| Popis: |
The aim of this article is to highlight the importance of parsimony in education research (pedagogy) by using factor analysis. It elaborates some requirements for the use of factor analysis in pedagogical research, such as: testing if data distribution is normal, ratios among sample and variable sizes and choosing the adequate scale. The postulates necessary for factor analysis stated above are often neglected in pedagogical research (but also in other social sciences) and therefore the results of the factor analysis are not statistically justified. Factor analysis, as a multivariate analysis, allows researchers when exploring the complexity of the areas dealt with pedagogy as critical and creative science of education to see the latent areas that are obtained from inter correlated set of manifest variables. The obtained latent dimensions (factors) that are operated with in further data processing, are scientifically a “higher” contribution to the field of cognitive phenomena that are investigated within pedagogy. Although factor analysis was first used in psychology, it now finds its application in many social and similar sciences. The emphasis of this article is the necessity of using the factor analysis in pedagogy as a substitute for the univariate approach which is currently more frequent in pedagogical research. |
