingofsocialinfluenceineating
behaviours.Themodellinginthecurrentstudywaslimitedtofastfoodconsumption.It
islikelythatthemoderatingeffectofinterpersonalrelationshipsisimportantto
considerintheexplorationofsocialinfluencesoneatingbehaviours.

5.7.4 Microenvironmentalinfluencesonfastfoodconsumption
Asidefromsocialinfluences,therewasalsoevidenceofmicroenvironmentalfactors
thatrelatedtointakeoffastfooditems.Theinitialstudyusingthesampleatthefair
indicatedthatthebrandoffastfoodconsumedwasamajorpredictorofenergyintake.
Evenaftercontrollingfordemographicandotherenvironmentalvariables,thebrandof
fastfoodeatenhadadirectinfluenceonenergyintakeinbothmenandwomen.This,
combinedwithresultsofthenutritionanalysisoffastfoodmeals,confirmstheideathat
notallbigbrandfastfoodsarelikelytohaveanequivalenteffectonintakeand,inthe
longterm,health.

AlthoughintheinitialFFS,itappearedasthoughlargerportionsmayexplaingreater
intakeatHungryJacksrelativetoMcDonaldsandDominosPizza,theKFCmeal
constructedinChapterTwohadasimilarportionsizerelativetoonefromMcDonalds.
ThereforeitwasinterestingthateatingKFCcouldpredictincreasedenergyintake
relativetoMcDonalds,inmalesthecurrentFFSsample.Examinationoftheitem
descriptionsoftheorderssuggeststhatapossiblereasonforthiscouldbethestyleof
mealsavailableatKFC.AlthoughthemealsreportedfromMcDonaldswerethetypical
burger,friesandsoftdrinkcombinations,themealsatKFChadmanymoreconstituents.
Theyfrequentlyincludedtwoformsoffriedchickenwithmultiplesideordersanda
168
drink.ThisisinlinewithsomemealdealsthatKFCadvertise.Forexample,theUltimate
BurgerMealconsistsofachickenburger,apieceofchicken,aPotato&Gravy,chips
andasoftdrinkwhichcanthenbeupsized(Yum!RestaurantsInternational,2007).
Therefore,itispossiblethatalthoughtheportionsizesatKFCarereasonable,the
purchasingenvironmentpromotesdealswhichencourageextraitemsand,
consequently,additionalenergyintake.

5.7.5 Sexdifferences,womenandsocialinfluenceduringfastfoodconsumption
Themostimportantfindingforfutureresearchonsocialinfluencestakenfromthe
currentresultsisthesexdifferencesthatwereobserved.Ultimately,althoughbothmen
andwomenwereinfluencedbythepresenceofothers,thewayinwhichthisinfluence
occurredwasdifferent.Theclearestexampleofthiswasthepositivetotaleffectof
othersontheamounteateninmenandthenegativetotaleffectseeninwomen.

Inthecontextofeatingbehaviours,differentinfluencesaffectingwomenandmenmay
resultfromvaryinginterestandengagementwiththeissueofhealth.Participant
recruitmentoutcomesherehavesuggestedthatwomenaremoreinterestedinhealth
thanmen.Thisheightenedinterestmaymeanthatdietandfoodchoicearedetermined
differentlyinwomenthanmen.Moreover,thehighfatnatureoffastfoodmayserveto
exacerbatesexdifferences.Researchsuggeststhatmostpeopleimplicitlyassociatefat
withnegativeaffect(Roefs&Jansen,2002)andthatwomenaregenerallymore
concernedwitheatingandtheirbodyweight(Pliner,Chaiken,&Flett,1990).Thereis
littledoubtthatmanypeopleknowthatfastfoodisgenerallyhighinfat;fastfood
companiesmarkethealthychoices,suggesting,indirectlythatotherchoicesarenot
healthy.Theperceptionthatfastfoodisnothealthyandagreaterawarenessofhealth
169
anddietingmaymakewomenmoresensitivetotheirenvironmentastheyattemptto
assesswhatisanappropriateamounttoeat.

Thefactthatthepresenceofotherscannegativelypredictintakewasfoundevenafter
controllingforamyriadofotherpotentiallyimportantfactorsinwomen,suggeststhat
womenarereactivetothepresenceofothers,possiblymoresothanmen.Itispossible
thefemaleintakeisbetterpredictedbyminimaleatingnormsratherthansocial
facilitation.Itistheorisedthatfollowingtheformerleadstoanassociationbetween
moderationoffoodintakeandfemininity(Mooney,Detore,&Malloy,1994).Ifthisnorm
isgenderspecific(beingmotivatedbysexroleexpectations),itispotentiallycrucial
whenconsideringhowsocialinfluencesaffectmenandwomendifferentlyandwarrants
furtherresearch.

5.7.6 Limitationsofthecurrentstudy
Thecurrentfocusoncrosssectionalsurveydatameansthatitisnotpossibleto
determinewhethertheamountoffastfoodeatenatasittingimpactedontheamountof
foodconsumedthroughouttheremainderoftheday.Consequently,itisnotpossibleto
linkhigherconsumptiontoobesityriskbecausebehavioursrelatedtooptimisingenergy
balancemayservetomediatethisrisk.

Generalisabilityoftheresultsiscompromisedbypossiblelimitationsofthesampling
process,particularlythefocusonInternetdatacollectionandthelowerrepresentation
ofmen.Recruitingmaleparticipantsisaproblemthathealthresearchfacesrepeatedly
withnoobvioussolution.Despitehavingareasonablenumberofrespondents,when
dividingthesamplebysex,thelackofmeninthesamplereducedtheabilityofthis
170
modeltoexploreinfluences;itimpededtheabilitytotestamodelofincreasedfastfood
intakeforpeopleeatingfastfoodinsocialgroups.Regardless,theregressionforwomen
gavesomedirectionforfutureresearch.Itshowedsimilartrendstothemodel
developedusingallwomen.

Inthisstudy,timewasdefinedastimespentengaginginaneatingbehaviourrather
thantimeactuallyeating.Althoughtheremaybesomeindividualvariationintherateof
eatingthatmayaltertimespentataneatingoccasion,thedefinitionoftimeusedhere
hasbeencommoninothersocialinfluenceresearch.Attemptsweremadetoimprove
thewayinwhichtimewasmeasured,however,therecalloftimeisstilllikelytolackthe
accuracyoftimedatathatisrecordedduringtheeatingoccasionbyanobserver.

Finally,thisstudywaslimitedtobigbrandfastfoods.Previousexperimentalstudies
haveofferedtheirparticipantsonlyoneortwotypesoffood;itisimportanttorecognise
thatthetypeoffoodeatenmayinfluencetheextenttowhichthepresenceand
behaviourofothershasaninfluenceonconsumptionamounts.Nevertheless,differences
betweenthestylesoffooditemsprovidedbythedifferentfastfoodchainsmayhave
affectedtherelationshipsinfluenced.Forexample,thedirectrelationshipbetweenthe
amountofenergyconsumedandeatingDominosPizzainwomenwasnegativewhereas
havingpizzaalsorelatedpositivelytothetimespenteating.Pizzarepresentsaneasy
andcheap
86
foodtoshareamongmanypeopleatasocialgathering.Furtherinspection
ofthedescriptiveinformationsupportsthisassertion.Acomparisonofthenumberof
peopleeatingalone,inapairorinagroupbetweenthedifferentfastfoodbrands,
showedthatfewerpeopleatepizzaalone(15.2%ofpeopleorderingDominosPizza),

86
InAustralia,DominosPizzaofferlargepizzasforaslittleasAUD4.90each.
171
comparedtootherformsoffastfood(21.4%to31.4%ofallpeopleeatingotherfast
foodsatealone).Socialinfluencemaybeeithermediatedormoderatedbycultural
normsaroundtheconsumptionofdifferenttypesoffood.deCastro,Bellisle,Feunekes,
DalixanddeGraaf(1997)reportedthattheamountoffoodconsumedoveraspecified
timeperiodvariedbetweencultures.Australianwomenspreferenceforeatingpizza
withothersmayreflectaculturalnormwherebypizza,becauseofthemealsizeandthe
wayitisserved,isviewedasamealtoshare.

5.7.7 Directionsforfutureresearch
Asissuggestedinpreviousmodelsofsocialinfluence,mostofthesocialeffectsoneating
wereindirect,meaningtheypredictedoutcomesthroughamediatingvariable.
Performingpathanalysisalloweddetectionofamediatingandsuppressing
relationshipsthatwouldhavebeenlostinsimplecorrelationalanalysesandrepresents
anexcellentmethodforassessingtheofteninterrelatedeffectsofsocialand
environmentalinfluencesoneatingbehaviour.Multiplepapersindicatethatminimal
totaleffectsshouldnotwarranttheexclusionofvariablesfrommultipleregression
analyses(Friedman&Wall,2005;Shieh,2006).Thiswasverymuchthecaseinthe
currentstudywherethecombinedindirectanddirecteffectsofthepresenceofothers
cancelledeachotheroutinwomen,creatingasmalltotaleffect.Althoughmodelling
oftenrequireslargersamplesizes,futurestudiesinsocialfacilitationshouldconsider
usingmodellingasastatisticalmethodtoassessitseffects.

Despitegoodevidenceforthetimeextensionhypothesis,socialfacilitationmaynotbe
asdemonstrableinfastfoodconsumptionasitisingeneraleatingbehaviours.Itwas
clearthatotherfactorsandsocialinfluencesmayalterbehaviourandthepresenceof
172
othersmaynotresultinincreasedintake,especiallyinfemalefastfoodeaters.More
needstobeunderstoodaboutthedifferentwaysinwhichsocialenvironmentsinfluence
menandwomen.Itappearsthatadherencetodifferentsocialnormsandimpression
managementmayunderliethesediscrepancies.Therefore,understandinghowpeople
doordonotconformtothesenormswilldeveloptheunderstandingofsocialinfluence
beyondtheeffectsofsocialfacilitation.

Timespenteatingwasarobustpredictoroffastfoodintakedespitetheinclusionof
differentvariables
87
.Itisthereforereasonabletoconcludethatoneofthemechanisms
behindsocialinfluenceisthetimespentataneatingoccasion.Whetherthiseffectcan
counteractotherinhibitorysocialinfluences(i.e.,thepresenceofothersforwomen)is
unclear.Nevertheless,ifspendinglongereatingcanincreasetheamountconsumed.
Therefore,itispossiblethatfactorsoutsideofsocialinfluencescouldactinasimilar
waytotimeextensioninthesocialfacilitationparadigm,prolongingthetimespent
eatingandtheamounteaten.SommerandSteele(1997)foundthatdinersinsocial
situationsextendedthetimespentataneatingoccasion.Theyalsoobservedthatpeople
diningalonecouldextendtheireatingtimethroughreading.Theysuggestthatreading
allowslonedinerstoreducethediscomfortcreatedbydiningalone.If,asisseeninthe
modelofsocialfacilitationinfastfoodconsumption,thetimespenteatingrelatesto
greaterconsumption,thenpeoplereadingmayalsoeatmore.Thismaybeawaythatthe
timeextensionhypothesiscanaffectlonedinersthusquestioningtheassociation
betweensocialvariablesandmealduration.Furtherresearchisneededtotestthis
possibility.


87
Thefactthatthisrelationshipwasnotfoundinthepreviousstudymaybearesultofthelimitedsample.
173
Giventheresultsofthecurrentstudy,theaimofthenextstudywastoexplorethe
effectsofsocialinfluenceinamorenaturalisticenvironmenttofurtherexaminethe
genderdifferencesthathavebeenrepeatedlydiscoveredintheexplorationoffastfood
consumptionandrelatedbehaviourstodate.
174

6 ChapterSix:Fastfoodconsumption,minimaleatingandimpression
management


Therehasbeensuggestioninpreviouschaptersofthisdissertationthatgenderroles
mayhaveanimportantinfluenceoneatingbehaviours,relatingtoincreasesand/or
decreasesintheamountoffastfoodeaten,dependingonboththesexoftheperson
eatingandthatofotherspresent.Feunekes,deGraafandvanStaveren(1995)didnot
accountforsexdifferencesinthedevelopmentoftheiroriginalmodelofsocial
facilitation(possibilityduetotheirsmallsamplesize).

Earlyresearchintosocialfacilitationineatingdidnotreportontheassessmentofany
potentialgenderdifferencesdespiteusingmixedsexsamples(deCastroandBrewer,
1992;deCastro,1990).Subsequentresearchassessedthesepotentialdifferences.Redd
anddeCastro(1992)foundthatmenandwomenatedifferentamountsoffoodbutdid
notseeanydifferencesbetweenmenandwomenindifferentsocialconditions(eating
withothersoralone).Clendenen,HermanandPolivys(1994)studyinvolvedonly
femaleparticipantsanddemonstratedevidenceofsocialfacilitation;thecoeaterswere
alsoexclusivelyfemale.deCastro(1994)presentstheonlystudyexploringdifferences
intheeffectsofsocialfacilitationbetweenwomenandmen.Hereportedthattheeffects
ofsocialfacilitationwereclearerinwomendiningwithmenthanmeneatingwithother
men.Theresultsfrommodellingofsocialfacilitationinfastfoodconsumers(discussed
inthepreviouschapter)suggestedsomeeffectsthatwerethecontrarytodeCastros
results;womenseemedtoeatlessinthepresenceofothers.Theseresultssuggestthe
175
activationofsomeformofminimaleatingnorminwomen.Thelargebodyofresearch
thathasexploredthisphenomenonwillbedetailedinthefollowingsection.

6.1 Womenandeatinglightly:Areviewoftheliterature
Theideathateatingsmallservesmightconveyafeminineimageforwomenhasbeen
hypothesisedintheliteraturefor20years(Chaiken&Pliner,1987;Mori,Pliner,&
Chaiken,1987).ChaikenandPliner(1987)amongstinitialgroupsofresearchersto
empiricallytestthissuggestion.Theygavestudentswrittendescriptionsofaperson,
manipulatingthegenderofthetargetpersonandtheamountoffoodtheyenjoyed
eating.Theresearchersthenaskedstudentstoratethetargetpersonaccordingtoa
numberofpersonalcharacteristics.Theyfoundthatwhenafemaletargetwasdescribed
aseatinglessfood,shewasratedasbeingmorefeminine.Ratingsofmaletargetsdidnot
changesignificantlywithmealsize.

Mori,ChaikenandPliner(1987)exploredthisfindingfurtherbytestingapossible
mechanismthatmightexplainthisresult.Morietal.hypothesisedthatwomenwho
wantedtoappearfeminineinasocialinteractionwouldrestricttheirintake.Inorderto
testthisthesis,theyplacedparticipantswithpartnersofvaryingdesirability(all
confederates)
88
.Participantsweretoldthattheexperimentwasabouthowpeople
becomeacquainted.Abowlofcandycoatedchocolates(i.e.,M&Ms)wasplacedonthe
tablewhiletheparticipantandconfederatetalkedandthedependentmeasurewasthe
numberconsumedduringtheconversation.Tocontrolforsocialfacilitationof
consumption,theconfederateatethesameamountinallconditions.Morietal.found

88
Tomanipulatedesirabilityandensurethatthepartnerwouldbeseenasattractivebeyondtheir
appearance,thepartnerspersonalprofilewasprovidedtoparticipantsbeforetheexperiment.This
outlinedpersonalinterestsandcharacteristicsthatwereratedasattractiveinoppositesexpeopleduring
piloting.
176
thatthepresenceofadesirableoppositesexpartnerresultedinfewerM&Ms
consumedbybothmaleandfemaleparticipants.Theysuggestedthatthedesireto
appearfemininemotivatedminimaleatinginwomenandtheorisedthatmenmayhave
reducedtheirintakeforreasonsbeyondgenderroleexpectations.

Inordertoconfirmtheirfemininityhypothesis,theresearchersconductedasecond
studywheretheythreatened(orreinforced)femaleparticipantsfemininity.They
hypothesisedthatwomenwhosefemininitywasthreatenedwouldhaveagreaterdesire
tocreateafeminineimpressionandconsequentlyeatless.Morietal.(1987)gavethe
participantsfeedbackindicatingthattheyhadbeenratedaslow(orhigh)infemininity
bytheirmalepartner.Ashypothesised,femaleparticipantsatesignificantlylessinthe
presenceofadesirablemalepartnerwhentheirfemininityhadbeenthreatened.The
authorsconcludedthataminimaleatingnorminwomen(i.e.,constrainedintake
quantity)mayresultfromconcernsaboutappearingfeminine.

Subsequentstudieshaveconfirmedtheminimaleatingnorm.Usingamethodsimilarto
ChaikenandPliner(1987),BockandKanarek(1995)gaveparticipantsdifferentmeal
descriptionsandaskedthemtoratethequalitiesofthefictionalmaleorfemalewho
consumedaparticulardiet.Theymanipulatedthegenderofthetargetperson,andthe
amountandthetypeoffoodstheyate.Peoplewhoatelargeramountsoffood(both
maleandfemale)weredescribedasmoremasculine.Peopledescribedaseatingsmaller
mealswereratedasneaterandmoreconcernedwiththeirappearance.Theauthors
reportednogenderbaseddoublestandard;thefemaletargetwasnotjudged
differentlyfromthemaletargetaccordingtotheamounteaten.Nonetheless,in
contradictiontothisassertion,examinationofratingsofgoodlooksfromthisstudy
177
indicatedthatfemaleswhoatelesswereratedassignificantlymoreattractivedespite
thefactthatthedescriptionofthetargetwasidenticalacrossconditions.Thissuggests
thattheremayhavebeendifferencesintheperceptionofthetargetrelativetothe
amounttheyateandtheirgender.PlinerandChaiken(1990)replicatedthisvignette
studyusinganexperimentalmanipulation.Theirresultsindicatedthatratingsof
femininityvariedinverselywithnumbersofcrackersconsumed.

Twootherexperimentalstudieshavedemonstratedhowdynamicsbetweengenderand
groupsizewithinthesocialenvironmentcaninfluencetheconsumptionbehavioursof
individuals.Klesges,Bartsch,Norwood,KautzmanandHaugrud(1984)observedpeople
eatinginvarioustypesofrestaurantsandrecordedthegroupsizeofthepeopleeating,
thegendercompositionofthegroup,theweightstatusofmembersofthegroupand
foodeaten.Theyreportedthatwomenateasimilaramounttothemenwheninasmall
groupinfastfoodrestaurants,butinlargergroups,femalesatesignificantlylessthan
males.

Usingamethodsimilartothoseusedinexperimentalresearchonsocialnorms,Salvy,
Jarrin,Paluch,IrfanandPliner(2007)invitedpeopletocometotheirlaboratoryfora
studyofsocialexchangesbetweenfriends,partnersandstrangers.Whilethe
participantsinteracted,snackfoodswereprovided.Theresearchersfoundthatmale
participantswiththeirfriendsatemorethanallothergroups.Therewasalsoatendency
forbothmaleandfemaleparticipantstodecreasetheirintakeinthepresenceof
unfamiliarpartners,suggestingeatingminimallymaynotbeaselfpresentational
concernexclusivetowomen.Matchingofintakewasonlyobservedwhenafemale
partnerwasinvolved.Theauthorssuggestedthatthehigherlevelofmatchingobserved
178
indyadswithafemaleparticipantoccurredbecausewomenweremoresensitiveto
theirpartnersbehaviourormoreconcernedwithselfpresentationwhile,ontheother
hand,menwantedtoappearindividual.Thisideaissupportedbypapersthathave
suggestedthatwomenaremoreresponsivetopressuretoconform(Eagly,1983)and
womenratesomeonewhofailstoconformtonormsmorenegativelythanmen
(Rudman,1998).

Recently,researchonsocialinfluence,genderandfoodconsumptionhasshiftedfrom
exploringnormativeinfluencesonconsumptionvolume,tofocusonfoodselectionand
gender.Thisworksuggeststhatminimaleatingnormsmaybeevidencedindifferences
intheenergycontentoffoodconsumed,regardlessofwhetherthesedifferencesare
determinedbyamountorenergydensity.Studiesdesignedtoidentifyfoodthatcouldbe
describedasfemininehavehighlightedthelightnatureofthesefoods;saladsanddiet
beveragesarecommonlyratedasfeminine(Basow&Kobrynowicz,1993;Mooneyetal.,
1994;Mooney&Lorenz,1997).Marketresearchdataconfirmtheseresultsshowing
significantdifferencesinconsumptionoflowenergyfoodsbetweenthesexes(Driskell,
Meckna&Scales,2006).

Genderstereotypingoffoodextendstoratingsoffemaleeaters;womenwhoeat
femininefoodsareratedmorefavourablythanwomeneatingothertypesoffood.Basow
andKobcynowicz(1993)foundthatfemaletargetsdescribedaseatingfemininemeals
wereratedasmoreappealingthanfemalesdescribedaseatingaverymasculinemeal.
Thisstudyconfoundedvolumeandenergydensitysoitwasnotpossibletoseparatethe
effectsofthesevariables.MooneyandLorenz(1997)keptvolumeconstantwhile
manipulatingthefemininityofthedietaryprofileandthegenderofthepersoneating
179
thediet
89
.Theirresultsindicatedthattheconsumptionoflighter(also,typicallymore
feminine)mealsledtomorefavourableimpressionsofbothmaleandfemaletargets.
Consistentwiththis,Mooney,deToreandMalloy(1994)foundthatfemalesdescribed
aseatinglowfatfoodwerejudgedmorepositively(i.e.,moreattractive,more
intelligent,calmerandhavingsmallerbodysize)thanfemaleseatinghighfatfood.The
researchonpersonalratings,femininityanddifferenttypesoffoodsuggeststhateating
foodthatisenergylight(frequentlylabelledlite)mayalsobeassociatedwithtypically
femininequalities.Beyondthis,italsosuggeststhatwomenwhoadheretogenderrole
expectationsarerewardedthroughpositiveinterpersonalratings.Thus,itmaynotbe
simplythevolumeeatenbutalsothetypeoffoodeaten,anditsinferredenergydensity,
thatmodifiessocialjudgements.

Itisclearfromempiricalresearchthattherearecertainaspectsofeatingthatare
genderorientated.Forexample,eatingminimallyisoftenseenasfeminine.Itisless
clearwhether,andifso,howwomenthenusethisinformationtoreducetheirintake,
andunderwhichcircumstancesthisislikelytooccur.Resultsfromlimitedexperimental
andobservationalstudiesprovidesomeevidencethismayoccur,butitislessclearhow,
andif,menchangetheirintake.Muchresearchhasfocusedsolelyonwomeneaters.
Considerablylessseemstobepublishedaboutmasculinityandhowratingsofmale
attractivenessmaybeaffectedbythevolumeandtypesoffoodmeneat.Nevertheless,
empiricalresearchemphasisestheneedtoconsidergenderandgenderinteractions
whenexploringeatingbehaviours.


89
Ininitialpiloting,subjectswereaskedtocreatelistsoftypicallymasculineandfemininefoods.Themost
frequentlymentionedfoodswereusedinthestudy.Thefeminineprofileincludeditemssuchasa
wholegraintunasandwich,atoastedbagelwithcreamcheeseandspaghetti.Themasculinemealincluded
atunasub,steakandflapjack.
180
6.1.1 Impressionmanagementandsexroles:Mechanismsbehindminimaleatingnorms
Theresearchoneatingnormsdescribedabovesuggeststhatindividualsmaymoderate
whattheyeatasanactofimpressionmanagement(Morietal.,1987;Rothetal.,2001;
Salvyetal.,2007).Specifically,theaimwouldappeartobetomodifyselfpresentation,
viamanipulationoffoodconsumed,insuchawayastoenhanceonesappearanceto
others.Baumeister(1982)suggestedthateffectiveselfpresentationinvolves
conformingtoexpectationsandsocialnorms.Thisisclearlythegoalofthose
participatinginthestudyundertakenbyMorietal.inwhichfemaleparticipants
attemptedtoenhancetheirimagebyeatinglessandconformingtofeminineideals.
Otherstudiesindicatethattheneedtocontrolselfpresentationchangesdependingon
thegendercompositionofgroupmembersandfamiliaritywithinthegroup(Learyetal.,
1994).

Itisimportanttoacknowledgethatminimaleatingnormsmayalsoexistforreasons
beyondtheneedtoenhanceonespresentationtoothers.Schlenker(1980)defined
impressionmanagementastheconsciousorunconsciousattempttocontrolimages
thatareprojectedinrealorimaginedsocialinteractions(p.6).Hesuggestedthat
impressionmanagementpermeatesavarietyofdifferentformsofinteractionsnot
onlythosewheretheneedforapositiveselfpresentationishigh.Impression
managementduringinteractionscanservetodefine,confirmandadjustselfidentity
androlesinsociety(Bozeman&Kacmar,1997).

Schenklers(1980)modelofimpressionmanagementsuggeststhatanindividuals
desiretoeatinconformancewithgenderstereotypesmayassistwomentoestablish
selfidentity.Research,inanumberoffields,hasestablishedanassociationbetweenthe
181
waypeopleeatandthewaytheyseethemselves(Bisogni,Connors,Devine,&Sobal,
2002;Mooney&Lorenz,1997).Vartanian,HermanandPolivy(2007)reviewedthe
literaturelinkingeatingandimpressionmanagement.Theyidentifiedtwotruismsthat
highlighttheassociation:youarewhatyoueatandyouarehowmuchyoueat.
Therefore,itislikelythateatingminimallycouldhelpwomentomaintaintheirfeminine
identity.

Otherresearchhasattemptedtoestablishwhetherimpressionmanagementandgender
stereotypingofbehaviourarereinforced.Wood,Christensen,HeblandRothgerber,
(1997)observedthatwhenpeopleconformtotheirgenderroles,theyfeelmore
positive.Effectiveimpressionmanagementalsoleadstomorepositivepersonal
evaluationsbyobservers(Wayne&Liden,1995).Ifthisisthecase,itwouldbeexpected
thatthedesiretoeatlightly(i.e.,conformtoaminimaleatingnorm)mayextend
beyondsituationswhereconcernsaboutselfpresentationareparamount,toavarietyof
eatinginteractions.ThiswasseeninthepreviousstudyinChapterFiveinwhichthe
presenceofpartnersnegativelypredictedenergyintakeinfemalerespondents.A
partnerbelongsfirmlywithintheprimaryreferencegroup,influencingmanyofour
behaviours,includingeating.Paradoxically,thefamiliarityhypothesissuggeststhat,as
interpersonalrelationshipsbecomecloser,theneedtocreateapositiveimpression
decreases(Tice,Butler,Muraven&Stillwell,1995).Therefore,awaythatthisinfluence
canbeunderstoodisnotbyexclusivelyconsideringthedriveforselfenhancementbut
byalsotakingintoaccountthedesiretosatisfysexroleexpectationssuchasappearing
feminine.

182
6.1.2 Fastfoodandimpressionmanagement
ParticipantsinFriesandCroyles(1993)studyindicatedthattheywouldratherbe
friendswithapersonwhoconsumedalowfatdietthanafastfoodconsumer
90
.Ratings
fromtherespondentsrevealedthatfastfoodconsumershadmorenegative
characteristics,includingbeinglessstudiousandmorelikelytobeoverweightthan
peopleonthelowfatdiet.SteinandNemeroff(1995)foundthatpeoplewhowere
describedaseatingfastfoodwerealsoseenaslessactive.Theredefinitelyseemstobe
evidenceofnegativestereotypessurroundingfastfoodconsumptionintheliterature.
Suchstereotypesmayincreasethedriveforimpressionmanagement.However,itis
unclearhowpervasivethestereotypessurroundingfastfoodconsumptionare.Asmore
peopleeatfastfoods,thestereotypesaboutfastfoodconsumersmayevolve.Thirty
yearsago,SadallaandBurroughs(1981)describedaverydifferentprofileoffastfood
consumers;theyweredescribedassomeonewhoisfamilyorientatedandconservative.

Itisdifficulttogaugehownegativethecurrentstereotypesoffastfoodconsumersare
and,consequently,theimpactthattheymayhaveonbehaviour.Nevertheless,recent
researchsuggeststhatnegativestereotypesstillsurroundpeoplewhoconsumefast
food(Dunn,Mohr,Wilson&Wittert,2008).Therefore,thereisalikelihoodthateating
fastfoods(relativetootherfoods)triggersimpressionmanagementstrategiesin
women,especiallyconsideringthatfemaletargetsdescribedaseatingenergydense
foodsareperceivedaslessfeminine(Mooneyetal.,1994;Mooney&Lorenz,1997).
Consequently,itislikelythatwomenwillfeelpressuretoeatminimallytoreinforce

90
Theexperimentaldescriptionofsomeonewhoeatsfastfoodalsomentionedthattheyavoideddiets.
Therefore,itcouldbearguedthatthisperceptionisnotrelatedsolelytofastfoodconsumption.
183
theirfeminineidentity.Moreimportantly,thismeansthatitislikelythatminimaleating
normsmaybeparticularlyrelevantinafastfoodenvironment.

6.1.3 Aimsofthecurrentstudy
ChapterFiveshowedthatpresenceofothershadanimpactontheamountconsumedin
afastfoodmeal;thenumberofpeoplepresentattheeatingoccasionpositively
predictedthetimespenteatingwhich,inturn,predictedtheamounteaten.Thisresult,
basedondatafromsurveys,wasinterpretedassupportforthetimeextension
hypothesis.Alimitationofthisstudywastherelianceonaccuraterecalloftheamountof
foodconsumedandthetimespenteatingfollowingthemostrecentconsumptionoffast
food.AlthoughthedatawereconsistentwithBellandPliners(2003)observationaldata
whichshowedthatthetimespenteatingincreasedinthepresenceofothers,neitherset
ofdataclearlyindicateswhethertheamountconsumedalsovaried.Thisiscriticaltothe
determinationofwhethermealdurationcanbealteredbythepresenceofothers.

GivenresultsofthepreviousFastFoodSurvey(FFS),itisexpectedthat,inanaturalistic
environment,peopleeatingingroupsinafastfoodrestaurantwilleatforlongerwhich
will,inturn,beassociatedwithgreaterenergyintake.Theserelationshipswerefound
consistentlyacrossmalesandfemalesinthepreviousstudy.Itisexpectedthatthe
relationshipbetweenthenumberofpeoplepresentandtheamounteatenwillbe
negativeforwomen.Inmales,itisexpectedthatrelationshipbetweenthenumberof
peoplepresentandtheamounteatenwillnotbesignificantastherewasnoevidenceof
thisrelationshipinanyofthemodelsoffastfoodconsumptioninthepreviouschapter.
Consequently,thefollowingarethehypothesestobetestedinanobservationalstudyof
foodconsumedinafastfoodrestaurant:
184
Hypothesis1.1:Fastfoodconsumerseatingingroupswilleatfor
longerthanpeoplediningalone.
Hypothesis1.2:Timespenteatingwillbepositivelyrelatedtothe
amountoffastfoodeatenandthenumberofpeoplepresent.
Hypothesis1.3:Thetotaleffectofothersontheamountoffastfood
consumedwillbenegativeinfemalesobserved.

Ithasbeenpreviouslyshownthatlonedinersmayincreasethetimetheyspendatan
eatingoccasionbyreading(Sommer&Steele,1997).ResultsfromthestudyinChapter
Fivesupportedtheideathattimespenteatingcanincreasetheamounteateninafast
foodrestaurant.Itisthereforepossiblethatiflonedinersreadandincreasethetime
theyspendeating,thismayalsoincreasetheamounteaten.Thispossibilityprovidesa
basisforthefollowinggroupofhypotheses:
Hypothesis2.1:Lonedinerswhoarereadingwilleatforlongerthan
lonedinerswhoarenotreading.
Hypothesis2.2:Therewillbepositiverelationshipbetweenthetime
spenteatingandtheamountconsumedbylonediners.

Hypotheses3and4weredesignedtoassesstheeffectofnormsonconsumptioninafast
foodenvironment(thusaddressingthethirdresearchquestionposedinSection3.5).
Therehasbeensomesuggestioninpreviousstudies(Rothetal.,2001)thatmatchingof
intakeoccurswheneatinginasocialcircumstance(i.e.,anintakenormisestablishedby
thoseparticipatinginthemealevent).Ithasbeensuggestedthatthepresenceofa
femaleeatercanemphasisetheimportanceofthismatchingnorm(Salvyetal.,2007).
Therefore,itisexpectedthattherewillbesomematchingofintakewithingroupsoffast
185
fooddiners.Itisunclearhowmatchingwilloccurwithingroupsthatdifferbygender
composition.
Hypothesis3:Therewillbeevidenceofmatchingwithingroupsof
fastfoodconsumersinanaturalisticsetting.

Thereiscurrentlygoodevidencefortheexistenceofaminimaleatingnormforfastfood
consumption.Thisisevidencedinthenegativetotaleffectofthepresenceofothersseen
inwhenmodellingfastfoodintakeinthepreviousstudyandmoregenerallybythe
previousstudiesdemonstratingtheeffect.Theunfemininenatureoffastfoodsmay
triggeradherencetothesenormswhenconsumingthisfood.Previousstudiesusingthe
FFSdidnotallowforassessmentofthesenormsbecausethesexofthepeoplepresent
wasnotrecordedinthesurveydata.Therefore,thefinalaimofthecurrentstudyisto
determinewhether,minimaleatingnormsareactivatedintheconsumptionoffastfood
andwhetherthese,inturn,alterthevolumeeaten(intermsofenergyintake)insocial
interactions.

