Effects of social interactions are reported in research on higher education to lead to positive outcomes such as higher levels of internalization, sense of community, academic achievement, metacognition, and student retention. The role of social networks has especially been emphasized in research due to the availability of theoretical foundations and analytic methods to investigate their effects in higher education. The increased use of technologies in education allows for the collection of large and rich datasets about social networks which call for the use of novel analytics methods. This talk will first give a brief overview of the existing work on and lessons learned from some well-known studies on social networks in higher education in diverse situations from face-to-face to massive open online courses. The talk will then identify critical challenges that require immediate attention in order for the study of social networks to make a sustainable impact on learning and teaching. The most important take away from the talk will be that
- computational aspects of the study of social networks need to be integrated deeply with theory, research and practice,
- novel methods for the study of critical dimensions (discourse, structure and dynamics) that shape network formation and network effects are necessary, and
- innovative instructional approaches are essential to address the changing conditions created by contemporary educational and technological contexts.
Learning with me Mate: Analytics of Social Networks in Higher Education
1. Learning with me mate
Analytics of social networks in higher education
Dragan Gasevic
@dgasevic
March 16, 2016
MCSHE, University of Melbourne
Joint work with Srecko Joksimovic, Vitomir Kovanovic, and many great collaborators
as cited in the presentation
4. The Strength of Weak Ties
Connections through
strong ties
Connections through
weak ties
Granovetter, M. S. (1973). The strength of weak ties. American journal of sociology, 1360-1380.
5. A common assumption
Higher social network centrality
leads to higher achievement
Burt, R. S. (2000). The network structure of social capital. Research in organizational behavior, 22, 345-423.
14. Theory-informed learning analytics
Gašević, D., Dawson, S., Rogers, T., & Gasevic, D. (2016). Learning analytics should not promote one size fits all: The
effects of instructional conditions in predicting academic success. The Internet and Higher Education, 28, 68-84.
15. Simmel’s theory of social interactions
Networks based on super strong ties
Triads as the unit of analysis
16. Study objective
Network
structural
properties
Learning
outcome
Social
dynamic
processes?
Tie dynamics:
• Homophily/
heterophily
• Reciprocity
• Triadic closure
Joksimović, S., Manataki, A., Gašević, D., Dawson, S., Kovanović, K., de Kereki, I. F. (2016). Translating network position
into performance: Importance of Centrality in Different Network Configurations. In Proceedings of the 6th International
Conference on Learning Analytics & Knowledge (LAK 2016), Edinburgh, Scotland, UK (in press).
17. Method (Data)
Code Yourself! (English), ¡A Programar! (Spanish)
Certificate: 50% for the coursework;
75% - distinction
0
10000
20000
30000
40000
50000
60000
70000
Enrolled Engaged Engaged with
forum
Course participants
Codeyourself Aprogramar
0
200
400
600
800
1000
1200
1400
1600
1800
Codeyourself Aprogramar
Obtained certificate
Normal Disctinction
20. Results of the multinomial regression analysis, * p<.05; ** p<.01; *** p<.001
In order to provide meaningful visualizations, estimates for betweenness centrality were
multiplied by 100 (only for the presentation purposes)
-0.15 -0.1 -0.05 0 0.05 0.1
Betweenness (normal)
Betweenness (distinct)
Closeness (normal)
Closeness (distinct)
W. Degree (normal)
W. Degree (distinct)
Aprgoramar Codeyourself
***
**
***
*
**
***
***
Results – centrality vs. performance
25. Learning and discourse
Graesser, A., Mcnamara, D., & Kulikowich, J. (2011). Coh-Metrix: Providing Multilevel Analyses of Text Characteristics.
Educational Researcher, 40(5), 223–234. http://doi.org/10.3102/0013189X11413260
26. Language and social ties
Granovetter, M. S. (1973). The strength of weak ties. American journal of sociology, 1360-1380.
27. Interaction strategy,
social networks, and performance
Kovanović, V., Gašević, D., Joksimović, S., Hatala, M., & Adesope, O. (2015). Analytics of communities of inquiry: Effects
of learning technology use on cognitive presence in asynchronous online discussions. The Internet and Higher
Education, 27, 74-89.
28. Method (data)
Courses: Delft Design Approach (DDA), Introduction to
Drinking Water (CTB), Functional Programming (FP)
Certificate: 60% for the coursework
730
135
645
281
1064
1962
0
500
1000
1500
2000
2500
Engaged with forum Obtained certificate
Forum participation & obtained
certificates
DDA CTB FP
11336 8484
316711397
1128
6560
0
10000
20000
30000
40000
50000
DDA CTB FP
Students overview
Enrolled Submitted
Joksimović, S., Kovanović, V., Milikić, N., Jovanović, J., Gasević, D., Zouaq, A., Dawson, S. (2016). Effects of discourse on
network formation and achievement in massive open online courses. Computers & Education (in preparation).
