Looks like no one added any tags here yet for you.
natural settings,
Study behaviors inā exploreĀ uniqueĀ orĀ rareĀ occurrences,Ā or sample personal information.
HighĀ inĀ internalĀ Validity
ā¢ Degree toĀ whichĀ aĀ researchĀ design allows us to make casualĀ statements
HighĀ inĀ ExternalĀ Validity
ā¢
ā¢ Generalizability or applicably to people andĀ situations outsideĀ theĀ research setting
naturall to controllingĀ
theĀ degreeĀ ofĀ manipulationĀ ofĀ antecedentĀ Ā conditions variesĀ fromĀ lowĀ toĀ highĀ (lettingĀ whateverĀ happensĀ ā toĀ ā it
responses
theĀ degreeĀ ofĀ impositionĀ ofĀ units
ā¢ theĀ extentĀ toĀ whichĀ theĀ researcherĀ limitsĀ theĀ ā aĀ subjectĀ may
contributeĀ toĀ theĀ data
theĀ degreeĀ ofĀ manipulationĀ ofĀ antecedentĀ Ā conditions
variesĀ fromĀ lowĀ toĀ highĀ (lettingĀ whateverĀ happensĀ naturallyĀ toĀ controllingĀ it
theĀ degreeĀ ofĀ impositionĀ ofĀ units
ā¢ theĀ extentĀ toĀ whichĀ theĀ researcherĀ limitsĀ theĀ responsesĀ aĀ subjectĀ may
contributeĀ toĀ theĀ data
Phenomenology
ā¢ theĀ descriptionĀ ofĀ anĀ individual'sĀ immediateĀ experience;
CaseĀ study
ā¢ descriptive record of a single individual's experiencesĀ or behaviorsĀ or both keptĀ by an outside observer
FieldĀ studies
ā¢ areĀ nonexperimentalĀ approachesĀ usedĀ inĀ theĀ fieldĀ orĀ inĀ real-lifeĀ settings.
ArchivalĀ study
ā¢ aĀ descriptiveĀ methodĀ inĀ whichĀ alreadyĀ existingĀ recordsĀ areĀ reexaminedĀ forĀ aĀ new purpose.
QualitativeĀ research
ā¢ reliesĀ onĀ wordsĀ ratherĀ thanĀ numbers;Ā studyĀ phenomenaĀ thatĀ areĀ contextual
SurveyĀ research
ā¢ useful way of obtaining information about peopleās opinions, attitudes, preferences, andĀ behaviorsĀ simply by asking
StepĀ 1
:Ā mapĀ outĀ yourĀ researchĀ objectives,Ā makingĀ themĀ asĀ specificĀ asĀ possible;
Step 2
design the surveyĀ items; decide how youĀ are going to addressĀ the imposition of units
StepĀ 3
Ā onceĀ theĀ questionsĀ haveĀ beenĀ designedĀ theyĀ needĀ toĀ beĀ pretested
LevelsĀ ofĀ Measurement
ā¢ theĀ kind of scaleĀ usedĀ toĀ measureĀ a responseĀ for a closedĀ question;Ā differentĀ statisticalĀ testsĀ are required for different levels of measurement;Ā four kinds of scales:
Nominal
ā¢ classifiesĀ responseĀ items intoĀ two or more distinctĀ categoriesĀ onĀ the basisĀ of some common feature.
Ordinal
1.
ā¢ RankĀ orderingĀ ofĀ responseĀ items.
Interval
ā¢ measuresĀ magnitudeĀ orĀ quantitativeĀ sizeĀ usingĀ measuresĀ withĀ equalĀ intervalsĀ between
theĀ valuesĀ andĀ hasĀ noĀ trueĀ zeroĀ point.
Ratio
ā¢ EqualĀ intervalsĀ betweenĀ allĀ valuesĀ andĀ aĀ trueĀ zeroĀ point.