Itwaspredictedthatfemaleseatingwithmalesinanaturalisticfastfoodenvironment
wouldeatlessthanfemaleseatingwithagroupofotherfemales.Itisunclearhow
differentgendercompositionsamongsocialdinerswillaffectmales.
Hypothesis4:Females[males]eatinginthepresenceofmixedsex
companywhiledininginafastfoodrestaurantwilleatdifferentlyto
females[males]insamesexgroups.

186
6.2 Method
6.2.1 Development
Observationalmethodsallowdatatobecollectedfromanaturalisticenvironment,and
thesedataaregenerallyviewedashigherinecologicalvaliditythandatacollectedinthe
laboratory.Traditionalmethodsutilisedinsociologicalobservationalstudieshave
involvedobserving,recording,classifyingandevaluatingbehaviours(Hennessy,Mabey,
&Warr,1998;Jorgensen,1989).Thesecontrastwithothermethodsofobservational
recording,developedfromabehaviouraltradition,thatinvolvethepreliminary
developmentofbehaviouralcategoriesthatareutilisedinthedevelopmentofachecklist
designedtocaptureonlydatarelevanttothepurposeofthestudy(Hennessyetal.,
1998).Thismethodhasbeenusedinanumberofobservationalstudiesofeating
behaviours(Klesgesetal.,1984;Sommer&Steele,1997).Inthesestudies,aresearcher
sitsandunobtrusivelyobservespeopleinaneatingenvironmentandrecordsthetime
thateatingstartsandstops(Bell&Pliner,2003;Sommer&Steele,1997)andinsome
cases,theitemsconsumed(Klesgesetal.,1984).Thesestudieswereusedtoguidethe
developmentofthemethodusedinthecurrentstudy.

Theobservationaldatacollectionmethodlimitedthedemographicdatathatcouldbe
recordedtosex,weightcategoryandagegroup.Thelattertworequiredsome
estimationskillsbutbothtypesofdatahavebeenreportedinpreviousobservational
studies.Forexample,SommerandSteele(1997)recordedageindecileswhileKlesgeset
al.(1984)trainedobserversinweightstatusobservations.Timespentatthetablecan
beeasilyobservedascantheitemseaten.Calculatingtheenergycontentofthefood
consumedwasmadeeasierbythestandardisedportionsizeandthedistinctive
187
packagingofitemsconsumedinafastfoodrestaurant.Forexample,McDonaldsemploy
acolouringsystemthatdefinesdifferentburgersaBigMaccomesinaboxwithred
colouring.

McDonaldsrestaurantswereselectedforobservationbecausetheyaccountforthe
highestproportionoffastfoodsalesinthequickservicerestaurantindustryinAustralia
(BISShrapnel,2003).Moreover,datafromtheFFSinthepreviousstudiesindicatedthat
McDonaldscustomersfrequentlydinedin.Behaviouralrecordingofmealcontentwas
facilitatedbyMcDonaldspackagingwhichallowedtheidentificationofdietandnon
dietbeveragesandburgertype.Finally,McDonaldsalsohasalargermenuwithmore
healthyoptionsthanequivalentfastfoodchainsallowingforgreatervariabilityin
energyintake,thecriticaldependentvariableofthestudy.

ThehumanethicscommitteeattheSchoolofPsychologyintheUniversityofAdelaide
approvedthecurrentstudy
91
.ThehumanethicscommitteeatCSIRO,HumanNutrition
endorsedthisapproval.

6.2.2 Piloting
PreliminaryvisitsweremadetoMcDonaldsrestaurantsaroundthecityofAdelaideand
itssurroundingsuburbs.Floorplansweremadeof11restaurantstoassesswhichhad
themostdesirablelayoutforoptimalobservations.Overtheperiodofoneweek,ten
twohourpracticeobservationsessionswereconductedatMcDonaldsrestaurantsin
thenorthandnortheasternareasofAdelaide.Thepurposeoftheseobservationswas

91
Ownersoftherestaurantswerenotinformedofthestudy.Theethicscommitteestipulatedthatto
justifyprolongedoccupationofaseatintherestaurant,theobserver/swouldhavetopurchaseitems.
188
to:refinethebehaviouralchecklist,estimatehowbusytherestaurantswere,and
developastrategyforoptimisingobservationalaccuracy.TheMcDonaldsrestaurantsin
thecitycentreweretheonlyrestaurantsthatwereconsistentlybusy
92
.

Doubleentry:Refiningmeasures.Onthreedatacollectionoccasions,asecondobserver
recordeddatatoassessandrefinethemeasuresbeingcollected.Thirtyonecaseswere
recordedoverthreeoccasions.Recordsoftheitemsconsumedandthetimespentatthe
tabledemonstratedhighconsistencyacrosstheprimaryobserverandthelesser
experiencedobserver.Thetotaltimespentatthetablerecordedbyeachresearcherwas
identicalforall31nodalparticipants.Onetimetherewasdiscrepancyinthedescription
oftheorderrecorded.Thiswasallresolvedpostdatacollectionandreflectedanerrorin
therecordingsofthelesserexperiencedobserver.

Thesexoftheparticipantmatchedperfectlybetweenthetwoobserversforthetarget
personandtheircoeaters(n=51)
93
.Forthreecases,thesecondobserverrecorded
overweightwhiletheprimaryresearcherrecordedtheweightasnormal.The
definitionfortheweightstatuscategorieswasclarifiedtomakerecordingtheweight
statusmoreaccurate.SilhouetteimagesofdifferentsizedwomenusedbyHoldsworth,
Gartner,Landais,MaireandDelpeuch(2004)wereusedtoconstructaguidefor
recordingweightstatus
94
(seeAppendix8).


92
Atsomeoftherestaurantsoutsideofthecity,datawereonlyrecordedforthreepeopleorgroupswho
mettheinclusioncriteriaduringthetwohourobservationperiod.Forexample,atoneoftherestaurants
thatwaspiloted,theclienteleconsistedpredominantlyofyoungfamilies.
93
Datafor31dinerswererecorded.Nineatewithoneotherperson,fouratewithtwootherpeopleone
atewiththreeothers.Asage,genderandweightstatuswererecordedforcoeaters,therewere51cases
intotaltocompare.
94
As these were perceived weights, they were made to be slightly less strict than clinical assessments of
overweight and obesity such as Body Mass Index (BMI) or waist circumference. Therefore, recordings of
normal weight status may technically represent people who are just overweight (e.g., BMI of 25-26).
189
Thecategorythatcausedthemostdifficultlyforrecordingwasage.Inoverhalfofthe
caseswhereageswererecorded(n=26)therewasafiveyeardiscrepancyinthe
estimatedageofaparticipant.Forfourcases,thedifferenceinageestimationwas
between10and15years.Thecaseswiththelargeageinconsistenciesexclusively
involvedpeopleovertheageof45.Giventhelackofconsistencyintheobservationof
participantage,themeasureofagewascollapsedintofouragegroups(1525;2635;
3645;over45).Thesegroupsencompassedallofthedifferencesinestimations
betweenobserversandallowedforgreateraccuracyinrecordingtheageofparticipants.

Followingobservationsmadebytwoobservers,itwasclearthatmeasurescouldbe
recordedsuccessfullyandaccuratelybytheprimaryresearcher.

6.2.3 Procedure
Inclusioncriteria.Datawerecollectedonlyforpeopleassessedasbeingofsenior
secondaryschoolingageandover(people15andover).Thereweretwoprimary
reasonsforthisrestriction:(1)childrenrequireconsiderablylessfoodthanadultsand
couldthereforeconfoundthedependentvariable,extratimespenteatingand,(2)
people15andoverhavegreaterindependenceandabilitytomaketheirownfood
choices.Thisrestrictionmeantthatanygroupofpeoplewithamemberunderthe
nominatedagewasexcludedfromobservation.

Samplingstrategy.Thesamplingstrategywaseventcontingentratherthantimebased
(Haynes&O'Brien,2000).Observationbeganwhentheeatingeventstarted.All
participantswhoenteredtherestaurantduringanobservationsession,andsatatoneof
thetablesinthetargetedobservationarea,wereobservedandinformationabouttheir
190
behaviourrecorded(c.f.Sommer&Steele,1997)
95
.Themethodfordecidingwhowasto
beincludedintheassessmentofsocialinfluencewassimilartooneappliedbyBelland
Pliner(2003).Ifmembersleftorjoinedthegroupbeforeallmembershadleftthetable,
thegroupsizewasadjustedaccordingly.Thosewhostartedeatingalonebutfinishedin
agroupwereexcludedfromobservationaspotentialsocialinfluenceontheir
consumptionchangedduringthecourseofthemealevent.

Timeofdatacollection.Observationsweremadebetween4pmand7pmbecausethis
wasidentifiedasthebusiesttimeattherestaurantsduringpiloting.Thedurationof
observationsrangedbetweenoneandahalfhoursandthreehours.Datawererecorded
directlyintoaPersonalDigitalAssistant(PDA).Imagesofthepackagingoftheburgers
werestoredwithintheExcelfileasareferenceforobservations(anexampleoftheform
usedfordataentryandthereferenceguidecanbeseeninAppendix8).Twentyfour
uniqueobservationsessionswereconductedbetweenJuneandSeptember2007.
Sessionsweredistributedequallyacrossweekdaysonly.

Datarecorded.Theamounteatenwasrecordedforallindividualsormembersofagroup
whosatatchosentablesduringthespecifiedtimes.Forgroups,datawererecordedfor
onerandomlyselectedrepresentativeofthoseatthetable.Thisgrouprepresentative
waschosenaccordingtoapredeterminedseatatthetableusuallytheonemostvisible
fromthepositionoftheresearcher.Forasmallnumberofobservations,datawere
recordedforallmembersofthegrouptoallowdescriptiveanalysisoftheeffectsofthe
compositionofselectedgroupsontheamountconsumed(Hypothesis3).


95
Theobserverorderedandconsumedfooditemssoasnottodrawattentionwhileobservingtables.
191
InformationcategoriescodedwithinthePDAconsistedof:thesexofthenominated
individual;theirageestimatedtothenearestfiveyears;theirweightstatus(normal,
overweightorobese);thecharacteristicsofotherspresentinthegroup(includingtheir
sex,ageandweightstatus);thetimeeatingstarted(fromthemomentthepersonstarted
eatingattheirtable);thetimeeatingended(recordedwhenapersonleftthetable)
96
;
thepresenceorabsenceofreadingmaterialsortechnologyusedforsimilaractivities
(e.g.,laptops);andthefoodseaten.Itwasnotpossibletodeterminewhetherdiners
consumeddietornondietbeverages,consequently,characteristicsrecordedwere
confinedtothesizeofthedrinksorderedandwhethertheywerethickshakes,soft
drinksorhotbeverages.

6.2.4 Calculationofenergyintake
Theitemseatenwererecordedaccordingtotheirtradenames(e.g.,BigMac,small
fries).TheitemseatenwereusedtocalculatethetotalKJconsumedattheeating
occasionobserved.NutritioninformationwastakenfromtheMcDonaldswebsite
(McDonald'sCorporation,2006)andusedinthecalculationofenergyconsumption.
Consistentwiththelimitationdescribedabove,drinkswerenotincludedinthis
calculation.Theenergycontentofsomeitemswasdifficulttodeterminebecauseof
variationintheconstituentingredients(e.g.,DeliChoiceRolls).Fortheseitems
97
,KJs
consumedwasestimatedasthemeanforthedifferentvarietiesavailable.Atthetimeof
datacollection,thedoublecheeseburger,triplecheeseburgerandthebeefandbacon
burgerallcameinthesamepackaging.Therefore,fortheseitems,ameanKJamount
wasalsoused.

96
Somepeopleremainedatthetablewhiletheircompanyvisitedthetoiletpriortoleaving.Forthese
cases,thetimethattheeatingoccasionstoppedwascalculatedaccordingtothetimethepersonleftto
visitthetoilet.
97
TheseitemsconsistedofDeliChoicesRolls,salads(excludingthegreensalad),andMcFlurries.
192

Energyintakeforincompleteportionswasrecalculatedaccordingtotheestimated
proportionoftheitemeaten.Therewereonlyfivepeoplewhodidnotconsumethe
entiremealitemtheyordered,withmostoftheunfinisheditemsbeingdrinks.Inavery
smallpercentageofcases(2.7%),observedpeoplereturnedtothecounterduringtheir
eatingoccasiontoorderanotheritem.Iftheythentookthisitemwiththemastheyleft
therestaurant,theitemwasnotaddedtothecalculationoftheKJtheyhadconsumed
andthetimetheylefttablewasrecordedasthetimetheygotuptocompletetheir
secondorder.Timespentorderingwassubtractedfromtheoveralltimespentatthe
eatingoccasion.

6.2.5 Datascreening
Datawereoriginallyrecordedfor329individualsorgroups.Thirteencaseswere
removedbecausetheyconsumedonlyadrink.Incompletedatawererecordedfora
furthersixcases.Ineightcasestherewerechangestogroupsizeorbehavioursthat
couldnotbeadjustedfor.Thepresenceofoutlierswasassessedforthedependent
variablesextratimespenteatingandkilojoulesconsumed.Thisrevealedfourunivariate
outliersforextratimespenteating(i.e.,aresultthatwasmorethan3.29standard
deviationsfromthemean;Tabachnick&Fidell,1989).Thesecaseswereremovedfrom
theanalysis.Inallofthesecases,additionalnoteswererecordedwhichindicatedthat
thesepeoplewereengagedintimeconsumingactivitiesintherestaurant;studying,
doingtaxesandwaitingtoleavefortheairport
98
.Therewerenooutliersforkilojoules
consumed.Afterremovalofoutliers,theremainingsamplewas298.


98
Thesenoteswereallmadefromconversationsoverheardbytheobserver.
193
6.3 Preliminarystatistics
6.3.1 Descriptivedata
DescriptivedataforthesamplecanbeseeninTable19.Thefinalsamplewascomprised
of195malesand103females.Amajorityofthesamplewasinthe15to25yearsoldage
group(n=200).Almosthalfofthepeopleobservedatealone(n=141).Mostofthe
samplewasratedasnormalweight(n=254).

Table19:Descriptionofthesampleasfrequencydata(n=298).
Variable % Variable %
Sex(Male) 65.4 Weightstatus


Agegroup

Normalweight 85.3
1525 67.1 Overweight 12.4
3035 14.8 Obese 2.3
4045 8.4 Groupsize


Over45 9.7 Lonerdiner 47.3
Reading/usingtechnology 25.2 Pair 35.6

Group3ormore 17.1

Giventhesmallpercentofthesampleratedasobese,theoverweightandobese
categorieswerecombinedinallanalyses.

Onlyfivepeople(1.7%)ateingroupsoffourofmore.Consequently,datawas
codedintogroupof3ormore.

TheamountoftimerequiredtoeatfastfoodmealsderivedinaSection5.3.3was
comparedtothetimethatlonediners,whowerenotreading,spenteatinginthecurrent
observationalstudy.Themeanforthisgroupwas8.88minutesandthemodaltime
spenteatingwas8minutes.Thisprovidedgoodvalidationofthe8minutetimederived
todescribetheamountoftimerequiredtoeatfastfoodpriortothelargerFFSsurvey.
Subsequently,thedependentvariableextratimespenteatingwascalculatedby
subtractingthe8minutetimerequiredtoeatafastfoodmealfromthedifference
betweenthetimeeatingstartedandthetimeitceased.

194
Themeanfortheextratimespenteatingwas6.91minutes(SD=7.15)
99
.Themean
energyconsumedwas3399(SD=1291).Thesmallestorderseatenwerethree
McNuggets(582KJ)andasoftservecone(613KJ).Thetwolargestordersconsistedofa
BigMac,aDoubleQuarterPounder,largefries(7500KJ)andaDoubleQuarterPounder,
Cheeseburger,largefries,andthreeMcNuggets(7260KJ).

6.3.2 Covariates
Analysesofvariance(ANOVA)wereperformedtoidentifypossibleconfoundsforthe
outcomemeasures(extratimespenteatingandenergyconsumed).Theresultingmeans
andstandarddeviationscanbeseeninTable20.

Table20:Comparisonofenergyconsumed(KJ)andtimespenteating
betweendifferentlevelsofthevariablesmeasured
*
.
Measure KJintake TimeSpentEating

n M SD M SD
Weightstatus


Normal 254 3348
a
1255 6.49
a
7.13
Overweight 44 3692
a
1463 9.34
b
6.86
AgeGroup

1525 200 3391
a
1305 6.41
a
6.76
2635 44 3965
b
1186 8.14
a
7.99
3645 25 3141
ab
1045 7.56
a
8.03
Over45 29 2814
ac
1246 7.97
a
7.71
Partofday

16001730 124 3189
b
1266 6.15
a
7.09
17311900 174 3548
a
1292 7.45
a
7.17
Dayofweek

Monday 39 3524
a
1400 8.08
a
9.16
Tuesday 55 3214
a
1151 6.33
a
6.79
Wednesday 89 3558
a
1416 6.70
a
6.84
Thursday 68 3500
a
1281 8.06
a
7.65
Friday 47 3061
a
1061 5.36
a
5.13
*
Numberswithsharedsuperscriptsforeachmeasureidentifyvaluesthat
arenotsignificantlydifferentfromeachotheratthe.05level.

Overweightandobesecategorieswerecombined.


99
Theunadjusted,meantimespenteatingwasclosetothosereportedintheFastFoodSurveysindicating
thatgeneralrecallofthetimespenteatingmayhavebeenreliable.
195
Therewerenosignificantdifferencesinenergy(KJ)intakebyweightstatus.Timespent
eatingdidvaryaccordingtodifferentweightstatus,t(296)=2.46,p<.05,with
overweightpeoplespendingsignificantlymoretimeeating.Therewasasignificant
effectforagegrouponKJsconsumed,F(3,294)=5.36,p=.001,withBonferroniposthoc
testingindicatingthatthoseinthe26to35yearoldagebracketatesignificantlymore
thanboththe15to25andover45consumers.Extratimespenteating(herein
referredtoastimespenteating)didnotvarysignificantlybetweenagegroups.

Asobservationsoccurredbetween4and7pm,itwassuspectedthatthistimeperiod
wasinclusiveoftwoseparateeatingoccasions:afternoonteaanddinner.Itwas
expectedthatthepartofdayinwhichtheitemswereeatenmightimpactuponthe
amounteatenandpossiblyuponthetimespenteatingbecausedinnerisgenerallya
bigger,moresociablemealinAustralia.Consistentwiththispossibility,attestrevealed
thatKJsconsumedvariedsignificantlyaccordingtotimeofday(t(296)=2.38,p<.05).
Althoughtheextratimespenteatingwasslightlyhigherinthelaterpartoftheday,this
differencefailedtoreachsignificance.Dayoftheweekdidnotsignificantlyinfluence
eitherenergyconsumedorextratimespenteating.

Giventhesignificantinfluenceofthetimeofdayandageonenergyintake,thesewere
usedascovariatesinfutureanalysesofKJintake.Likewise,weightstatuswasusedasa
covariateinanalysesoftimespenteating.Priortohypothesistesting,allindependent
variableswerescreenedtoensuretheysatisfiedtheassumptionsofeachoftheanalyses.

196
6.4 Hypothesistesting
6.4.1 Analysisoftheeffectsofthepresenceofothers
Hypothesis1.1.AnANCOVA(analysisofcovariance)indicatedasignificantdifferencein
amountoftimespenteatingbetweenlonedinersandthoseeatingwithothers,
F(1,296)=4.98,p<.05.Thoseeatingwithothersateforsignificantlylonger(M=7.71,
SD=6.86)relativetothoseeatingalone(M=6.01,SD=7.38).

Hypothesis1.2.1.WithoutthesamedescriptiveinformationthatwasobtainedintheFFS,
itwasimpossibletoreplicatethemodelsoffastfoodconsumptiondevelopedinthe
previouschapter.Inanefforttoexplorewhetherthereweresimilartrendstothose
observedinthepreviousstudy,bivariatecorrelationswerecalculatedbetweenthetime
spenteatingandtheamounteatenandthenumberofpeoplepresent.Therewasa
significantpositiverelationshipbetweentheamounteatenandthetimespenteating,
r(298)=.20,p<.001.Therelationshipbetweenthetimespenteatingandthenumberof
peoplepresentwasalsosignificant,r(298)=.15,p<.01.

Hypothesis1.3.Toassesstherelationshipbetweenthenumberofpeoplepresentatthe
eatingoccasionandtheenergyconsumedandhowthiswouldvarybetweensexes,the
samplewassplitaccordingtosex.Thetotaleffectbetweenthenumberofotherspresent
andtheamounteatenfailedtoreachsignificanceinmales,r(195)=.10,p>.05.The
associationbetweenthenumberofotherspresentandtheamounteatenwasalso
negligibleinfemales,r(103)=.05,p>.05butinapositivedirection.

197
6.4.2 Lonedinersandreading
Restrictinganalysestothoseeatingaloneresultedinatotalof141participants(109
male,32female)fromtheoriginalsample
100
.Approximately51.1%ofthelonediners
werereadingwhileeatinginthefastfoodrestaurant.Thelonedinerswereaslightly
heterogeneoussubset.Thereweremorepeoplethanexpectedintheover45group
whowerereadingandlessfromthe15to25agegroupsthatwereobservedreading,

2
(3,158)=17.47,p<.001.Therewerenodifferencesbetweenlonereadersandnon
readersforsex(
2
(1,140)=2.17,p>.05),weightstatus(
2
(1,140)=2.82,p>.05)orpart
ofthedaytheitemswereconsumed(
2
(1,140)=.14,p>.05).

Hypothesis2.1.AnANCOVAwasusedtocomparethetimespenteatingbylonediners,
withandwithoutreadingmaterial,controllingfortheirweightstatus.Lonedinerswho
read(M=9.44,SD=8.12)spentsignificantlylongereatingthanthosewhodidnotread
(M=2.43,SD=4.24),F(1,139)=37.03,p<.001.

Hypothesis2.2.Analysisoftherelationshipbetweenthetimespenteatingandthe
amounteatenwasperformedforalllonediners
101
.Thisassociationfailedtoreach
significance,r(141)=.14,p>.05.


100
Lonedinersdidshowslightlyheterogeneousqualitiescomparedtogroupsofdiners.Peopleinthe15
25agegrouphadahigherpercentageofpeoplediningingroups,
2
(3,295)=40.17,p<.001andmore
malesthanwouldbeexpecteddinedalone,
2
(1,297)=16.67,p<.001.Therewerenodifferencesbetween
weightstatus
2
(1,297)=1.08,p>.051andthepartofthedaytheitemswereeaten,
2
(1,
297)=.16,p>.05.
101
Astheassociationbetweenthetimespenteatingtheamounteatenwasconsistentlypositiveinall
analysesforbothmalesandfemales,therewasnoneedtoseparatetheanalysesbygender.
198
6.4.3 Intragroupmatchingofintake
Previousstudiessuggestthatpeoplemayengageinmatchingwhenchoosingwhatand
howmuchtoeatinasocialsituation(Leoneetal.,2007;Rothetal.,2001;Salvyetal.,
2007).Matchingenergyconsumptionisadifficultbehaviour;peoplearenotnecessarily
awareoftheenergycontentofvariousmealconstituents(seediscussionofportionsize
datainAppendix7).Itismorefeasibleforfastfoodconsumerstooperationalise
matchingbehaviourbyorderinglikeproducts(e.g.,aburger).Totestthisproposition,
themenuitemsthatwereconsumedduringtheobservationswerecodedaccordingto
thetypeoffoodordered.Elevenfooditemcategorieswerecreated:small,mediumor
largefriesordrink,burger,healthychoiceitem,coffee,nuggetsanddessertitems.The
aimofthiscodingwastoassesswhetherpeopleeatingtogetherorderedthesametypes
offood.

Hypothesis3.Theordersofallthepeoplepresentattheeatingoccasionwererecorded
for38groups:18mixedsexdyads,7samesexdyads,4mixedsexgroups
102
,and8
samesexgroups.Matchingofintakewasrecordedwhenthesamecategoriesofitems
(ofthe11codedcategories)wereconsumedbybothoralleaters.Thenumberofgroups
thatshowedevidenceofmatchingisdisplayedinTable21.

Table21:Frequencyofmatchingordersforgroups
ofdifferentsizesanddifferentsexcomposition.
Groupdescription
n
Matched
orders
MixedSexDyad 18 9
MixedSexGroup 4 0
SameSexDyad 8 6
SameSexGroup 7 2


102
Groups had three or members at the eating occasion.
199
Inmixedsexdyads,ahighlevelofmatchingoccurred;malesandfemalestendedto
orderthesametypesandcombinationsofitems.Ineightoftheninegroupswhere
matchingwasnotevident,themalesatemoreitemsthanthefemales.Inthelastgroup,
themembersordereddifferenttypesofitems(amealversusadessert).

Thesamesexdyadsobservedalsoshowedevidenceofmatching.Thismatchingwas
clearerinsamesexgroupsoffemalescomparedtomales.Allofthefourdyadsof
femalesmatchedtheirordersidenticallywhileonlyoneofthethreedyadsofmales
orderedthesameitems.Thereseemedtobenoconsistencyinthematchingbehaviour
ofgroupsofsameormixedsex.Onlyonegroupofmalesandonegroupoffemales
matchedtheirorderswithinthegroup(bothweregroupsofthree).Inallothergroups
therewasnoconsistencyinmatchingand,asgroupsgotlarger,anytrendsweremuch
lessapparent.

Alogisticregressionwasperformedtoassesswhetherthesizeandthegender
compositionofthegrouppredictedordermatching.HosmerandLemehowsTest
revealedthatthemodelhadanappropriatefit(p>.10).CoxandSnellsR
2
indicatedthat
thetwopredictorsaccountedforapproximately19.3%ofthetotalvarianceinmatching.
TheWaldstatisticindicatedthatthesexcompositionofthegroupdidnotsignificantly
predictordermatching(Wald=2.39,p>.10).Thesizeofthegroupdidpredictmatching
(Wald=5.49,p<.05).Theoddsratiowas.25(95%ConfidenceIntervals:upper=1.45,
lower=.04)
103
indicatingthateatinginagroupdecreasedtheprobabilitythatmatching
occurredduringfastfoodconsumption.Thesmallsamplesizelimitedassessmentof
interactioneffectsinthemodel.

103
Theconfidenceintervalsindicatealargerangeofresponsesasaresultofthesmallsamplesize.
200
6.4.4 Minimaleatingnorms
Hypothesis4.A2X2X2ANCOVAwasperformedtotestfordifferencesinenergy
consumption.Theindependentvariableswere:participantsex(male/female),group
composition(samesex/mixedsex)andgroupsize(pair/largergroup).Ageandthepart
ofthedayinwhichtheobservationsoccurredwereusedascovariates.

Resultsindicatedamaineffectforparticipantsex,F(1,156)=14.51,p<.001,signifying
thatmalesconsumedsignificantlymore(M=3561,SD=1302)thanfemalesinthe
sample(M=3092,SD=1219).Nomaineffectswereobservedforthesexcompositionof
thegroup(F(1,156)=.02,p>.05)orthegroupsize(F(1,156)=2.96,p>.05).Therewere
twosignificantinteractioneffects.Thefirstwasatwowayinteractionbetween
participantsexandgroupsize(F(3,153)=3.99,p<.01).Thesecondwasathreeway
interactionbetweensex,groupcompositionandgroupsize(F(7,147)=2.87,p<.01).
PosthoctestingoftheseeffectswasdoneusingonewayANOVA.Themeansand
standarddeviationsforthekilojoulesconsumedforinthevariouscomparisongroups
aredescribedinTable22(onthefollowingpage).

Posthoctestingofthesignificantparticipantsexbysexcompositioncomparisons
indicatedthatmalesineithermixedorsamesexgroupsatesignificantlymorethan
femalesinmixedsexgroups.Thesignificantthreewayinteractioneffectalsoindicated
thatmalesinmixedsexgroupsconsumedthemostkilojoules,whilefemalesinmixed
sexgroupsatethelowestamountofkilojoules.Bonferronicomparisonsrevealedthat
malesinmixedsexgroups(threeormoremembers)consumedsignificantlymore
energythanfemalesinmixedcompanyeatinginadyadoragroup.

201
Table22:MeansandstandarddeviationsforKJconsumptionfromfastfood
itemsforbothtwoway(sexbysexcomposition)andthreewayinteraction
effects(sexbysexcompositionbygroupsize)
*
.
Sex/Sexcomposition n M SD
Female/Same 32 3275
ab
1264
Female/Mixed 39 2820
a
957
Male/Same 44 3551
b
1173
Male/Mixed 42 3692
bc
1431
Sex/Sexcomposition/Groupsize n M SD
Male/Mixed/3ormore 8 4555
a
1575
Male/Mixed/Pair 34 3489
ab
1340
Male/Same/3ormore 17 3721
ab
753
Male/Same/Pair 27 3443
ab
1377
Female/Mixed/3ormore 14 2672
b
805
Female/Mixed/Pair 25 2903
bc
1039
Female/Same/3ormore 12 3651
ab
1593
Female/Same/Pair 20 3050
ab
998
*
Numberswithsharedsuperscriptsidentifyvaluesthatarenotsignificantly
differentfromeachotheratthe.05level.

ThethreewayinteractionisillustratedinFigure17 (on the following page).Inthisfigure


itisclearthatasgroupsizeincreases,sodoestheamounteatenforallgroups,withthe
notedexceptionoffemalesinmixedsexgroups,wheretheamountseemstodecrease,
whichisconsistentwiththeminimaleatingnorm.
202


6.5 Discussion
Thisstudytestedseveralhypothesessurroundingfastfoodconsumption.Thesewere
concernedwithexploringtheextenttowhichtimeextension,socialfacilitation,gender
rolesandimpressionmanagementmightimpactontheenergyconsumedatafastfood
eatingoccasion.Datawerecollectedbyobservationofconsumersatanumberof
McDonaldsrestaurants.Statisticalanalysesrevealedvaryinglevelsofsupportforthe
hypotheses.

Figure17: Interactioneffectsforsubjectsex,groupcompositionandgroupsizeonKJ
intakefromfastfoods.
203
6.5.1 Timeextensioninafastfoodrestaurant
Thetimeextensionhypothesispredictsthatpeoplewhoeatingroupsconsumemore
andeatforlonger(e.g.,J.M.deCastro,1990).BellandPliner(2003)reportedthatthe
timespenteatingfoodcouldbeincreasedthroughthepresenceofothers.Thecurrent
resultsshowedthatthepeopleeatingfastfoodingroupsateforlongerthanlonediners,
whichsupportsthisliterature.Moreover,therewerepositivecorrelationsbetweenthe
timespenteatingand,boththeamounteatenandthenumberofpeoplepresentatan
eatingoccasion.Bothoftheserelationshipswereweakbutsignificant.

Thefailuretoobserveadirectassociationbetweentheamounteatenandotherspresent
isconsistentwithresultsfromthepreviousFFSstudy(reportedinChapterFive).
Withoutusingmodellingtechniques,itisunclearhowdirectandindirecteffects
combinedtocreatethetotaleffect.

6.5.2 Matchingnormsandtheirinfluenceonfastfoodconsumption
Overall,thesexcompositionofthegroupfailedtopredictordermatching.Partofthe
reasonforthismaybethelimitedsamplesizewhichrestrictedtheabilitytoassess
interactioneffects(betweenparticipantsexandgroupcomposition,forexample).The
descriptivedataindicatedthatinalmostallofthegroupsoffemalesobserved,therewas
clearevidenceofmatching.Themotivationforthiscouldnotbedetermined.Leone,
HermanandPliners(2008)datasuggeststhatadesiretobeseenpositivelybyothers
mightunderliethisbehaviourandthatthisdesiremayservetooutweighpressureto
conformtothefeminineminimaleatingnorm.Fastfoodconsumptionbyawoman
eatingwithotherwomenobservedinthestudyreportedhereofferssomesupportfor
thissuggestion;womeneatingwithotherwomendemonstratedmatchingintakesmore
204
consistentlythanminimaleatingbehaviour.Theextenttowhichnormativematchingof
foodconsumptionisgenderspecific(see,Salvyetal.,2007)requiresfurtherexploration
withlargersamplesizes.

Pliner,Bell,HirschandKinchla(2006)suggestedthatasgroupsizeincreases,matching
normsbecomemoredifficulttodiscern.Theyarguedthatitismoredifficultto
determinetheamountothersareeatingwhentherearemorepeopletomonitor.People
eatinginpairsweremorelikelytomatchtheirordersthanpeopleinlargergroups.The
descriptivedataindicatedthisfindingwastrueinmixedsexpairsaswellassamesex
pairs.Therefore,theresultsreportedhereareconsistentwithPlineretal.scontention.

Overall,itisdifficulttodeterminewhatnormsunderliematching.Forexample,
whetherpeopleatethesametypeofmealsasaresultoftheotherpeoplepresentorasa
resultoftheeatingoccasionwasnotdiscoveredinthecurrentstudy.Giventhefactthat
McDonaldslabeltheirmealsassuch,itcouldonlybeexpectedthatpeoplevisitingthe
restaurantfordinnerwouldallorderameal.Determiningthemechanismbehind
matchingisanimportantstepforfurtherresearch.

