Analytical methods for measurement error estimation in survey data This dissertation aims at the domain of measurement error in social science survey data. To conceptualize and estimate measurement errors it employs the analytical and theoretical framework that stems from the analytical method of structural equation modeling (SEM) and Classical Test Theory (CTT), extended with the component of the systematic measurement error. This thesis has two goals that may contribute to development and extension of Czech social science methodology. The first goal is to illustrate methods of measurement error estimation, which has not been used for analysis of Czech data yet, and to point out to some problematical aspects of these methods. The second goal is to employ presented methods to obtain new findings regarding the quality of data from Czech surveys. The dissertation presents three empirical studies, each of which uses one of the methods defined within the presented theoretical and analytical framework. First study presents an analysis of reliability of measurement with the Quasi Simplex Model (QSM). It illustrates how to use the model and brings optimistic results regarding the reliability of the Czech EU SILC panel data. In the second study the confirmatory factor analysis model, operationally called...
| Identifer | oai:union.ndltd.org:nusl.cz/oai:invenio.nusl.cz:353506 |
| Date | January 2017 |
| Creators | Chylíková, Johana |
| Contributors | Vinopal, Jiří, Kreidl, Martin, Buriánek, Jiří |
| Source Sets | Czech ETDs |
| Language | Czech |
| Detected Language | English |
| Type | info:eu-repo/semantics/doctoralThesis |
| Rights | info:eu-repo/semantics/restrictedAccess |
Page generated in 0.0022 seconds