Monitoring online collaborative learning networks to guide an informed intervention
Mohammed Saqr, Uno Fors, Matti Tedre, Jalal Nouri
PLOS ONE
Network design
Nodes
students and teachers participating in three online medical courses
Ties
directed discussion interactions recorded by the learning platform before and after the mid-course change
Methods
The study combined network visualizations, density, degree and clustering measures with participation-role classifications, then compared the pre-midterm and post-midterm network periods.
Source: PLOS ONE
Reviewed summary
Interaction logs from three medical courses with 82 students and three teachers were mapped before and after a mid-course intervention. Participation roles, density, in-degree, and clustering increased after five targeted teaching actions, but the observational time comparison cannot isolate the intervention from normal course development. The paper demonstrates how interpretable network diagnostics can inform support while a course is still running, instead of waiting for final grades or using raw message counts alone.
Active network roles increased from 15 to 40 while non-participatory roles fell from 67 to 32, and the later networks showed higher density, in-degree, and clustering. The three courses were not randomized and had no concurrent untreated control. Instructor actions, assessment timing, maturation, and changing course demands coincide with the before-after contrast, so the observed network shift cannot be attributed causally to the intervention.
SNA
How SNA was used
The study defined students and teachers participating in three online medical courses as nodes and directed discussion interactions recorded by the learning platform before and after the mid-course change as ties. The study combined network visualizations, density, degree and clustering measures with participation-role classifications, then compared the pre-midterm and post-midterm network periods.
Source
PLOS ONE, 13(3), e0194777
DOI: 10.1371/journal.pone.0194777