Finally,theextenttowhichmatchingcouldbeconfoundedbythenatureorclosenessof
therelationshipbetweenthepeopleeatingneedstobeconsidered.Inthelargestgroup
observed,thereappearedtobethreegenerationsofpeoplepresentincludinganelderly
male/femalepair,amiddleagefemaleandtwopeopleintheirtwenties.Thesedifferent
relationshipsmaychangethenormssurroundingthesituation.

205
6.5.3 Evidenceofminimaleatinginfastfoodconsumption
Literatureidentifyingminimaleatinghassuggestedthat,inthepresenceofmen,women
decreasetheirintake(Chaiken&Pliner,1987;Morietal.,1987;Pliner&Chaiken,1990).
Therewasindicationofminimaleatinginthemodellingoffastfoodintakereportedin
thepreviouschapter.Specifically,thenumberofpeoplepresenthadadirect,negative
associationwiththeamounteateninthefemalesobserved.Giventhisevidence,itwas
expectedthat,wheninthepresenceofamalecoeater,womenwouldeatless.The
resultsofthestudysupportedthishypothesisinanaturalisticenvironment.

Eatingwithmixedversussamesexgroupsmayhaveinfluencedtheintakeofbothmales
andfemalesobserved.Theeffectofthesexcompositionofthegroupwasonly
significantininteractionwiththeeaterssex.Thereasonforthiswasmostlikelythefact
thatthemalesandfemalesobservedbehavedoppositelyinthepresenceofmixed
company;malesinthepresenceofmixedcompanyatethemostwhilefemalesinmixed
companyatetheleast.Thisfindingsupportstheexistenceofandadherencetominimal
eatingnorms.

Despitethefactthatfemalesatelessthanmalesoverall,insamesexgroupsmalesand
femalesatesimilaramounts.Thisindicatesthatitmaybethepresenceofmalesthat
reducedfemalesintakeandnotsimplythedifferentenergyrequirementsofmalesand
femalesthatdeterminedthisdifference.Minimaleatingintheobservationsinthefast
foodenvironmentmayhavebeenmotivatedbythedesiretomeetgenderrole
expectationsratherthangeneralselfpresentationalconcerns.Ifwomenwererestricting
theireatingpurelyinthedesiretolookgood,thenitwouldbeexpectedthatwomenin
allgroupswouldlimittheirintake.Yet,femalesobservedheredidnoteatdifferent
206
amountstomaleswheneatinginsamesexgroups,whichindicatesthatperhapseating
insamesexgroupsliberateswomenfromgenderroleexpectations.

Someauthorshavesuggestedthateatinglargeramountsoffoodisperceivedasa
masculinebehaviour(Bock&Kanarek,1995).Paradoxically,othershaveshownthat,
wheninfemalecompany,malescanalsoeatless(Morietal.,1987).Inthisstudy,males
inthepresenceoffemalesatethemostfood.Theamountstheyateweredetermined
largelyattheorderingstage(fewmadesecondordersoratetheireatingcompanions
leftovers).Thisresultreinforcestheideathatmalesmayhavebeendrivenbyadesireto
appearmasculinewheninthepresenceoffemalecompany.Yet,malesintakedidnot
varymuchbetweensamesexandmixedsexgroups,whichcouldindicatethattheir
behaviourwasnotdrivenbyexternalnorms.Salvyetal.(2007)andEagly(1983)
suggestedthatmenarelesslikelytoadheretonormsbecausetheyhaveagreaterdesire
toappearindependent.Furthermore,Salvyetal.(2007)foundthatmeninthepresence
ofothermalefriendsatemore.Itislikelythatmenwerenotactinginawaytoincrease
theirintakeinordertoappearmasculine.Withoutassessingthesebehavioursinan
experimentaldesign,itremainsunclearwhyandhowmenchangetheirintake.
Unfortunately,therearefewstudiesinvestigatinghowdifferentenvironmentsinfluence
eatingbehavioursinmenorgroupsofpeople,thereforethereislimitedprecedent.

Thethreewayinteractionobserveddemonstratedthatthedifferencesinintake
betweenmalesandfemalesinmixedsexgroupsvariedwithgroupsize.Thisfinding
supportsresultsfromKlesgesetal.s(1984)study,whichshowedthatfemaleeatersina
fastfoodenvironmentinmixedsexgroupsatelessasgroupsizeincreased.Inthat
study,however,itwasfoundthatfemalesinsmallmixedsexgroupsateasimilar
207
amounttomalesinmixedsexgroups.Inthecurrentstudy,femalesinmixedsexpairs
atesignificantlylessthanmalesinmixedsexgroups,butnotmalesinmixedsexpairs.

Themodelofsocialfacilitationsuggeststhatasthenumberofpeoplepresentincreases,
sodoestheamounteaten.Thethreewayinteractioninthecurrentdataindicatedthat
increasesingroupsizemayalsoemphasiseminimaleatingnormsinfemaleeatersand,
moreindirectly,thedesiretoeatmoreinmen(maleswithtwoormorepeopleatethe
mostwhilewomeninlargergroupsatetheleast).Itislikelythatwhenmorepeopleare
present,womensdesireforimpressionmanagementincreases.Otherauthorshave
suggestedthatwomenaremoreconcernedwiththeirappearanceandalsomore
interestedinfood(Klaczynski,Goold&Mudry,2004;Pliner,Chaiken&Flett,1990),and
itmaybethecasethatwomenaremotivatedtousethewaytheyeattocontrolthe
impressiontheycreateortodefineorreinforcetheirselfidentity.

6.5.4 Lonedinersandreading
Intheirobservationsofdinersatcafes,SommerandSteele(1997)reportedthatpeople
frequentlyreadwhendiningalone.Theysuggestedthatpeoplediningaloneused
readingmaterialstoreducethediscomfortassociatedwitheatingaloneinpublic;itis
clearfromtheresultsofotherstudiesthateatingaloneisstigmatised(Redd&deCastro,
1992).

Onaverage,peopleeatingfastfoodalonespentalmostthreetimesasmuchtimeatthe
tableiftheyread,supportingthesuggestionthatreadingcanbeusedtoreducethe
discomfortassociatedwithdiningalone.Consideringtheconsistentrelationship
betweenthetimespenteatingandtheamounteaten,itispossiblethatextendingthe
208
timespenteatingalsoincreasestheamounteatenbylonediners.Intheobservational
study,timespenteatingwhilereadingdidnotrelatetotheamounteaten.

Otherresearchhasconfirmedthatincreasedtimespenteatingdoesnotinevitablylead
toincreasedamountsofintake(e.g.,Mathey,Zandstra,deGraaf&vanStaveren,2000).
Thediscrepancyinfindingsofthedataforloneconsumersascomparedtogroupsof
consumerssuggeststhatincreasedconsumptioningroupsituationscannotbeexplained
solelybyincreasedtimeintherestaurant.Therewasevidencethatlonedinerscould
prolongthetimetheyspendeatingwithoutincreasingtheamounttheyconsume.
Althoughtheresultsreportedhereindicatethattherelationshipbetweenthenumberof
peoplepresentandtheamounteatenmayoperateindirectlythroughthetimespent
eating,thesocialnatureofrelationshipmaybecrucialtoincreasedintakeasthisis
absentforlonediners.Withoutexternalinfluencesanddirect,socialnorms,foodchoice
inalonedinermaysimplybedrivenbypersonalpreferenceorhabit.

6.5.5 Conclusions
Thisstudyshowsthatthepresenceofotherscanhaveasignificantinfluenceoneating
behavioursinafastfoodrestaurant.Theinfluenceofthepresenceofotherswas
demonstratedthroughevidenceofminimaleatingnormsaswellasthroughfurther
supportforthetimeextensionhypothesis.Itisinterestingthat,incontrasttosocial
facilitationtheory(whichsuggeststhatthepresenceofothersincreasesintake),
minimaleatingnormsoperatesothatthepresenceofotherscandecreaseintakein
certaincircumstances.Thesetwoinfluencesdirectlyconflictwitheachother,andyet,
mayexisttogetherinaneatingenvironment.Althoughfuturestudieswillneedto
209
replicatethisobservation,itemphasisestheimportanceofnotassumingthatinfluences
onconsumptioninmenandwomenareuniform.

Itisdifficulttodeterminehowexactlyminimaleatingnormsoperateinthefastfood
environment.Partofthereasonforthisisthecomplexnatureofimpression
managementtheory.Authorshavenotedthattheprocessofimpressionmanagementis
dynamicandchangesdependingonthetarget,theaudienceandthesituation(Bozeman
&Kacmar,1997).Twopeopleusingsimilarimpressionmanagementstrategiescanbe
judgeddifferentlydependingontheirpredeterminedroleinsociety(Rudman,1998;
Westphal&Stern,2007).Thefamiliarityhypothesissuggeststhatthereisalsolikelyto
besomeinfluenceofthetypeofrelationshipbetweenpeopleandhowthismoderates
impressionmanagement(Learyetal.,1994;Ticeetal.,1995).Themotivationsbehind
minimaleatingcannotbeconclusivelydeterminedthroughobservationalmethods.

Eatingfastfoodmayemphasisethedesiretoconformtosocialnormsbecausefastfood
consumptioncanbeassociatedwithnegativejudgements(Stein&Nemeroff,1995).de
Castro(1994)foundthatwomeneatingwithmenmoreclearlydemonstratedsocial
facilitation,whereasthedatafromthecurrentstudysuggestthatwomeneatingwith
menrestricttheirintake.ThefactthatdeCastroassessedgeneraleatingwhilethe
currentresearchfocusesonlyonfastfoodsmayexplainthisdifference;fastfoodsmay
triggerminimaleatingnorms.Unfortunately,withoutobservingpeopleeatingbothfast
andnonfastfood,thereisnopointofcomparisontoallowsuchconclusiontobedrawn.

Theresultsofthisstudysuggestthatfastfoodeatersaredrivenbysimilarsocial
influencesasthosereportedforotherformsofeating.Thereasonthatthenorms
210
surroundingconsumptionmaybesopervasiveandimpactonfastfoodconsumption,as
wellasregulareating,mayoriginatefromtheuncertaintyindividualsexperiencewhen
tryingtodetermineappropriateeatingbehaviourssuchastheamounttoeat(Herman,
Fitzgerald,&Polivy,2003).Compliancewithnormshelpstoeliminatebehavioural
uncertainty.Bycontrast,normviolationcanresultinpooroutcomesforanoncompliant
individual(Herman&Polivy,2007;Ohbuchietal.,2004).Forexample,inaprofessional
context,womenwhofailtomeetsexroleexpectationscanreceivenegativejudgment
ratings(Rudman,1998).Furthermore,peoplewhosatisfynormativeexpectations
experiencepositivefeelings(W.Wood,Christensen,Hebl,&Rothgerber,1997).

Insummary,thecurrentobservationalstudyexploredfastfoodconsumersinanatural
environmentandovercameseverallimitationsofsurveymethods
104
.Keyamongtheseis
thevalidityofboththedependentandindependentvariables.Itishighlylikelythat
socialdesirabilitycontaminatesselfreportdataonfastfoodconsumptionanditisalso
conceivablethatpeopleareunawareofhowtheirownbehaviourisinfluencedbythe
presenceofothers,therebycreatingproblemsformeasurement(Vartanian,Herman&
Wansink,2008).Althoughobservationalmethodsprovideecologicallyvalidsnapshot
data,theyarelimitedbythescopeoftheinformationthatcanberecorded.Theissueof
thegeneralisabilityofobservationaldatatoothercontexts,whetherthesebedefinedby

104
RecalloftimespenteatingwasidentifiedasanissueforthepreviousstudiesusingtheFastFood
Surveydata.Observationalmethodsovercametheseproblemsastimeandintakecouldbehighly
monitored.Itwasreassuringthattheaveragesofthetimespenteatingandtheamountseatenwerefairly
consistentacrossmethods.Previoussocialinfluencestudieshavebeencriticisedforcomparinggroupsof
dinersunlikelytoexistnaturally,forexample,groupsofstrangersdiningtogether.Thefactthatthe
groupsinthisstudywerenaturallyformedrepresentsastrengthofthecurrentstudy.Itwasinterestingto
observethatcertaincombinationsofgroupsseemedtobelesscommonamongstfastfooddiners.Large
groupswereveryuncommoninthefastfoodenvironment.Furthermore,itwasinterestingtofindthat,in
thisobservationalstudyandincontrasttothepreviousFastFoodSurveys,moremaleswereobserved
eatingfastfoodwhilstmorefemalesrespondedtothesurveys.Thissuggeststhatthesurveydatamaynot
accuratelyrepresentthepopulationoffastfoodconsumers,butdoesshowwomenshigherinterestin
healthrelatedresearch.
211
franchise,timeofday,geographiclocation,orfoodtype,isanissuethatremainstobe
addressed.

Giventheimportanceofculturalnormsandhowthesemaymodifyintake(Devine,
Sobal,Bisogni,&Connors,1999),itisimportantthatfutureresearchexploreswhether
culturesoutsideofAustralia,demonstratesimilarpatternsofbehaviour.Inorderto
comparedifferentculturesandwhethertheywereinfluencedbysimilarnormstothe
Australiansampleasmall,crossculturalobservationalstudywascompleted.

6.6 ACrossCulturalComparisonofFastFoodBehavioursinaNaturalistic
Environment

Resultsreportedintheprecedingchaptersindicatethatenergyconsumedatafastfood
mealisinfluencedbythetimespenteatingandthatincreasinggroupsizecorrelates
indirectlywithincreasingintake(i.e.,timeextension).Inadditionthesestudies
indicatedthatfemaleconsumersareinfluencedbytheminimaleatingnorms.Theextent
towhichthisresultisgeneralisablebeyondMcDonaldsrestaurantsinSouthAustralia
hasyettobeconfirmed.Dietarybehaviourispotentiallyabehaviourlikelytobe
amenabletonormativeinfluences(Herman,Rothetal.,2003)anddisplaycultural
variation(Devineetal.,1999;Marshall,1995).Thereisagrowingbodyofresearch
exploringthesecultural,normativedifferences.Comparisonsbetweenthefood
consumptionbehavioursofAmericansandEuropeansareamongthemostcommon.de
Castroetal.(1997)founddifferencesinthetimespenteatingandtheamountseaten
betweenDutch,FrenchandNorthAmericanpeople.Giventheimportantinfluenceof
socialcontextontimespentandamountconsumedinthecurrentstudy,these
differencesbetweencultureintypicaleatingpatternsmaylimitthegeneralisabilityof
212
theresultsreportedinthepreviousstudy.Cultureandethnicidentityhelpdefinethe
normsofbehaviour,includingdietarychoices.SiwikandSenf(2006)reportedthat
ethnicitypredictedadifferentialtendencytoeatoutamongAmericansofdisparate
ethnicbackgrounds.Inaseriesofinterviews,Devine,Sobal,BisogniandConnors(1999)
exploredhowethnicityshapedfoodchoice.Theyconcludedthatdifferentcultureshave
varyingnormsthatdeterminethewayindividualsviewfoodandeating.Foodwas
describedasamechanismforconveyingculturalidentitywithbothpreparationand
consumptionoffoodtelegraphingvalues.Thus,interpretationofthelinkbetweenfood
andethnicityorculturalidentityisconsistentwithargumentsintheimpression
managementliterature;theneedtofulfilculturallyacceptablerolesisplayedoutinthe
foodconsumptionsetting.

Itisunclearwhethertherigidformulathatdefinesfastfoodproductionand
disseminationoverridesculturalconstraintsoneatingbehaviours.Thereisevidence
thatsuchisnotthecase;Rozin,Kabnick,Pete,FischlerandShields(2003)notedthatthe
portionsizesfromMcDonaldsinFranceweresmallerthanintheUS,andyet,theFrench
alsospentlongereating.Thissuggeststhatuniformityinthebrandmaynotoverride
normativeculturalinfluences.

Theaimofthisstudywastocomparethefastfoodeatingbehavioursobservedinthe
AustraliansampletoeatersinanenvironmentoutsideofAustralia.Inordertoassess
eatingbehavioursinaculturedifferenttoAustralia,McDonaldsrestaurantsincentral
Oslowereobserved.OsloisthelargestcityinNorway.Ithasapopulationsimilarinsize
213
tothatofAdelaide
105
.UnlikethepopulationofAdelaide,thepopulationofOslois
estimatedtobeamongthefastestgrowingintheworld(StatisticsNorway,2009a).

NorwegianandAustralianeatingpatternshavesomesimilarities.InAustralia,people
eattheirmainmealsintheevening.ThisisalsotrueofpeopleinNorwaywhoeata
large,hotmealormiddagintheevening.TherearealsoaspectsofeatinginNorway
thatdifferwhencomparedtotypicalAustralianeatingbehaviours.Thefirstisthe
seasonalvariationineatingpatternsinNorway.Thisispartlyaproductofthemore
markedseasonalvariationsobservedinNorwaycomparedtoAustralia(Perry,Silvera,
Rosenvinge,Neilands,&Holte,2001).Theseconddifferenceisbetweentraditionalfoods
withineachculture.Regardlessofthesedifferences,bothcultureshaveembracedthe
fastfoodculture.ThepopularityofAmericanfastfoodchainsisblossominginNorway
(EuromonitorInternational,2008).Furthermore,Norwegianshavereducedthetime
theyspendpreparingfoodby30minutesoverthepast20years(StatisticsNorway,
2009b).Thissuggeststhattheremayhavebeenaculturalshiftinattitudestofoodand
thatNorwegiansareasvulnerabletotimescarcityandthedesireforconvenienceas
Australians(seeSection1.3.2fordiscussionofprevalenceoffastfoodconsumptionin
Australia).

Thisstudywasexploratoryandnohypothesesweremaderegardingthenatureofany
differencesbetweenthebehavioursobservedinAustralianandNorwegiancultures.The
earlierworkofdeCastro(1997)suggestedthatdifferencesmightbeobservedon

105
ThepopulationofOslowas1,283,533in2004(StatisticsNorway,2009a).Adelaidespopulationwas
1,158,259in2007(AustralianBureauofStatistics,2007).
214
measuresoftimespenteatingandquantityconsumed,howeverthisworkdidnot
includeaNorwegiansample.

Therefore,thecurrentanalysesaimedtoassessthefollowingquestions:
1. Istheamounteatenand/orthetimespenteatingdifferentbetweenAustralian
andNorwegianMcDonaldsconsumers?
2. Aretherelationshipsbetweenthenumberofpeoplepresentattheeating
occasion,thetimespenteatingandtheamounteatensimilarbetweenthe
AustralianandNorwegiansamples?

6.6.1 AssessingeatingbehavioursinaNorwegiansample
106

Intotal,threerestaurantsinOslocitywerevisitedoverthreedays.Theserestaurants
weresimilartothoseobservedinAdelaide;theywerenearthecentreofthecityandnot
partofafoodcourt.Therestaurantswereobservedoverperiodsoftimecomparableto
thoseusedinthepreviousstudybetween5pmand8pm.Allobservationsoccurredin
lateJune,2007.

Theprocedureusedforthisstudywasidenticaltothatofthepreviousstudy(detailedin
Section6.2.3).TheresearchersatintheMcDonaldsrestaurantsandunobtrusively
observeddiners.Thetimeparticipantsbeganeating,thetimetheystopped,theirsex,
weightstatus,theitemstheyconsumedanddetailsofthecoeaterswererecorded
directlyintoaPDA.Additionaldescriptiveinformationregardinganybehavioural
differencesbetweenthediningbehavioursofpeopleintheNorwegianandAustralian

106
This study was an extension of the previous observational study and was approved by the ethics committee as
an amendment to the existing application.
215
restaurantswasalsocollected.Asinthepreviousstudy,thepersonsittinginthemost
visibleseatwasnominatedtobeincludedindatacollection.Thisindividualsintakewas
chosentorepresenttheintakeofthegroup.

Beforetheitemseatencouldbeconvertedintodietaryinformation,correctnutrition
detailswerecollectedfromtheMcDonaldswebsiteinNorway(McDonald's
Corporation,2007)
107
.ThisinformationwascollectedinearlyJuly.Energycontentwas
providedinKcalandconvertedintoKJbymultiplyingitby4.184.

6.6.2 Results
Descriptivedata.Datawererecordedfor50groupsorindividuals.Twoofthe50cases
weredeletedbeforeanalysisbecausetheparticipantconsumedonlyadrink.The
descriptionofeachsampleisdetailedinTable23.

Table23:Demographiccharacteristicsofthesamplesof
consumersobservedinOslo(n=48)andAdelaide(n=298).
Variable Oslo
*
Adelaide

Sex(Male) 58.3 65.3


Agegroup

1525 58.3 67.1
3035 25.0 14.1
4045 10.4 8.7
Over45 6.2 10.1
Weightstatus

Normal weight 91.7 85.5
Overweight 6.2 12.4
Obese 2.1 2.3
Groupsize

Lonerdiner 43.8 47.3
Pair 41.7 35.6
Group3ormore 14.6 17.1


107
The nutritional composition of the McDonalds items sold in Norway is slightly different and the correct
information is only available from the Norwegian website.
216
ThedemographicsofthesampleobtainedinOslowerecomparedtotheAustralian
sample.ChiSquaretestsrevealednodifferencesatthep<.05levelinthedemographic
characteristicsofthesamplesforsex,
2
(1,345)=0.91,p>.05;age,
2
(3,343)=4.40,
p>.05;weightstatus,
2
(2,344)=1.57,p>.05;orgroupsize,
2
(2,344)=0.69,p>.05.

6.6.3 DiningatMcDonaldsinNorwaycomparedtoAustralia
ServiceattheNorwegianrestaurantswasnotasfastasthatwhichwaswitnessedinthe
Australianrestaurants.Therestaurantsobservedwerealso,onaverage,busierthanthe
restaurantsobservedinAdelaide.ThetypicalprocessforeatingfastfoodinAustralia
involvedorder,sit,eatandleave;intheNorwegianrestaurants,patronsarrived,
selectedatable,satdown,orderedandreturnedtotheirtable(and,inmanycases,their
foodwasboughtouttothem).Thismeantthatthetimerecordedhadtobeobservednot
fromthetimetheparticipantsatattheirtable,butfromthetimethefoodarrived.None
oftheNorwegiansreturnedtothecountertoorderseconds
108
.Ifpeopleordereda
dessert,moretimesthannot,itwasbroughttothetableatthesametimeasthefood.

FastfoodeatersinAdelaideateanaverageof3399KJ(SD=1291),whereasthe
consumersinOsloateanaverageof3311KJ(SD=1387,range565to6171KJ).The
differencebetweentheseenergyintakeswasnotsignificant,t(344)=.43,p>.05.

Therewasnoopportunitytoassesstheaveragetimerequiredtoeatfastfooditemsin
Oslo(aswasdoneforAustraliansampleinSection5.3.3),thereforethetimespent
eatingwasusedwithoutanyadjustment.Onaverage,theNorwegiansamplespent20.77

108
IntheobservationalstudyinAdelaide,only9peopleofover330observed(priortoremovalof
outliers)returnedtothecounterforasecondorder.
217
minutes(SD=11.77)eating.ThemeanunadjustedtimespenteatingintheAustralian
samplewas14.39minutes(SD=7.75).Thedifferenceintimespenteatingwas
significant,t(53.75)=4.88,p<.01.

TherewaslittleevidenceoftimeextensioninthepeopleobservedintheNorwegianfast
foodrestaurant.Therelationshipbetweenthetimespenteatingandtheamounteaten
wasnegligible,r(48)=.04,p>.05.Therewasevidenceofapositiveassociationbetween
thenumberofpeoplepresentandtheamountoftimespenteating,r(48)=.38,p<.01.
Forwomen,therelationshipbetweenotherspresentandtheamounteatenwas
negligible,r(20)=.09,p>.05.Althoughthisassociationwasstrongerinmen,itfailedto
reachsignificance,r(28)=.17,p>.05.

TherewasasignificantdifferencebetweenthetotalKJintakeofmenoveralland
women,t(46)=2.52,p<.05.AsintheAustraliandata,women(M=2745,SD=1150)ate
lessthanmen(M=3715,SD=1419).Giventhesmallsamplesize,therewerenot
enoughpeopletorunthesamemultivariatemodelcomparingdifferentgroup
compositionsaswasperformedfortheAustraliansample(seeSection6.4.4);therewas
onlyonefemalewhowasobservedeatinginamixedsexgroup.Therefore,itwas
impossibletoassesspossibledifferencescreatedbypotentialminimaleatingnorms.

6.6.4 Discussionandconclusions
Therehasbeensomesuggestionthatculturaldifferencesexistineatingbehaviours(de
Castroetal.,2007)andthatthesemayextendtofastfood(Rozinetal.,2003).Thisstudy
setouttoexplorethedifferencesbetweenthebehavioursoffastfoodconsumersin
218
restaurantsinNorwayandAustralia.Therewerebothsimilaritiesanddifferences
betweenthebehavioursoffastfoodconsumersineachculture.

Inthecurrentstudy,therewerenodifferencesintheamountoffoodconsumedby
patronsinMcDonaldsinOsloandAdelaide,althoughtheNorwegianconsumersatefor
significantlylonger.Thisdifferencemaybeculturallydetermined.deCastroetal.(2007)
reportedthatDutchpeopleatemorefrequent,smallermeals,butalsoateforlonger
thanFrenchandNorthAmericanpeople.Rozinetal.(2003)reportedthatFrenchpeople
spentlongereatingMcDonaldsthanNorthAmericanpeople.Itmaythereforebeamore
typicallyEuropeanbehaviourtospendlongereating,regardlessoffoodtype.Theeating
styleofAustraliansmaybemoresimilartothatofNorthAmericans,butwithoutany
directcomparisons,thisassertionispurelyspeculative.

Therewasnodirectrelationshipbetweentheamounteatenandthenumberofpeople
presentformalesorfemalesindataforconsumersineithercountry.Theassociation
betweenthenumberofpeoplepresentandthetimespenteatingwassignificantinboth
populations.However,timespenteatingdidnotrelatetotheamounteateninthepeople
observedintheNorwegianfastfoodrestaurants.Thisisacrucialassociationinthetime
extensionhypothesis;withoutit,thetheoryremainsunsupported.Thefactthatthe
Norwegiansampleatelonger,andthateatinglongerwasnotassociatedwithgreater
intake,suggeststhatthetimeextensionhypothesisinfastfoodconsumersmaybe
culturallymediated.

Inbothcountries,theamountoffastfoodeatenbymenandwomenwassignificantly
different.Thesmallsamplesizerestrictedexplorationofminimaleatingnormsinthe
219
Oslosample.ThereisverylittleresearchonminimaleatingnormsoutsideofNorth
America.GivenotherdifferencesintheeatingbehavioursofEuropeansandNorth
Americans,itwouldbeinterestingtoexplorewhetherthesenormsareapplicablein
Europeanpopulations.Alargersamplewillneedtobeobservedforthistobeassessed.

Thisstudyshowsthattherearedifferencesbetweentheeatingbehavioursofconsumers
inNorwegianandAustralianMcDonaldsrestaurants.Thebroaddemographicprofileof
peopleobservedineachrestaurantwasnotdifferent,suggestingthatthesedifferences
maybeculturallydetermined.Thismeansthatalthoughthisresearchhaslargely
supportedthetimeextensioninAustraliansamples,theseresultsmaynotbe
generalisabletootherWesterncultures.

Futurestudiesareneededtofurtherexploretheculturalinvarianceofnormative
influencesincludingtimeextension,socialfacilitationandminimaleating.Additionally,
thesestudiesmightexploreotherpotentialnormativeinfluencesoneatingbehaviour
thatmightvaryintheirimportancebetweenculturalgroups.Futureresearchcould
assesswhethergroupswithmembersofmixedculturalbackgroundsretainthe
normativeinfluencesspecifictothecultureoftheconstituentgroupmembers.Thelevel
ofidentificationanindividualfeelswiththeirculturecanchangeadherencetonorms
(Jetten,Postmes,&McAuliffe,2002).

AsWesternisedfastfoodspenetratetheAsianmarkets(Park,2004),itwouldbe
interestingtoexplorethefactorsthatinfluencetypicalconsumptionbehaviourinfast
foodsettingsintheseAsiancountries.Thisgroupwouldalsobeaninterestingpointof
comparisongiventhatWesternstylefastfoodscanbemuchlesshealthythantradition
220
Asiantakeoutlunches(Kamei,Ki,Kawagoshi,&Kawai,2002).Differingcultural
ideologiesaffectmanysocialbehaviours,includingchildrearing,foodpreparationand
eatingbehaviours(Triandis,1989).Furthermore,thevalueswhichareplacedon
feminineidealsmaydifferbetweeneachWesternandEasternculture(Beardsley&
Pedersen,1980).Therefore,itispossiblethatthesecultureswillcreatedifferencesin
eatingpatterns.

Inconclusion,thecurrentdatasuggestsomeconsistentinfluencesinamountoffood
eaten.Forexample,energyconsumeddidnotvarybetweencountries.Incontrast,the
failuretoobservetimeextensionsuggeststhatsomesocialinfluencesmaybeculturally
dictated.
221


7 ChapterSeven:Asummaryofthefindingsandimplicationsofthecurrent
research



Traditionaldietsarebeingslowlyreplacedwithdietscontainingahighproportionof
foodspreparedawayfromthehome(Binkley,Eales&Jekanowski,2000;Kant&
Graubard,2004).Consistentwithpublichealthadvocates,Stanton(2006)hasargued
thatconsumptionoftheseconveniencefoodsisassociatedwithincreasingenergy
consumption,particularlythroughfat.Oneofthelargestconstituentsoffoodsprepared
awayfromhomeinAustraliaisfastfood(AustralianBureauofStatistics,2000).

Severalstudieshaveshownarelationshipbetweentheconsumptionoffastfoodsand
higherBodyMassIndex(BMI;e.g.,Duffey,GordonLarsen,Jacobs,Williams&Popkin,
2007)andothershavedemonstratedthatincreasesinfastfoodconsumptionresultin
weightgain(French,Harnack&Jeffery,2000).Theassociationbetweenincreasingrates
ofobesityandmorefrequentconsumptionoffastfoods,andthefactthatmanyofthese
foodsarelesshealthythanthosepreparedfromhome,hasledtothesuggestionthatfast
foodsmaybeacontributortotheobesityepidemic.

Asidefrompotentialhealthimplicationsoffastfoodconsumption,littleisknownabout
fastfoodconsumptionbehavioursandwhethertheyaresubjecttothesameinfluences
asothereatingbehaviours.Thereisevidencethatthepresenceofotherscaninfluence
eatingbehaviours,withresearchonsocialnormssuggestingthatthisinfluencecan
encourageorinhibitfoodintakedependingonthesocialcontext(Pliner,Bell,Hirsch&
Kinchla2006;Rosenthal&McSweeney,1979;Roth,Herman,Polivy&Pliner,2001).
222
Socialfacilitationstudieshaveshownthatthenumberofcoeaterspresentcanincrease
thetimespenteating,whichincreasestheamounteaten(calledthetimeextension
hypothesis;deCastro,1990;1994).Therearealsostudiesdemonstratingthatthetype
ofpeoplepresentmaymoderatesocialinfluences(Hetherington,Anderson,Norton&
Newson,2006;Learyetal.,1994;Salvy,Jarrin,Paluch,Irfan&Pliner,2007).Although
eachapproachhasshownthatthepresenceofotherpeoplemayaltergeneralfood
intake,nonehasconsideredhowthepresenceofothersmayinfluencefastfood
consumption.Theresearchpresentedinthisdissertationaimedtoexplorehowsocial
influencescouldaffecttheconsumptionoffastfoodatasingleeatingoccasion.

7.1 Overviewofcurrentresearch
Previousresearchonfastfoodconsumptionhasexploredthecharacteristicsofpeople
whoconsumefastfoods(e.g.,Mohr,Wilson,Dunn,Brindal&Wittert,2007;Schroderet
al.,2007)andthequalitiesthatattractconsumerstofastfood(Bryant&Dundes,2008;
Driskell,Meckna&Scales,2006;M.J.A.Schroder&McEachern,2005).Overall,thereare
limitedotherdataonfastfoodconsumption.Therefore,intheinitialstagesofthis
dissertationdatawerecollectedonfastfoodbehavioursandsocialinfluencesusingan
exploratoryapproach.Inordertocaptureinformationonfastfooditemsconsumed,a
varietyofcorrelatesoffastfoodconsumption,andtoassesstheviabilityofthesurvey,
theinitialFastFoodSurvey(FFS)wasdevelopedandadministeredtoasmallsampleat
theRoyalAdelaideShow(anagriculturalfair).

TheinitialFFSdemonstratedthatsocialfactorscouldbeassessedthroughtheFFSand
wereabletopredictfastfoodconsumption.Consequently,theFFSwasadministeredtoa
largersampleinordertoinvestigatefurthersocialinfluencesonfastfoodconsumption.
223
deCastrohasreportedontheeffectsofsocialfacilitationinmanyeatingcontexts,but
notfastfoodconsumption(J.M.deCastro,1990,1994;Redd&deCastro,1992;
Stroebele&deCastro,2004).Anexistingmodelofsocialfacilitation(basedongeneral
foodintakeanddevelopedbyFeunekes,vanStaverenanddeGraaf,1995)was
developedandrefinedusingdatafromthelargerFFS.Contrarytotheframeworkof
socialfacilitation,therefinedmodelofsocialinfluenceforfastfoodconsumption
indicatedthathavingpeoplepresentatafastfoodeatingoccasionmayhavean
inhibitoryeffectonintake,atrendparticularlyapparentinthesampleofwomenwho
respondedtotheFFS.

