Multilinear decomposition application into students' evaluation of teaching effectiveness

Abstract

This paper presents a comparative analysis between a bilinear decomposition (Principal Component Analysis - PCA) and a multilinear decomposition (Parallel Factor - PARAFAC) on the data obtained by the students' evaluation of teaching effectiveness in an engineering course. It is known that the teaching and learning relationship is a nonlinear process related to many fundamental subjects, which are highlighted into students' evaluation of the teaching effectiveness methodology. The comparison was made in order to know which additional knowledge is obtained when the analysis takes into account the multilinear process model of this. The results shown the importance of the latent characteristics in a students' professional formation in accordance with the results presented in the application of PARAFAC decomposition, that has been reveled as a powerful tool for data analysis in the assessment area.