#3 Four Validities

octopirecipes's version from 2016-02-22 07:38


Question Answer
construct validityhow well the operational definition of a variable captures it's conceptural definition.
statistical validitythe degress to which a study's statistical conclusions are accurate.
external validityhow well the conclusions drawn from the study apply to other people and situations (generalize.)
internal validityhow well a study can rule out alternative explanations for the results.


Question Answer
construct validity measuresfrequency claims, association claims, causal claims.
statistical significancethe probability that the obtained results were due to chance alone. (p value p<.05.)
effect sizethe strenght/degree of the relationship indicates how meaninful it is in the real worls.
stastical validaity measuresfrequency claims, association claims, causal claims.
external validity measuresfrequency claims, association claims only.
internal validity measurescausal claims.


Question Answer
how can we reduse confounding variableshold them constant, balance them across conditions, use random assigment.
random assignmentrequired that each participant has an equal chance of being in any IV group/level. (participants are put into groups at randoms.)
what random assigment doesbalances individual differences across groups.


Question Answer
commom types of measurementself report, observation/behavioral, physiologial (brain scans, biological data.).
face validityis it a plausible measure of constact (Can be used to inflence behavior, problem is subjectivity (plausible to you but not to others).)
content validitydoes the operational definition capture all relevant parts (content) of the concept? (depends again of subjectives subjects.)
criterion validityis the operation definition consistent with real world outcomes? if strong = high leves of validity. Also known as known group paradigm.
known groups paradigndoes our operational identify differencent between groups we already know are different?
convergent and divergent validityis the operational definition specifically and only measuring what we expect?
divergent validityaka discriminate validity, does the operation definition correlate with mesures of dissimilar concepts? no associations are expected between measures.
convergent validitystrong possitibe associations should exist between similar meassures.


Question Answer
test-retest realibilitydoes the test produce the same scores under identical circumstances? (most stdies, self report/physio/observational.)
interrater realiabilityis the operational definition clear enough that different people will come to the same conclusions based on the same data? (critical in observational studies.)
internal reliabilityis each item in our measure assesing the same construct? (surveys/questionnaires.)
split half realibityinvolves splitteing the test in half randomly, then comparing the average scores for the halves.
cronbach's alphacomputes the average correlation among all the items in the test.