Thereisagrowingbodyofresearchindicatingthatsocialnormsmaycreategender
baseddifferencesineatingbehaviours(Basow&Kobrynowicz,1993;Chaiken&Pliner,
1987;Mori,Pliner&Chaiken,1987;Pliner,Chaiken&Flett,1990).Consequently,the
primaryaimofthenextstudywastoaddresswhethertheinhibitoryeffectthatthe
presenceofothersappearedtohaveonwomensintakeinthelargerFFScouldbe
explainedbytheactivationofminimaleatingnorms(thetendencyforwomentoeatless
inthepresenceofacoeater)andimpressionmanagement.Inthisstudy,sexdifferences
infastfoodconsumptionweremeasuredbyobservingdiners,theitemseatenandtime
spenteatinginafastfoodrestaurant.

Finally,theobservationalstudywasrepeatedusingasampleofpeopleobservedina
NorwegianrestaurantinordertoassesswhetherthetrendsobservedinAustraliawere
alsoobservableinothercultures.Previousstudieshaveindicatedthateatingcanbe
stronglygovernedbyculturaldifferencesthatalterthetimingandstylesoffoodeaten(J.
224
M.deCastro,Bellilse,Feunekes,Dalix&deGraaf,1997;Prescott,Young,ONeill,Yau&
Stevens,2002).


7.2 Socialinfluenceandfastfoodconsumptionbehaviours
7.2.1 Socialfacilitationandthetimeextensionhypothesis
Therewaspreviouslyonetestedsocialfacilitationmodelthatwasusedtoguidethe
currentresearch(i.e.,Feunekesetal.,1995).Themodelsdevelopedandpresentedhere
providefurtherunderstandingofhowsocialinfluencemayoperateinaneating
environmentbytestingandredevelopingthemodelofsocialfacilitationpresentedby
Feunekesetal.Thisextendedthepreviousmodelbyseparatingmodelsformenand
women,includingdifferentinterpersonalrelationshipsandaccountingforanumberof
fastfoodspecificfactors.

Overall,therewaslittlesupportforthemodelofsocialfacilitationforfastfood
consumptionintheFFSdata.Contrarytoexpectationsbasedontheoriesofsocial
facilitation(J.M.deCastro,1990;J.M.deCastro&E.S.deCastro,1989),womensfast
foodconsumptiondecreasedinthepresenceofothers.Theeffectofthepresenceof
otherdinerswassmallformaleparticipants,alsoindicatinglittlesupportforthemodel
ofsocialfacilitationinfastfoodconsumption.Intheobservationalstudy,therewas
someevidencethatthepresenceofothersincreasedintake(inallgroupsexceptfor
womeneatinginmixedcompany).Yet,thedirectrelationshipbetweenthenumberof
peoplepresentandtheamounteatenstillfailedtoreachsignificance.Althoughthe
presenceofothershasbeenshowntoincreasemealsizeinmanysocialfacilitation
studies,therewasnotstrongevidencethatthepresenceofothersdirectlyincreased
intakeinthecontextoffastfoodconsumption.
225
Inthemodelsdeveloped,therewasevidencethatthepresenceofotherscouldpredict
thetimespenteating,whichconsequentlypredictedintakefromfastfooditems,evenin
thepresenceofadirect,negativeinfluenceofthenumberofotherpeoplepresenton
consumptionbyfemaleconsumers.Beforethesocialinfluencemodelswereconstructed,
itwasunclearwhetherthetimeextensionhypothesiswouldbeobservableinfastfood
consumptionbehavioursbecausefastfoodsaredesignedtobeeatenquickly.Tothe
contrary,theresultsfrommodellingsupportedtheutilityofdeCastros(1990)time
extensionhypothesiswhilealsopartiallysupportingthemodelpresentedbyFeunekes
etal.(1995).

Althoughthisresearchhasdemonstratedtimeextensioninthecontextoffastfood,it
doesnotaccountforwhataspectsofsocialinfluenceprolongthetimespenteating.
Therewassomeindicationthatitisnotpurelydurationwhichincreasesintakeandthat
itmaybesomethingaboutthepresenceofotherpeoplethatfacilitatestherelationship
betweenthetimespenteatingandtheamounteaten.Datafromtheobservationalstudy
indicatedthat,althoughlonedinerswhowerereadingateforlongerthanlonediners
whowerenotreading,thetimespenteatingdidnotrelatetotheamounteatenbythese
consumers.Furtherresearchisneededtoclarifywhatitisaboutthepresenceofother
peopleataneatingoccasionthatcouldincreasethetimespenteating.Hetherington,
Anderson,NortonandNewson(2006)assessedthisintheirlaboratorystudy.They
reportedthatwhenpeopleateingroups,40%ofthetimespenteatingwasactually
spenttalking.deCastro(1994)alsosuggestedthatprolongedeatingtimemaybe
explainedbysocialinteraction.Futureresearchonthetimeextensionhypothesiscould
incorporatenaturalisticobservationalmethodswhichalsorecordtheamountoftime
eatingversusdoingotheractivities.
226
7.2.2 Minimaleatingnormsandgender
Theresultsfromtheobservationalstudysuggestedthatwomeninmixedsexcompany
conformedtominimaleatingnorms(i.e.,restrictedtheirintakeinordertoappearmore
feminine)wheneatinginafastfoodenvironment.Therefore,impressionmanagement
(motivatedbyselfpresentationalconcernsand/orgenderidentity)maybean
importantaspectofsocialinfluenceinafastfoodenvironment.

Ifselfpresentationunderliesimpressionmanagementandminimaleating,gendermay
notbetheonlyfactorthatcanactivatetheseconcerns.Forexample,obesepeoplemay
bemotivatedbyselfpresentationalconcernstoeatminimallyinsocialsituationswith
normalweightdiners(deLuca&Spigelman,1979).Furthermore,ifselfpresentation
motivatesminimaleating,theremaybesomethinguniqueaboutfastfoodconsumption
thatactivatesthesenormsasfastfoodconsumptioncanbeaccompaniedbynegative
stereotypes(e.g.,Dunn,Mohr,Wilson&Wittert,2008).Furthermore,otherauthorshave
suggestedthatcertainfoodqualitiescanbeassociatedwithgender(i.e.,appearing
feminine;Mooney&Lorenz,1997).Atthistime,itisunclearexactlyhowthefastfood
contextmoderatestheinfluencesonamountconsumedwhichhavebeenidentifiedin
previousresearchinotherfoodenvironments.

Asidefrommaintainingselfpresentation,minimaleatingmayhelpwomentodefine
themselvesasitreinforcestheirfeminineidentity.Ifthisisthecase,minimaleatingmay
becomeinternalisedandextendbeyondsocialeatingoccasions.Theinternalisationof
thisnormmayresultinrestrictiveeatingstyles.Inthisevent,theremaybean
associationbetweenadherencetominimaleatingnormsanddietarydisinhibitionlater
inthedayorevendisorderedeatingand/orbodyimageconcerns.Althoughspeculative
227
atthisstage,ifthisassociationexists,ithasimportanthealthimplicationsandwarrants
futureresearch.

Itisessentialtoincorporatethegenderofthefocalparticipantandtheotherpeople
presentattheeatingoccasioninfuturesocialinfluenceresearch.Womensufferhigher
ratesofdisorderedeatingandbodyimageconcernsthanmen(Plineretal.,1990).
Womenalsoseemtobemoreinterestedinnutritionandhealthresearchthanmen;
AustralianBureauofStatistics(2007)datashowsthattherearealmost3.5timesmore
womenmajoringinthehealthfieldthanmen
109
.Despitethefactthatthereisabodyof
researchfocussingongenderdifferencesineatingwhichiscentredprimarilyon
minimaleatingnorms,thewayinwhichthesecouldaltereatingbehaviourshasbeen
overlookedinsomesocialinfluenceresearch(e.g.,deCastro1990;Feunekesetal.,
1995).Socialinfluenceresearchersneedtobeawarethatmenandwomenmaybe
affecteddifferently(especiallyinthecontextofnutritionresearch)andaccountfor
thesedifferencesinstudydesign.

Priortotheobservationalstudy,theFFSfailedtoaccountforthesexoftheotherpeople
presentatthefastfoodeatingoccasionrecalled.Failingtorecordthesexofthecompany
presentattherecalledfastfoodeatingoccasionintheFFSmeantthatthepotential
effectsofminimaleatingnormscouldnotbecontrolledforinthisstudy.Itispossible
thatthesampleintheFFScontainedahighproportionofwomenadheringtominimal
eatingnorms(giventhehighpercentageofwomen,thisiscertainlypossible).Ifwomen
eatinginmixedsexgroupswereoverrepresented,thismayhaveobscuredtheabilityto

109
ThisdifferencewasalsoclearinthefemaledominatedsamplesobtainedthroughtheFFSs.
228
detectanysocialfacilitationeffectsintheFFSsample.Withoutdetailsofthegenderof
theotherpeoplepresent,therewasnowayofdeterminingthis.

Asidefromevidenceofminimaleatingnorms,therewasanindicationthattherewere
differentinfluencesatafastfoodeatingoccasionformenandwomen.Forexample,
perceivedatmospherewasanimportantpredictorforthetimespenteatinginwomen
butnotmeninthemodelofsocialinfluencedevelopedforfastfoodconsumption(from
theFFSdata).Differingphysiologymayexplaindifferencesinbasicenergyrequirements
betweenthesexes,butitdoesnotexplainwhyvaryingfactorsinafastfood
environmentinfluencemenandwomendifferently.Theoriesongenderrolesandnorms
offerexplanationastowhydifferencesinthetypeandvolumeoffoodeatenexist
betweensexes.Thesetheoriesdonotspecificallyaccountforwhydifferentaspectsof
theeatingenvironmentcouldaffectmenandwomendifferently.Itmaybethecasethat
thetwoinfluencesarerelated;thedifferentgenderrolesthatsurroundeatingmayalso
determinehowoneperceivestheirenvironment.

Finally,thecurrentresearchhassupportedpreviousstudiesbyshowingthatadherence
tominimaleatingnormscanbewitnessedusingobservationalmethods(e.g.,Klesges,
Bartsch,Norwood,Kautzman&Haugrud,1984).Inthecurrentresearch,theuseof
observationalmethods,althoughnotascontrolledasmethodsusedinexperimental
settings,alloweddifferentsocialinfluencestobewitnessedsimultaneously(e.g.,
minimaleatingandmatchingnorms).Inthefuture,dietdiarydatamayalsobeusedto
assessevidenceofminimaleatingnormsandinwhatcircumstancestheymaybe
witnessed(e.g.,wheneatingout,wheneatingwithcertaintypesofpeopleorwhen
eatingcertaintypesoffood).
229
7.2.3 Wheredonormsstartandfinish?
Normativetheoryprovidesausefulperspectiveforappliedresearchasitattemptsto
considertheinfluencesofnaturallyoccurringenvironmentsonbehaviour.Theextentto
whichoneparticularnorminfluencesbehaviourisoftendifficulttodiscern,withnorms
potentiallyactingbothadditivelyandinteractivelyandvaryingintheirinfluence
accordingtovariableslikeparticipantsex.Researchreportedheresuggeststhatthe
differentialinfluenceofthevariousnormsmaybecontextuallydetermined;itislikely
thatminimaleatingnormsmaydriveconsumptionforsomepeople(womeneatingin
mixedsexgroups)whileothernormsareimportantinothercircumstances.For
example,matchingtheintakeofthecodiner(i.e.,thematchingnorm)wasimportantfor
peopleeatinginpairsregardlessofthegendermixofthegroupintheobservational
study.Atamacrolevel,crossculturalcomparisonsoffastfoodintakereportedhere
indicatethatculturemayalsobeanimportantnormativeinfluenceonintake.
ConsumersatMcDonaldsinOsloatefordifferenttimesthanthoseobservedin
Adelaide.Furthermore,therewasnorelationshipbetweenthetimespenteatingandthe
amounteatenintheNorwegiansample,therebycastingdoubtontheapplicabilityofthe
timeextensionhypothesisinthissample.

Theinterrelatednatureofnormsmeansthatitisdifficulttodistinguishwhichwillbe
dominantinanygivensituation.Thismaybethereasonwhyotherresearchershave
foundinterpretingtheirresultsdifficultwhenassessingtheinfluenceofmultiplenorms
oneatingbehaviour(Plineretal.,2006).Someoftheambiguityassociatedwiththe
interpretationofnormativeinfluencesisevidentatthetheoreticallevel.TheBoundary
ModelofOvereating(Herman&Polivy,1984)suggeststhateatingand,more
specifically,theappropriateamounttoeat,issurroundedbyuncertaintyandthat
230
peoplelooktothebehaviourofothersintheenvironmenttodefineacceptable
behaviour(describedasmatching).Incontrast,minimaleatingresearchsuggeststhat
therearesomeaspectsofeatingwherethereareclearrulesandstereotypesabouthow
muchfoodshouldbeeaten(e.g.,minimaleatingnorms),presumablyregardlessofthe
environment.Underthesecircumstances,behaviourmaybecircumscribedbymore
broadlyandfundamentallydefinednorms,suchasgendernorms.Althougheating
behavioursmaybeuncertain,genderrolestendtobeclearandareingrainedinmany
socialinteractions(Eagly,1987).

7.2.4 Thewhofactorofsocialinfluence
deCastro(1994)reportedthateatingbehavioursweredifferentbetweenmealswith
familyandcoworkers.Feunekes,deGraaf,MeyboomandvanStaveren(1998)reported
thatsocialnetworkscaninfluenceintakedifferentlydependingupontheirnature.Other
researchershavealsosuggestedthatfriendscanincreasetheintakeofdifferenttypesof
food(Clendenenetal.,1994).Thecurrentresearchhasshownthatthepresenceof
differentpeople(i.e.,family,friendsandapartner)canalsohavevaryinginfluenceson
theconsumptionofasingletypeoffood(fastfood).Forexample,thenumberoffriends
presentwhenconsumingfastfoodhadlittleinfluenceonconsumptionbymalesinthe
largerFFS.Thus,despitethefactthatfastfoodconsumptionhasitsownrulesand
norms,withinthesethepeoplepresentmaystillbeimportantfordefiningtheeating
environment.

Thereareanumberofothercharacteristicsoutsideofinterpersonalrelationshipsthat
mayhelptoaccountforthedifferentinfluencesthatvaryingtypesofpeoplecreate.For
example,measuringtheclosenessofarelationshipsharedwithacoeatermayprovide
231
moreinformationthanasimpledescriptionoftheotherpeoplepresent(i.e.,partner).
VerleghandCandel(1999)suggestedthatprimaryreferencegroups,suchasfamilies,
canstronglyinfluenceeating.Butevenwithinfamilies,theremaybefactorssuchas
dominance,closenessandthewaythatindividualsinteractwitheachotherthat
moderatetheseinfluences.Investigatingtheinfluenceofthesevariablesmaycontribute
toknowledgeabouthowthetypeofpeoplepresentataneatingoccasioninfluences
eatingbehavioursbeyondfastfoodconsumption.

Includingtheweightstatusoftheotherpeoplepresentwhileeatingmayhelpto
determinehowwhoispresentaffectseatingenvironments.Intheirexperimentalstudy,
deLucaandSpigelman(1979)foundthatobesepeopleatemorelollieswheneating
withanobeseconfederate.Thecurrentresearchhasnotfocussedonweightstatus
becausetherewerelimitednumbersofpeopleconfidentintheirselfreportedheights
andweightsintheFFSandlittlevariationintheestimatedweightstatusofparticipants
recordedintheobservationalstudy.Nevertheless,weightstatusmayprovean
interestingcomparisoninthefuture.

Newapproachestotheanalysisofsocialinfluencescouldhelptoexplainhowother
peopleinfluenceeatingbehaviour.Thiscouldincludestructuralequationmodelling,in
whichatheorythatdescribestherelativecircumstancesunderwhichvariousnormsare
differentiallyinfluentialistested.Anothersuitableapproachforfutureresearchislag
sequentialanalysis,inwhichthebehaviouraleventsimmediatelyprecedingfastfood
orderingorconsumption(e.g.,arrivalofadesirablepartner),arecodedanddifferential
probabilitiesforatargetbehaviour(e.g.,orderingasmallmeal)examined.This
approachtoobservationaldataanalysisaimstodetecttherecurringsequential
232
patternsinastreamofcodingcategoriesdescribingsocialinteraction.Thesetechniques
canbeemployedtostudytherepertoiresofindividualsandofdyadsandgroups
(Gottman&Roy,1990:p.2).Finally,recentadvancesinsocialnetworkanalysis
techniques(Butts,2008),particularlywhenutilisedlongitudinally,providethescopeto
testwhetherfastfoodconsumptionbehavioursareatleastpartiallysociallydetermined.

7.2.5 Dosocialinfluencesaffectfastfoodconsumption?
Therewasevidenceinthecurrentresearchthatsocialinfluencesongeneraleating
behaviours(i.e.timeextensionandminimaleatingnorms)couldbeobservedinafast
foodeatingcontextdespitethelimitedavailabilityoffooditemsandemphasisonspeed
ofserviceandconsumption.Inotherwords,theregimentedandwellcontrolledportion
sizesprovidedinfastfoodrestaurantsdidnotpreventtheeffectsofsocialinfluence
beingwitnessed
110
.Therefore,socialfactorsmayhaveapervasiveinfluenceon
consumptionamountacrossavarietyofeatingcontexts.

Thefindingthatfastfoodeatingcontextsmaybesubjecttosimilarinfluencestoother
eatingsituationsmeansthattheknowledgegainedfrompreviousresearchonsocial
influenceisrelevanttotheunderstandingoffastfoodconsumption.Italsomeansthat
anyinsightsintothewayinwhichtheseinfluencescanbeusedtopositivelyaffect
behavioursmightalsobeapplicableinafastfoodcontext.Itwillbechallengingtodesign
healthpromotionapproachesthatutilisesocialinfluencetoimprovedietarybehaviours.
Advisingwomentoeatinmixedsexgroupsmayresultinlowertotalfastfoodintakeat
asingleeatingoccasionbutmaynotchangetheamountmaleseat.Therefore,

110
Acrossmanycircumstances,theaverageenergyamountsconsumedfromfastfooditemswerefairly
consistent,yettherewasenoughdifferenceintheintakeofwomeninmixedgroupstoindicatesignificant
differences.
233
understandingtheapplicabilityofminimaleatingnormstofastfoodconsumptionalone
provideslimiteddirectionregardingintervention.Yet,overall,knowledgeofnormsmay
helpguideinterventionprograms.Previousresearchhasdirectlyappliedageneral
normativeapproachtomodifyhealthbehaviours(Berkowitz,2003,2005).Thishas
involvedalteringsocietalexpectationsorbycorrectingnormativemisperceptions,and
thusindividualbehaviourandhasprovensuccessful.

7.3 Unravellingfastfoodconsumption
Therewereseveralresultsfromthecurrentfindingsthatsupportedthesuggestionthat
fastfoodconsumptioncouldresultinweightgainandthereforeisabehaviour
potentiallyrelatedtoobesity.Nutritionprofilingrevealedthattraditionalbigbrandfast
foodswerehighinfat.DatafromtheinitialFFSfurtherindicatedthatfewpeople
orderedlowfatoptionswhentheywereavailable.Finally,therewassomeindication
thatBodyMassIndexwasrelatedtofrequencyoffastfoodconsumptionintheFFS.

AlthoughtheFFStendedtoattractparticipantsthatwerenotrepresentativeofthe
widerAustralianpopulation,recruitmenttargetedonlypeoplewhohaveactually
consumedfastfood.ItisdifficulttodeterminehowrepresentativetheFFSsampleswere
relativetothepopulationofAustralianfastfoodconsumers.Thereissomeindicationin
previousstudiesthatcertaindemographiccharacteristicsarerelatedtofrequencyof
fastfoodconsumption(e.g.,highincomeandbeingyounger;Mohretal.,2007).
Therefore,thefactthatthesamplestendedtorepresentpeopleofhighincome(and
fromareaswithlowsocioeconomicdisadvantage)maybeattributedtothefastfood
natureofthesurveysandnotpurelysamplingbias.Furthermore,behavioursreported
bytheinitialFFSsamplerecruitedatthefairandthelargersamplerecruitedviathe
234
Internetwerelargelyconsistent,providinggoodvalidityoftheFFS.Therefore,someof
theinsightgainedaboutfastfoodbehavioursandsocialinfluencesinthecurrent
researchcanbeusedtoprovidedirectionforconsumeradviceandthedevelopmentof
interventionprogramsthatattempttoreducetheassociationbetweenfastfood
consumptionandweightgain.


7.3.1 Theculturesurroundingfastfoodconsumption
Thereareaspectsofbigbrandfastfoodthatareculturallydistinct.Fastfoodisoften
seenasaniconofWesternisationandglobalisation(Park,2004)andisregularly
presentedasasliceofNorthAmericanculture(Brailsford,2003).

DatafromtheFFSindicatedthatsomeoftheculturalaspectsoffastfoodswere
motivatorsforconsumption.Conveniencewasreportedasadominantdriveroffast
foodconsumptioninbothsurveys.Thedriveforconvenienceinfastfoodconsumersis
alsosupportedbythehighfrequencyoftakeawayorders,andisevidentamongthose
whopurchaseditemswhiletravellingbetweenactivitiesintheFFS.Therecentchanges
inactivitypatterns,timedemandsandineatingbehaviours(e.g.,thetendencytoeat
whiledoingotheractivities)maymeanthatmuchmoderneatingismotivatedby
convenience.

ItisclearthatcurrentWesternculturehaselementsthatresultinfeelingsoftime
scarcityand,ifconvenienceistheprimarydriverofconsumption,increasingtime
scarcityislikelytofurtherincreasedemandforconveniencefoods,includingfastfood.
Asthemodeloffoodcognitionsuggests,limitedtimeisalsolikelytochangefoodchoices
aspeoplehavelesstimetomakedecisionsthatsatisfydesiresfortaste,cost,healthand
235
convenience(Scheibehenne,Miesler&Todd,2007).Lifestylemaythereforebeabig
contributortotheincreaseinfastfoodconsumptionoverrecentdecades.

Thepervasivenessoftimespenteatingasapredictoroftheamounteatensuggeststhat
theremaybeareversemechanisminwhichlifestylefactorscoulddecreasefastfood
intake.Ifconsumersspentlesstimeeating,theymayconsumelessfastfoodthereby
reducingthelikelihoodofoverconsumption(andnegativehealthoutcomesinthelong
term).Limitedtimeandconvenienceincreasinglyattractconsumerstofastfoods,butif
theseconsumersateforlesstime,theymightalsoconsumelessfood.Inthissense,time
mayhaveaparadoxicaleffectonfastfoodintakebyattractingconsumerstofastfood
whilealsodictatingthattheyeattheirfoodquickly.

Therearefewwaystoofferpeoplemoretimeintheirdaytodaylivesandtherefore
changetheirfoodchoices.Yet,theculturalshifttowardsconvenientfoodconsumptionis
beingopposedbyasmallbutvocalsegmentinthecommunity;theslowfood
movement
111
.Theaimofproponentsofslowfoodistoreintroducefoodasanimportant,
valuableaspectofourculture.Thismovementisrelativelysmallcomparedtothefast
foodmarketandhasobviouscontrasttofastfoodculture.Nevertheless,itsuggeststhat
ifconsumersvaluedfoodandfoodpreparationmore,theywouldbelesslikelytoallow
timepressurestointerferewiththeirfoodchoices.


111
Seehttp://slowfoodaustralia.com.au/.TheslowfoodmovementinAustraliaispartonaninternational
associationandclaims2000membersasofJune2008onitswebsite.
236
7.3.2 Fastfoodconsumptionbehavioursandthepotentialforweightgain
Someoftheuniqueaspectsofbigbrandfastfoodsarethefoodsofferedandthewayin
whichtheyaresoldtoconsumers.Theseareaspectsthatcouldbetargetedtomakefast
foodconsumptionlessunhealthy.

FollowingthenutritionprofilinginChapterTwo,itwassuggestedthatchoicescouldbe
madetoalterthepotentialeffectoffastfoodconsumptiononenergybalance.
Specifically,mealsfromsomecompaniesprovidedmoretotalenergythanothers.When
actualconsumerbehaviourswereassessed,theaverageintakesupportedtheresultsof
thenutritionprofiling
112
.ThetypicalHungryJacksmealwasthelargestamongstthe
traditionalfastfooditemsbasedonthenutritionanalysis.Subsequentdatafromthe
initialFFSindicatedthatpeopleeatingitemsfromHungryJacksatemorethanpeople
eatingatotherchains(McDonaldsorDominosPizza).Themodelsofsocialinfluence
fromthelargerFFSalsoindicatedthateatingfromHungryJackswasasignificant
predictorofincreasedintake,evenwhencontrollingforavarietyofotherbehaviours
andindividualdemographics.Thus,independentofwhatbehavioursaretypically
displayedbydinersataneatingoccasionwithintherestaurant,theactualchoiceoffast
foodestablishment,byitself,isasignificantpredictorofintake.Thiswaslargely
determinedbythesizeandcookingmethodsofthefoodprovided.Theseareissuesthat
companiesthemselvesmaybeabletoaddresstoreducetheassociationbetweenfast
foodconsumptionandweightgain.


112
Typicalmediummealsrangedfromabout3200KJ(DominosPizza)to5100KJ(HungryJacks).The
meanenergyconsumedfromeachofthefastfoodchainsrangedfrom2184(DominosPizza)to4338KJ
(HungryJacks)inthefirstFFS.
237
Itisnotthefoodalonethatcanincreaseintakebutalsothemarketingwithinthe
purchasingenvironment.ThenutritionalprofileofthetraditionalmealatKFCwas
similartotheonefromMcDonalds,yethavingKFCpositivelypredictedenergyintake
(relativetoMcDonalds)inthelargerFFS.Anassessmentoftheordersofpeoplewhoate
atKFCshowedthatmealswereconstructedfrommoreitemsthanthetraditional
burger,friesandsoda.Itislikelythatthesemealsactedtopromoteincreasedintake
becauseconsumersareattractedtothevalueoflargerportions(Vermeer,Steenhuis&
Seidell2009).

Althoughtherewaslimitedabilitytoassessenergybalancing(i.e.,theextenttowhich
peoplebalancedietaryenergyconsumedinonedayacrossallmealsandsnacks)
throughoutthedayintheFFS,theenergycontentofthemealssuggestedthat,asamain
mealreplacement,fastfoodintakemaynotupsetenergybalance.Thiswassupportedby
dataindicatingthatroughly90%ofpeopleclassifiedtheirfastfoodeatingoccasionasa
meal(breakfast,lunchordinner).However,thosewhoreportedhavingsnacksdidnot
eatsignificantlylessthanthosewhoatefastfoodasameal,andthismaytherefore
impactontheirenergyequilibrium.Itisthosewhoeatfastfoodasasnackwhoare
probablyatgreatestriskforweightgain.

Frequencyofconsumptiontrendedtowardsignificanceinthepredictionoftheamount
offastfoodeatenatasingleoccasionintheinitialFFS.Thisfindingfailedtobe
corroboratedinthelargerFFSsample.Moreneedstobeunderstoodaboutthe
consumptionbehavioursofpeoplewhofrequentlyconsumefastfoods.Existing
literatureindicatesthatfastfoodconsumptionrelatestootherunhealthylifestyle
behaviours(Bowman,Gortmaker,Ebbeling,Pereira&Ludwig,2004;French,Story,
238
NeumarSztainer,Fulkerson&Hannan,2001;Jeffery,Baxter,McGuire&Linde,2006).
Someauthorssuggestthateatingenergydensefoodencouragestheconsumptionof
otherenergydensefoods(Prentice&Jebb,2003).

Neitheroftheapproachestodatacollectionandanalysisusedinthisresearchallowed
assessmentofgeneraldietarybehaviours.Tounderstandtheeffectsthatfastfoodintake
hasonenergybalance,andinthelongertermobesity,otherlifestylebehavioursneedto
beassessed.Understandingtheimpactoffastfoodconsumptiononriskforobesity
requiresmappingpotentialcompensatorybehaviours.Theseincludeexercise
participationandenergyintakethroughouttherestoftheday,weekormonth.
Currently,thereislimiteddataonthepatterningofenergyinputandoutputandhow
thebalanceisaffectedbytheconsumptionofafastfoodmeal.

Thegreaterunderstandingoffastfoodbehavioursthatthisdissertationprovidescanbe
usedtoguidethedevelopmentofdietdiarystyleassessmentoftheimpactoffastfood
intakeinthefuture.Thesediariesmayalsobeabletoincorporatemeasurementofother
obesogenicbehavioursassociatedwithfastfoodconsumption(e.g.,participationin
sedentarybehaviourssuchasvideogamingandtelevisionwatching).Gainingaclearer
pictureofhowthepatternsofbehaviourrelatetooneanotherwillprovideaclearer
focusforbehaviourswhichwarrantattentionfromthoseconcernedwithimproving
publichealthoutcomes.Changingsomebehavioursinanegativepatternofhealth
behavioursmayfacilitatechangeinotherareasbutonlytheimplementationand
assessmentoffutureinterventionscandeterminethis.

239
7.3.3 Integratingknowledgeoffastfoodbehavioursandsocialinfluenceforhealth
promotion

Feelingsoftimescarcitymaychangefoodchoices,butnotthegeneralinfluenceson
eatingbehaviour.Ifconsumersareeatingfastfoodsmoreoften,ifthesefoodsarehigher
infat,andifsocialfactorscanincreaseconsumption,thereisastronglikelihoodthatfast
foodcouldbeassociatedwithoverconsumptionofdietaryenergyandnegativehealth
outcomes.Itisthereforeimportanttoconsiderhowknowledgedevelopedfromthe
currentstudiescanbeusedtoreducethepotentiallydamagingimpactoffastfood
consumption.

Inaccordancewiththesocialnormsapproach,therearevariousmessagesthatcouldbe
conveyedtothepublictoaltersocialnorms(Berkowitz,2005)andpotentiallyreduce
associationsbetweenfastfoodconsumptionandweightgain.Thesemessagesneedto
becarefullyconstructed.Peoplehaveatendencytoreducefoodsintothemostbasic
categories;foodsareessentiallygoodorbad(Rozin,Ashmore,&Markwith,1996).
Furthermore,peopleinherentlybelievethathighfatfoodsarebad(Roefs&Jansen,
2002).Manyhealthprofessionalsshareandpromotethisviewwhendiscussingfast
food.Therearenumerouspsychologicaltheoriesthatsuggestthatpurelydemonisingor
banninganitemcanincreaseitsdesirability.Reactancetheorysuggeststhatpeoplewill
reacttohavingtheirfreedomtakenawaybyseekingtheverythingtheyhavehadtaken
awayfromthem(Brehm,1989;Rains&Turner,2007).Socialdisinhibitionexperiments
haveshownthatwhenrestrainedeaterseatsomethingbad,theycanlosetheir
inhibitionfortherestoftheday,consumingmorebadfoodsthanusualbecausethey
240
havebrokentheirdiet(Herman,Polivy&Leone,2006)
113
.Providingpublichealth
messagesthatsimplytellconsumersthatfastfoodsarebadandthattheyshouldnot
consumethemisanarrowandpotentiallyacounterproductiveapproach.

Combininganormativeinterventionapproachwithanunderstandingoffastfood
consumptionbehavioursprovidesamorepromisingdirectionfordevelopingmessages
toaltersocialnorms.Healthmessagescouldbedesignedtoremindconsumersthatfast
foodmealsrepresentamainmealandnotanextrafoodaddedtothedailylistoffood
sins.Simultaneously,fastfoodconsumersneedtoberemindedofsomeofthe
overconsumptionnormssurroundingfastfoodconsumption.Forexample,thefactthat
anindividualburgercanrepresentanentiremealsintake,eventhoughitisnotsold
individuallyasameal.Finally,messagescouldalsobecommunicatedtoeducate
consumersthatsomebigbrandfastfoodsaremorelikelytoincreaseenergyintakethan
others.

7.4 Conclusion
Thisdissertationhasexploredmultipleformsofsocialinfluenceonfastfood
consumption.Thecurrentstudiescontributetothegrowingbodyofliteratureregarding
thesocialinfluencesoneatingbehaviourbyshowingthatminimaleatingnorms,time
extensionand,tosomeextent,matchingnorms,canbeassociatedwithfastfood
consumption.Theseresultshavebeenobservedinstudiesutilisingnovelapproachesto
datacollectionthatrequirefurtherrefinement.Ultimately,thisdissertationrepresents
thefirstcomprehensivesteptowardabetterunderstandingoffastfoodconsumption

113
Researchrecentlypublishedconteststhetheoryofdietarydisinhibition(seeTomiyama,Moskovich,
ByrneHaltom,JuandMann,2009).
241
anditssocialandenvironmentalinfluences.Amorethoroughunderstandingoffastfood
consumptionanditsassociatedbehaviourscanbeusedtodeveloppublichealth
messagesandeveninterventionsthataimtoreduceanyweightgainassociatedwiththe
intakeoffastfood.

