TEACHING MULTIVARIATE DATA ANALYSIS IN CHEMISTRY USING SCIENTIFIC PAPERS

Multivariate data analysis constitutes an important part of chemometrics, which is a very useful tool for a chemist, especially to optimize chemical methodologies as well as to interpret and classify the obtained results. However, very frequently, university students find this subject very difficult to understand and to recognize their utility for their future work.
As an innovative educational experience we have used selected scientific papers for teaching multivariate data analysis in the last courses of the degree in chemistry. In the very beginning, magisterial lessons were given in order to state the basis, significance and interpretation of these chemometric tools which included principal component analysis, cluster analysis, artificial neural networks, etc. These lessons also incorporated the use of suitable software for data analysis. Afterwards, students were required to prepare at home and present in the class a short oral communication (approx. 10 min) on a selected paper. Further discussion took place between the professor and the students. Results showed that the proposed experience highly contributed to improve the knowledge of the students on this subject and to solve their doubts as well as to improve their skills on oral communications and results presentation. This innovative experience is highly related to the new objectives according to the European Space of Higher Education. A detail presentation of the methodology and discussion of the results will be provided.