A bifactor exploratory structural equation modeling framework for the identification of distinct sources of construct-relevant psychometric multidimensionality
| Autor: | Morin, Alexandre J. S., Arens, A. Katrin, Marsh, Herbert W. |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2016 |
| Předmět: |
050103 clinical psychology
Educational measurement Sociology and Political Science Psychometrics Example of application General Decision Sciences 050109 social psychology Psychologische Forschung Test validity Educational research Psychological research Structural equation modeling ddc:150 Germany Psychology 0501 psychology and cognitive sciences Relevance (information retrieval) Bildungsforschung Set (psychology) Deutschland 05 social sciences Strukturgleichungsmodell Confirmatory factor analysis Psychologie Modeling and Simulation Anwendungsbeispiel Construct (philosophy) Psychometrie Psychometry General Economics Econometrics and Finance Social psychology Cognitive psychology |
| Zdroj: | Structural equation modeling 23 (2016) 1, S. 116-139 |
| Popis: | This study illustrates an overarching psychometric approach of broad relevance to investigations of 2 sources of construct-relevant psychometric multidimensionality present in many complex multidimensional instruments routinely used in psychological and educational research. These 2 sources of construct-relevant psychometric multidimensionality are related to (a) the fallible nature of indicators as perfect indicators of a single construct, and (b) the hierarchical nature of the constructs being assessed. The first source is identified by comparing confirmatory factor analytic (CFA) and exploratory structural equation modeling (ESEM) solutions. The second source is identified by comparing first-order, hierarchical, and bifactor measurement models. To provide an applied illustration of the substantive relevance of this framework, we first apply these models to a sample of German children (N = 1,957) who completed the Self-Description Questionnaire (SDQ-I). Then, in a second study using a simulated data set, we provide a more pedagogical illustration of the proposed framework and the broad range of possible applications of bifactor ESEM models. (DIPF/Orig.) |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|