242
References

Aikman,S.N.,&Crites,S.L.(2005).Hashbrownsforbreakfast,bakedpotatoesfor
dinner:Changesinfoodattitudesasafunctionofmotivationandcontext.
EuropeanJournalofSocialPsychology,35(2),181198.
Ajzen,I.(1985).Fromintentionstoactions:Atheoryofplannedbehavior.InJ.Kuhl&J.
Beckman(Eds.),Actioncontrol:Fromcognitiontobehavior(pp.1139).
Heidelberg,Germany:Springer.
Allport,F.H.(1920).Theinfluenceofthegroupuponassociationandthought.Journalof
ExperimentalPsychology,3(3),159182.
Asch,S.E.(1955).Opinionsandsocialpressure.ScientificAmerican,193(5),3135.
AustralianBureauofStatistics(2000).Householdexpendituresurvey,Australia:Detailed
expenditureitems.(No.6535.0).Canberra:TheAustralianGovernment.
AustralianBureauofStatistics(2007).2006Census:Communityprofileseries(No.
2001.0).Canberra:TheAustralianGovernment.
AustralianBureauofStatistics(2008a).Censusofpopulationandhousing:Socio
EconomicIndexesforAreas(SEIFA),AustraliaDataonly.Canberra:The
AustralianGovernment.
AustralianBureauofStatistics(2008b).AnintroductiontoSocioEconomicIndexesfor
Areas(SEIFA)(No.2039.0).Canberra:TheAustralianGovernment.
AustralianBureauofStatistics(2008c).SocioEconomicIndexesforAreas(SEIFA)
Technicalpaper(No.2039.0.55.001).Canberra:TheAustralianGovernment.
AustralianInstituteofHealth&Welfare[AIHW]&NationalHeartFoundationof
Australia(2004).Therelationshipbetweenoverweight,obesityandcardiovascular
disease.(No.Cat.CVD29).Canberra:AIHW.
243
Bandura,A.(1977).Sociallearningtheory.EnglewoodCliffs,N.J:PrenticeHall.
Baron,R.M.,&Kenny,D.A.(1986).Themoderatormediatorvariabledistinctionin
socialpsychologicalresearch:Conceptual,strategic,andstatistical
considerations.JournalofPersonalityandSocialPsychology,51,11731182.
Barr,E.L.M.,Magliano,D.J.,Zimmet,P.,Polkinghorne,K.R.,Atkins,R.,Dunstan,D.,
Murray,S.G.&Shaw,J.(2006).AusDiab2005theAustralianDiabetes,Obesityand
LifestyleStudy:Trackingtheacceleratingepidemic:itscausesandoutcomes.
Melbourne:TheInternationalDiabetesInstitute.
Baskin,M.L.,Ard,J.,Franklin,F.,&Allison,D.B.(2005).Prevalenceofobesityinthe
UnitedStates.ObesityReviews,6,57.
Basow,S.A.,&Kobrynowicz,D.(1993).Whatissheeatingtheeffectsofmealsizeon
impressionsofafemaleeater.SexRoles,28(56),335344.
Batada,A.,Seitz,M.D.,Wootan,M.G.,&Story,M.(2008).Nineoutof10food
advertisementsshownduringSaturdaymorningchildren'stelevision
programmingareforfoodshighinfat,sodium,oraddedsugars,orlowin
nutrients.JournaloftheAmericanDieteticAssociation,108(4),673678.
Baumeister,R.F.(1982).Aselfpresentationalviewofsocialphenomena.Psychological
Bulletin,91(1),326.
Beardsley,L.,&Pedersen,P.(1980).Healthandculturecenteredintervention.InJ.Berry,
M.Segall&C.Kagitcibasi(Eds.),HandbookofCrossCulturalPsychology(2nd
editioned.,Vol.3:Socialbehaviorandapplications,pp.415442).Boston:Allyn
andBacon.
Bell,R.,&Pliner,P.L.(2003).Timetoeat:therelationshipbetweenthenumberof
peopleeatingandmealdurationinthreelunchsettings.Appetite,41(2),215218.
244
Bellisle,F.,&Dalix,A.M.(2001).Cognitiverestraintcanbeoffsetbydistraction,leading
toincreasedmealintakeinwomen.AmericanJournalofClinicalNutrition,74(2),
197200.
Bellisle,F.,Dalix,A.M.,&Slama,G.(2004).Nonfoodrelatedenvironmentalstimuli
induceincreasedmealintakeinhealthywomen:comparisonoftelevision
viewingversuslisteningtoarecordedstoryinlaboratorysettings.Appetite,
43(2),175180.
Berkowitz,A.D.(2003).Applicationsofsocialnormstheorytootherhealthandsocial
justiceissues.InH.W.Perkins(Ed.),Thesocialnormsapproachtopreventing
schoolandcollegeagesubstanceabuse:Ahandbookforeducators,counsellors,
andclinicians(1sted.,pp.259279).SanFrancisco:JosseyBass.
Berkowitz,A.D.(2005).Anoverviewofthesocialnormsapproach.InL.C.Lederman&L.
P.Stewart(Eds.),Changingthecultureofcollegedrinking:Asociallysituated
healthcommunicationcampaign(1sted.,pp.29).NewJersey:HamptonPress.
Best,S.J.,Krueger,B.,Hubbard,C.,&Smith,A.(2001).Anassessmentofthe
generalizabilityofInternetsurveys.SocialScienceComputerReview,19(2),131
145.
Billon,S.,Lluch,A.,Gueguen,R.,Berthier,A.M.,Siest,G.,&Herbeth,B.(2002).Family
resemblanceinbreakfastenergyintake:theStanislasFamilyStudy.European
JournalofClinicalNutrition,56(10),10111019.
Bingham,S.A.,Nelson,M.,Paul,A.A.,Haraldsdottir,J.,Loken,E.B.,&vanStaveren,W.A.
(1988).Methodsfordatacollectionatanindividuallevel.InM.E.Cameron&W.A.
vanStaveren(Eds.),Manualonmethodologyforfoodconsumptionstudies(pp.
53105).Oxford:OxfordUniversityPress.
245
Binkley,J.K.,Eales,J.,&Jekanowski,M.(2000).Therelationbetweendietarychangeand
risingUSobesity.InternationalJournalofObesity,24(8),10321039.
BISShrapnel(2003).FastfoodandeatingoutinAustralia20032005[pressrelease].
RetrievedOctober4,2005,from
<http://www.bis.com.au/verve/_resources/FastFoodPressRelease[1].pdf>
Bisogni,C.A.,Connors,M.,Devine,C.M.,&Sobal,J.Y.(2002).Whoweareandhowwe
eat:Aqualitativestudyofidentitiesinfoodchoice.JournalofNutritionEducation
andBehavior,34(3),128139.
Blundell,J.E.&MacDiarmid,J.I.(1997).Fastasariskfactorforoverconsumption:
Satiation,satiety,andpatternsofeating.JournaloftheAmericanDietetic
Association,97(7),S63S69.
Bock,B.C.,&Kanarek,R.B.(1995).Womenandmenarewhattheyeat:Theeffectsof
genderandreportedmealsizeonperceivedcharacteristics.SexRoles,33(1/2),
109119.
Bond,C.F.,&Titus,L.J.(1983).Socialfacilitation:Ametaanalysisof241studies.
PsychologicalBulletin,94(2),265292.
Bonifacj,C.,Gerber,M.,Scali,J.,&Daures,J.P.(1997).Comparisonofdietaryassessment
methodsinaSouthernFrenchpopulation:useofweighedrecords,estimateddiet
recordsandafoodfrequencyquestionnaire.EuropeanJournalofClinical
Nutrition,51(217231).
Bowman,S.A.,Gortmaker,S.L.,Ebbeling,C.B.,Pereira,M.A.,&Ludwig,D.S.(2004).
Effectsoffastfoodconsumptiononenergyintakeanddietqualityamong
childreninanationalhouseholdsurvey.Pediatrics,113(1),112118.
246
Bowman,S.A.,&Vinyard,B.T.(2004).FastfoodconsumptionofUSadults:Impacton
energyandnutrientintakesandoverweightstatus.JournaloftheAmerican
CollegeofNutrition,23(2),163168.
Bozeman,D.P.,&Kacmar,K.M.(1997).Acyberneticmodelofimpressionmanagement
processesinorganizations.OrganizationalBehaviorandHumanDecision
Processes,69(1),930.
Brailsford,I.(2003).USimagebutNZventure:Americanaandfastfoodadvertisingin
NewZealand,19711990.AustralasianJournalOfAmericanStudies,31,1024.
Brehm,J.W.(1989).Psychologicalreactancetheoryandapplications.Advancesin
ConsumerResearch,16,7275.
Brinkworth,G.D.,Noakes,M.,Keogh,J.B.,Luscombe,N.D.,Wittert,G.A.,&Clifton,P.M.
(2004).Longtermeffectsofahighprotein,lowcarbohydratedietonweight
controlandcardiovascularriskmarkersinobesehyperinsulinemicsubjects.
InternationalJournalofObesity,28(9),11871187.
Bryant,R.,&Dundes,L.(2008).Fastfoodperceptions:Apilotstudyofcollegestudents
inSpainandtheUnitedStates.Appetite,51(2),327330.
Bucher,H.(2002,May2225).UsabilityCorefeatureofinteractivityempiricalresultsof
audienceresearchonInternetandebusinesscommunication.Paperpresentedat
theConferenceWWDU2002WorldWideWork,Berchtesgaden.
Burke,K.(2007,October17).Aussiesgivepizzathechop.TheSydneyMorningHerald,
from<http://www.smh.com.au/cgi
bin/common/popupPrintArticle.pl?path=/articles/2007/2010/2016/11923008
20287.html>.
Butts,C.T.(2008).SocialNetworks:AMethodologicalIntroduction.AsianJournalof
SocialPsychology,11(1),1341.
247
Byrne,B.M.(2001).StructuralequationmodelingwithAMOS:basicconcepts,
applications,andprogramming.Mahwah,NewJersey:LawrenceErbaum
Associates.
Caildini,R.B.(1993).Influence:Scienceandpractice(3rded.).NewYork:HarperCollins
CollegePublishers.
CameronSmith,D.,Bilsborough,S.A.,&Crowe,T.C.(2002).UpsizingAustralia's
waistline:thedangersof"mealdeals".MedicalJournalofAustralia,177(1112),
686686.
Carron,A.V.,Burke,S.M.,&Prapavessis,H.(2004).Selfpresentationandgroup
influence.JournalofAppliedSportPsychology,16(1),4158.
Chaiken,S.,&Pliner,P.(1987).Women,butnotmen,arewhattheyeattheeffectof
mealsizeandgenderonperceivedfemininityandmasculinity.Personalityand
SocialPsychologyBulletin,13(2),166176.
Chapman,K.,Nicholas,P.,&Supramaniam,R.(2006).Howmuchfoodadvertisingis
thereonAustraliantelevision?HealthPromotionInternational,21(3),172180.
Cheang,M.(2002).Olderadultsfrequentvisitstoafastfoodrestaurant:Nonobligatory
socialinteractionandthesignificanceofplayinathirdplace.JournalofAging
Studies,16,303321.
Christakis,N.A.,&Fowler,J.H.(2007).Thespreadofobesityinalargesocialnetwork
over32years.NewEnglandJournalofMedicine,357(4),370379.
Cialdini,R.B.,&Trost,M.R.(1998).Socialinfluence:socialnorms,conformity,and
compliance.InD.T.Gilbert,S.T.Fiske&G.Lindzey(Eds.),TheHandbookofSocial
Psychology(4thed.,Vol.2,pp.151192).Boston:McGrawHill.
Clendenen,V.I.,Herman,C.P.,&Polivy,J.(1994).Socialfacilitationofeatingamong
friendsandstrangers.Appetite,23(1),113.
248
CodexAlimentariusCommission(2006).Guidelinesonnutritionallabelling.Retrieved
January2,2007,from<www.codexalimentarius.net>
Colagiuri,S.,&Caterson,D.(2008).KFCsponsorshipofcricket.MedicalJournalof
Australia,189(7),415416.
Comley,P.(1996,2008).TheuseofInternetasadatacollectionmethod.RetrievedJune,
2008,from
<http://www.websm.org/index.php?fl=2&lact=1&bid=2493&avtor=29&parent=
12>
Conger,A.J.(1974).Areviseddefinitionforsuppressorvariables:Aguidetotheir
identificationandinterpretation.EducationalandPsychologicalMeasurement,34,
3547.
Conger,J.C.,Conger,A.J.,Costanzo,P.R.,Wright,K.L.,&Matter,J.A.(1980).Theeffectof
socialcuesontheeatingbehaviorofobeseandnormalsubjects.Journalof
Personality,48(2),258271.
Cook,C.,Heath,F.,&Thompson,R.L.(2000).AmetaanalysisofresponseratesinWeb
orInternetbasedsurveys.EducationalandPsychologicalMeasurement,60(6),
821836.
Coon,K.A.,Goldberg,J.,Rogers,B.L.,&Tucker,K.L.(2001).Relationshipsbetweenuse
oftelevisionduringmealsandchildren'sfoodconsumptionpatterns.Pediatrics,
107(1),E7.
Coopera,T.V.,Klesges,L.M.,DeBonc,M.,Klesges,R.C.,&Shelton,M.L.(2006).An
assessmentofobeseandnonobesegirls'metabolicrateduringtelevision
viewing,reading,andresting.EatingBehaviors,7(2),105114.
249
Cottone,E.,&ByrdBredbenner,C.(2007).Knowledgeandpsychosocialeffectsofthe
filmSuperSizeMeonyoungadults.JournaloftheAmericanDieteticAssociation,
107(7),11971203.
Cowan,G.S.M.,Cowan,K.B.,Hiler,M.L.,Smalley,M.D.,&Sehnert,W.(1992).Obesity
stereotypes.ProblemsinGeneralSurgery,9(2),218226.
Cox,C.,&Foster,R.(1985).What'saheadfortheU.S.foodprocessingindustry?
Discussion.AmericanJournalofAgriculturalEconomics,67(5),11551157.
Datamonitor(2005).Globalfastfood:IndustryProfile[summary].NewYork:
DatamonitorUSA.
deCastro,J.M.(1990).Socialfacilitationofdurationandsizebutnotrateofthe
spontaneousmealintakeofhumans.Physiology&Behavior,47,11291135.
deCastro,J.M.(1991).Weeklyrhythmsofspontaneousnutrientintakeandmealpattern
ofhumans.Physiology&Behavior,50(4),729738.
deCastro,J.M.(1994).Familyandfriendsproducegreatersocialfacilitationoffood
intakethanothercompanions.Physiology&Behavior,56(3),445455.
deCastro,J.M.(1997).Inheritanceofsocialinfluencesoneatinganddrinkingin
humans.NutritionResearch,17(4),631648.
deCastro,J.M.(2000).Eatingbehavior:Lessonsfromtherealworldofhumans.
Nutrition,16(10),800813.
deCastro,J.M.,Bellisle,F.,Dalix,A.M.,&Pearcey,S.M.(2000).Palatabilityandintake
relationshipsinfreelivinghumans:characterizationandindependenceof
influenceinNorthAmericans.Physiology&Behavior,70(34),343350.
deCastro,J.M.,Bellisle,F.,Feunekes,G.I.J.,Dalix,A.M.,&deGraaf,C.(1997).Culture
andmealpatterns:AcomparisonofthefoodintakeoffreelivingAmerican,
Dutch,andFrenchstudents.NutritionResearch,17(5),807829.
250
deCastro,J.M.,&Brewer,E.M.(1992).Theamounteateninmealsbyhumansisa
powerfunctionofthenumberofpeoplepresent.Physiology&Behavior,51,121
125.
deCastro,J.M.,&deCastro,E.S.(1989).Spontaneousmealpatternsofhumans:
influenceofthepresenceofotherpeople.AmericanJournalofClinicalNutrition,
50,237247.
deLuca,R.V.,&Spigelman,M.N.(1979).Effectsofmodelsonfoodintakeofobeseand
nonobesefemalecollegestudents.CanadianJournalofBehavioralScience,11(2),
124129.
Deutskens,E.,deRuyter,K.,Wetzels,M.,&Oosterveld,P.(2004).Responserateand
responsequalityofInternetbasedsurveys:Anexperimentalstudy.Marketing
Letters,15(1),2136.
Devine,C.M.,Sobal,J.,Bisogni,C.A.,&Connors,M.(1999).Foodchoicesinthreeethnic
groups:Interactionsofideals,identities,androles.JournalofNutritionEducation,
31(2),8693.
Dietz,W.H.,Bandini,L.G.,Morelli,J.A.,Peers,K.F.,&Ching,P.L.Y.H.(1994).Effectof
sedentaryactivitiesonrestingmetabolicrate.AmericanJournalofClinical
Nutrition,59(3),556559.
DiPietro,R.B.,Roseman,M.,&Ashley,R.(2004).Astudyofconsumers'responseto
quickservicerestaurants'healthymenuitems:attitudesversusbehaviors.
JournalofFoodserviceBusinessResearch,7(4),5977.
Dixon,T.,&Waters,A.M.(2003).Agrowingproblem:trendsandpatternsinoverweight
andobesityamongadultsinAustralia,1980to2001.Canberra:AIHW.

251
DoctorsAssociatesInc.(2002).Subwaydietguymaintainsweightandcelebritystatus
[pressrelease].RetrievedJune12,2004,from
<http://www.subway.com/Publishing/PubRelations/PressRelease/pr
030502.pdf>
Dodd,C.J.,Welsman,J.R.,&Armstrong,N.(2008).Energyintakeandappetitefollowing
exerciseinleanandoverweightgirls.Appetite,51(3),482488.
Domino'sPizzaEnterprisesLtd.(2005).Informationaboutfranchising.RetrievedAugust
10,2005,from<http://dominos.com.au/corporate/franchise/Franchise.aspx>
Driskell,J.A.,Meckna,B.R.,&Scales,N.E.(2006).Differencesexistintheeatinghabitsof
universitymenandwomenatfastfoodrestaurants.NutritionResearch,26(10),
524530.
Dube,L.,Paquet,C.,Ma,Z.,McKenzie,D.S.A.,Kergoat,M.J.,&Ferland,G.(2007).
Nutritionalimplicationsofpatientproviderinteractionsinhospitalsettings:
evidencefromawithinsubjectassessmentofmealtimeexchangesandfood
intakeinelderlypatients.EuropeanJournalofClinicalNutrition,61(5),664672.
Duffey,K.J.,GordonLarsen,P.,Jacobs,D.R.,Williams,O.D.,&Popkin,B.M.(2007).
Differentialassociationsoffastfoodandrestaurantfoodconsumptionwith3y
changeinbodymassindex:theCoronaryArteryRiskDevelopmentinYoung
AdultsStudy.AmericanJournalofClinicalNutrition,85(1),201208.
Duffy,B.,Smith,K.,Terhanian,G.,&Bremer,J.(2005).Comparingdatafromonlineand
facetofacesurveys.InternationalJournalofMarketResearch,47(6),615639.
Dunn,K.I.,Mohr,P.B.,Wilson,C.J.,&Wittert,G.A.(2007,July).Amixedmodelanalysis
ofattitudestowardsfastfoods.Paperpresentedatthe10thEuropeanCongressof
Psychology,Prague,CzechRepublic.
252
Dunn,K.I.,Mohr,P.B.,Wilson,C.J.,&Wittert,G.A.(2008).Beliefsaboutfastfoodin
Australia:Aqualitativeanalysis.Appetite,51(2),331334.
Dunstan,D.,Zimmet,P.,Welborn,T.,Sicree,R.,Armstrong,T.,Atkins,R.,Cameron,A.,
Shaw,J.&Chadban,S.(2001).DiabesityandassociateddisordersinAustralia
2000.Melbourne:InternationalDiabetesInstitute.
Eagly,A.H.(1983).Genderandsocialinfluenceasocialpsychologicalanalysis.
AmericanPsychologist,38(9),971981.
Eagly,A.H.(1987).Sexdifferencesinsocialbehavior:asocialroleinterpretation.New
Jersey:LawrenceErlbaumAssociates,Inc.
Earle,M.D.(1997).Innovationinthefoodindustry.TrendsinFoodScience&Technology,
8(5),166175.
Edwards,J.S.,Engstrm,K.,&Hartwell,H.J.(2005).Overweight,obesityandthefood
serviceindustry.FoodServiceTechnology,5(24),8594.
Epstein,L.H.,Wing,R.R.,&Valoski,A.(1985).Childhoodobesity.PediatricClinicsof
NorthAmerica,32(2),363379.
EuromonitorInternational(2008).ConsumerfoodserviceinNorway[summary].
RetrievedApril10,2009,from
<www.euromonitor.com/Consumer_Foodservice_in_Norway>
Feunekes,G.I.J.,deGraaf,C.,Meyboom,S.,&vanStaveren,W.A.(1998).Foodchoice
andfatintakeofadolescentsandadults:Associationsofintakeswithinsocial
networks.PreventiveMedicine,27(5),645656.
Feunekes,G.I.J.,deGraaf,C.,&vanStaveren,W.A.(1995).Socialfacilitationoffood
intakeismediatedbymealduration.Physiology&Behavior,58(3),551558.
FOODweek(2008).Fastfoodquizgetsmixedresults.FOODweek,August22,56.
253
French,S.A.,Harnack,L.,&Jeffery,R.W.(2000).Fastfoodrestaurantuseamongwomen
inthePoundofPreventionstudy:dietary,behavioralanddemographic
correlates.InternationalJournalofObesity,24(10),13531359.
French,S.A.,Story,M.,NeumarkSztainer,D.,Fulkerson,J.A.,&Hannan,P.(2001).Fast
foodrestaurantuseamongadolescents:associationswithnutrientintake,food
choicesandbehavioralandpsychosocialvariables.InternationalJournalof
Obesity,25,18231833.
Friedman,L.,&Wall,M.(2005).Graphicalviewsofsuppressionandmulticollinearityin
multiplelinearregression.AmericanStatistician,59(2),127136.
Fries,E.,&Croyle,R.T.(1993).Stereotypesassociatedwithalowfatdietandtheir
relevancetonutritioneducation.JournaloftheAmericanDieteticAssociation,
93(5),551555.
Galcheva,S.V.,Iotova,V.M.,&Stratev,V.K.(2008).Televisionfoodadvertisingdirected
towardsBulgarianchildren.ArchivesofDiseaseinChildhood,93(10),857861.
Gerber,M.,&Corpet,D.(1999).Energybalanceandcancers.EuropeanJournalofCancer
Prevention,8(2),7789.
Gillis,L.J.,&BarOr,O.(2003).Foodawayfromhome,sugarsweeteneddrink
consumptionandjuvenileobesity.JournaloftheAmericanCollegeofNutrition,
22(6),539545.
Gillman,M.W.,RifasShiman,S.L.,Frazier,A.L.,Rockett,H.R.H.,Camargo,C.A.,Field,A.
E.,Berkey,C.S.&Colditz,G.A.(2000).Familydinneranddietqualityamongolder
childrenandadolescents.ArchivesofFamilyMedicine,9(3),235240.
Glanz,K.(2002).Convenience.AmericanDemographics,March,3031.

254
Goldman,J.,&Borrud,L.(1997).AnoverviewoftheUSDA's19941996ContinuingSurvey
ofFoodIntakesbyIndividualsandtheDietandHealthKnowledgeSurvey.
Retrieved5Oct,2005,from
<http://www.amstat.org/sections/srms/Proceedings/papers/1997_136.pdf>
Goldman,S.J.,Herman,C.P.,&Polivy,J.(1991).Istheeffectofasocialmodeloneating
attenuatedbyhunger?Appetite,17,129140.
Goris,A.H.C.,&Westerterp,M.R.(2008).Physicalactivity,fatintakeandbodyfat.
Physiology&Behavior,94(2),164168.
Gosling,S.D.,Vazire,S.,Srivastava,S.,&John,O.P.(2004).Shouldwetrustwebbased
studies?AcomparativeanalysisofsixpreconceptionsaboutInternet
questionnaires.AmericanPsychologist,59(2),93104.
Gottman,J.M.,&Roy,A.K.(1990).Sequentialanalysis.Aguideforbehavioralresearchers.
Cambridge:CambridgeUniversityPress.
Granello,D.H.,&Wheaton,J.E.(2004).Onlinedatacollection:Strategiesforresearch.
JournalofCounselingandDevelopment,82(4),387393.
Grant,J.F.,Taylor,A.W.,Ruffin,R.E.,Wilson,D.H.,Phillips,P.J.,Adams,R.J.T.,Price,K.
&theNorthwestAdelaideHealthTeam(2008).CohortProfile:TheNorthWest
AdelaideHealthStudy(NWAHS).InternationalJournalofEpidemiology,Dec.11(E
pub).
Gruber,T.,Szmigin,I.,Reppel,A.E.,&Voss,R.(2008).Designingandconductingonline
interviewstoinvestigateinterestingconsumerphenomena.QualitativeMarket
Research:AnInternationalJournal,11(3),256274.
Grundy,S.M.(2004).Obesity,metabolicsyndrome,andcardiovasculardisease.The
JournalofClinicalEndocrinology&Metabolism,89(6),25952600.
255
Gumbel,A.(2004,July23).McDonald'scelebrateitssaladdaysassalesofhealthmeals
rocket,TheIndependent,from
<http://www.independent.co.uk/news/world/americas/>.
Guthrie,J.F.,Lin,B.H.,&Frazao,E.(2002).Roleoffoodpreparedawayfromhomeinthe
Americandiet,197778versus199496:Changesandconsequences.Journalof
NutritionEducationandBehavior,34(3),140150.
Harris,M.B.,&Walters,L.C.(1991).Alteringattitudesandknowledgeaboutobesity.
JournalofSocialPsychology,131(6),881884.
Haslam,D.,&Wittert,G.(2009).Fastfacts:Obesity.Oxford:HealthPress.
Haynes,S.,&O'Brien,W.H.(2000).Principlesandpracticeofbehaviouralassessment.
NewYork:Springer.
Hennessy,J.,Mabey,B.,&Warr,P.(1998).Assessmentcentreobservationprocedures:
Anexperimentalcomparisonoftraditional,checklistandcodingmethods.
InternationalJournalofSelectionandAssessment,6(4),222231.
Herman,C.P.,Fitzgerald,N.E.,&Polivy,J.(2003).Theinfluenceofsocialnormson
hungerratingsandeating.Appetite,41(1),1520.
Herman,C.P.,Olmsted,M.P.,&Polivy,J.(1983).Obesity,externality,andsusceptibility
tosocialinfluence:Anintegratedanalysis.JournalofPersonalityandSocial
Psychology,45(4),926934.
Herman,C.P.,&Polivy,J.(1984).Aboundarymodelfortheregulationofeating.InA.J.
Stunkard&E.Stellar(Eds.),Eatinganditsdisorders(1sted.,pp.141156).New
York:RavenPress.
Herman,C.P.,&Polivy,J.(2005).Normativeinfluencesonintake.Physiology&Behavior,
86(5),762772.
256
Herman,C.P.,&Polivy,J.(2007).Normviolation,normadherence,andovereating.
CollegiumAntropologicum,31(1),5562.
Herman,C.P.,Polivy,J.,&Leone,T.(2006).Thepsychologyofovereating.InD.Mela(Ed.),
Food,Diet,andObesity(pp.115136).Cambridge,UK:WoodheadPublishing.
Herman,C.P.,Roth,D.A.,&Polivy,J.(2003).Effectsofthepresenceofothersonfood
intake:Anormativeinterpretation.PsychologicalBulletin,129(6),873886.
Hertzler,A.A.,&Frary,R.B.(1996).Familyfactorsandfatconsumptionofcollege
students.JournaloftheAmericanDieteticAssociation,96(7),711714.
Heshka,S.&Allison,D.B.(2001).Isobesityadisease?InternationalJournalofObesity,25,
14011404.
Hetherington,M.M.,Anderson,A.S.,Norton,G.N.M.,&Newson,L.(2006).Situational
effectsonmealintake:Acomparisonofeatingaloneandeatingwithothers.
Physiology&Behavior,88(45),498505.
Holdsworth,M.,Gartner,A.,Landais,E.,Maire,B.,&Delpeuch,F.(2004).Perceptionsof
healthyanddesirablebodysizeinurbanSenegalesewomen.International
JournalofObesity,28,15611568.
Horst,P.(1941).Theroleofpredictorvariableswhichareindependentofthecriterion.
SocialScience&ResearchBulletin,48,431436.
Hu,F.B.,Stampfer,M.J.,Manson,J.E.,Rimm,E.,Colditz,G.A.,Rosner,B.A.,Hennekens
C.H.&WillettW.C.(1997).Dietaryfatintakeandtheriskofcoronaryheart
diseaseinwomen.NewEnglandJournalofMedicine,337(21),14911499.
Jabs,J.,&Devine,C.M.(2006).Timescarcityandfoodchoices:Anoverview.Appetite,
47(2),196204.
Jackson,M.,Ball,K.,&Crawford,D.(2001).Beliefsaboutthecausesofweightchangein
theAustralianpopulation.InternationalJournalofObesity,25,15121516.
257
James,W.T.P.,Leach,R.,Mhurchu,C.N.,Kalamara,E.,Shayeghi,M.,Rigby,N.J.,Nishida,
C.&Rodgers,A.(2004).Overweightandobesity(highbodymassindex).InM.
Ezzati,A.D.Lopez,A.Rodgers&C.J.L.Murray(Eds.),Comparativequantification
ofhealthrisks:Globalandregionalburdenofdiseaseattributabletoselected
majorriskfactors.(pp.497596).Geneva:WorldHealthOrganization.
Jeffery,R.W.,Baxter,J.,McGuire,M.,&Linde,J.(2006).Arefastfoodrestaurantsan
environmentalriskfactorforobesity?InternationalJournalofBehavioral
NutritionandPhysicalActivity,3(2),from
<http://www.ijbnpa.org/content/3/1/2>.
Jeffery,R.W.,&French,S.A.(1998).EpidemicobesityintheUnitedStates:Arefastfoods
andtelevisionviewingcontributing?AmericanJournalofPublicHealth,88(2),
277280.
Jeffery,R.W.,Rydell,S.,Dunn,C.L.,Harnack,I.J.,Levine,A.S.,Pentel,P.,Baxter,J.,&Walsh,
E.M.(2007).Effectsofportionsizeonbodyweight.InternationalJournalof
BehavioralNutritionandPhysicalActivity,4(27),from
<http://www.ijbnpa.org/content/4/1/27>.
Jetten,J.,Postmes,T.,&McAuliffe,B.J.(2002).'We'reallindividuals':groupnormsof
individualismandcollectivism,levelsofidentificationandidentitythreat.
EuropeanJournalofSocialPsychology,32(2),189207.
Johnson,M.J.(2005,December5).Fastfoodchainditchessugarladensoftdrink.The
NewZealandHerald,from<www.nzherald.co.nz>.
Johnson,R.K.(2002).DietaryintakeHowdowemeasurewhatpeoplearereally
eating?ObesityResearch,10,63S68S.
Jorgensen,D.L.(1989).Participantobservation:Amethodologyforhumanstudies(Vol.
15).NewburyPark,Calif:SagePublications.
258
Judd,C.M.,&Kenny,D.A.(1981).Estimatingtheeffectsofsocialinterventions.NewYork:
CambridgeUniversityPress.
Kahan,D.,Polivy,J.,&Herman,C.P.(2003).Conformityanddietarydisinhibition:Atest
oftheegostrengthmodelofselfregulation.InternationalJournalofEating
Disorders,33(2),165171.
Kalmijn,S.,Launer,L.J.,Ott,A.,Witteman,J.C.M.,Hofman,A.,&Breteler,M.M.B.(1997).
DietaryfatintakeandtheriskofincidentdementiaintheRotterdamStudy.
AnnalsofNeurology,42(5),776782.
Kamei,M.,Ki,M.,Kawagoshi,M.,&Kawai,N.(2002).NutritionalevaluationofJapanese
takeoutlunchescomparedwithwesternstylefastfoodssuppliedinJapan.
JournalofFoodCompositionandAnalysis,15(1),3545.
Kant,A.K.,&Graubard,B.I.(2004).EatingoutinAmerica,19872000:trendsand
nutritionalcorrelates.PreventiveMedicine,38(2),243249.
Keith,T.Z.(2006).Multipleregressionandbeyond.Boston:PearsonEducationInc.
Kellett,E.,Smith,A.,&Schmerlaib,Y.(1998).TheAustralianguidetohealthyeating.
Canberra:CommonwealthDepartmentofHealthandFamilyServices.
Kelly,B.,King,L.,Bauman,A.,Smith,B.J.,&Flood,V.(2007).Theeffectsofdifferent
regulationsystemsontelevisionfoodadvertisingtochildren.AustralianandNew
ZealandJournalofPublicHealth,31(4),340343.
Kim,D.M.,Ahn,C.W.,&Nam,S.Y.(2005).PrevalenceofobesityinKorea.Obesity
Reviews,6,117121.
Klaczynski,P.A.,Goold,K.W.,&Mudry,J.J.(2004).Culture,obesitystereotypes,self
esteem,andthe"thinideal":Asocialidentityperspective.JournalofYouthand
Adolescence,33(4),307317.
259
Klauer,K.C.,Herfordt,J.,&Voss,A.(2008).Socialpresenceeffectsonthestrooptask:
Boundaryconditionsandanalternativeaccount.JournalofExperimentalSocial
Psychology,44(2),469476.
Klesges,R.C.,Bartsch,D.,Norwood,J.D.,Kautzman,D.,&Haugrud,S.(1984).Theeffects
ofselectedsocialandenvironmentalvariablesontheeatingbehaviourofadults
inthenaturalenvironment.InternationalJournalofEatingDisorders,3(4),3541.
Klesges,R.C.,Shelton,M.L.,&Klesges,L.M.(1993).Effectsoftelevisiononmetabolic
ratepotentialimplicationsforchildhoodobesity.Pediatrics,91(2),281286.
Kline,R.B.(2005).Principlesandpracticeofstructuralequationmodeling(2nded.).New
York:TheGuilfordPress.
Kubik,M.Y.,Lytle,L.,&Fulkerson,J.A.(2005).Fruits,vegetables,andfootball:Findings
fromfocusgroupswithalternativehighschoolstudentsregardingeatingand
physicalactivity.JournalofAdolescentHealth,36(6),494500.
Larter,P.(2008,June20).AustraliabeatsUStotitleofmostobesenation,reportfinds.
TimesOnline,from
<http://www.timesonline.co.uk/tol/news/world/article4171160.ece>.
Latner,J.D.&Stunkard,A.J.(2003).GettingWorse:TheStigmatizationofObeseChildren.
ObesityResearch,11,452456.
Leary,M.R.,Nezlek,J.B.,Downs,D.,Radforddavenport,J.,Martin,J.,&Mcmullen,A.
(1994).Selfpresentationineverydayinteractionseffectsoftargetfamiliarity
andgendercomposition.JournalofPersonalityandSocialPsychology,67(4),664
673.
Leatherdale,S.T.,Brown,K.S.,Cameron,R.,&McDonald,P.W.(2005).Socialmodeling
intheschoolenvironment,studentcharacteristics,andsmokingsusceptibility:A
multilevelanalysis.JournalofAdolescentHealth,37(4),330336.
260
Leone,T.,Herman,C.P.,&Pliner,P.(2008).Perceptionsofundereaters:Amatterof
perspective?PersonalityandSocialPsychologyBulletin,34(12),17371746.
Leone,T.,Pliner,P.,&Herman,C.P.(2007).Influenceofclearversusambiguous
normativeinformationonfoodintake.Appetite,49(1),5865.
Levin,B.E.(2004).Thedrivetoregainismainlyinthebrain.AmericanJournalof
PhysiologyRegulatoryIntegrativeandComparativePhysiology,287(6),R1297
R1300.
Levin,B.E.(2007).Whysomeofusgetfatandwhatwecandoaboutit.Journalof
PhysiologyLondon,583(2),425430.
Levin,B.E.,&DunnMeynell,A.A.(2000).Defenseofbodyweightagainstchroniccaloric
restrictioninobesityproneandresistantrats.AmericanJournalofPhysiology
RegulatoryIntegrativeandComparativePhysiology,278(1),R231R237.
Lichtenstein,A.H.,Erkkila,A.T.,Lamarche,B.,Schwab,U.S.,Jalbert,S.M.,&Ausman,
L.M.(2003).InfluenceofhydrogenatedfatandbutteronCVDriskfactors:
remnantlikeparticles,glucoseandinsulin,bloodpressureandCreactive
protein.Atherosclerosis,171(1),97107.
Lin,B.H.,&Frazao,E.(1999).Awayfromhomefoodsincreasinglyimportanttoqualityof
Americandiet.(AgricultureInformationBulletinNo.749).WashingtonDC:USDA.
Livingstone,B.(2000).EpidemiologyofchildhoodobesityinEurope.EuropeanJournal
ofPediatrics,159,S14S34.
Lobera,I.J.,Fernandez,M.R.,Gonzalez,M.T.M.,&Millan,M.T.M.(2008).Theinfluence
ofstereotypesonobesityperception.NutricionHospitalaria,23(4),319325.
Ludwig,D.S.,Peterson,K.E.,&Gortmaker,S.L.(2001).Relationbetweenconsumption
ofsugarsweeteneddrinksandchildhoodobesity:aprospective,observational
analysis.Lancet,357(9255),505508.
261
Lumeng,J.C.,&Hillman,K.H.(2007).Eatinginlargergroupsincreasesfood
consumption.ArchivesofDiseaseinChildhood,92(5),384387.
MacKinnon,D.P.,Krull,J.L.,&Lockwood,C.M.(2000).Equivalenceofthemediation,
confoundingandsuppressioneffect.PreventionScience,1(4),173182.
Maffeis,C.,Talamini,G.,&Tato,L.(1998).Influenceofdiet,physicalactivityandparents'
obesityonchildren'sadiposity:afouryearlongitudinalstudy.International
JournalofObesity,22(8),758764.
Malouf,N.M.,&Colagiuri,S.(1995).TheeffectsofMcDonalds,KentuckyFriedChicken
andPizzaHutmealsonrecommendeddiets.AsiaPacificJournalofClinical
Nutrition,4,265269.
Marshall,D.(1995).Eatingathome:Mealsandfoodchoice.InD.Marshall(Ed.),Food
choiceandtheconsumer(pp.264291).London:BlackieAcademic&
Professional.
Marshall,D.,O'Donohoe,S.,&Kline,S.(2007).Families,food,andpesterpower:Beyond
theblamegame?JournalofConsumerBehaviour,6,164181.
Martin,Y.,Pliner,P.,&Lee,C.(2004).Theeffectsofmealsizeandbodysizeon
individuals'impressionsofmalesandfemales.EatingBehaviors,5,117132.
Mathey,M.F.A.M.,Zandstra,E.H.,deGraaf,C.,&vanStaveren,W.A.(2000).Socialand
physiologicalfactorsaffectingfoodintakeinelderlysubjects:anexperimental
comparativestudy.FoodQualityandPreference,11(5),397403.
Mathy,R.M.,Kerr,D.L.,&Haydin,B.M.(2008).Methodologicalrigorandethical
considerationsinInternetmediatedresearch.Psychotherapy:Theory,Research,
Practice,Training,40(12),7785.
262
Mathy,R.M.,Schillace,M.,Coleman,S.M.,&Berquist,B.E.(2002).Methodologicalrigor
withInternetsamples:Newwaystoreachunderrepresentedpopulations.
CyberPsychology&Behavior,5(3),253266.
Mauer,M.M.,Harris,R.B.S.,&Bartness,T.J.(2001).Theregulationoftotalbodyfat:
lessonslearnedfromlipectomystudies.NeuroscienceandBiobehavioralReviews,
25(1),1528.
McCabe,M.P.,&Ricciardelli,L.A.(2001a).Parent,peer,andmediainfluencesonbody
imageandstrategiestobothincreaseanddecreasebodysizeamongadolescent
boysandgirls.Adolescence,36(142),225240.
McCabe,M.P.,&Ricciardelli,L.A.(2001b).Thestructureoftheperceivedsociocultural
influencesonbodyimageandbodychangequestionnaire.InternationalJournalof
BehavioralMedicine,8(1),1941.
McCabe,M.P.,&Ricciardelli,L.A.(2003).Socioculturalinfluencesonbodyimageand
bodychangesamongadolescentboysandgirls.JournalofSocialPsychology,
143(1),526.
McDonald'sCorporation(2006).Nutritioninformation[download].RetrievedJune2,
2006,from<http://mcdonalds.com.au/PDFs/Master_Nutrition_Information.pdf>
McDonald'sCorporation(2007).McDonald'sNorge[website].RetrievedAugust15,
2007,from<http://www.mcdonalds.no/Meny?prev=0>
McIntosh,W.A.,Fletcher,R.D.,Kubena,K.S.,&Landmann,W.A.(1995).Factors
associatedwithsourcesofinfluenceinformationinreducingredmeatbyelderly
subjects.Appetite,24(3),219230.
McLean,T.(2008a,June19).Australiaisworld'sfattestnation.TheAustralian,from
<http://www.theaustralian.news.com.au/story/0,25197,23890071
23812377,23890000.html>.
263
McLean,T.(2008b,September15).Preschoolerswantidealbodystudy.TheAdvertiser,
p.28.
Meiselman,H.L.(1992).Obstaclestostudyingrealpeopleeatingrealmealsinreal
situationsresponse.Appetite,19(1),8486.
Milewicz,A.,Jedrzejuk,D.,Lwow,F.,Bialynicka,A.S.,Lopatynski,J.,Mardarowicz,G.&
ZahorskaMarkiewicz,B.(2005).PrevalenceofobesityinPoland.Obesity
Reviews,6,113114.
Miller,K.E.,&Ginter,J.L.(1979).Aninvestigationofsituationalvariationinbrand
choicebehaviorandattitude.JournalofMarketingResearch,16(1),112123.
MinistryofHealth(2004).Trackingtheobesityepidemic:NewZealand19772003.
Wellington:MinistryofHealth.
Mohr,P.B.,Wilson,C.J.,Dunn,K.I.,Brindal,E.A.,&Wittert,G.A.(2007).Personaland
lifestylecharacteristicspredictiveoftheconsumptionoffastfoodsinAustralia.
PublicHealthNutrition,10(12),14561463.
Mooney,K.M.,Detore,J.,&Malloy,K.A.(1994).Perceptionsofwomenrelatedtofood
choice.SexRoles,31(78),433442.
Mooney,K.M.,&Lorenz,E.(1997).Theeffectsoffoodandgenderoninterpersonal
perceptions.SexRoles,36(910),639653.
Moray,J.,Fu,A.,Brill,K.,&Mayoral,M.S.(2007).Viewingtelevisionwhileeatingimpairs
theabilitytoaccuratelyestimatetotalamountoffoodconsumed.Bariatric
NursingandSurgicalPatientCare,2(1),7176.
Mori,D.,Pliner,P.,&Chaiken,S.(1987).Eatinglightlyandtheselfpresentationof
femininity.JournalofPersonalityandSocialPsychology,53(4),693702.

