Research Design and Analysis

Early Positivism (Comte)belief that society can be studied scientificallySocial Darwinismsurvival of the fittest (most adaptable) -application of Darwin's principles to human beings (class/gender/race)Conflict Paradigmfundamental frame of reference for human beings is conflict -history is about competing groupsSymbolic Interactionismmicrolevel, focuses on communication and meaning -how interactions affect us on an individual level -primary and secondary groupsEthnomethodologists-methods of the people -people create social structures and this explains why what is and what was does not stay the same -what is good/bad is determined through human interactionStructural Functionalismeverything has a function -organizations and societies are viewed as organismsFeminist Paradigmsissues of gender are at the heart of political behaviorCritical Race theoryrace is at the heart of political behaviorCritical theorycritique conventional wisdom -race, gender, class theories -conventional wisdom doesnt measure up to realityPost modernismsubjectivity of human beings will always confound the study of human behaviorAssumptions of science1- Nature is orderly 2- We can know nature 3- All natural phenomenon have natural causes 4- Nothing is self evident, it must be demonstrated 5- knowledge is based on experience 6-knowledge is better than ignorance3 Ethical Considerations-voluntary participation -do no harm -confidentialityindependent variablecauseDependent variableThe measurable effect, outcome, or response in which the research is interested.Intervening variablealternative cause4 rules of causation1-explanation must make sense 2-cause must precede effect 3-variable must be empirically correlated 4-cannot be explained away by an intervening variablenecessary conditionin order for an effect to come out, this condition must be presentsufficient causeenough to bring about an effect, but not necessary, could come about without itCross Sectional Studiesstudies based on observations that represent a single point in time -evidence becomes less useful overtimeLongitudinal Studiescollect data over different points in time trend: characteristics of a population over time cohort: study a specific subset overtime panel: study same people in multiple points of time -expensive (money and time)Steps of Research Process-Introduction -Lit Review -Theoretical Framework -Methods -Results -Analysis -Discussion -ConclusionConceptualizationadding precision to your concepts -interest, idea, theoryOperationalizationhow are you going to measure your conceptsLevels of Measurement-Nominal: variable whos attributes have only 2 characteristics (exhaustiveness, mutual exclusiveness) -Ordinal: variable has attributes that can be ranked -Interval: variable has attributes that can be ranked and we can measure the difference between rankings -Ratio: have a true zero pointReliabilitydo you get the same results every timeValidityare you measuring what you think you're measuringTypes of Validity-Face: on its face it seems reasonable -Criterion Related: when there is an alternative measure that supports -Construct: does your measure relate to variables as expected within a theoretical framework -Content: degree to which your measure covers the whole concept