30. Discussion forum
extract
Weighted,
directed graph
Statistical
network analysis
Exponential random graph models
Homophily
Achievement
Transition count
Post count
Reciprocity
Popularity
Expansiveness
Simmelian ties
student, post, timestamp
post => keywords Alchemy API
post_id, parent_post_id, student_id, keywords
Block HMM
Dominant topics Topic coherence
Interpretation
Paul, M. J. (2012). Mixed membership Markov models for unsupervised conversation modeling. In Proc. 2012 Joint Conf.
on Empirical Methods in Natural Language Processing and Computational Natural Language Learning (pp. 94-104).
31. Discussion forum
extract
Weighted,
directed graph
Statistical
network analysis
Exponential random graph models
Homophily
Achievement
Transition count
Post count
Reciprocity
Popularity
Expansiveness
Simmelian ties
student, post, timestamp
post => keywords Alchemy API
post_id, parent_post_id, student_id, keywords
Block HMM
Dominant topics Topic coherence
Association?
Interpretation
Regression analysis
Interpretation
Transition count
Post count
Replies count
Betweenness centrality
Closeness centrality
Degree centrality
33. Common ground as a key factor in
shaping network structures
Clark, H., & Brennan, S. E. (1991). Grounding in communication. In L. B. Resnick, J. M. Levine, & S. D. Teasley (Eds.),
Perspectives on socially shared cognition (pp. 127–149). Washington, DC, US: American Psychological Association.
34. The principle of least effort in
communication
Clark, H., & Krych, M. A. (2004). Speaking while Monitoring Addressees for Understanding. Journal of Memory and
Language, 50(1), 62–81.
35. DDA topics
Topic 11: Video concept
- video making,
- upload
- particular assignment that included
video making
Topic 5: Course information
- resources,
- readings,
- discussions
Topic 7: Design thinking
- thinking about design process,
- different approaches to design
36. -8 -6 -4 -2 0 2 4 6
Expansiveness
Popularity
Assortative mixing
Simmelian ties
Simmelian cliques
Reciprocity
Post count
Transition count
Achievement
Edges
CTB DDA FP
***
***
***
***
***
*
***
Analysis of the estimates for the three ERG models
Note: * p<.05; ** p<.01; *** p<.001
***
***
***
***
***
***
**
***
***
***
***
***
***
***
Results - network characteristics
37. Results
(centrality vs. performance)
-0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12
Betweenness
Closeness
W. Degree
Post count
Replies count
Transition count
CTB DDA FP
R2
CTB = .17
R2
DDA = .21
R2
FP = .08
Results of the three regression analysis
Note: * p<.05; ** p<.01; *** p<.001
***
***
*
***
***
***
***
***
39. One size fits all does not work in
learning analytics
Gašević, D., Dawson, S., Siemens, G. (2015). Let’s not forget: Learning analytics are about learning. TechTrends, 59(1),
64-71.
40. Theory as a driver of the study of
networked learning
43. Teaching to recognize structural
wholes in networks
Burt, R. S., & Ronchi, D. (2007). Teaching executives to see social capital: Results from a field experiment. Social
Science Research, 36(3), 1156-1183.
44. Social presence in
network formation
Kovanovic, V., Joksimovic, S., Gasevic, D., & Hatala, M. (2014). What is the source of social capital? The association
between social network position and social presence in communities of inquiry. Proceedings of 7th International
Conference on Educational Data Mining – Workshops, London, UK, 2014
45. Scaling up qualitative
research methods
Kovanović, V., Joksimović, S., Waters, Z., Gašević, D., Kitto, K., Hatala, M., Siemens, G. (2016). Towards Automated
Content Analysis of Discussion Transcripts: A Cognitive Presence Case In Proceedings of the 6th International Conference
on Learning Analytics & Knowledge (LAK 2016), Edinburgh, Scotland, UK (in press).
46. To what extent instructional design
can affect network structures?
Class size as an important factor
Skrypnyk, O., Joksimović, S., Kovanović, V., Gašević, D., & Dawson, S. (2015). Roles of course facilitators, learners, and
technology in the flow of information of a cMOOC. The International Review of Research in Open and Distributed
Learning, 16(3).
50. Tie building approach less important
than experience in networks
Burt, R. S., & Ronchi, D. (2007). Teaching executives to see social capital: Results from a field experiment. Social
Science Research, 36(3), 1156-1183.
51. Ideally suited
method
Not ideally suited
method
Ideally suited method,
but context dependent
Azevedo, R. (2015). Defining and measuring engagement and learning in science: Conceptual, theoretical,
methodological, and analytical issues. Educational Psychologist, 50(1), 84-94.
