LEARNING ANALYTICS: MEASURING STUDENTS’ LEARNING PERFORMANCE WITH A PROBABILITY MODEL FOR DISCRETE CHOICES
National Institute for Documentation, Innovation and Educational Research (ITALY)
About this paper:
Conference name: 14th International Technology, Education and Development Conference
Dates: 2-4 March, 2020
Location: Valencia, Spain
Abstract:
Educational research faces an important challenge: the need to manage a huge amount of data about the actors involved in the educational context. The Learning Analytics topic and the tools used to apply it are part of this scenario. This article presents the Unitary Programming Management System as a tool for collecting, managing and monitoring the National Operational Program 2014/2020 “For the School: Skills and Environments of Learning”. More specifically, in order to measure the learnings performance of students who participated in an educational activity focussed on basic skills, a probability model for binary choices was implemented to measure performance improvement in four subject areas:
1) Mother tongue;
2) Mathematics;
3) Foreign Languages; and
4) Scientific Area.
The data refer to the grades attributed to students in the four subject areas before and after the delivery of the educational activities. Results show that the probability of registering a training success triggered by the training course is greater for the foreign languages area, generating inclusion and social integration mechanisms, as well as mediation and intercultural understanding.Keywords:
Learning Analytics, Students’ learning performance, National Operative Program, Discrete choices model, Education.