Chapter 07 Quantitative Research Designs

Quantitative Research DesignsChapter 7The experimental frameworkUsed to determine causationConducted in lab or other controlled setting To control for extraneous influencesIndependent variable is manipulated by researcherParticipants randomly assigned to conditions or treatmentsClassical experimentResearcher controls Selection of treatment & control groups – based on theoryRandom sampleRandom assignment of participants to conditions – ensures participants in each condition are equivalent before any treatmentManipulation checksPosttest experimental designPretest-posttestFactorial designTreatment groups based on two or more independent variablesDependent variable measured once after treatment givenCan test for main and interaction effectsLongitudinal experimental designMultiple measurements of the dependent variable across timeTime 1→ Time 2 → Time 3Time between measurements depends on the nature of the communication phenomenon studiedUsed for studying training effects or degree of retentionEvaluating experimental designsstrengthslimitationsResearcher controls manipulation of IV and random assignment to conditionAssumes that any effect on DV is the cause of the IVPrecisionNot all communication can be studied using experimentsOther external influence may be the cause of DV changeMay not reflect realityQuasi-experimental designsResearcher relies on natural variation of independent variablePosttest, pretest-posttest, factorial, and longitudinal designs can be usedIncludes field experiments in which communication is studied in contextField experimentsResearcher controls manipulation of IVSituated in real and natural environmentsAllows examination of complex and unfolding interactionEvaluating quasi-experimental designs strengths limitationsCan address real world problemsDesign based on naturally existing variationConditions may not be equivalentCannot determine cause and effectDescriptive designsAlso called cross-sectional or non-experimental studiesResearcher does not control manipulation of IVParticipants not randomly assigned to conditionsPredictor and criterion are better labels for IV and DV as they do not imply causalityOnline experimentsUsed for experimental, quasi-experimental, and descriptive designsCan design a logic for how participants respond to stimuli or answer itemsHelps to reach diverse audience of participantsCan embed audio and video elementsSampling is problematicResearcher effects and procedural biasIntroduced in interaction with participantsSimilarity of researcher to participantsHow researcher communicates to participantsResearcher unknowingly encourages desired responsesDemand characteristics created when topic has socially desirable elementsUse research protocol to detail procedural stepsComparing research designsEvaluating descriptive designsStrengths limitationsConducted in environments with participants who are like the researcher wants to generalize to.Can be used to explore complex processesNo direct control over manipulation of IVResearchder does not assign participants to experimental conditionsLimits interpretation of results