Statistical treatment of multivariate constructs in social psychology
Loading...
Files
Date
Authors
Publisher
Philipps-Universität Marburg
Abstract
This dissertation argues that multivariate constructs require multivariate statistical analyses. Collapsing distinct measures into index variables renders subsequent statistical analyses uninterpretable. This problem is illustrated with commonly used index variables for the constructs of relative ingroup prototypicality and attitude ambivalence. A multivariate approach to testing hypotheses about multivariate constructs is proposed and compared with the univariate approach in several empirical studies and computer simulations.
Review
Metadata
License
This item has been published with the following license: In Copyright