ContinuousĀ Dimension
when variablesĀ lend themselvesĀ to differentĀ levels ofĀ measurement;Ā traits,Ā attitudes,Ā and preferences are all continuous
SemanticĀ Differential
q evaluatingĀ variable onĀ a numberĀ ofĀ dimensions;Ā two adjectives
Likert
q presentĀ aĀ positivelyĀ wordedĀ statementĀ withĀ aĀ negativelyĀ wordedĀ statement
ResponseĀ Styles
ā¢ areĀ tendenciesĀ toĀ respondĀ toĀ questionsĀ orĀ testĀ itemsĀ inĀ specificĀ ways,Ā regardless
ofĀ theĀ content
Reliability
q theĀ extentĀ toĀ whichĀ theĀ surveyĀ isĀ consistentĀ andĀ repeatable
Validity
q theĀ extentĀ toĀ whichĀ aĀ surveyĀ actuallyĀ measuresĀ theĀ intendedĀ topic
Sampling
q
q decidingĀ whoĀ orĀ whatĀ theĀ subjectsĀ will beĀ and,Ā then,Ā selectingĀ them
Population
q
q allĀ people, animals,Ā orĀ objects thatĀ haveĀ atĀ leastĀ oneĀ characteristicĀ in common; ex: all undergraduate student
SampleĀ ofĀ subjects
q
q aĀ groupĀ thatĀ isĀ aĀ subsetĀ ofĀ theĀ populationĀ ofĀ interest
Representativeness
q
howĀ closelyĀ theĀ sampleĀ mirrorsĀ theĀ largeĀ population
ProbabilityĀ Sampling
q
ā¢ involvesĀ selecting subjects in such a way that the oddsĀ of their being in the study are known or can be calculated.
SimpleĀ randomĀ sampling
1.
ā¢ aĀ portionĀ ofĀ theĀ wholeĀ populationĀ is selectedĀ inĀ anĀ unbiasedĀ way
SystematicĀ randomĀ sampling
ā¢ allĀ membersĀ ofĀ theĀ populationĀ areĀ knownĀ andĀ canĀ beĀ listedĀ inĀ anĀ unbiasedĀ way
StratifiedĀ randomĀ sampling
1.
ā¢ obtainedĀ byĀ randomlyĀ samplingĀ fromĀ peopleĀ inĀ eachĀ subgroupĀ inĀ theĀ same
proportionsĀ asĀ theyĀ existĀ inĀ theĀ population
ClusterĀ sampling
1.
ā¢ sampleĀ entireĀ clustersĀ orĀ naturallyĀ occurringĀ groupsĀ thatĀ existĀ withinĀ theĀ population
CorrelationalĀ Designs
ā¢ used to establish relationships among preexisting behaviors and can be used to predict one set of behaviorsĀ from others
Quasi-ExperimentalĀ Design
q CalledĀ ānatural experimentā,Ā butĀ they
lack one or more of the essential elements, such as manipulation of antecedents or randomĀ assignment to treatmentĀ conditions.
CorrelationalĀ Study
designedĀ to determineĀ theĀ correlation,Ā or degree ofĀ relationship,Ā between two traits,Ā behaviors, or events.
SimpleĀ Correlation
ā¢
ā¢ RelationshipsĀ betweenĀ pairs ofĀ scoresĀ fromĀ eachĀ subject
Scatterplot
ā¢
ā¢ visualĀ representationsĀ ofĀ theĀ scoresĀ belongingĀ toĀ eachĀ subjectĀ inĀ theĀ study
RegressionĀ lines
ā¢
ā¢ directionĀ ofĀ lineĀ correspondsĀ toĀ theĀ directionĀ ofĀ theĀ relationship
CoefficientĀ OfĀ DeterminationĀ (R2)
ā¢ estimates theĀ amountĀ of variability in scores onĀ oneĀ variable thatĀ can beĀ explainedĀ byĀ theĀ other variable; how muchĀ one variable can explain theĀ variability in scores of theĀ otherĀ variable
PathĀ Analysis
ā¢ method that can be used when subjects are measured on several related behaviors
Cross-laggedĀ PanelĀ Design
ā¢
ā¢ usesĀ relationshipsĀ measuredĀ over
timeĀ toĀ suggestĀ theĀ causalĀ path
ExĀ PostĀ FactoĀ Studies
q
ā¢ ExploresĀ characteristics,Ā behaviors,Ā or effects of naturally occurring eventsĀ in preexisting groups of subjects.
NonequivalentĀ GroupĀ Designs
q
ā¢ Compares the effects of differentĀ treatmentĀ conditions on preexisting groups of subjects.
LongitudinalĀ Design
q
ā¢ InvestigatesĀ changes across timeĀ by measuring behaviorĀ of same groupĀ of subjects at differentĀ points in time.
Cross-sectionalĀ Studies
q
ā¢ InvestigatesĀ changes across timeĀ by comparing groups of subjects alreadyĀ at differentĀ stages atĀ aĀ single pointĀ in time.
Pretest/PosttestĀ Design
q
ExploresĀ the effects of an eventĀ (or treatment) by comparing behaviorĀ before and after the event (or treatment).
SolomonĀ 4-groupĀ Design
q
number of comparison groups are needed --> a nonequivalentĀ control group (tookĀ both pre and post tests but was not exposedĀ to the "treatment",Ā a group that receivedĀ theĀ treatmentĀ andĀ tookĀ onlyĀ theĀ posttest,Ā andĀ a posttest-onlyĀ group