Construct validity

Construct validity ( construct validity engl. ) referred to in the multivariate statistical one aspect of validity and is present when the measurement of a construct is neither corrupted by systematic errors or by other constructs. It is in a sense the question of whether the solutions adopted to measure the construct indicators behave in a manner which allows them to collectively as an " intellectual whole" interpreted.

Target

The construct validation aims to analyze a test psychological and is therefore contrary to the criteria -related validity. The focus of the construct validity lies in the theoretical clarification of what a test measures. For example, recorded an intelligence test or an ability test adopted properties or constructs.

This is a derived and not directly tangible operational unit complex. Due to lack of intelligibility, the operational construct validation is associated with considerable difficulties and large technical and economic effort. Here, theory and empiricism to each other in an interaction ratio.

And the test and the construct each other in an interaction, so that the test may alter the construct and the construct affect the structure of the test. Construct validity is therefore anchored in a much higher degree than criterion-related validity and logical in the psychology of personality fundamental research

History

The concepts of convergence and discriminant validity as aspects of construct validity were introduced by Campbell and Fiske (1959). Since then, other aspects of the construct validity have been proposed.

Statement

Construct validity is given when content validity, Konvergenzvalidität, discriminant and nomological validity can be determined and methods distortion (common -method bias) can be excluded. While content validity closes the gap between a mentally - theoretical construct and its measurement by an existing indicators of scale, convergence and discriminant validity are often determined by objective, statistically measurable indicators. Content, convergence and discriminant validity can be improved but also with the aid of jurors. Such a subjective process in which jurors on index cards noted indicators on the one hand to yourself to be selected and to be named even categories ( thus to constructs ), and secondly to assign into predetermined categories, has been introduced to improve the construct validity of Moore and Benbasat.

Criticism

The mere consideration of convergence and discriminant validity for the detection of construct validity is criticized especially by John R. Rossiter, by citing that the construct must be obtained independently of other constructs. He emphasizes the importance of content validity and sets them equal even with construct validity. Thus, measures to improve the discriminant and Konvergenzvalidität can cause indicators are removed and statistically measurable properties of the measurement models be improved by it, but to remove the measurement models simultaneously by the semantic content of their constructs. The indicators are circling construct a certain extent and the unilateral removal of indicators to improve statistical metrics might remove the construct of the measurement. John G. Wacker (2004) highlights the importance of formal conceptual definitions as the most important step before any traditional statistical validity test is performed. He describes such definitions as a necessary condition for construct validity, while statistical tests are sufficient conditions. The overall conclusion is that measures in the past to define a construct, and in particular to improve the content validity is often not necessary attention were paid while to improve purely objective statistical quality criteria such as Cronbach's alpha or the model quality of a structural equation model often at the expense of construct hasty indicators have been deleted.