264
Morley,B.,Chapman,K.,Mehta,K.,King,L.,Swinburn,B.,&Wakefield,M.(2008).
Parentalawarenessandattitudesaboutfoodadvertisingtochildrenon
Australiantelevision.AustralianandNewZealandJournalofPublicHealth,32(4),
341347.
Motl,R.,McAuley,E.,Birnbaum,A.S.,&Lytle,L.A.(2004).Relationshipsamong
longitudinalchangesintelevisionwatching,videogameplaying,andphysical
activityduringadolescence.MedicineandScienceinSportsandExercise,36(5),
S298S298.
NationalHealthandMedicalResearchCouncil(2006).NutrientReferenceValuesfor
AustraliaandNewZealandIncludingRecommendedDietaryIntakes.Australia:
DepartmentofHealthyandAging.
NationalPreventativeHealthTaskforce(2008).Australia:thehealthiestcountryby2020
(No.2.Technicalreport).Canberra:CommonwealthofAustralia.
NeumarkSztainer,D.,Hannan,P.J.,Story,M.,Croll,J.,&Perry,C.(2003).Familymeal
patterns:Associationswithsociodemographiccharacteristicsandimproved
dietaryintakeamongadolescents.JournaloftheAmericanDieteticAssociation,
103(3),317322.
Nisbett,R.E.,&Storms,M.D.(1974).Cognitiveandsocialdeterminantsoffoodintake.In
H.London&R.E.Nisbett(Eds.),ThoughtandFeeling:Cognitivealterationsof
feelingsstates(pp.190208).Chicago:Aldine.
NSWCentreforPublicHealthNutrition,&NSWDepartmentofHealth(2003).NSW
reportontheweightstatusofNSW:2003.Sydney:NSWCentreforPublicHealth
Nutrition,NSWDepartmentofHealth.
OBrien,K.,&Webbie,K.(2001).Health,wellbeingandbodyweight:characteristicsof
overweightandobesityinAustralia.Canberra:AIHW.
265
Ohbuchi,K.O.,Tamura,T.,Quigley,B.M.,Tedeschi,J.T.,Madi,N.,Bond,M.H.&
Mummendey,A.(2004).Anger,blame,anddimensionsofperceivednorm
violations:Culture,gender,andrelationships.JournalofAppliedSocial
Psychology,34(8),15871603.
Olds,T.S.,&Harten,N.R.(2001).Onehundredyearsofgrowth:Theevolutionofheight,
mass,andbodycompositioninAustralianchildren,18991999.HumanBiology,
73(5),727738.
Paeratakul,S.,Ferdinand,D.P.,Champagne,C.M.,Ryan,D.H.,&Bray,G.A.(2003).Fast
foodconsumptionamongUSadultsandchildren:Dietaryandnutrientintake
profile.JournaloftheAmericanDieteticAssociation,103(10),13321338.
Park,C.(2004).Efficientorenjoyable?Consumervaluesofeatingoutandfastfood
restaurantconsumptioninKorea.InternationalJournalofHospitality
Management,23(1),8794.
Pedhazur,E.J.,&Schmelkin,L.P.(1991).Measurement,design,andanalysis:An
intergratedapproach.London:LawrenceErbaumAssociates.
Pereira,M.A.,Kartashov,A.I.,Ebbeling,C.B.,VanHorn,L.,Slattery,M.,Jacobs,D.R.&
Ludwig,D.S.(2005).Fastfoodhabits,weightgain,andinsulinresistance(the
CARDIAstudy):15yearprospectiveanalysis.Lancet,365(9453),3642.
Perry,J.A.,Silvera,D.H.,Rosenvinge,J.H.,Neilands,T.,&Holte,A.(2001).Seasonal
eatingpatternsinNorway:Anonclinicalpopulationstudy.ScandinavianJournal
ofPsychology,42(4),307312.
Petrini,C.(2003).SlowFood:thecasefortaste(W.McCuaig,Trans.).NewYork:
ColumbiaUniversityPress.
266
Petty,R.E.,&Cacioppo,J.T.(1986).Theelaborationlikelihoodmodelofpersuasion.InL.
Berkowitz(Ed.),Advancesinexperimentalsocialpsychology(Vol.19,pp.123
205).NewYork:AcademicPress.
Pfeffer,J.(1985).Organizationsandorganizationtheory.InG.Lindzey&E.Aronson
(Eds.),HandbookofSocialPsychology(3rded.ed.,Vol.1,pp.379441).New
York:RandomHouse.
Pliner,P.,Bell,R.,Hirsch,E.S.,&Kinchla,M.(2006).Mealdurationmediatestheeffectof
"socialfacilitation"oneatinginhumans.Appetite,46(2),189198.
Pliner,P.,&Chaiken,S.(1990).Eating,socialmotivesandselfpresentationinwomen
andmen.JournalofExperimentalSocialPsychology,26,204254.
Pliner,P.,Chaiken,S.,&Flett,G.L.(1990).Genderdifferencesinconcernwithbody
weightandphysicalappearanceoverthelifespan.PersonalityandSocial
PsychologyBulletin,16(2),263273.
Pliner,P.,&Mann,N.(2004).Influenceofsocialnormsandpalatabilityonamount
consumedandfoodchoice.Appetite,42(2),227237.
Poirier,P.,Giles,T.D.,Bray,G.A.,Hong,Y.,Stern,J.S.,PiSunyer,F.X.,&Eckel,R.H.
(2006).Obesityandcardiovasculardisease:pathophysiology,evaluation,and
effectofweightloss.Arteriosclerosis,Thrombosis,andVascularBiology,26(5),
968976.
Popkin,B.M.(2001).Thenutritiontransitionandobesityinthedevelopingworld.
JournalofNutrition,131(3),871s873s.
Popkin,B.M.,&Nielsen,S.J.(2003).Thesweeteningoftheworld'sdiet.Obesity
Research,11(11),13251332.
Popkin,B.M.,SiegaRiz,A.M.,Haines,P.S.,&Jahns,L.(2001).Where'sthefat?Trendsin
USdiets19651996.PreventiveMedicine,32,245.
267
Preacher,K.J.,&Hayes,A.F.(2004).TestingofmediationmodelsinSPSSandSAS.
BehaviorResearchMethods,Instruments,andComputers,36,717731.
Preacher,K.J.,&Leonardelli,G.J.(2001,08/10/2006).CalculationfortheSobelTest.
RetrievedNovember1,2008,from
<http://people.ku.edu/~preacher/sobel/sobel.htm>
Prentice,A.M.,&Jebb,S.A.(2003).Fastfoods,energydensityandobesity:apossible
mechanisticlink.ObesityReviews,4,187194.
Prescott,J.,Young,O.,O'Neill,L.,Yau,N.J.N.,&Stevens,R.(2002).Motivesforfood
choice:acomparisonofconsumersfromJapan,Taiwan,MalaysiaandNew
Zealand.FoodQualityandPreference,13(78),489495.
PublicHealthServiceNationalInstitutesofHealthNationalHeartLungandBlood
Institute(1998).Clinicalguidelinesontheidentification,evaluation,andtreatment
ofoverweightandobesityinadults:Theevidencereport(No.984083;NIH
Publication)from<http://www.nhlbi.nih.gov/guidelines/obesity/ob_gdlns.pdf>:
U.S.DepartmentofHealthandHumanServices.
Puhl,R.&Brownell,K.D.(2001).Obesity,biasanddiscrimination.ObesityResearch,9,
788805.
PuhlR.&Brownell,K.D.(2003).Psychosocialoriginsofobesitystigma:towardchanging
apowerfulandpervasivebias,ObesityReviews,4,213227.
Raben,A.,Vasilaras,T.H.,Moller,A.C.,&Astrup,A.(2002).Sucrosecomparedwith
artificialsweeteners:differenteffectsonadlibitumfoodintakeandbodyweight
after10wkofsupplementationinoverweightsubjects.AmericanJournalof
ClinicalNutrition,76(4),721729.
268
Rains,S.A.,&Turner,M.M.(2007).Psychologicalreactanceandpersuasivehealth
communication:Atestandextensionoftheintertwinedmodel.Human
CommunicationResearch,33(2),241269.
Redd,M.,&deCastro,J.M.(1992).Socialfacilitationofeatingeffectsofsocial
instructiononfoodintake.Physiology&Behavior,52(4),749754.
Reidpath,D.D.,Burns,C.,Garrard,J.,Mahoney,M.,&Townsend,M.(2002).Anecological
studyoftherelationshipbetweensocialandenvironmentaldeterminantsof
obesity.Health&Place,8(2),141145.
Rennie,K.L.,&Jebb,S.A.(2005).PrevalenceofobesityinGreatBritain.ObesityReviews,
6,1112.
Ricciardelli,L.A.,&McCabe,M.P.(2001).Children'sbodyimageconcernsandeating
disturbance:Areviewoftheliterature.ClinicalPsychologyReview,21(3),325
344.
Rice,S.,McAllister,E.J.,&Dhurandhar,N.V.(2007).Fastfood:friendly?International
JournalofObesity,31(6),884886.
Rodriguez,G.,&Moreno,L.A.(2008).Isdietthefuelforobesityinchildrenand
adolescents?ObesityandMetabolismMilan,4(3),183188.
Roefs,A.,&Jansen,A.(2002).Implicitandexplicitattitudestowardhighfatfoodsin
obesity.JournalofAbnormalPsychology,111(3),517521.
Roehling,M.V.(1999).Weightbaseddiscriminationinemployment:psychologicaland
legalaspects.PersonnelPsychology,52,9691017.
Rosenheck,R.(2008).Fastfoodconsumptionandincreasedcaloricintake:asystematic
reviewofatrajectorytowardsweightgainandobesityrisk.ObesityReviews,9(6),
535547.
269
Rosenthal,B.,&McSweeney,F.K.(1979).Modelinginfluencesonbehavior.Addictive
Behaviors,4,205214.
Roth,D.A.,Herman,C.P.,Polivy,J.,&Pliner,P.(2001).Selfpresentationalconflictin
socialeatingsituations:anormativeperspective.Appetite,36(2),165171.
RoyMorgan(2006).MajorityofAustralianshaveboughtfastfoodinlastfourweeks(No.
555).Melbourne:RoyMorganResearch.
RoyalAgricultural&HorticulturalSocietyofSA(2006).RoyalAdelaideShow
[homepage].RetrievedNovember17,2006,from
<http://www.adelaideshowground.com.au/showground/royaladelaide
show/royaladelaideshowhome.jsp)>
Rozin,P.,Ashmore,M.,&Markwith,M.(1996).LayAmericanconceptionsofnutrition:
Doseinsensitivity,categoricalthinking,contagion,andthemonotonicmind.
HealthPsychology,15(6),438447.
Rozin,P.,Kabnick,K.,Pete,E.,Fischler,C.,&Shields,C.(2003).Theecologyofeating:
smallerportionsizesinFranceThanintheUnitedStateshelpexplaintheFrench
paradox.PsychologicalScience,14(5),450454.
Rudman,L.A.(1998).Selfpromotionasariskfactorforwomen:Thecostsandbenefits
ofcounterstereotypicalimpressionmanagement.JournalofPersonalityandSocial
Psychology,74(3),629645.
Sadalla,E.,&Burroughs,J.(1981).Profilesineatingsexyvegetariansandotherdiet
basedsocialstereotypes.PsychologyToday,15(10),5155.
Salvy,S.J.,Jarrin,D.,Paluch,R.,Irfan,N.,&Pliner,P.(2007).Effectsofsocialinfluenceon
eatingincouples,friendsandstrangers.Appetite,49(1),9299.

270
Satia,J.A.,Galanko,J.A.,&SiegaRiz,A.M.(2004).Eatingatfastfoodrestaurantsis
associatedwithdietaryintake,demographic,psychosocialandbehavioural
factorsamongAfricanAmericansinNorthCarolina.PublicHealthNutrition,7(8),
10891096.
Savige,G.S.,Ball,K.,Worsley,A.,&Crawford,D.(2007).Foodintakepatternsamong
Australianadolescents.AsiaPacificJournalofClinicalNutrition,16(4),738747.
Schachter,S.(1968).Obesityandeating.Science,161(3843),751756.
Schachter,S.(1971).Someextraordinaryfactsaboutobesehumansandrats.American
Psychologist,26,129144.
Scheen,A.J.(2002).Resultsofobesitytreatment.AnnalesDEndocrinologie,63(2),163
170.
Scheibehenne,B.,Miesler,L.,&Todd,P.M.(2007).Fastandfrugalfoodchoices:
Uncoveringindividualdecisionheuristics.Appetite,49(3),578589.
Schenkler,B.R.(1980).Impressionmanagement:theselfconcept,socialidentity,and
interpersonalrelations.Monterey,California:Brooks/ColePub.Co.
Schillewaert,N.,&Meulemeester,P.(2005).Comparingresponsedistributionsofoffline
andonlinedatacollectionmethods.InternationalJournalofMarketResearch,
47(2),163178.
Schmidt,W.C.(1997).WorldWideWebsurveyresearch:Benefits,potentialproblems,
andsolutions.BehaviorResearchMethodsInstruments&Computers,29(2),274
279.
Schroder,H.,Fito,M.,Covas,M.I.,&Investigators(2007).Associationoffastfood
consumptionwithenergyintake,diet.quality,bodymassindexandtheriskof
obesityinarepresentativeMediterraneanpopulation.BritishJournalofNutrition,
98(6),12741280.
271
Schroder,M.J.A.,&McEachern,M.G.(2005).Fastfoodsandethicalconsumervalue:a
focusonMcDonald'sandKFC.BritishFoodJournal,107(45),212224.
Seidell,J.C.(2000).Obesity,insulinresistanceanddiabetesaworldwideepidemic.
BritishJournalofNutrition,83,S5S8.
Seidell,J.C.,&Flegal,K.M.(1997).Assessingobesity:Classificationandepidemiology.
BritishMedicalBulletin,53(2),238252.
Shah,A.K.,&Oppenheimer,D.M.(2008).Heuristicsmadeeasy:Aneffortreduction
framework.PsychologicalBulletin,134(2),207222.
Shai,I.,Schwarzfuchs,D.,Henkin,Y.,Shahar,D.R.,Witkow,S.,Greenberg,I.,Golan,R.,
Fraser,D.,Bolotin,A.,Vardi,H.,TangiRozental,O.,ZukRamot,R.,Sarusi,B.,
Brickner,D.,Schwartz,Z.,Sheiner,E.,Marko,R.,Katorza,E.,Thiery,J.,Fiedler,
G.M.,Blher,M.,StumvollM.,Stampfer,M.J.&DietaryInterventionRandomized
ControlledTrial(DIRECT)Group(2008).Weightlosswithalowcarbohydrate,
Mediterranean,orlowfatdiet.NewEnglandJournalofMedicine,359(3),229241.
Shankar,K.,Harrell,A.,Gilchrist,J.M.,Ronis,M.J.J.,&Badger,T.M.(2007).Maternal
obesityatconceptionprogramsobesityintheoffspring.EarlyHuman
Development,83,S164S164.
Sheehan,K.B.,&Hoy,M.G.(1999).UsingemailtosurveyInternetusersintheUnited
States:Methodologyandassessment.JournalofComputerMediated
Communication,4(3),from
<http://www.ascusc.org/jcmc/vol4/issue3/sheehan.html>.
Shieh,G.(2006).Suppressionsituationsinmultiplelinearregression.Educationaland
PsychologicalMeasurement,66(3),435447.
Sibbald,B.(2002).ObesitymaysoonbeleadingcauseofpreventabledeathinUS.
CanadianMedicalAssociationJournal,166(5),642642.
272
Simmons,D.,McKenzie,A.,Eaton,S.,Cox,N.,Khan,M.A.,Shaw,J.&Zimmet,P.(2005).
Choiceandavailabilityoftakeawayandrestaurantfoodisnotrelatedtothe
prevalenceofadultobesityinruralcommunitiesinAustralia.International
JournalofObesity,29,703710.
Siwik,V.P.,&Senf,J.H.(2006).Foodcravings,ethnicityandotherfactorsrelatedto
eatingout.JournaloftheAmericanCollegeofNutrition,25(5),382388.
SlowFood(2008).OurPhilosophy.RetrievedAugust13,2006,from
<http://www.slowfood.com/about_us/eng/philosophy.lasso>
Smith,B.(2006,November1).OilchangeasMcDonald'smovestocutfats.TheAge,from
<http://www.theage.com.au/news/national/anoilchangefor
mcdonalds/2006/2010/2031/1162278145195.html>.
Smith,B.,Smith,T.C.,Gray,G.C.,Ryan,M.A.K.,&Team,M.C.S.(2007).When
epidemiologymeetstheInternet:Webbasedsurveysinthemillenniumcohort
study.AmericanJournalofEpidemiology,166(11),13451354.
Smith,R.L.,Ager,J.W.,&Williams,P.L.(2006).Suppressionsituationsinmultiplelinear
regression.EducationalandPsychologicalMeasurement,52,1730.
Snoek,H.M.,vanStrien,T.,Janssens,J.M.A.M.,&Engels,R.C.M.E.(2006).Theeffectof
televisionviewingonadolescents'snacking:Individualdifferencesexplainedby
external,restrainedandemotionaleating.JournalofAdolescentHealth,39(3),
448451.
Sobel,M.E.(1982).Asymptoticintervalsforindirecteffectsinstructuralequationsmodels.
InS.Leinhart(Ed.),Sociologicalmethodology(pp.290312).SanFrancisco:
JosseyBass.
SocialDevelopmentCommittee(2007).Fastfoodsandobesityinquiry.Adelaide:The
SouthAustralianParliament.
273
Sommer,R.,&Steele,J.(1997).Socialeffectsondurationinrestaurants.Appetite,29(1),
2530.
Sommer,R.,Wynes,M.,&Brinkley,G.(1992).Socialfacilitationeffectsinshopping
behavior.EnvironmentandBehavior,24(3),285297.
Spencer,E.H.,Frank,E.,&McIntosh,N.F.(2005).Potentialeffectsofthenext100billion
hamburgerssoldbyMcDonald's.AmericanJournalofPreventiveMedicine,28(4),
379381.
Stanton,R.J.(2006).Nutritionproblemsinanobesogenicenvironment.MedicalJournal
ofAustralia,184(2),7679.
StatisticsNorway(2009a).FiguresonOsloMunicipality.RetrievedMarch12,2009,from
<http://www.ssb.no/english/municipalities/0301>
StatisticsNorway(2009b).ThisisNorway:Useoftime.RetrievedMarch9,2009,from
<http://www.ssb.no/vis/english/subjects/00/norge_en/tidsbruk_en/main.html>
Stein,R.I.,&Nemeroff,C.J.(1995).Moralovertonesoffoodjudgmentsofothersbased
onwhattheyeat.PersonalityandSocialPsychologyBulletin,21(5),480490.
Stender,S.,Dyerberg,J.,&Astrup,A.(2007).Fastfood:unfriendlyandunhealthy.
InternationalJournalofObesity,31(6),887890.
Steptoe,A.,Pollard,T.M.,&Wardle,J.(1995).Developmentofameasureofthemotives
underlyingtheselectionoffoodtheFoodChoiceQuestionnaire.Appetite,25(3),
267284.
Stewart,S.,Tikellis,G.,Carrington,C.,Walker,K.,&O'Dea,K.(2008).Australiasfuture
fatbomb:AreportonthelongtermconsequencesofAustraliasexpanding
waistlineoncardiovasculardisease.Melbourne,Australia:BakerHeartResearch
Institute.
274
Stroebele,N.,&deCastro,J.(2003).Televisionviewinginfluencespeople'sfoodintake.
ObesityResearch,11,A78A78.
Stroebele,N.,&deCastro,J.M.(2004).Effectofambienceonfoodintakeandfood
choice.Nutrition,20(9),821838.
Stroebele,N.,&deCastro,J.M.(2006).Influenceofphysiologicalandsubjectivearousal
onfoodintakeinhumans.Nutrition,22(10),9961004.
Swinburn,B.A.,Caterson,I.,Seidell,J.C.,&James,W.P.T.(2004).Diet,nutritionandthe
preventionofexcessweightgainandobesity.PublicHealthNutrition,7(1),123
146.
Swinburn,B.A.,Egger,G.,&Raza,F.(1999).Dissectingobesogenicenvironments:The
developmentandapplicationofaframeworkforidentifyingandprioritizing
environmentalinterventionsforobesity.PreventiveMedicine,29(6),563570.
Tabachnick,B.,&Fidell,L.(1989).Usingmultivariatestatistics(2nded.).NewYork:
HarperCollinsPublishers.
Tay,J.,Brinkworth,G.D.,Noakes,M.,Keogh,J.,&Clifton,P.M.(2008).Metaboliceffects
ofweightlossonaverylowcarbohydratedietcomparedwithanisocalorichigh
carbohydratedietinabdominallyobesesubjects.JournaloftheAmericanCollege
ofCardiology,51(1),5967.
Tice,D.M.,Butler,J.L.,Muraven,M.B.,&Stillwell,A.M.(1995).Whenmodestyprevails
differentialfavorabilityofselfpresentationtofriendsandstrangers.Journalof
PersonalityandSocialPsychology,69(6),11201138.
Tomiyama,A.J.,Moskovich,A.,ByrneHaltom,K.,Ju,T.,&Mann,T.(2009).Consumption
afteradietviolation:Disinhibitionorcompensation?PsychologicalScience,
20(10),12751281.
275
Triandis,H.C.(1989).Theselfandsocialbehaviorindifferingculturalcontexts.
PsychologicalReview,96(3),506520.
U.S.DepartmentofAgriculture(1997).Designandoperation:TheContinuingSurveyof
FoodIntakesbyIndividualsandtheDietandHealthKnowledgeSurvey,19941996.
(No.961).Washington:AgricultureResearchService.
U.S.DepartmentofHealthandHumanServices(2008).WeightCycling[NIHPublication
No.013901].RetrievedApril,2008,from
<http://win.niddk.nih.gov/publications/cycling.htm>
Unger,J.B.,Reynolds,K.,Shakib,S.,SpruijtMetz,D.,Suit,P.,&Johnson,C.A.(2004).
Acculturation,physicalactivity,andfastfoodconsumptionamongAsian
AmericanandHispanicadolescents.JournalofCommunityHealth,29(6),467481.
Utter,J.,NeumarkSztainer,D.,Jeffery,R.,&Story,M.(2003).Couchpotatoesorfrench
fries:Aresedentarybehaviorsassociatedwithbodymassindex,physicalactivity,
anddietarybehaviorsamongadolescents?JournaloftheAmericanDietetic
Association,103(10),12981305.
Uziel,L.(2007).Individualdifferencesinthesocialfacilitationeffect:Areviewandmeta
analysis.JournalofResearchinPersonality,41(3),579601.
VandenBulck,J.,&VanMierlo,J.(2004).Energyintakeassociatedwithtelevision
viewinginadolescents,acrosssectionalstudy.Appetite,43(2),181184.
Vartanian,L.R.,Herman,C.P.,&Polivy,J.(2007).Consumptionstereotypesand
impressionmanagement:Howyouarewhatyoueat.Appetite,48(3),265277.
Vartanian,L.R.,Herman,C.P.,&Wansink,B.(2008).Areweawareoftheexternal
factorsthatinfluenceourfoodintake?HealthPsychology,27(5),533538.
276
Verlegh,P.W.J.,&Candel,M.J.J.M.(1999).Theconsumptionofconveniencefoods:
referencegroupsandeatingsituations.FoodQualityandPreference,10(6),457
464.
Vermeer,W.M.,Steenhuis,I.H.M.,&Seidell,J.C.(2009).Fromthepointofpurchase
perspective:Aqualitativestudyofthefeasibilityofinterventionsaimedat
portionsize.HealthPolicy,90(1),7380.
Wadden,T.A.,Brownwell,K.D.,&Foster,G.D.(2002).Obesity:respondingtotheglobal
epidemic.JournalofConsultingandClinicalPsychology,70(3),510525.
Wadden,T.A.,&Stunkard,A.J.(1985).Socialandpsychologicalconsequencesof
obesity.AnnalsofInternationalMedicine,103(2),10621067.
Wadden,T.A.,Vogt,R.A.,Kuehnel,R.H.,Andersen,R.E.,Bartlett,S.J.,Foster,G.D.,
Kuehnel,R.H.,Wilk,J.,Weinstock,R.,Buckenmeyer,P.,Berkowitz,R.I.&Steen,S.
N.(1997).Exerciseinthetreatmentofobesity:Effectsoffourinterventionson
bodycomposition,restingenergyexpenditure,appetite,andmood.Journalof
ConsultingandClinicalPsychology,65(2),269277.
Wansink,B.,Painter,J.E.,&North,J.(2005).Bottomlessbowls:Whyvisualcuesof
portionsizemayinfluenceintake.ObesityResearch,13(1),93100.
Wansink,B.,&Park,S.B.(2001).Atthemovies:howexternalcuesandperceivedtaste
impactconsumptionvolume.FoodQualityandPreference,12(1),6974.
Wayne,S.J.,&Liden,R.C.(1995).Effectsofimpressionmanagementonperformance
ratingsalongitudinalstudy.AcademyofManagementJournal,38(1),232260.
Weber,A.J.,King,S.C.,&Meiselman,H.L.(2004).Effectsofsocialinteraction,physical
environmentandfoodchoicefreedomonconsumptioninamealtesting
environment.Appetite,42(1),115118.
277
Weinsier,R.L.,Hunter,G.R.,Heini,A.F.,Goran,M.I.,&Sell,S.M.(1998).Theetiologyof
obesity:Relativecontributionofmetabolicfactors,diet,andphysicalactivity.
AmericanJournalofMedicine,105(2),145150.
Wells,N.M.,Ashdown,S.P.,Davies,E.H.S.,Cowett,F.D.,&Yang,Y.Z.(2007).
Environment,designandobesityOpportunitiesforinterdisciplinary
collaborativeresearch.EnvironmentandBehavior,39(1),633.
Westphal,J.D.,&Stern,I.(2007).Flatterywillgetyoueverywhere(especiallyifyouare
amalecaucasian):Howingratiation,boardroombehavior,anddemographic
minoritystatusaffectadditionalboardappointmentsatUScompanies.Academy
ofManagementJournal,50(2),267288.
Whitaker,R.C.,Wright,J.A.,Pepe,M.S.,Seidel,K.D.,&Dietz,W.H.(1997).Predicting
obesityinyoungadulthoodfromchildhoodandparentalobesity.NewEngland
JournalofMedicine,337(13),869873.
White,M.(2007).Foodaccessandobesity.ObesityReviews,8(supp.1),99107.
Wiggins,J.A.,Wiggins,B.B.,&VanderZanden,J.(1994).SocialPsychology(5thed.).New
York:McGrawHill,Inc.
Wing,R.R.(1999).Physicalactivityinthetreatmentoftheadulthoodoverweightand
obesity:currentevidenceandresearchissues.MedicineandScienceinSportsand
Exercise,31(11),S547S552.
Wood,M.D.,Read,J.P.,Mitchell,R.E.,&Brand,N.H.(2004).Doparentsstillmatter?
Parentandpeerinfluencesonalcoholinvolvementamongrecenthighschool
graduates.PsychologyofAddictiveBehaviors,18(1),1930.
Wood,W.,Christensen,P.N.,Hebl,M.R.,&Rothgerber,H.(1997).Conformitytosex
typednorms,affect,andtheselfconcept.JournalofPersonalityandSocial
Psychology,73(3),523535.
278
WorldHealthOrganisation(1995).Physicalstatus:theuseandinterpretationof
anthropometry(No.TechnicalReportSeriesNo.854.).Geneva:WorldHealth
Organization.
WorldHealthOrganisation,&InternationalAssociationfortheStudyofObesity
InternationalObesityTaskForce(2000).TheAsiaPacificperspective:Redefining
obesityanditstreatment.Sydney:HealthCommunications.
YackobovitchGavan,M.,Nagelberg,N.,Demol,S.,Phillip,M.,&Shalitin,S.(2008).
Influenceofweightlossdietswithdifferentmacronutrientcompositionson
healthrelatedqualityoflifeinobeseyouth.Appetite,51(3),697703.
Yum!RestaurantsInternational(2007).KFCMenu.RetrievedDecember2,2007,from
<www.kfc.com.au/Default.asp?page=/menu>
Zajonc,R.B.(1965).Socialfacilitation.Science,149(3681),269274.
Zhang,Y.(1999).UsingtheInternetforsurveyresearch:Acasestudy.Journalof
EducationforLibraryandInformationScience,40(4),A9A9.
Zimmerman,R.S.,&Connor,C.(1989).HealthPromotioninContexttheEffectsof
SignificantOthersonHealthBehaviorChange.HealthEducationQuarterly,16(1),
5775.





