G. Castilla1
, A. Durán1 and L. Isasi1
1 Dpto. de Ingeniería Mecánica. Escuela Politécnica Superior. Universidad Carlos III de Madrid.
Keywords: Moodle, Blackboard Collaborate, log analysis, methodology, production.
1. Introduction
The analysis of the student interactions with the online systems such as Learning Management Systems (LMS) [1] and videoconferencing system can help to understand the student behaviour and improve the course design in order to foster student learning. This study presents the results of an integrated, statistical analysis of the log records extracted from Moodle Learning Management System and Blackboard Collaboration (BBC) used in Universidad Carlos III de Madrid (in the case of BBC, since the COVID-19 lockdown), for the “Design and Simulation of Production Systems” (DSSP) course during the 2019/20 academic year.
1.1. Design and Simulation of Production Systems
This is a mandatory subject in the third year of the Bachelor’s Degree in Industrial Engineering.
The main competences are to provide the students the knowledge to use simulation and linear programming to analyze and optimise production systems.
A set of 17 didactic materials was offered to the student in the Moodle throughout the course, and from the lockdown, online classes were offered in BBC (4 hours/week).
2. Statistical analysis of student activity
Analyzing the Moodle and BBC log files, it is possible to detect behavioural patterns that make possible to improve the learning process, predict students’ performance or early detect students experiencing difficulties, not motivated or at risk of dropping out the course [2].
It is possible to estimate the correlation between students’ academic performance and the use of the offered LMS resources and attendance to online classes, with the following linear model.
may = a0 + a1 midterm + a2 time + a3 recPre + a4 `1`+ a5 `2`+ a6 `6`+ a7 `14`
Multiple R-squared= 0.33, Adjusted R-squared R2adj = 0.3
Being, midterm = grade of the midterm exam (midterm) prior to the lockdown; time = time spent connected to online classes in BBC; recPre = the number of times the student accesses (before the midterm exam) the Course Presentation slides; and `N’ = number of times course resource N is accessed in Moodle (N=’1’ to ‘17’).
We conclude that students with better marks are the one who access more frequently the Presentation slides, both before (recur) and after the midterm (resource `14`), and the Continuous Linear Programming and Simulation slides, with better midterm grade, and who have attended the online classes for a longer time, obtain better marks.
3. Conclusions
The log data currently accessible on the LMS and videoconference platforms used allow us to analyse both the degree of acceptance of the different teaching resources and the relationship between their use and the final grades obtained, which allows lecturers to fine-tune their design. However, these analyses are currently limited by insufficient integration between platforms and insufficient data capture by some of them, so it is recommended, now that the universities are in the process of choosing the future platforms on which to base its new teaching model, that the mentioned criteria be explicitly included among the selection criteria.
References
- M. Zabolotniaia, Z. Cheng, E. M. Dorozhkin, and A. I. Lyzhin, “Use of the LMS Moodle for an effective implementation of an innovative policy in higher educational institutions,” Int. J. Emerg. Technol. Learn., vol. 15, no. 13, pp. 172–189, 2020.
- Y. Zhang, A. Ghandour, and V. Shestak, “Using Learning Analytics to Predict Students Performance in Moodle,” Int. J. Emerg. Technol. Learn., vol. 15, no. 20, 2020.
