Abstract:

The ability to solve a completely new issue is key to the development of the professional activity of a university degree, particularly when continuous innovation is demanded to achieve a more effective production model and adapted to a changing global market. These ideas are at the core foundation of the European Higher Education Area and should be practiced actively by students during the university learning. This is the goal of the learning activity described below.

Inside the framework of the teaching innovation project funded by University of Burgos (2010) an e-learning resource made up by 12 groups of documents (texts and presentations) were developed with a conceptual map that relates them. The resource is made available in a comprehensive resource of a community in the teaching of the UBU on-line platform. The purpose of this resource is incorporate interdisciplinarity in different teaching-learning activities, focusing on the specific content "molecular spectroscopy in the UV/visible, IR and NIR."

As the resource is already active, a project was designed for "Numerical computation and applied statistics" (fourth semester in the Chemistry degree) to make students learn about 'Non-linear regression" in an autonomous way.
The project is given to each student after completing the related issues concerning linear univariate and multivariate regression and experimental design. In the project it is explicitly stated the resources of the on-line platform that they should at least use, namely: i) A document about mathematical modelling with differential equations of chemical kinetics (that they have already studied in the first course), ii) a PowerPoint presentation describing the handling of Statgraphics (specific software) for non-linear regression by using experimental data of a kinetic with spectroscopic monitoring. With these elements, the student must solve and validate several problems about non-linear regression, all of them with experimental data. The problems are posed in an itemized guide with specific questions that should be reasonably answered.

Additionally, the student is asked to explicitly write i) three differences and three similarities between linear and non-linear regression both fitted by least squares; ii) three questions among those posed in the project that the own student can not clearly explain, and iii) three questions that he can not even understand. From the responses and comments on the 12 questions posed to each student in the work done and delivered to the teacher, a seminar about non-linear regression is scheduled.

From the learning point of view, this approach implies to learn without a teaching programming based on a logic-formal structure of the subject that traditionally would be exposed along some theoretic sessions prior to some practical sessions before entering in the questions that arise when solving a real problem. By using related skills (multivariate linear regression and validation) and computational tools (non-linear regression with Statgraphics) the aim is to solve a completely new problem that shares the methodological foundation: to make minimum a sum of squares.

The project given to the students end with seven questions (scored from 1 to 5) to evaluate the learning methodology.

Acknowledgments: We thank University of Burgos for the funding granted under the 2011 aids to educational innovation projects.

Keywords:

e-learning, European Higher Educaton Area, autonomous learning, interdisciplinary, Chemistry Degree.