Brindal, E., Mohr, P., Wilson, C. and Wittert, G. (2008) Obesity and the effects of
choice at a fast food restaurant.
Obesity Research and Clinical Practice, v.2 (2), pp. 111-117, July 2008





NOTE: This publication is included in the print copy of the thesis
held in the University of Adelaide Library.

It is also available online to authorised users at:

http://dx.doi.org/10.1016/j.orcp.2008.03.004

VIII
Appendix2:NutritionalinformationfollowingFastFoodSurvey



You ate ___ KJ in the fast food item/s you ate.
Energy
Most of the foods we eat and drink provide us with energy (otherwise referred to as kilojoules
[KJ] or calories [cal] 1 calorie=4.2kilojoules). The kilojoules in foods and drinks are supplied
through protein, carbohydrates (including sugar), alcohol and fat. Throughout the day we use
these kilojoules in all of the activities we do (even when we are sleeping). When we eat more
kilojoules than we use, we put weight on and vice versa.

The amount of kilojoules needed varies depending on things such as your size, age, height,
sex, activity levels and whether or not you are pregnant. The Australian Guide to Healthy
Eating (1998) provides some rough guidelines for how many kilojoules you may need. These
guidelines are provided below. It is important to remember that these values are generalised
and if you have further questions regarding your personal energy requirements you should
contact a health professional.

You may also want to visit the following website www.healthyactive.gov.au

Women aged 19-60
7200-11300KJ per day*
Men aged 19-60
9000-13700KJ per day*
*The lower energy values are applicable to older and smaller adults with low activity levels while the
higher values are based on larger, younger adults with a light physical activity level.

You ate ___g of protein in fast food item/s
Protein
Protein is provided in foods such as meat, fish, dairy products and legumes. For each gram of
protein we eat, we get 17KJ of energy. Protein is an important nutrient to the human body and
is involved with things such as muscle development and general body repair. There are many
diets that recommend different amounts of protein are needed in the diet. However, a way in
which you can work out a rough estimate of the minimum amount of protein you need is by
taking you body weight (in kilograms) and multiplying it by .75. Again, your protein needs will
vary on a number of factors including how active you are. But this calculation can be used as a
rough guide for the amount of protein you should eat.


IX

You ate __g of fat and ___g of saturated fat in the fast food item/s
Fat
Every gram of fat contains just over twice the kilojoules of protein. For each gram of fat eaten,
we get about 37KJ of energy. This is why fat is referred to as energy dense. There are many
different types of fat including saturated, polyunsaturated, monounsaturated and trans fats.
There has been a debate surrounding the effect that these different fats have on the body.
Research has suggested that eating a high amount of saturated and trans fats increases
cholesterol levels in the blood and the risk heart disease in some people. The Australian Guide
to Healthy Eating suggests that no more than 30% of the total amount of kilojoules you eat
should come from fat. It also suggests that no more than a third (33.3%) of the total fat you eat
should be saturated fat. So, if for example, you ate 9000KJ of energy, then the maximum
amount of fat you should eat per day is around 73g which means that there should be no more
than 24g of saturated fat.

You ate ___g of carbohydrate from the fast food item/s.
Carbohydrates
Carbohydrates (also referred to as carbs) are the main source of kilojoules in our diet. There
are lots of sources of carbohydrate including sugar, rice, pasta, fruit, milk and grainy foods. All
carbohydrates provide 17KJ of energy for each gram eaten. Different sources of carbohydrates
release this energy into the body at different speeds. Simple carbohydrates (such as sugar)
release the energy quickly while less refined carbohydrates (such as pasta) take longer to
release this energy. The more slowly that energy is released to the body, the better our body
can use it. As with protein, there are a variety of diets with different recommendations for the
amount of carbohydrate you should eat. The amount of activity you do will again vary the
amount you need. It is recommended by the Australian Guide to Health Eating that 52-57% of
the energy you eat should be provided by less refined carbohydrates. If we use the example of
a person who is not very active and eats 9000KJ of energy per day, then we could estimate
that they would need about 275g of carbohydrate a day. Choosing carbohydrates as
wholegrain breads and cereal, milk and fruit are the best options.
Sugar
Refined sugars are added to a lot of foods and drinks we buy including non-diet soft drinks, ice
cream, chocolates and fruit drinks. There are also foods that naturally contain sugar including
honey, fruits and milk. Diets high in sugar from drinks can result in weight gain because refined
sugars do not satisfy hunger despite the kilojoules they provide.

You ate ___mg of sodium in the fast food item/s
Sodium
The most obvious source of sodium in our diet is table salt, but we actually get more sodium
from processed foods such as bread and processed meat which are likely to be high in sodium.
Unlike protein, fat and carbohydrates, sodium does not provide any kilojoules to the body. It is
a micronutrient that helps the cells and nerves of our body function and is needed in the diet for
this purpose. However, only a small amount of sodium is required by our bodies. The
Australian Guide to Healthy Eating suggests that between 900 and 2300mg of sodium is
required daily. To achieve this it is necessary not only to avoid adding salt to food but to keep
many salty processed foods to a minimum. A single standard teaspoon of table salt contains
2000mg of sodium. Too much sodium can increase blood pressure which is a risk factor for
stroke.
X
Appendix3:FastFoodSurveyQuestions+InformationSheet

Pleasetakeamomenttorecallthelasttimethatyouateanitemfrom:McDonalds,
HungryJacks,KFC,RedRoosterorDominosPizza.Inthefirstsectionofthesurveyyou
willbeaskedtoselecttheitem/syouatefromoneoftheselocationsonthemostrecent
timeyouatefastfood.Ifyouatefromanyofthesefastfoodrestaurantsonmorethanone
occasionthroughoutthedayyouarerecalling,pleaserefertoonlythemostrecentofthese
whenansweringthefollowingquestions.

Onceyouhaveselectedtheappropriatefastfoodrestaurant,eachscreenhasnavigation
buttonsandaccesstothehelpmenu.Thetutorialtakesonlyafewminutesanditis
recommendedthatyoudothisbeforeanswering.

Followingisthesecondsectionofthesurvey.Itwillaskyouquestionsthataimtogaina
furtherunderstandingofyourfastfoodpurchasebehaviourthelasttimeyouhadfastfood.
Recordyourresponsestothequestionsbyselectingfromtheavailableoptionsorentering
textwhereindicated.

Q1.Howdidyoupurchasetheitemsyouconsumedthelasttimeyouatefastfood?
[takeaway][drivethru][dinein]

Q2.Wheredidyoueatthefood?
[inthefastfoodrestaurant][inthecar][athome][otherdescribe]

Q3.Brieflydescribewhatwereyoudoingbeforeyoudecidedtogetfastfood?
[entertext]

Q4.Approximatelyhowlongdidyouspendeatingonthislastoccasion?
[0999mins]

Q5.Pleaseusethescalebelowtoindicatehowsociabletheatmospherewaswhenyou
atethisfastfood.
[1=unsociable7=sociable]

Q6.Pleaseusethescalebelowtoindicatehowpleasanttheatmospherewaswhenyou
atethisfastfood.
[1=unpleasant7=pleasant]

Q7.Pleaseindicatehowmanyofthepeoplebelowwerewithyouwhenyouatethisfood.
[Parents][099]
[Siblings][099]
[Children][099]
[Otherrelative][099]
[Partner][099]
[Friends][099]
[Colleagues][099]
[Membersofastructuredsocialgroup(sporting,religiousetc.)][099]
[Other,pleasespecify______][099]
XI

Q8.Onthisoccasion,didyoupurchaseamealdeal(multipleitemspackagedtogether
forasingleprice)?
[Yes][No]

Q9.Whatdaywasitwhenyouatetheseitems?
[Monday,Tuesday,Wednesday,Thursday,Friday,Saturday,Sunday]

Q10.Onaverage,howmuchdidyouenjoythealloftheitemsyouate?
[1=notatall7=extremely]

Q11.Brieflydescribewhyyouatefastfoodonthisoccasion?
[entertext]

Q12.Whichmealoccasionbelowbestdescribesyourfastfoodconsumptionthisday?
[breakfast]
[lunch]
[dinner]
[inbetweenmealssnack]

Q13.Usethescalebelowtoshowwhetheryouthinkyouateabigger,smallerortypical
sized[insertQ11.response]thisday.
[1=muchsmaller,4=typical,7=muchbigger]

Q14.Whichpartofthedaydidyoueattheseitems?
[511.59morning]
[121.59midday]
[24.59afternoon]
[58.29evening]
[8.3011.59night]
[124.59latenight]

Q15.Usethescalebelowtoshowwhetheryouthinkthetotalofallthemealsandin
betweenmealsnacksyouatethroughoutthedaywhenyoulastatefastwasmore,less
orthesamethanistypicalforyouinaday.
[1=muchless,4=typical,7=muchmore]

Q16.Indicatebelowifyoureatingpatternwasdifferentfromusualonthisdayby
describingitbelow?Ifitwastypicalleaveblankorwritetypical
e.g.,IskippedbreakfastorIhad2lunches
[inserttext]

Q17.Howawarewereyouofthosepeoplearoundyouwhenyouwereeating?
[1=notaware7=veryaware]

Q18.Howmuchfastfooddoyouthinkthosepeoplearoundyouate?
[1=notmuch7=alot]

Q19.Howawareoftheamountyouatedoyouthinkthosepeoplearoundyouwere?
[1=notaware7=veryaware]

XII
Q20.Howconcernedwereyouaboutwhatthosepeoplearoundyouthoughtaboutyou?
[1=notconcerned7=veryconcerned]

Theupcomingthirdsectionofthesurveycontainsquestionssurroundingyourperceptions
ofsomefooditemsandquestionsaboutsomegeneralfastfoodbehaviours.
Pleasereadtheinstructionssuppliedwitheachquestionandindicateyouranswersusing
theoptionssupplied.

Q21.Howbigdoyouthinkthefollowingitemsareintermsoftheamountoffoodyou
get?
[Allrated1=notenoughfood7=toomuchfood]

Q21.1Ham&lettucesandwichwithmayonnaise
Q21.2Tbonesteakwithtomatosauceandbroccoli,carrots&potato
Q21.3BigMac
Q21.4MediumBigMacmealdeal(BigMac,mediumfries&medium[nondiet]cola)
Q21.5BigMacwithmediumfries
Q21.6Chickensaladwithnodressing
Q21.74Slicesofham&pineapplepizza
Q21.8 Quarterofroastedchicken(wing)withchips

Q22.Approximatelyhowoftendoyoueatfastfood?
[099][day/fortnight/monthoryear]

Q23.Doyouusuallyorderthesameitems?
[1=notatalltrueofme1=verytrueofme]

Q24.Howappealingistheconceptofupsizingtoyou?
NBUpsizingreferstoincreasingthesizeofamealdealforasmallprice.
[1=notatallappealing7=veryappealing]

Mostitemsweeatordrinkcontainenergyotherwisereferredtoaskilojoules(KJ).This
energyissuppliedpredominantlythroughcarbohydrates(includingsugar),proteinsand
fats.

Forthefollowingfooditemspleaseindicatehowmanykilojoules(orenergy)youthink
eachcontainsbyclickingonthebarbelowitsdescription.

Theamountofkilojoulesneededthroughoutawholedayhasbeensetat9000KJ(2152
calories)inthiscase.Manypeoplehavesuggestedthatanaveragesizedmealconsistsof
approximatelyathirdofdailykilojoulerequirements.

Q24.1Ham&lettucesandwichwithmayonnaise
Q24.2Tbonesteakwithtomatosauceandbroccoli,carrots&potato
Q24.3BigMac
XIII
Q24.4MediumBigMacmealdeal(BigMac,mediumfries&medium[nondiet]cola)
Q24.5BigMacwithmediumfries
Q24.6Chickensaladwithnodressing
Q24.74Slicesofham&pineapplepizza
Q24.8Quarterchicken(wing)withchips

[Allperceptionofportionsizequestionsareratedusingtheabove.Userscanput
indicationofsizeonbaritisprogrammedwith2.5%increments]

Thefourthandfinalsectionofthesurveyasksyoutoprovidesomedemographic
information.Theinformationyouprovideisconfidentialandwillnotdirectlyidentifyyou
inanyway.

Q25.Sex
[male/female]

Q26.Age
[16100]yearsold

Q27.EstimatedHeight
[centimetres]
Q27.1Howconfidentareyouthatyourestimatedheightiscorrect?
[1=notatallconfident4=veryconfident]

Q28.EstimatedWeight
[kilograms]
Q28.1Howconfidentareyouthatyourestimatedweightiscorrect?
[1=notatallconfident4=veryconfident]

Q29.Postcode
[numbers]

Q30.Highestlevelofeducationobtained
[BelowSecondary][Secondary][Trade/Apprenticeship][Certificate/Diploma][Bachelor
degreeorhigher]

Q31.Grossannualhouseholdincome
[20,000andless][20,00140,000][40,00160,000][morethan60,000]

Q32.AreyouofAboriginalorTorresStraitIslandorigin?
[yes][no]

Guideline Daily Energy Intake


(9000KJ ; 2151cal)
XIV
Q33.Countryofbirth
[Australia][UKorIreland][Other]

Q34.MartialStatus
[Marriedorlivingwithpartner][Separated/Divorced][Widowed][NeverMarried]

Q35.WorkStatus
[Fulltimeemployed][Parttime/Casualemployed][Unemployed]
[Homeduties/Retired][Student][Other]

Q36.PensionerStatus
[Receiveapensionfromsocialsecurity][Donotreceiveapension][DontKnow]

Q37.Numberofpeopleinyourimmediatefamily,excludingyourself
[099]

XV


HUMAN ETHICS COMMITTEE APPROVAL NUMBER: 06-96

INTRODUCTION
Fast food consumption is increasing dramatically in Australia and internationally. The
tendency for people to generally consume foods away from home has grown dramatically in
the last 20 years. Part of the increase in the amount of these foods eaten is accounted for by
the expanding number of fast food outlets. Although many people acknowledge the
importance of fast food in eating patterns, little is actually known about fast food eating
behaviours. Even less is know about these behaviours in the Australian community.
This study is being conducted as part of a PhD being undertaken at the University of Adelaide
and CSIRO Human Nutrition. None of the fast food companies included in this study are
associated with the study, development of the program used in the study, or CSIRO. Care has
been taken to indicate trade marks where appropriate. To view the relevant trade mark
information you can click icons in the first section of the survey. The program has been
developed solely by the primary researcher of the study. All graphics and programming are
copyrighted to the primary researcher.
WHAT IS THE AIM OF THIS STUDY?
To obtain data surrounding behaviours and perceptions relevant to fast food consumption.
WHAT IS INVOLVED?
Completion of an interactive online survey that will take less than 10 minutes of your
time
You will need to recall the most recent time you ate McDonalds, Hungry J acks, Red
Rooster, Dominos or KFC
You will also be asked to answer questions surrounding your general fast food
purchasing and consumption behaviour
WHO CAN TAKE PART?
Any person who:
o Is at least 16 years of age,
o Has consumed fast food purchased from:
1. McDonalds, Hungry J acks, Red Rooster, Dominos or KFC
2. In Australia
3. Within the last 12 months
4. Can accurately recall the details (location, people present, time eating etc.) of
their last visit to one of these fast food restaurants
INFORMATION SHEET


Unravelling fast food consumption in everyday living
XVI

WHAT ARE THE BENEFITS OF PARTICIPATING IN THE STUDY?
You will get feedback from the program regarding the nutritional content of the items you ate.
This will be included with more details surrounding the guidelines for ideal consumption.
You will also get your unique Fast Food Profile. This is designed purely for fun and is loosely
based on data from the previous fast food survey. Here you can read which of the five fast
food consumer profiles you match the most.
ARE THERE ANY RISKS INVOLVED?
There are no anticipated negative effects for those individuals wishing to complete this
survey. However, if you do have any concerns regarding this study, you are welcome to
contact the researcher to discuss them (see, If You Have Any Further Questions).
HOW WILL MY PRIVACY BE PROTECTED?
This survey is designed to maintain your anonymity and confidentiality. There will be
questions regarding demographic information (e.g., sex, age). These details will not directly
identify you. The data collected from the survey will only be used for the purposes of the
study as required by the NHMRC National Statement on Ethical Conduct in Research
Involving Humans.
WHAT IF I WISH TO WITHDRAW?
Participating in this study is completely voluntary. At any time you wish, you may withdraw
from the study (without prejudice). Exiting the survey can be done simply by closing your
web browser or clicking on the exit button in the bottom left of the screen.
IF YOU WOULD LIKE TO TAKE PART
Follow the icon at the bottom of the page which will take you directly to the survey. Please
note that clicking this link indicates to the researcher that you have agreed to take part and
have read the information provided to you on this sheet.
This will take to a Flash Program. If you do not have a Flash Player on your computer, you
can download it by clicking the Flash icon. This software is free and will help you viewing
other Flash websites in the future.
The survey will take a minute to download if you have a broadband connection, but longer if
you are using dial-up. While you are in the survey, you should not push back on your web-
browser, because this will take you back to the very beginning of the program and your data
will be lost. There are back and continue buttons in the survey screen which you should click
if you need to go back.
Push the red submit button when you have finished with the program and this will shut your
window and submit your data. But only hit submit when you have read your profile and
viewed the nutritional information if this is what you wish to do.




XVII

IF YOU HAVE FURTHER QUESTIONS
Contact information:
Primary Researcher/PhD Candidate Supervisor
MsEmilyBrindal Dr Carlene Wilson
Phone: 83050633 Email: Carlene
Email: Emily

This study has been approved by the University of Adelaide School of Psychology Human
Research Ethics Subcommittee.

If you would like to speak with someone with respect to ethical matters please contact Paul
Delfabbro on83034936, or email him.
XVIII

Appendix4:ExamplescreenshotsofFastFoodSurvey(FFS)

Screen1(below):Introduction(appearsafterinformationsheetanddownload)

Screens23:Instructionsabouthowtousetheprogram
Screen4(below):Nominationofbrand/soffastfoodconsumed.

XIX

Menuscreens:Afterselectingfastfoodbrand,participantsweredirectedtothe
relevantmenu.Theycouldnavigatethesetoselectdifferentitems.Thefollowing
threescreensshowselectionofMcChicken,mediumfriesandaDietCoke.

XX

Screen5(below):Participantscheckandcanadjustamountsofitem/sselected

XXI

Screen6(below):Participantsdetailanyadjustmentstoorders

Screens7to13:Questionnaireitems(seeScreen7below)

XXII

Screens14to22:Nutritionalfeedback.Includesnutritionalcontentofitemseaten
(seeScreen14above)andfurtherinformationaboutmacronutrientsandsodium
(seeScreen18below).

XXIII

Appendix5:Completedpilotingform
Pilotformforinternetsurvey 2006

Instructions:
Please fill out the following form by entering your answers into the Word
document and sending it back to me via email emily.brindal@csiro.au by next
Tuesday evening (the 22nd of August). Or you can print this out, fill it
in, attention it to Emily Brindal and fax it back to me on 8303 8899.

1. How of t en do you use i nt er net ? Descr i be t he pr i mar y r eason f or your
i nt er net use.
Dai l y. For wor k mai nl y r esear ch and emai l .

2. Fol l ow t he l i nk t o t he home page www. f ast f oodst udy. com. au
Read t he i nf o sheet and cont i nue wi t h t he sur vey. I f you not i ce any
t ypos i n t he i nf o sheet or somet hi ng i s uncl ear , pl ease wr i t e
comment s bel ow:
No obvi ous er r or s/ pr obl ems.

3. Now coul d you pl ease wr i t e down t he t i me i t t akes you t o downl oad t he
pr ogr am:
Under 1 mi nut e.

4. Once t he pr ogr ami s downl oaded, ent er t he passwor d pi l ot and pr ess
cont i nue. Read t he i nst r uct i ons at t he begi nni ng of t he pr ogr am. Was
t her e anyt hi ng t hat di d not make sense or any t ypos? Pl ease comment
bel ow:
Was ok.

5. Use t he pr ogr amt o sel ect a quar t er ( wi ng) wi t h chi ps and a 390mL
Coke f r omRed Roost er , 3 sl i ces of Aussi e Bob on Puf f ect i on cr ust
f r omDomi no s and a chocol at e sundae f r omMcDonal d s. Use t he hel p
menu i f you need t o. Di d you have any di f f i cul t i es doi ng t hi s? Di d
you use t he hel p? Comment bel ow:
None at al l , I l ove t hat t he i cons ar e so easy t o i dent i f y and use. No use of
hel p menu.

6. Cont i nue t o t he r est of t he sur vey and answer t he r emai ni ng
quest i ons. Answer t hese quest i ons however you wi sh. I f t her e i s a
quest i on t hat doesn t make sense, comment bel ow.
Al l ok.

7. You can choose t o ent er t he compet i t i on or not ( you wi l l not act ual l y
be ent er ed i nt o t he compet i t i on) . Now cl i ck Submi t or Submi t &
vi ew nut r i t i onal i nf or mat i on. Pl ease i ndi cat e whi ch you cl i cked and
any pr obl ems t hat occur r ed bel ow:
Went t o compet i t i on and nut r i t i onal i nf o. Ther e was one t ypo i n t he end
i nf or mat i on sect i on under Pr ot ei n wher e t her e was a r ef er ence t o you
i nst ead of your ot her t han t hat , not hi ng out st andi ng.

8. How l ong do you t hi nk t he whol e sur vey wi l l t ake t o compl et e?
5- 10 mi nut es.

I di d f ol l ow t he st eps wi t h anot her ver si on, and f ound t hat t he over al l
usabi l i t y was out st andi ng, ver y easy t o under st and and kept me i nt er est ed.
Over al l wel l done.

XXIV

Summary:PilotingoftheinternetbasedFastFoodSurvey

Theeightpeoplewhorespondedtothepilotvariedinfamiliaritywithwebbased
technology.Fourpeoplewerehighlyfamiliarwithwebbasedtechnologiesandindicated
thattheyworkedinoccupationswheretheinternetwasuseddaily.Twoparticipants
wereretiredandindicatedthattheyusedtheircomputersafewtimesaweek,primarily
tocheckemail.Theothertwoparticipantsusedcomputersregularlyfornonworkbased
activities.

Usershadaccesstoavarietyofinternetconnectionsrangingfromhighspeed
broadbandtoregulardialup.Thosewithbroadband(6users)allreportedadownload
114

timeofapproximatelyoneminute.Downloadingtheprogramusingdialuptook
considerablylongerwithreportedtimesrangingfrom10to12minutes.Thetimeto
completethesurvey,followingdownload,rangedfrom8minutesto20minutes.
Participantswhotooklongertofinishthesurveytestedallaspectsofthesurvey
includingviewingthenutritioninformationandenteringthecompetition.

Theresponsesfromthepilottestingwerealsousedtotestthebackendoftheonline
database.Allresponsesweresuccessfullyandaccuratelysavedtothedatabaseby
theprogramusingthe.phplinkingfile.Thistypeoffileisaserversidescripting
languagethatexecutescodeonaserverallowingdynamicstorage,manipulationor
retrievalofdata.Theresponsesrecordedinthedatabaseindicatedthatparticipants
were100%accurateatchoosingandadjustingtheitemsrequestedbyfollowingthe
instructions.Nonereferredtothehelpscreenorthetutorial.Noissueswith
functionalitywerereported.

114
AstheprogramwasprogrammedinFlash,ratherthandownloadingeachpagetobeviewed
individually,theentiresurveyisdownloadedinitially.Thismeansthattheinitialdownloadtimeisslower
butthattheprogramworksfasterduringresponding.
XXV

Generalpositivecommentsweresentregardingthedesignofthebuttonsandtempoof
thesurvey.Oneparticipantsuggestedminorchangesinspacingandformattingfor
clarity.Theonlyothercommentsthatweregivenpertainedtotheidentificationof
typographicalmistakes.
XXVI

Appendix6:RequestforparticipationinFastFoodSurveytoNWAHSmembers

DearNorthWestAdelaideHealthStudyParticipant

Canyoupleasehelpus?IfyouhaveeatenfastfoodfromMcDonalds,HungryJacks,RedRooster,
DominosorKFCwithinAustraliainthelast12monthsandcanreasonablyrecallthedetailsofyour
lastvisittooneofthesefastfoodrestaurants,wewouldverymuchappreciateassistancewithan
onlinesurveytohelpusunderstandperceptionsandbehaviourssurroundingfastfoodconsumption.
Thestudyinvolvesansweringsomefairlysimplequestionssurroundingaspectsofthelasttimeyou
atefastfoodfromoneoftheserestaurants,yourgeneralfastfoodpurchasingbehaviourandyour
perceptionsofsomefooditems.

Itshouldonlytakeapproximately15minutestocompleteandwillgiveyousomeinteresting
informationabouttheparticularfastfoodyounominated.

Bycompletingthesurvey,youcanchoosetogointoadrawforaFREEdoublemoviepass.Ifyou
wouldliketotakepartinthissurveypleasegotothefollowinglinkwww.fastfoodstudy.com.auand
followtheinstructions.Ifyouhaveanyquestions,pleasedonothesitatetocalltheresearcher,Emily
Brindalon(tel)83050633orJanetGrant,NorthWestAdelaideHealthStudyCoordinatoron(tel)
82266054.

ThankyouforyourongoingparticipationintheNorthWestAdelaideHealthStudy.Your
contributionisverymuchappreciated.
TheNorthWestAdelaideHealthStudyTeam

XXVII

Appendix7:SummaryofExplorationofPortionSizeItems
Theresultspresentedbelowwerepresentedaspartofasymposiumatthe2009Annual
MeetingoftheInternationalSocietyforBehaviouralNutritionandPhysicalActivity,June20,
Lisbon,Portugal.SymposiumtitlePortionSizeGuidanceforHealthyEatingIntervention
Issues,chairedbyProfessorAnnieAnderson.


Perceptionsofportionsizeoffastfooditems
Numerousstudieshavedemonstratedthatlargerportionsizesresultinincreased
energyintake(Diliberti,Bordi,Conklin,Roe,&Rolls,2004;Ledikwe,ElloMartin,&
Rolls,2005;Jefferyetal.,2007;Rolls,Roe,Kral,Meengs,&Wall,2004;Wansink,Painter,
&North,2005;Wansink&Park,2001)whichmayovertimeleadtoweightgainand
obesity.Portionsizeisthereforeconsideredtobeanimportantcontributorto
overeatingandobesity(Herman,Roth,&Polivy,2003).Thevalueformoneythatlarge
portionsofferremainsappealingforsomeconsumersdespitepotentialhealthoutcomes
(Vermeer,Steenhuis,&Seidell,2009).

Therearefewpublishedstudiestriallingportionsizeinterventionsandthesehave
generallybeenunsuccessful(Steenhuis&Vermeer,2009).Oneofthelimitingfactorsin
thesestudiesisthepaucityofinformationrelatingtoconsumersperceptionsand
understandingofportionsizes.Portionsizeinformationhasmanycomponents.Inorder
touseportionsizetocontrolintake,consumersneedtobeawareofthevolumeoffoods
theyareeatingandhowlargetheseservingsarerelativetotheirdailyintake.The
cognitiveprocessestoachievethisrequirethatanindividualisreadilyabletolocate
appropriatenutritionalinformation,understandstheirdailyenergyrequirements,and
canintegratethisinformationintoacoherentapproachtomanagingintakeforthe
remainderoftheday.Previousresearchsuggeststhatmostpeoplefindareunableto
accomplishtheprocesseslistedabove,irrespectiveoflevelofeducation(Schapira,
XXVIII

Kumar,Lyman,&Mcmillan,1990).Nutritionstudentshavebeenshowntohavepoor
accuracyinchoosingmediumservingsizesaccordingtorecommendations(Young&
Nestle,1998),andinanotherstudyapproximately,75%ofdoctorsandnurseswere
unabletoestimatetheenergycontentoffooddishestotheclosest250kcal(Ashley,
Davidson,Wilkins,&Thompson,2004).Interestingly,dietersappeartobemore
accurateincalorieestimationtasks(Carels,Konrad,&Harper,2007),whereasobese
peoplearemorelikelytounderestimatetheirfoodintake(Rasmussen,Matthiessen,
BiltoftJensen,&Tetens,2007;Rennie,Coward,&Jebb,2007;Waling&Larsson,2009).

