Fostering Evidence-Based Education with Learning Analytics: Capturing Teaching-Learning Cases from Log Data

Evidence-based education has become more relevant in the current technology-enhanced teaching-learning era. This paper introduces how Educational BIG data has the potential to generate such evidence. As evidence-based education traditionally hooks on the meta-analysis of the literature, so there are...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Educational Technology & Society Jg. 23; H. 4; S. 14 - 29
Hauptverfasser: Kuromiya, Hiroyuki, Majumdar, Rwitajit, Ogata, Hiroaki
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Palmerston North International Forum of Educational Technology & Society 01.10.2020
International Forum of Educational Technology & Society
Schlagworte:
ISSN:1176-3647, 1436-4522, 1436-4522
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Evidence-based education has become more relevant in the current technology-enhanced teaching-learning era. This paper introduces how Educational BIG data has the potential to generate such evidence. As evidence-based education traditionally hooks on the meta-analysis of the literature, so there are existing platforms that support manual input of evidence as structured information. However, such platforms often focus on researchers as end-users and its design is not aligned to the practitioners' workflow. In our work, we propose a technology-mediated process of capturing teaching-learning cases (TLCs) using a learning analytics framework. Each case is primarily a single data point regarding the result of an intervention and multiple such cases would generate an evidence of intervention effectiveness. To capture TLCs in our current context, our system automatically conducts statistical modelling of learning logs captured from Learning Management Systems (LMS) and an e-book reader. Indicators from those learning logs are evaluated by the Linear Mixed Effects model to compute whether an intervention had a positive learning effect. We present two case studies to illustrate our approach of extracting case effectiveness from two different learning contexts-one at a junior-high math class where email messages were sent as intervention and another in a blended learning context in a higher education physics class where an active learning strategy was implemented. Our novelty lies in the proposed automated approach of data aggregation, analysis, and case storing using a Learning Analytics framework for supporting evidence-based practice more accessible for practitioners.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1176-3647
1436-4522
1436-4522
DOI:10.30191/ETS.202010_23(4).0002