El análisis multivariado de datos como herramienta en la exploración de modelos en la comprensión de variables psicológicas

Abstract

Psychology, as a behavioral science in a preparadigmathic state, needs the support of theoretical models that validate the constructs that it uses for explaining the phenomena it studies. For sustaining the advances in the construction of models that arise in different domains of the Psychology, we present the application of multivarate data analysis tools such as Principal Component Analysis, Correspondence Analysis and Discriminant Analysis. We study variables such as prediction of university performance according to classical variables in Psychometrics (Seijas 1998a) and sognoscitive variables (seijas 1999b), learning style (Seijas 1998b), autoesteem (Seijas 1999a), and finally transcultural variables such as individualism and collectivism (Seijas 1999c). We want to show the usefulness that multivariate data analysis can give to behavioral sciences for making theoretical constructs with power to explain the reality, in such a way that the can be operational tools in the people interested in analyzing and handling the behavior.