Indicators for predicting performance in Problem-Based Learning teamwork: analyzing Moodle data

Evangelia Triantafyllou, Eoin Ivan Rafferty, Emmanouil Xylakis

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

Abstrakt

Learning Analytics (LA) aims to improve the learning process by analyzing learning data, and communicating the results of this analysis to both educators and learners. LA has been employed in a few cases for improving Problem-Based Learning (PBL) collaborative work but the literature has yet to discuss in detail the incorporation and potential of LA in this context. In this paper, we describe our approach that aimed at developing a platform in Moodle for monitoring progress and facilitating communication during PBL collaborative work. By employing Moodle LA plugins, statistical analysis on quantitative learning data, and an inductive approach for qualitative analysis on forum discussion data, we attempted to find indicators for individual final grades. Our analysis showed that the student engagement with the platform and supervisor activity in the discussion forum are correlated with the final grades. However, there was a negative correlation with the amount of posts on the supervisor forum and grades. This suggests that teams who struggle communicate more with their supervisors. There was also a strong correlation between the total activity in the forum and the percentage of topics started by the supervisor. This indicates that an active forum is driven more by the supervisor than by the students. One drawback from this analysis is that there was a relatively small sample size, and the data was only collected across one semester. Future research should therefore address these limitations by developing LA algorithms to operate in larger datasets, but employing the indicators suggested by this study.
OriginalsprogEngelsk
TidsskriftTechnology, Knowledge and Learning
ISSN2211-1662
StatusAfsendt - 2020

Fingeraftryk Dyk ned i forskningsemnerne om 'Indicators for predicting performance in Problem-Based Learning teamwork: analyzing Moodle data'. Sammen danner de et unikt fingeraftryk.

Citationsformater