factors shaping learners interactions in networked learning
TRANSCRIPT
Factors shaping learners’ interactions in networked learning context
Srećko Joksimović[email protected]
@s_joksimovic
Dragan Gašević[email protected]
@dgasevic
Vitomir Kovanović[email protected]
@vkovanovic
C X- Open design- Learner centered- Use of social media- Distributed communication
- Fixed design- Focused on learner-content Interaction- Video lectures- Peer assessment
2023-04-15 Slide 2 of 23
2023-04-15
cMOOC study context
Social capital
Slide 3 of 23
2023-04-15
cMOOC study context
Social capital
Social network analysis
Slide 4 of 23
2023-04-15
cMOOC study context
Information exchange
Slide 5 of 23
2023-04-15
cMOOC study context
• Language and the quality of social ties are mutually dependent.
Slide 6 of 23
xMOOC study context
2023-04-15 Slide 7 of 23
2023-04-15
cMOOC study approachCCK11 (12 weeks)
CCK12 (12 weeks)Week1
Learner 1
Learner 2
Learner N
Week2
Learner 1
Learner N
…
SNA
Degree centralityEigenvalue centralityBetweenness centralityCloseness centrality
Coh-Metrix
NarrativitySyntactic SimplicityWord ConcretenessReferential CohesionDeep Cohesion
MLM
<<extract>
>
<<extract>>
<<analyze>>
<<an
alyz
e>>
<<measure>>
<<measure>>
Slide 8 of 23
1755 posts
2483 posts
1473 posts
61 posts
2266 posts
624 posts
2023-04-15
xMOOC study approach
Slide 9 of 23
NGIx (8 weeks)SNA
Degree centralityEigenvalue centralityBetweenness centralityCloseness centrality
<<extract>>
<<analyze>> <<measure>>
Learner 1
Learner 2
…
Learner N
Coh-Metrix<<
anal
yze>
>
<<extract>> NarrativitySyntactic SimplicityWord ConcretenessReferential CohesionDeep Cohesion<<measure>>
MLMActive learners
All learners
Grades
MLMActive learners
All learners
2023-04-15
SNA metricsDegree Centrality
Closeness Centrality
Betweenness Centrality
Eigenvalue Centrality
Slide 10 of 23
2023-04-15
Linguistic properties• Narrativity
• Deep cohesion
• Referential cohesion
• Syntactic simplicity
• Word concreteness
Slide 11 of 23
2023-04-15
Statistical analyses
Slide 12 of 17
Dependent Independent Random
cMOOC and xMOOC cMOOC and xMOOC cMOOC
Degree centrality Narrativity Learner within a course
Eigenvalue centrality Deep Cohesion Course slope
Betweenness centrality Referential Cohesion
Closeness centrality Syntax Simplicity xMOOC
xMOOC Word Concreteness Learner
Final grade cMOOC Word count
Media
Time
Activity
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06
Degree models
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.1 -0.05 0 0.05 0.1
*
*
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.1 0 0.1 0.2 0.3 0.4 0.5
All learners Active learners
**
*
**
**
*
***
**
cMOOC
xMOOC
Betweenness models
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.05 0 0.05 0.1
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.1 0 0.1 0.2 0.3 0.4 0.5
All learners Active learners
*
***
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.05-0.04-0.03-0.02-0.01 0 0.01 0.02 0.03
*
*
cMOOC
xMOOC
Closeness models
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4
All learnersActive learners
**
**
**
*
**
**
cMOOC
xMOOC
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.03 -0.02 -0.01 0 0.01 0.02
2023-04-15
cMOOC Eigenvalue model
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.08 -0.06 -0.04 -0.02 0 0.02 0.04
***
Slide 16 of 23
Low referential cohesion Higher eigenvalue centrality
2023-04-15
Week 1
Week 2
Week 3
Week 4
Week 5
Week 6
Week 7
Week 8
Week 9
Week 10
Week 11
Week 12
00.10.20.30.40.50.60.70.80.9
1
cMOOC Contextual factors
Slide 17 of 23
Media Twitter vs. Facebook vs. Blogs
Differ in their affordances
Time Negative association?
Activity More active -> more likely to grow
influence
2023-04-15
xMOOC Performance models
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.25 -0.2 -0.15 -0.1 -0.05 0 0.05 0.1
Slide 18 of 23
Syntax Simplicity
Deep Cohesion
Referential Cohesion
Word Concreteness
Narrativity
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6
All learners Active learners
****
**
**
**
**
*
**
**
What do we know…?
• Contextual, as well as linguistic and discourse features of written artefacts, are important determinants of learning in a cMOOC environment.
• cMOOC: Course participants who tend to use more narrative and informal style, nevertheless still manage to maintain a deeper cohesive structure in their communication will have more ties.
• xMOOC: – Better performance – more expository style discourse– Higher centrality – more narrative style, with less overlap between
words and ideas
2023-04-15 Slide 19 of 23
So what…?• Practice
– Effective use of language to communicate and share knowledge – Sharing novel information, using concrete and coherently
structured language– Traditional academic performance vs. social centrality
• Research– Are learners able to develop all the necessary skills to learn in
distributed settings?– Changes in linguistic features as indicators of learning progress
2023-04-15 Slide 20 of 23
What we are missing?
• cMOOC– temporal dimension– 72 undirected weighted graphs
• xMOOC– cross-sectional analysis– limited set of interactions
2023-04-15 Slide 21 of 23
What we are missing?
L1
L2
L3
FB post - fi
TW post - ti
L4Blog post - bi
TW post - tj
SM post - smn
post
comment
like
comment
post
mention
retweet
post
post
post
commentreference fa
vorit
e
What we are missing?
L2Temporal properties:- Linguistic and discourse features – t1- Linguistic and discourse features – t2- …- Linguistic and discourse features – tn
Overall- GPA- Previous activities- Demographics
2023-04-15 Slide 23 of 23
Further research• (Temporal) Exponential Random Graph
Models– Mathematical vs. Statistical model
Henry and Dietz (2011)2023-04-15 Slide 24 of 23
Factors shaping learners’ interactions in networked learning context
Srećko Joksimović[email protected]
@s_joksimovic
Dragan Gašević[email protected]
@dgasevic
Vitomir Kovanović[email protected]
@vkovanovic
References• Joksimović, S., Dowell, N. M., Skrypnyk, O., Kovanović, V., Gašević, D., Dawson, S., Graesser,
A.C. - Exploring the Development of Social Capital in cMOOC through Language and Discourse, Journal of Educational Data Mining, 2015 (submitted).
• Dowell, N. M., Skrypnyk, O., Joksimović, S., Graesser, A. C., Dawson, S., Gašević, D., Hennis, T. A., de Vries, P., Kovanović, V. – Modeling Learners’ Social Centrality and Performance through Language and Discourse, The 8th International Conference on Educational Data Mining, Madrid, Spain, 26-29 June, 2015 (accepted).
• Henry, A. D., Dietz, T. - Information, networks, and the complexity of trust in commons governance. International Journal of the Commons, [S.l.], v. 5, n. 2, p. 188-212, sep. 2011. ISSN 1875-0281. Available at: <http://www.thecommonsjournal.org/index.php/ijc/article/view/312/231>. Date accessed: 17 May. 2015.