Abstract:

As a complementary tool for learning Calculus, we have been using SIACUA (see [1,2]), a computer system designed to help autonomous learning, which is based on the use of parameterized questions with detailed solutions, also parameterized, with a Bayesian user model (see [3]) for feedback. This system has proven to be effective in keeping the students working during the classes’ period, as the recent data usage we present in this article shows.

A difficulty in flipped learning is to guarantee that students study before the classes. We address this issue by trying to provide the best possible conditions for this autonomous work to occur. Hence, we propose an expansion of SIACUA, by combining it with the model MOTRAC (see [4]), a model for creating learning objects and learning trajectories for meaningful learning, based on Meaningful Learning Theory and Cognitive Load Theory, in order to achieve good guidance, together with the already existing Bayesian feedback, in a flipped learning set, with active learning occurring in the classes.

In the proposed learning set we also use an assessment computer system: PmatE (see [5,6]). System PmatE is being used from 1989, mainly in the yearly science competitions, that nowadays join in our University, in the three days of the competitions, about ten thousand students, from all ages, from basic to secondary schools.

In order to achieve some extrinsic motivation, and to further guarantee the students study before the classes, we use system PmatE, which imports contents from SIACUA, for assessment in the end of each class.

Past experiences shows two main advantages of using two completely different computer systems, one for learning and another one for assessment are: (i) students are more motivated to use the learning system because they know its contents or similar are going to be used for assessment; (ii) students are not afraid of answering questions in the learning system because they know that the diagnosis on this system is not going to be used for assessment and so it is a useful and safe feedback.

References:
[1] SIACUA - Interactive Computer Learning System, University of Aveiro. http://siacua.web.ua.pt, Accessed 26.04.17.
[2] L. Descalço, P. Carvalho, J.P. Cruz, P. Oliveira, D. Seabra. Using Bayesian Networks and Parametrized Questions in Independent Study, EDULEARN15 Proceedings, , International Academy of Technology, Education and Development (IATED), 3361-3368, 6th - 8th July 2015, 2015, Barcelona, Spain.
[3] Eva Millán, L. Descalço, Gladys Castillo, Paula Oliveira and & Diogo, Using Bayesian networks to improve knowledge assessment, Computers and Education, Volume 60, Issue 1, January 2013, Pages 436-447.
[4] Canto Filho, A.B. Do (2015). MOTRAC - Conceptual learning trajectory model (Doctoral dissertation, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil).
[5] PmatE (Projeto Matemática Ensino), University of Aveiro. http://pmate.ua.pt. Accessed 26.04.17.
[6] J. Camejo, A. Silva, L. Descalço, P. Oliveira. ModelMaker, a Multidisciplinary Web Web Application to Build Question Generator Models From Basic to Higher Education. EDULEARN16 Proceedings, Academy of Technology, Education and Development (IATED), 5095-5103, 4-6 July, 2016, Barcelona, Spain.

Keywords:

Bayesian model; parametrized; trajectories; flipped learning