Mealcharacteristicssuchassizeandhealthinessmaybebetterdeterminantsof
consumersaccuracyatkilojouleestimation(Wansink&Chandon,2006a,2006b).Ithas
beenshownthatstudyparticipantsassignedcaloriecontentwithgreateraccuracyto
smallmeals(Wansink&Chandon,2006b)andforindividualfooditems(Chandon&
Wansink,2007)ascomparedtolargemeals.

WhilestudieshaveinvestigatedtheabilityofconsumersintheUSAtoestimatethe
caloriecontentoffastfooditems(Chandon&Wansink,2007;Wansink&Chandon,
2006b),itisunclearwhetherthisisapparentelsewhere,andtheabilityofconsumersto
interpretmealsrelativetodailyintakeremainsunknown.

Theaimsofthisstudywere:(a)todeterminewhetherfastfoodconsumerscould
accuratelyestimateandinterprettheKJinformationforfastfooditems,(b)toexplore
howfastfooditemsareperceivedrelativetostandardfoodsand(c)totestanovel
energyestimationformat.

XXIX

Method
PerceptionofportionsizewasmeasuredaspartofthelargerFastFoodSurvey(FFS).
TheFFSaskedparticipantstorecalltheitemsconsumedattheirpreviousvisitto
McDonalds,HungryJacks(akaBurgerKing),KFC,RedRoosterorDominosPizza.This
recallwasfollowedby38questionsaboutthecontextofconsumption,fastfood
behaviours(includingfrequencyofconsumption)andparticipantdemographics.This
surveyrecruitedtwocohortsoffastfoodconsumerswhocompletedacomputer
assistedsurveyregardingthebehavioursurroundingtheirmostrecentconsumptionof
afastfooditem.

StudyPopulation
TwoindependentsamplescompletedtheFFS.Thefirstsampleconsistedofpeople
attendingtheannualRoyalAdelaideShow(RASsample;n=116).Thesepeoplewere
askedtocompletethesurveyatacomputerstationwhileengaginginaninteractive
displayforanationalscientificorganisation.Thesecondsamplewasrecruited
exclusivelythroughtheinternet(internetsample;n=419).Inthiscase,recruitment
methodsincludedpromotingthewebaddressofthesurveywidelythroughthe
community(usingmassmedia,wordofmouthandsnowballemailing).

InadditiontotheFFSrespondents,asmallsampleofdieticianswasaskedtocomplete
theKJestimationtaskandprovidefeedbackonthedifficultyofthetasks(forthemselves
andforthegeneralpopulation).Theseitemswereratedusingasingleitem6point
scalewhere,1=impossible,2=verydifficult,3=difficult,4=reasonable,5=easy,
6=veryeasy.Thiswasdoneinordertoobtainabenchmarkforhowdifficultthetask
wasforanexpertsampleaswellastheiraccuracyatcompletingtheitems.
XXX

PortionsizemeasureswithintheFFS
Subjectiveratingsofamountoffood:Theinitialsetofportionsizequestionsasked
participantstosubjectivelyrateoftheamountoffoodintheeightitems:sixfastfood
itemsandtwostandardmeals(describedbelow).Subjectiveratingsofamountoffood
wereusedtoassesswhetherconsumersthoughtthefooditemsdescribedprovided
enoughfood.Thefooditemswereratedusingascalewithscoresrangingfrom1(not
enoughfood)to7(toomuchfood).

Kilojoule(KJ)estimation:Thesecondsetofportionsizequestionsaskedparticipantsto
estimatetheKJcontentofeachoftheitemsrelativetoadailyintake.Inthistask,avisual
scalerepresentinga9000KJrecommendeddailyintakewasused.Participantscould
indicateaKJamountatanypointonthebar.Thebarcouldbedraggedatincrementsof
225KJ(2.5%ofthedailyintake).Thustheminimumvaluethatcouldbeindicatedwas
225KJandthenextwas450JKandsoon.Faintlinesindicate25%intervalsofthedaily
intake.Participantsweretoldbeforethetaskthataregularmealcouldbeconceivedof
asroughlyathirdofdailyintake.Theresponsebarextended20%beyondthedaily
intake(indicatedbyalineacrossthebar).Ascreenshotoftheitemscanbeseenin
Figure1.

XXXI

Figure1:ScreenshotofcompletedestimationofKJitems

Thefastfooditemsratedineachsetofportionsizeitemsrepresentedtypicalburger,
chicken,andpizzafastfoodmeals.McDonaldsBigMacwasusedtorepresentburger
options.ThereasonforthiswasMcDonaldspreeminentmarketstatusandtheiconic
natureoftheBigMac.Theburgerwaslistedaloneandinamediumsizedvaluemeal
(withmediumfriesandamediumnondietdrink).Aquarterchickenwithchipswas
chosenforthechickenmeal.Fourslicesofpizza(halfalarge)wasusedtodescribea
pizzamealbecausethiswasanequivalentweighttotheothermeals.Hamandpizza
(Hawaiian)toppingwaschosenbecauseitisavailableacrossarangeofdifferentbrands
(www.calorieking.com).Asthechickenandpizzaitemswerelistedwithoutadrink,a
variationoftheBigMacmealwithoutadrinkwasalsoincludedintheportionsize
measures.Finally,ahealthymealavailablefromfastfoodrestaurantswasalso
nominatedachickensalad.Thedescriptionofthesaladspecifiedthatitcontainedno
dressingtomaintainthehealthinessofthisitem.

XXXII

ThedescriptionsofthetwostandardmealsusedcommonAustralianfooditemssothey
wouldbeeasilyrecognisablefortheparticipants.Eachmealwasdesignedtorepresent
bothalargeandalightmeal.Torepresentthelargemeal,atbonesteakmealincluding
tomatosauce,broccoli,carrotsandpotatowasconstructed.Thelightmealconsistedofa
hamandlettucesandwichwithmayonnaise.

Kilojoule(KJ)information
TheactualKJcontentoffooditemswasgatheredfromcompanywebsitesinmid2006.
ThenutritionalKJcontentforthechickensaladandburgeritemswasobtainedfrom
McDonaldswebsite.TheRedRoosterandDominosPizzawebsiteswereusedforthe
chickenandchipsandpizzarespectively.ThesebrandswereusedtorepresentKJ
contentastheyareamongthemostpopularquickservicerestaurantsinAustralia(see,
Brindal,Mohr,Wilson,&Wittert,2008).Kilojouleinformationforthestandardmeals
wascalculatedinFoodWorks
115
usingstandardservingsizes.

OncetheactualKJinformationfortheitemshadcollectedfromtherelevantsources,
therewerethreeresultingversionsofKJinformation:theFFSsamplesestimations,the
expertsampleestimationsandtheactualKJcontentoftheitems.Fortheratingsofthe
amountoffood,therewereonlytheratingsfromtheFFSsamples.


115
FoodWorksisthetrademarknameofacomputerprogramthatbreaksdownthenutritionalcontentof
variousfooditems.
XXXIII

Results
Samples
Aftertheportionsizemeasureswerescreenedforoutliersandmissingresponses,there
wereatotalof395participantsintheinternetsampleand104intheRASsample.The
demographicsofeachsamplearedescribedinTable1.

Table1:FrequencyofdemographicsfortheRoyalAdelaideShow(RAS;n=104)and
internetsamples(Net;n=395).
Variable %RAS %Net Variable %RAS %Net
Martialstatus Employmentstatus
Married/livingwith
partner
47.1 50.1 Fulltime 39.4 57.2
Nevermarried 49.0 44.1 Student 26.9 21.0
Other 3.8 5.8 Parttime/casual 22.1 13.7
Householdincome Homeduties/retired 7.7 4.7
20000andbelow 14.4 7.8 Unemployed 3.8 1.8
20001to40000 17.3 9.9 Frequencyofpatronage*
40001to60000 23.1 26.6 <1/half 4.6
60000andabove 45.2 55.7 1/half<1/month 16.3 16.7
Highestlevelof
education
1/month 14.4 21.5
Bachelor 36.5 57.7 1/fortnight<1/week 25.0 23.0
Trade/certificate 24.0 22.5 1/week 21.2 19.0
SecondarySchool 34.6 18.7 twiceaweek 17.3 9.4
BelowSecondary
School
4.8 1.0 >twiceaweek 5.8 5.8
Australianborn 83.7 79.7
*Onlyfastfoodconsumersparticipatedinthesurvey

Thefinalsampleswerepredominantlyfemalewith67.1%oftheinternetsampleand
61.5%oftheRASsamplebeingfemale.Themeanagesofthesampleswere31.51
(SD=11.73)and29.96(SD=12.11)fortheinternetandRASsamplerespectively.The
XXXIV

internetsamplehadmoreparticipantspossessingaBachelorsDegree,
2
(3,496)=
23.01,p<.001,morepeopleinthehighestincomebracket,
2
(3,496)=9.93,p<.05,and
morefulltimeemployedpeople,
2
(4,495)=11.81,p<.05,thantheRASsample.

Nineregisteredand/orpracticingdieticianscompletedthesurvey.Allthedieticians
foundthetaskreasonabletoveryeasyindifficulty.TwodieticiansdescribedtheKJ
estimationtaskasofreasonabledifficultyforpotentialparticipants;theremaining
respondentsbelievedthetasktobedifficulttoverydifficultforthegeneralpopulation.

Demographicdifferencesinportionsizeratings
Therewerenosystematicdifferencesintheratingsoftheamountoffoodin,orKJ
estimationof,thefooditemsbetweentheRASandinternetsamples.Thereforeresults
fromtheseitemswerecombinedforbothcohortsthesedatawereusedtorepresent
responsesfromalaysample(n=499).

Therewerenosignificantrelationshipsbetweenage,levelofeducationorselfreported
BMI
116
andanyoftheportionsizeitems.Sexdifferenceswereobservedintheratingsof
theamountoffoodinallfooditems(p.001).TheKJestimationsformalesandfemales
forthethreeBigMacitemsandthepizzawerealsosignificantlydifferentaccordingto
independentsamplesttests(p<.05fortheBigMac;p<.01fortheBigMacandfriesand
thepizza;p<.001fortheBigMacmeal).Inlightofthesedifferences,analyseswere
performedseparatedformalesandfemalesinthelaysamples.


116
Onlypeoplewhoreportedconfidenceintheselfreportedheightandweightswereusedinthese
correlations(n=414).
XXXV

Subjectiveratingsofamountoffood
Manipulationcheck.Toassesswhetherparticipantsperceivedthehamsandwichtobea
lightmealrelativetothetbonemeal,apairedsamplesttestwasperformed.Themean
ratingoftheamountoffoodinthesandwichwasbelowthemidpointwhilethetbone
mealwasrankedabovethemidpoint.Thedifferencebetweentheseitemswas
significantinbothmales,t(169)=14.63,p<.001,andfemales,t(328)=24.74,p<.001.

Results:Themeansandstandarddeviationsfrommalesandfemalesinthelaysample
forthesubjectiveratingsoftheamountofeachfooditemcanbeseeninTable2.

Table2:Meansandstandarddeviationsforratingsofamountoffoodinthefast
foodandstandardfooditemsformalesandfemales.
Items Actual Males
*
Females
**

grams M SD M SD
Hamsandwich 147 3.13 0.90 3.59 1.18
BigMac 201 3.21 1.38 3.97 1.50
BigMac,mediumfries 299 3.98 1.29 4.84 1.39
MediumBigMacMeal 299 4.46 1.35 5.29 1.50
Chickensalad 300 3.25 0.95 3.58 1.23
4slicesHawaiianpizza 316 4.18 1.49 5.02 1.40
Quarterchicken&chips 441 4.30 1.16 5.02 1.31
Tbonemeal 447 4.82 1.07 5.29 1.16
*
n=170;
**
n=329

Overall,malesratedthefooditemsassubjectivelylessfoodthanthefemales.The
rankingofamountoffoodforeachitemcorrespondedtotheactualsizeoftheitems.On
average,thetbonemealwasratedasthelargestamountoffoodandthehamsandwich
wasratedasthesmallestbymales.Infemales,thetbonemealandBigMacmealwere
ratedequallargestandthehamsandwichandchickensaladwereratedasthesmallest
items.
XXXVI

AsitemswereaddedtotheBigMac(firstthemediumfriesthenthedrink),ratingsof
theamountoffoodincreased.Apairedsamplesttestrevealedthatthedifference
betweentheBigMacaloneandtheBigMacwithfrieswassignificantinbothsexes,
t(169)=9.92,p<.001(males);t(328)=17.23,p<.001(females).TheratingfortheBig
MacwithfriesanddrinkwashigheragainwhencomparedtotheBigMacandfries,
t(169)=7.39,p<.001(males);t(169)=10.96,p<.001(females).

Conclusion:Mostconsumersthoughtthatthefastfooditemsdescribedprovidedan
adequatevolumeoffoodalthoughthisdifferedslightlybysex.Asthefastfooditem
containedmoreitems,ratingsoftheamountoffoodprovidedincreased;evenwhen
onlyadrinkwasaddedtotheburgerandfries.Thisindicatedthatconsumersdid
perceivedifferencesintheamountoffoodprovidedandadjustedtheirratings
appropriately.

Canconsumersaccuratelyestimatekilojoulesinfastfooditems?
ThemeanKJestimationsforthedieticianandlaysamplesalongwiththeactual
estimationofKJfromthefastfoodcompaniesandFoodWorksaredescribedinTable3.
Thepercentageofthe9000KJdailyintakethattheseestimationsrepresentispresented
inFigure2.

XXXVII

Table3:Kilojoules(KJ)offooditems(Actual)withestimationsfromdieticians,malesand
femalesinthelaysamples.
Actual Dieticians Males Females
KJ
1
M(SD) M(SD) M(SD)
Chickensalad 972 1102(471) 3229(1896) 2912(1604)
Hamsandwich 1179 1252(493) 3630(1854) 3565(1857)
BigMac 2010 2304(387) 5701(2553) 6211(2544)
Tbonemeal 2157 2404(676) 5136(2003) 4976(1858)
4slicesHawaiianpizza 2928 3731(1009) 6353(2167) 7069(2277)
BigMac,mediumfries 3270 3531(374) 6507(2347) 7133(2266)
Quarterchicken&chips 3741 3205(907) 6217(2238) 6586(2228)
MediumBigMacmeal 3866 4282(465) 7309(2432) 8174(2299)
1
AsestimatedviaFoodWorks;
*
Allvaluesweresignificantlydifferentbetweenthelaysamples(both
malesandfemales)andthedieticiansatthep<.001level.

Theestimationsthatthedieticiansgavecorrespondedcloselytotheactualdaily
amountsthatthemealsaccountedfor.Onaverage,malesandfemalesinthelaysamples
0 25 50 75 100
Chickensalad
Hamsandwich
BigMac
Tbonemeal
4slicesHawaiianpizza
BigMac,mediumfries
Quarterchicken&chips
MediumBigMacmeal
Percentof9000KJdailyallowance
Females
Males
Dieticians
Actual
Figure2Percentageofthe9000KJdailyintakeaccountedforbyfooditemsincludingestimations
fromthedieticiansampleandmalesandfemalesinthelaysamples.
XXXVIII

estimatedlargeKJvaluesforallthefooditems(bothfastfoodandstandarditems).All
participantsestimatedthattheBigMacmealaccountedforahighportionofthedaily
intake,withfemalesandmalesindicatingthatthemealaccountedforover90%and
80%ofthedailyintakerespectively.

ThepercentageoferrorestimationfromtheactualKJamountsisdepictedinFigure3.
Thedifferenceintheoverallerrorsbetweenthelayanddieticiansampleswashighly
significantforallitems(p<.001).Inallsamplesoverestimationsweremorecommon
thanunderestimations.Theonlyitemthatwasmorefrequentlyunderestimatedthan
overestimatedbythedieticiansamplewasthequarterofachickenandchips.

1000 500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Chickensalad
Hamsandwich
BigMac
Tbonemeal
4slicesHawaiianpizza
BigMac,mediumfries
Quarterchicken&chips
MediumBigMacmeal
ErrorinKJestimationcomparedtoactual
Females
Males
Dieticians
Figure3:ErrorinKJestimationforallfooditemsbydieticians,andmalesandfemalesinthelay
samples.
XXXIX

Howarefastfooditemsperceivedrelativetostandardfooditems?
Aratioforeachofthefastfooditemsrelativetothelargemealwascreatedbydividing
theparticipantsKJestimationoftheitembytheirestimationofthestandarditems.
Thusratioscorrespondedtothehamsandwich(lightmeal)orthesteakmeal(large
meal).Resultingratiosabove0butbelow1representaKJestimationsmallerthanthat
forthestandarditemwhilevaluesabove1representaKJestimationgreaterthanthatof
thestandarditem.Afinalratioof1indicatesthattheestimationsforthefastfooditem
andstandardmealswereidentical.Theadvantageoftheratioisthatitcontrolsforeach
individualserrorinestimation.Theresultingaverageratiosforthelaysamplescanbe
seeninTable4.

Table4:Averageratiosforestimationsoffooditemsrelativetoalargemeal(SteakMeal)and
alightmeal(HamSandwich).
SteakMeal HamSandwich
Males Females Males Females
Actual M(SD) M(SD) Actual M(SD) M(SD)
BigMac 0.93 1.23(0.72) 1.35(0.72) 1.70 1.75(0.78) 2.03(1.16)
MediumBigMacMeal 1.79 1.63(0.86) 1.82(0.82) 3.28 2.10(0.95) 2.43(1.47)
BigMac,mediumfries 1.52 1.44(0.79) 1.58(0.75) 2.77 2.42(1.16) 2.83(1.64)
4slicesHawaiianpizza 1.36 1.39(0.70) 1.56(0.72) 2.48 2.09(0.98) 2.41(1.25)
Quarterchicken&chips 1.73 1.35(0.72) 1.44(0.68) 3.17 2.02(0.94) 2.21(1.14)
Chickensalad 0.45 0.65(0.33) 0.60(0.30) 0.82 0.94(0.41) 0.90(0.45)

Withtheexceptionofthehealthyfastfooditem(thechickensalad),allofthefastfood
itemswereratedmorecloselytothelargemealthantothelightmealbybothmalesand
females.Itisnoteworthythatwhencomparingfastfoodtoasteakmealthattheratios
fortheestimationswerereasonablyclosetotheactualratios.

XL

Discussion
Manyofthedieticianswhocompletedtheportionsizeitemsindicatedthattheythought
thatthetaskwouldbedifficultforuntrainedconsumers.Thelargeinaccuraciesin
energyestimationbythelaysamplesconfirmedthattheydidhavedifficultywiththe
task.Errorsinestimationindicatedthattheoverestimationinofenergyinthefood
itemswasnotlimitedtothefastfooditems.Whenaratiooftheestimationsoffastfood
itemsrelativetothestandardfooditemswascreated(thuscontrollingforgeneralerror
inestimation),theestimationsappearedtobemuchmoreaccurate.Inotherwords,it
appearedasthoughconsumerscouldratethefastfooditemsenergycontentrelativeto
standardmealitemsbutoverall,theirabilitytoestimateenergycontentwaspoor.

Inthepast,closeto70%ofconsumershaveindicatedthattheywouldlikenutrition
labelsthatareeasiertounderstand(Kristal,Levy,Patterson,Li,&White,1998).Non
numericalstylesofpresentation,suchasthetrafficlightsystemintheUK,aredesigned
toimprovetheuseabilityofnutritionalinformationbysimplifyingdetailsandproviding
adailycontextforinterpretingtheinformation.Whiledebateremainsastowhatthe
mosteffectiveformatforpresentingnutritionalinformationis(Bussell,2005),theneed
tochangefromthecurrentnumericallybasedsystemisclear.Providingadailyenergy
allowancedidnotappeartobeadvantageousintheportionsizetaskusedhere.
AlthoughtheaverageratingoftheBigMacmealwasalmostanentiredaysintake,the
samegroupofconsumersratedtheamountoffoodinthemealasslightlymorethan
enoughfood.Thisfurthersuggeststhatconsumershaddifficultywiththetask.
DifficultywastheKJestimationtaskmaybetheresultofalackofunderstandingof
dietaryenergy.

XLI

NutritioneducationmayimproveconsumersunderstandingofKJinformationand
thereforehelpthemusethisinformationtoguidetheirunderstandingofportionsize.
Caloriecountingisaskillthatcanbetaught(Martinetal.,2007).Translatingthisskill
intobehaviourchangepresentsafurtherchallenge.Forexample,44to57%people
reportthattheywouldnotbelikelytousenutritioninformationifitwasavailablein
restaurants(Krukowski,HarveyBerino,Kolodinsky,Narsana,&DeSisto,2006).
Furthermore,itmaytakeconsiderableeffortandresourcetoeducateconsumersuntil
theyaresufficientinenergyestimation.Newinnovationsforimprovingconsumers
portionsizeestimationsarebeingdeveloped(Ayala,2006;Godwin,Chambers,
Cleveland,&Ingwersen,2006;Hernandezetal.,2006;Riley,Beasley,Sowell,&Behar,
2007).Nevertheless,Hernandezetal.reportthatprecisioninportionsizeestimationis
notyetarealisticexpectation(p.S22).

Thepublicneedguidanceabouthowtoeatfastfoodsinawaythatrestrictsanydamage
totheirfuturehealth.Amoreefficientwayofcommunicatingthisinformationmaybe
bytranslatingtheKJcontentoffastfooditemsrelativetostandardfooditems;by
informingpeoplethataBigMacisclosetoasteakmealinthedietaryenergyitprovides,
forexample.Presentinginformationinthiswaymayplayonthestrengthsthat
consumerspossesstheyhaverelativelygoodperceptionofthesizeofitemsrelativeto
otheritems.

Thecurrentsamplewaslimitedinitsrepresentativeness.Itconsistedsolelyoffastfood
consumerswhomaypossessvaryingskillsininterpretingnutritioninformationthan
nonfastfoodconsumers.Godwinetal.(2006)reportedthathalfthepeoplewho
reportedeatinghighcaloriefoodsdidnotengagewithnutritioninformation.
XLII

Nevertheless,previousstudieshavelookedatfastfoodconsumersestimationsoffast
fooditems(Chandon&Wansink,2007;Wansink&Chandon,2006b).Therewerealso
technicallimitstothetaskthatmaybealterabletomakethetaskmoreuserfriendlyin
thefuture.Anexamplethatgivesfeedbackaboutaccuracycouldprecedetheactual
items.Inthecurrentformat,theminimumamountofKJthatpeoplecouldenterwas
approximatelyaquartertheKJsofthesmallestitemwhereastheestimationbar
extendedtoanamountofKJthreetimesthelargestitemlisted.Thismayhavecreateda
flooraffectwithfewpeopleestimatinglowerKJamounts.Havingaresponsebarthatis
displayedlargerandhassmallerincrementsmaybealsobeusefulforparticipants.

ThisstudyhasprovideduniquedataonthewayinwhichAustralianconsumers
perceiveportionsize.Furtherdataonhowconsumersviewportionsizeisneededif
interventionsinvolvingportionsizewillbesuccessfulinthefuture.Giventhepotential
implicationofportionsizeintheobesityepidemic,betterunderstandingthebestwayto
adviseconsumersaboutportionsizewillbeimportantforfutureresearch
XLIII

References

Abbott,R.(1997).Foodandnutritioninformation:astudyofsources,uses,and
understanding.BritishFoodJournal,99,4349.
Armitage,C.J.,&Conner,M.(2001).Efficacyofthetheoryofplannedbehaviour:Ameta
analyticreview.BritishJournalofSocialPsychology,40,471499.
Ashley,B.D.,Davidson,B.,Wilkins,J.,&Thompson,N.P.(2004).Countingcalorieshow
goodarehealthcareprofessionals?Nutrition,20(9),817820.
Ayala,G.X.(2006).Anexperimentalevaluationofagroupversuscomputerbased
interventiontoimprovefoodportionsizeestimationskills.HealthEducation
Research,21(1),133145.
Brindal,E.,Mohr,P.,Wilson,C.,&Wittert,G.(2008).Obesityandtheaffectsofchoiceata
fastfoodrestaurant.ObesityResearchandClinicalPractice,2(2),111117.
Bussell,G.(2005).Nutritionalprofilingvsguidelinedailyamountsasameansofhelping
consumersmakeappropriatefoodchoices.Nutrition&FoodScience,35(5),337
343.
Carels,R.A.,Konrad,K.,&Harper,J.(2007).Individualdifferencesinfoodperceptions
andcalorieestimation:Anexaminationofdietingstatus,weight,andgender.
Appetite,49(2),450458.
Chandon,P.,&Wansink,B.(2007).Isobesitycausedbycalorieunderestimation?A
psychophysicalmodelofmealsizeestimation.JournalofMarketingResearch,
44(1),8499.
Colagiuri,S.,&Caterson,D.(2008).KFCsponsorshipofcricket.MedicalJournalof
Australia,189(7),415416.
Diliberti,N.,Bordi,P.L.,Conklin,M.T.,Roe,L.S.,&Rolls,B.J.(2004).Increasedportion
sizeleadstoincreasedenergyintakeinarestaurantmeal.ObesityResearch,
12(3),562568.
Godwin,S.,Chambers,E.,Cleveland,L.,&Ingwersen,L.(2006).Anewportionsize
estimationaidforwedgeshapedfoods.JournaloftheAmericanDietetic
Association,106(8),12461250.
Herman,C.P.,Roth,D.A.,&Polivy,J.(2003).Effectsofthepresenceofothersonfood
intake:Anormativeinterpretation.PsychologicalBulletin,129(6),873886.
Hernandez,T.,Wilder,L.,Kuehn,D.,Rubotzky,K.,MoserVeillon,P.,Godwin,S.,etal.
(2006).Portionsizeestimationandexpectationofaccuracy.JournalofFood
CompositionandAnalysis,19,S14S21.
Jeffery,R.W.,Rydell,S.,Dunn,C.L.,Harnack,I.J.,Levine,A.S.,Pentel,P.,Baxter,J.,&Walsh,
E.M.(2007).Effectsofportionsizeonbodyweight.InternationalJournalof
BehavioralNutritionandPhysicalActivity,4(27),from
<http://www.ijbnpa.org/content/4/1/27>.
Kristal,A.R.,Levy,L.,Patterson,R.E.,Li,S.S.,&White,E.(1998).Trendsinfoodlabel
useassociatedwithnewnutritionlabelingregulations.AmericanJournalofPublic
Health,88(8),12121215.
Krukowski,R.A.,HarveyBerino,J.,Kolodinsky,J.,Narsana,R.T.,&DeSisto,T.P.(2006).
Consumersmaynotuseorunderstandcalorielabelinginrestaurants.Journalof
theAmericanDieteticAssociation,106(6),917920.
Ledikwe,J.H.,ElloMartin,J.A.,&Rolls,B.J.(2005).Portionsizesandtheobesity
epidemic.JournalofNutrition,135(4),905909.

XLIV

Martin,C.K.,Anton,S.D.,YorkCrowe,E.,Heilbronn,L.K.,VanSkiver,C.,Redman,L.M.,
etal.(2007).Empiricalevaluationoftheabilitytolearnacaloriecountingsystem
andestimateportionsizeandfoodintake.BritishJournalofNutrition,98(2),439
444.
Rasmussen,L.B.,Matthiessen,J.,BiltoftJensen,A.,&Tetens,I.(2007).Characteristicsof
misreportersofdietaryintakeandphysicalactivity.PublicHealthNutrition,
10(3),230237.
Rennie,K.L.,Coward,A.,&Jebb,S.A.(2007).Estimatingunderreportingofenergy
intakeindietarysurveysusinganindividualisedmethod.BritishJournalof
Nutrition,97(6),11691176.
Riley,W.T.,Beasley,J.,Sowell,A.,&Behar,A.(2007).Effectsofawebbasedfood
portiontrainingprogramonfoodportionestimation.JournalofNutrition
EducationandBehavior,39(2),7076.
Rolls,B.J.,Roe,L.S.,Kral,T.V.E.,Meengs,J.S.,&Wall,D.E.(2004).Increasingthe
portionsizeofapackagedsnackincreasesenergyintakeinmenandwomen.
Appetite,42(1),6369.
Schapira,D.V.,Kumar,N.B.,Lyman,G.H.,&Mcmillan,S.C.(1990).Thevalueofcurrent
nutritioninformation.PreventiveMedicine,19(1),4553.
Steenhuis,I.H.M.,&Vermeer,W.M.(2009).Portionsize:reviewandframeworkfor
interventions.InternationalJournalofBehavioralNutritionandPhysicalActivity,
6(58).
Vermeer,W.M.,Steenhuis,I.H.M.,&Seidell,J.C.(2009).Fromthepointofpurchase
perspective:Aqualitativestudyofthefeasibilityofinterventionsaimedat
portionsize.HealthPolicy,90(1),7380.
Waling,M.U.,&Larsson,C.L.(2009).EnergyIntakeofSwedishOverweightandObese
ChildrenIsUnderestimatedUsingaDietHistoryInterview.JournalofNutrition,
139(3),522527.
Wansink,B.,&Chandon,P.(2006a).Healthhalos:Hownutritionclaimsinfluencefood
consumptionforoverweightandnormalweightpeople.FasebJournal,20(5),
A1008A1008.
Wansink,B.,&Chandon,P.(2006b).Mealsize,notbodysize,explainserrorsin
estimatingthecaloriecontentofmeals.AnnalsofInternalMedicine,145(5),326
332.
Wansink,B.,Painter,J.E.,&North,J.(2005).Bottomlessbowls:Whyvisualcuesof
portionsizemayinfluenceintake.ObesityResearch,13(1),93100.
Wansink,B.,&Park,S.B.(2001).Atthemovies:howexternalcuesandperceivedtaste
impactconsumptionvolume.FoodQualityandPreference,12(1),6974.
Young,L.R.,&Nestle,M.(1998).Variationinperceptionsofa'medium'foodportion:
Implicationsfordietaryguidance.JournaloftheAmericanDieteticAssociation,
98(4),458459.
XLV

Appendix8:Datacollectioninstrumentfortheobservationalstudy

Below is an example of form completed in PDA for the observational study

Case 1 2 3 4 5 6
Location Adel, H Adel, H Adel, H Adel, H Adel, H Adel, H
Day (M/ TU/ W/ T/ F) T T T T T T

Time Seated (24h) 1727 1729 1732 1739 1755 1802

Burgers
Hamburger
Cheeseburger Double 1 1
Quarter Pounder
McOz
Big Mac 1
Lean Beef Burger
McChicken Deluxe
Double Quarter Pounder
Fillet-o-fish
Type unknown or specify Name It
Fries
Small 1 1
Medium 1 1 1
Large 1
Drinks
small unsure 1 1
med unsure 1 1 1 1
large unsure 1
bottled juice
coffee
Others
3 McNuggets
6 McNuggets
10 McNuggets
Pasta Zoo Happy Meal
Type unknown
Salads
Garden
Crispy cut
Herb fusion
Type unknown
Deli-Choice Roll
Desserts
Soft-serve cone
chocolate sundae
strawberry sundae
caramel sundae
flake
muffin
apple pie
fresh apple
cookies
mcflurry
XLVI

Thickshake
small unsure
med unsure
large unsure

Sex (M/ F) m m m m m f

Age
15-25 1 1 1 1
26-35 1 1
36-45
46 and over

Weight Status
(n/ ov/ ob) n ov n n n n

Reading material (y/ n) n n n y n y

Technology (y/ n) n n n n n n

Leftovers (if any,
detail)

Group # for session 1 2
Total others in group 1 1
Details of first other
Sex m f
Age 20 25
Weight Status (n/ov/ob) n n
Details of second other
Sex
Age
Weight Status (n/ov/ob)
Details of third other
Sex
Age
Weight Status (n/ov/ob)
Details of fourth other
Sex
Age
Weight Status (n/ov/ob)

Time leave table 1742 1751 1746 1752 1803 1823
XLVII

Below are photos of McDonalds items for reference during observation















Blackwithplasticbowl
Crispyorherbfusionsalads
Smallburgerwrappedin
paperwithyellowlogos
Cheeseburger
Cubeshapedboxwithpics
Happymeal
PastaZooorCheeseburger
Rollwrappedwithgreenend
Delichoicesroll(white)
Unspecifiedvariety
Rollwrappedwithbrownend
Delichoicesroll(multigrain)
Unspecifiedvariety
Smallburgerwrappedin
paperwithpurplelogos
Hamburger
Friesinpaperpackage
Smallfries
Friesincardboardpackage
Mediumfries
Friesincardboardpackage
Largefries
Cupwithcolouredimages
Smalldrink
Cupwithredbottom
Mediumdrink
Cupwithredbottom
Largedrink
XLVIII















Boxwithred
BigMac
Boxwithdarkbrown
QuarterPounder
Boxwithmaroon/red
Double/Triplecheeseburger
DoubleQuarterPounder
Boxwithlightblue/pictures
Leanbeefburger
Boxwithlightblue
FilletOFish
Boxwithgreen
McFeast(formerlyMcOz)
Boxwithlightbrown/dark
mustard
Premiumchickenburger
Boxwithorange
McChicken
YellowSquareBox
10McNuggets
Boxwithmustardy
6McNuggets
XLIX




Below is the weight status indicator used to classify normal, overweight or obese consumers:




Normal Overweight Obese

Das könnte Ihnen auch gefallen