Analytics and predictive models of student's activity in off/on-line learning environments

Information and Communication Technologies can be used not only to provide knowledge and education, but also to collect and manage useful data that facilitate effectiveness and efficiency in instructional decision making. The aim of this study is to identify the relevant factors that influence most in teaching and the academic outcomes in engineering and, in particular, in a Higher Education Computer Technology course. For this purpose, on the one hand, online tasks developed within an online learning environment through virtual laboratories, interactive activities and game-based learning techniques have been analysed. On the other hand, participation in the face-to-face activities has also been evaluated. This study uses statistical and data mining tools to find patterns and regularities. The correlation of different parameters has been studied and models that allow predicting whether a student will pass or fail have been applied with the aim of identifying learners at risk of failure. The results obtained in this study will allow personalizing the educational experience and help in making decisions to improve learning. Furthermore, findings suggest a strong correlation among online practical activities, activities developed in the classroom and course outcomes.