Classical test theory is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Read the full entry →Research Wiki · Methods Lab
Reliability, validity and measurement science
Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation.
Methods family
Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation.
Use methods only after defining the estimand, data structure, assumptions, validation plan and decision consequence.
Reliability analysis is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Read the full entry →Cronbach’s alpha summarises internal consistency under assumptions that are often stronger than users realise. A high alpha does not prove unidimensionality, validity or item quality.
Read the full entry →McDonald’s omega estimates scale reliability from a factor model and can be more appropriate than alpha when item loadings differ. Its interpretation depends on a defensible measurement model.
Read the full entry →Split-half reliability is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Test-retest reliability is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Read the full entry →Inter-rater reliability is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Cohen’s kappa is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Read the full entry →Fleiss’ kappa is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Intraclass correlation is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Construct validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Convergent validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Discriminant validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Criterion validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Predictive validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Concurrent validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Face validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Content validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Nomological validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Item analysis is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Item-total correlation is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Scale purification is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Item-response theory is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Rasch modelling is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Two-parameter logistic IRT is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Three-parameter logistic IRT is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Graded-response models is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Differential item functioning is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Measurement invariance is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Common-method bias is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Response-style analysis is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Acquiescence adjustment is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Extreme-response-style analysis is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Browse linked method entries
Classical test theory
Classical test theory is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodReliability analysis
Reliability analysis is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodCronbach’s alpha
Cronbach’s alpha summarises internal consistency under assumptions that are often stronger than users realise. A high alpha does not prove unidimensionality, validity or item quality.
Open methodMcDonald’s omega
McDonald’s omega estimates scale reliability from a factor model and can be more appropriate than alpha when item loadings differ. Its interpretation depends on a defensible measurement model.
Open methodSplit-half reliability
Split-half reliability is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodTest-retest reliability
Test-retest reliability is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodInter-rater reliability
Inter-rater reliability is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodCohen’s kappa
Cohen’s kappa is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodFleiss’ kappa
Fleiss’ kappa is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodIntraclass correlation
Intraclass correlation is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodConstruct validity
Construct validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodConvergent validity
Convergent validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodDiscriminant validity
Discriminant validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodCriterion validity
Criterion validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodPredictive validity
Predictive validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodConcurrent validity
Concurrent validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodFace validity
Face validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodContent validity
Content validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodNomological validity
Nomological validity is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodItem analysis
Item analysis is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodItem-total correlation
Item-total correlation is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodScale purification
Scale purification is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodItem-response theory
Item-response theory is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodRasch modelling
Rasch modelling is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodTwo-parameter logistic IRT
Two-parameter logistic IRT is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodThree-parameter logistic IRT
Three-parameter logistic IRT is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodGraded-response models
Graded-response models is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodDifferential item functioning
Differential item functioning is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodMeasurement invariance
Measurement invariance is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodCommon-method bias
Common-method bias is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodResponse-style analysis
Response-style analysis is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodAcquiescence adjustment
Acquiescence adjustment is a statistical or analytical concept within reliability, validity and measurement science. It should be selected for the data-generating process and decision question rather than because software makes it available.
Open methodExtreme-response-style analysis
Extreme-response-style analysis is a method within reliability, validity and measurement science. Methods for evaluating whether scales and measures are consistent, valid, invariant and fit for the intended interpretation. Its usefulness depends on data structure, assumptions, validation and whether the output answers the intended decision.
Open methodNo entries match this search.