Abstract:

E-learning platforms used in Higher Education Institutions store valuable information that can be analysed. Data mining is a multidisciplinary technique that integrates computer science, education, and statistics, and that could serve to interpret results and predict academic performance through the virtualized subjects.

Objective:
To identify the use of the interactive platform based on Moodle LMS (Learning Management System) in the Occupational Health compulsory subject in the curriculum of the podiatry degree and its relationship with academic success from 2017 to 2019 courses.

Materials and methods:
Logs (files of the interactions between a system and the students in the virtual system) for analyzing a variety of information from different activities carried out through the virtual campus on a Moodle platform were extracted, depurated and prepared for analysis. Finally, 33,776 (13,818 in 2017 and 19,958 in 2018) logs were used to perform the statistical analysis using the RStudio program and the SPSS v.22 software. A descriptive analysis, Pearson correlations and continuous variable decision trees diagrams were performed to determine the use of the Moodle classroom activities and resources and their relationship with academic results obtained in this subject.

Results:
62 students enrolled in the academic course 2017-18 and 59 in the 2018-19 were studied. In the academic course of 2017-18, 62.9% were women, and mean and SD of academic results was 7.5±1.09. In 2018-19 academic course, 76.3% were women, and the mean results was 7.2±1.06. The highest peak of activity registered on the virtual subject was 200 visits in each academic course, with differences by months in relation to the distribution of tasks. The highest activity recorded was on Tuesdays and Sundays, in both years, but with more activity in the 2018-19 academic year. Lessons were the most used tools in both courses (40.3% in 2017 vs 46.2% in 2018) followed by the participation in forums (32.7% in 2017 vs 12.8% in 2018). Participation in the forums was 100% vs 93.1% and URLs entries 71.43% vs 87.93%; comparing both academic courses. In 2018, 3 new tools were introduced with high participation: Self-assessment Test 96.55%, glossary 87.93% and a wiki activity with a 48.28% of participation. Tools that significantly correlated with better test scores in year 2017 was the participation in the forums (p=0.016), while in 2018 test scores were significant correlated with the participation in the tasks (p=0.041) and self-assessment tests (p=0.044) carried out on the virtual classroom. Tree-like graphs identified two clusters of students related with forums and URLs entries, in the virtual classroom.

Conclusion:
This study reveals the importance of identifying and selecting tools with the capacity to improve and stimulate active and significant learning.

Keywords:

Higher education, Moodle platform, e-learning, student performances, Occupational health, learning analytics.