Scientific theories organize and link observed factsthey imply hypotheses that offer testable predictionsGetting Ideas -> forming hypotheses –> choosing a research design –> choose subjects and measures –> analyze data –> report results -> replicateResearch DesignsDescriptive Studiesfield studiesnaturalistic observationwatching and recording the behavior in the organsm’s natural environmentPROSunobtrustivebeginning step of further researchdisconfirm a hypothesisCONSno control – can’t figure out cause/effect relationshipsobserver bias – see what you want to seecan lead to overblown claimscase studiesintensive study of single individual over a period of timePhineas Gagefrontal lobes play a part in personality/reasoningPROdepth of informationuseful for very rare populationsCONSatypical, non-representative caseoverblown claimssubjective interpretationcause/effect relations cannot be determinedSelf Report StudiesSurveyscan get a lot of information about peoplelearn about attitudesopinionsdemographicsmake sure you’re getting a representative sample of the population you want to look atCorrelational Researchcorrelation coefficient –> rpositive (same directions) or negative/inverse (one goes up and another goes down) relationship“The longer children are breast-fed, the greater their later academic achievement”“Global warming rates have increase since pirates have starting shinking in numbers” – NOCorrelation does not infer causationLurking variable could existthe more strongly correlation two variables are, the more one variable predicts the otherExperimental ResearchDetects cause-efffect relationshipsIndependent Variable (IV)ManipulatedPresumed causeDependent Variable (DV)What we measurePresumed effectDifferences in the dependent variable are thought to be caused by the independent variable“Do infants who are breast-fed have higher intelligence scores in later childhood than children who are bottle-fed with infant formula?”Experimental termsOperational definition – clearly define how we are measuring our variables“what do you mean by breast fed?” – amount? Time?“what do you mean by intelligence?” – grades? IQ? As a child? As an adult?Experimental Group vs. Control GroupExperimental group – do something different – can be many, many, but you’re comparing it to the controlControl – inertTo see if experimental is different than the controlRandom assignmentRepresentative sample of population in both groupsAre you going to use mothers who are all from PhD programs? Or mothers who all haven’t had a high school education?Confounding variablesEmpirical ResearchWhat are the variables we’re interested in?Feeding methodIntelligenceControlling for other factorsFirst, we must operationally define our variablesFeeding method – breast vs. bottleIntelligence defined as IQ score at age 8Feeding method would be the IVIf effects the outcomeIntelligence score is the DVFindingsOn intelligence tests taken at age 8, the children nourished with breast milk had significantly higher intelligence scores than their formula-fed counterpartsConfounding – breast feeding mothers more likely to be nurturing? Stuff in breast milk that isn’t in formula?Experiments aim to manipulate an independent variable, measure changes in the dependent variable, and control all other variablesRandom assignement works to equate the experimental control groups by minimizing preexisting differences between those assigned to the different groupsPsychology - Research Methods 18/04/2012 17:03:00← Scientific theories organize and link observed facts- they imply hypotheses that offer testable predictions←← Getting Ideas -> forming hypotheses –> choosing a research design –>choose subjects and measures –> analyze data –> report results -> replicate←← Research Designs- Descriptive Studieso field studies naturalistic observation watching and recording the behavior in the organsm’s natural environmentPROS unobtrustive beginning step of further research disconfirm a hypothesisCONS no control – can’t figure out cause/effect relationships observer bias – see what you want to see can lead to overblown claimso case studies intensive study of single individual over a period of time Phineas Gage- frontal lobes play a part in personality/reasoningPRO depth of information useful for very rare populationsCONS atypical, non-representative case overblown claims subjective interpretation cause/effect relations cannot be determined- Self Report Studieso Surveys can get a lot of information about people learn about attitudes opinions demographics make sure you’re getting a representative sample of thepopulation you want to look at- Correlational Researcho correlation coefficient –> ro positive (same directions) or negative/inverse (one goes up and another goes down) relationshipo “The longer children are breast-fed, the greater their later academic achievement”o “Global warming rates have increase since pirates have starting shinking in numbers” – NO Correlation does not infer causation Lurking variable could existo the more strongly correlation two variables are, the more one variable predicts the other- Experimental Researcho Detects cause-efffect relationshipso Independent Variable (IV) Manipulated Presumed causeo Dependent Variable (DV) What we measure Presumed effecto Differences in the dependent variable are thought to be caused by the independent variableo “Do infants who are breast-fed have higher intelligence scoresin later childhood than children who are bottle-fed with infant formula?”o Experimental terms Operational definition – clearly define how we are measuring our variables “what do you mean by breast fed?” – amount? Time? “what do you mean by intelligence?” – grades? IQ? As a child? As an adult? Experimental Group vs. Control Group Experimental group – do something different – canbe many, many, but you’re comparing it to the control Control – inert To see if experimental is different than the control Random assignment Representative sample of population in both groups- Are you going to use mothers who are all from PhD programs? Or mothers who all haven’t had a high school education? Confounding variableso Empirical Research What
View Full Document
Unlocking...