市场调查方法(英文版)第十章

PowerPointPresentationbyCharlieCookTheUniversityofWestAlabamaWilliamG.ZikmundBarryJ.BabinPart3Measurement3rdEdition©2007Thomson/South-Western.Allrightsreserved.Chapter10MeasurementandAttitudeScaling©2007Thomson/South-Western.Allrightsreserved.10–2LEARNINGOUTCOMES1.Explainwhatneedstobemeasuredtoaddressaresearchquestionorhypothesis2.Defineoperationalization3.Distinguishlevelsofscalemeasurement4.Explaintheneedforindexorcompositemeasures5.Listthethreecriteriaforgoodmeasurement6.Explainthesignificanceofscalereliabilityandvalidity7.DescribehowmarketingresearchersthinkofattitudesAfterstudyingthischapter,youshouldbeableto©2007Thomson/South-Western.Allrightsreserved.10–3LEARNINGOUTCOMES(cont’d)8.Identifybasicapproachestomeasuringattitudes9.Discusstheuseofratingscalesformeasuringattitudes10.Representalatentconstructbyconstructingasummatedscale11.SummarizewaystomeasureattitudeswithrankingandsortingtechniquesAfterstudyingthischapter,youshouldbeableto©2007Thomson/South-Western.Allrightsreserved.10–4WhatDoIMeasure•MeasurementTheprocessofdescribingsomepropertyofaphenomenonofinterest,usuallybyassigningnumbersinareliableandvalidway.•ConceptAgeneralizedideaaboutaclassofobjects,attributes,occurrences,orprocesses•OperationalDefinitionSpecifieswhattheresearchermustdotomeasuretheconceptunderinvestigation©2007Thomson/South-Western.Allrightsreserved.10–5EXHIBIT10.1AreThereAnyValidityIssueswiththisMeasurement?©2007Thomson/South-Western.Allrightsreserved.10–6OperationalDefinitions•OperationalizationTheprocessofidentifyingscalesthatcorrespondtovarianceinaconcepttobeinvolvedinaresearchprocess.•ScalesAdeviceprovidingarangeofvaluesthatcorrespondtodifferentvaluesinaconceptbeingmeasured.•CorrespondencerulesIndicatethewaythatacertainvalueonascalecorrespondstosometruevalueofaconcept.©2007Thomson/South-Western.Allrightsreserved.10–7OperationalDefinitions(cont’d)•VariableAnythingthatvariesorchangesfromoneinstancetoanother;canexhibitdifferencesinvalue,usuallyinmagnitudeorstrength,orindirection.•ConstructsConceptsmeasuredwithmultiplevariables.©2007Thomson/South-Western.Allrightsreserved.10–8EXHIBIT10.2MediaSkepticism:AnOperationalDefinition©2007Thomson/South-Western.Allrightsreserved.10–9LevelsofScaleMeasurement•ScaleAseriesofitemsarrangedalongacontinuousspectrumofvaluesforthepurposeofquantification.Properties:uniquelyclassify,preserveorder,setequalintervals,andhaveanaturalzero.•TypesofScalesNominal:classifies/identifiesbyaqualityoftheobjectOrdinal:classifiesusingarankingorderofobjectsInterval:quantifiesobjectsinorderonacontinuumRatio:classifiesbycomparisontoastandardCategorical:classifiesbyaspecificcharacteristic©2007Thomson/South-Western.Allrightsreserved.10–10EXHIBIT10.3Nominal,Ordinal,Interval,andRatioScalesProvideDifferentInformation©2007Thomson/South-Western.Allrightsreserved.10–11EXHIBIT10.4FactsAbouttheFourLevelsofScales©2007Thomson/South-Western.Allrightsreserved.10–12EXHIBIT10.4FactsAbouttheFourLevelsofScales(cont’d)©2007Thomson/South-Western.Allrightsreserved.10–13IndexMeasures•AttributesSinglecharacteristicsorfundamentalfeaturesthatpertaintoanobject,person,orissue.•IndexMeasuresAssignavaluebasedonhowmuchoftheconceptbeingmeasuredisassociatedwithanobservation.Indexesoftenareformedbyputtingseveralvariablestogether.•CompositeMeasuresAssignavaluetoanobservationbasedonamathematicalderivationofmultiplevariables.©2007Thomson/South-Western.Allrightsreserved.10–14ThreeCriteriaforGoodMeasurementSensitivityReliabilityValidityGoodMeasurement©2007Thomson/South-Western.Allrightsreserved.10–15ReliabilityversusValidity•ReliabilityThedegreetowhichmeasuresarefreefromrandomerrorandthereforeyieldconsistentresults.Anindicatorofameasure’sinternalconsistency.•ValidityTheaccuracyofameasureortheextenttowhichascoretruthfullyrepresentsaconcept.Doesascaletomeasurewhatwasintendedtobemeasured?©2007Thomson/South-Western.Allrightsreserved.10–16EXHIBIT10.5ReliabilityandValidityonTarget©2007Thomson/South-Western.Allrightsreserved.10–17MeasurementAccuracy:Sensitivity•SensitivityAmeasurementinstrument’sabilitytoaccuratelymeasurevariabilityinstimuliorresponses.Compositemeasuresallowforagreaterrangeofpossiblescores,theyaremoresensitivethansingle-itemscales.Sensitivityisgenerallyincreasedbyaddingmoreresponsepointsoraddingscaleitems.©2007Thomson/South-Western.Allrightsreserved.10–18AttitudesinMarketingResearch•AttitudeAnenduringdispositiontoconsistentlyrespondinagiventovariousaspectsoftheworld.•AttitudesasHypotheticalConstructs“Hypotheticalconstruct”describesalatentorunobservablevariablethatismeasurableonlybyanindirectmeanssuchasverbalexpressionorovertbehavior—attitudesareconsideredtobesuchvariables.©2007Thomson/South-Western.Allrightsreserved.10–19ComponentsofanAttitudeBehavioralAffectiveCognitiveAttitude©2007Thomson/South-Western.Allrightsreserved.10–20ComponentsofanAttitude•AffectiveComponentThefeelingsoremotionstowardanobject•CognitiveComponentKnowledgeandbeliefsaboutanobject•BehavioralComponentPredispositiontoactionIntentionsBehavioral