研究设计之测量.ppt

测量的准确度Towhatextentareyoumeasuringwhatyouintendedtomeasure?Ideal:Measurementerroris0Measurement=Intendedcharacteristic+Systematicerror+RandomerrorSystematicerrorisdurableRandomerroristemporaryPre-test=MathAbility+____+_____ 准确度:信度Reliability:producesameresultsoverrepeatedapplications.StabilityandConsistencyTest-Retestmethod再测信度 *Twotimes *maturation *sensitivityAlternativeformat复本信度 *twotimes *maturationSplit-Halvesmethod(multiplemeasures,oneconcept)折半信度Sub-samplemethod亚样本信度4Participant:knowledge,traumaticexperience,reluctant;fatigue,boredon,anxiety,distractionshungerimpatienceSF:anonymity;curbsiderointerceptinterivewsvs.homeinterviewsMeasurer:rewording,parapharing;stereotypesinappearance;carelesswriting;machineprocessing;incorrectcodingInstrument:confusing,ambiguous,pooritemsParticipant:knowledge,traumaticexperience,reluctant;fatigue,boredon,anxiety,distractionshungerimpatienceSF:anonymity;curbsiderointerceptinterivewsvs.homeinterviewsMeasurer:rewording,parapharing;stereotypesinappearance;carelesswriting;machineprocessing;incorrectcodingInstrument:confusing,ambiguous,pooritems研究设计之

测量SchoolofPublicAdministrationPolicyDr.KaifengYangIssues概念操作化概念化指标化测量的层次复合指数测量测量的好坏：效度与信度提高效度与信度的途径概念操作化沟通绩效?目标清晰度?沟通绩效?目标清晰度?例子从名义/概念定义到操作定义菜谱Adefinitionthatissimplyassignedtoatermwithoutanyclaimthatthedefinitionrepresentsa“real”entity.ItisA)definedbyothers;b)whatitisvs.whatitisnot;c)shared;d)realisticandsensible.Nominal(Conceptual)Definition01SpecifiespreciselyhowaconceptwillbemeasuredOperational(Working)Definition02概念化 Theprocess…wespecifywhatwemeanwhenweuseparticularterms.Itinvolvesdescribingtheessentialfeaturesoftheterms.Sometimes,itinvolvesindicatorswe’llbeusingtomeasureourconceptandthedifferentaspect