Capturing and
measurement of
engagement-
related processes
As the field of learning analytics continues to mature, there is a corresponding evolution and sophistication of the associated analytical methods and techniques. In this regard social network analysis (SNA) has emerged as one of the cornerstones of learning analytics methodologies. However, despite the noted importance of social networks for facilitating the learning process, it remains unclear how and to what extent such network measures are associated with specific learning outcomes. Motivated by Simmel's theory of social interactions and building on the argument that social centrality does not always imply benefits, this study aimed to further contribute to the understanding of the association between students' social centrality and their academic performance. The study reveals that learning analytics research drawing on SNA should incorporate both – descriptive and statistical methods to provide a more comprehensive and holistic understanding of a students' network position. In so doing researchers can undertake more nuanced and contextually salient inferences about learning in network settings. Specifically, we show how differences in the factors framing students' interactions within two instances of a MOOC affect the association between the three social network centrality measures (i.e., degree, closeness, and betweenness) and the final course outcome.
As the field of learning analytics continues to mature, there is a corresponding evolution and sophistication of the associated analytical methods and techniques. In this regard social network analysis (SNA) has emerged as one of the cornerstones of learning analytics methodologies. However, despite the noted importance of social networks for facilitating the learning process, it remains unclear how and to what extent such network measures are associated with specific learning outcomes. Motivated by Simmel's theory of social interactions and building on the argument that social centrality does not always imply benefits, this study aimed to further contribute to the understanding of the association between students' social centrality and their academic performance. The study reveals that learning analytics research drawing on SNA should incorporate both – descriptive and statistical methods to provide a more comprehensive and holistic understanding of a students' network position. In so doing researchers can undertake more nuanced and contextually salient inferences about learning in network settings. Specifically, we show how differences in the factors framing students' interactions within two instances of a MOOC affect the association between the three social network centrality measures (i.e., degree, closeness, and betweenness) and the final course outcome.
As the field of learning analytics continues to mature, there is a corresponding evolution and sophistication of the associated analytical methods and techniques. In this regard social network analysis (SNA) has emerged as one of the cornerstones of learning analytics methodologies. However, despite the noted importance of social networks for facilitating the learning process, it remains unclear how and to what extent such network measures are associated with specific learning outcomes. Motivated by Simmel's theory of social interactions and building on the argument that social centrality does not always imply benefits, this study aimed to further contribute to the understanding of the association between students' social centrality and their academic performance. The study reveals that learning analytics research drawing on SNA should incorporate both – descriptive and statistical methods to provide a more comprehensive and holistic understanding of a students' network position. In so doing researchers can undertake more nuanced and contextually salient inferences about learning in network settings. Specifically, we show how differences in the factors framing students' interactions within two instances of a MOOC affect the association between the three social network centrality measures (i.e., degree, closeness, and betweenness) and the final course outcome.
As the field of learning analytics continues to mature, there is a corresponding evolution and sophistication of the associated analytical methods and techniques. In this regard social network analysis (SNA) has emerged as one of the cornerstones of learning analytics methodologies. However, despite the noted importance of social networks for facilitating the learning process, it remains unclear how and to what extent such network measures are associated with specific learning outcomes. Motivated by Simmel's theory of social interactions and building on the argument that social centrality does not always imply benefits, this study aimed to further contribute to the understanding of the association between students' social centrality and their academic performance. The study reveals that learning analytics research drawing on SNA should incorporate both – descriptive and statistical methods to provide a more comprehensive and holistic understanding of a students' network position. In so doing researchers can undertake more nuanced and contextually salient inferences about learning in network settings. Specifically, we show how differences in the factors framing students' interactions within two instances of a MOOC affect the association between the three social network centrality measures (i.e., degree, closeness, and betweenness) and the final course outcome.
Benefits of network centrality are possible in networks with weak ties
Network centrality does (not) necessarily imply less constraints and more benefit otherwise
Simmel’s theory of social interactions allows for investigation whether networks are based on super strong ties. This can be analyzed with statistical methods for social network analysis (SNA) such as exponential random growth models (ERGMs) that allow for the use of triads as units of analysis, unlike commonly used dyads in descriptive SNA
This diagram shows the results of the analysis conducted with the ERG models. It shows a significant effect of homphilic relationships between students (e.g., those who are similar in terms of achievement, country of origin) are more likely to connect. Likewise, there is a high level of reciprocity in both networks (e.g., if student A ask something student B, student B replied back to student A).
The most striking results was that the social network in CodeYourself! was based on Simmelian ties (i.e., super strong ties), while this was not the case for the social network in Aprogramar. Therefore, theoretically, we expected that network centrality would be a significant predictor of achievement (i.e., grades and completion) in Aprogramar as its network was based on weak ties, while this was not the case for CodeYourself due to the nature of its network.
The hypothesized statements from the previous slide were confirmed: network centrality was a significant predictor of achievement (i.e., grades and completion) in Aprogramar as its network was based on weak ties, while this was not the case for CodeYourself due to the nature of its network.
Psychological models of discourse comprehension and learning, such as the construction-integration, constructionist, and indexical-embodiment models, lend themselves nicely to the exploration of learning related phenomena in computer-mediated educational environments. These psychological frameworks have identified the representations, structures, strategies, and processes at multiple levels of discourse (Graesser & McNamara, 2011; Kintsch, 1998; Snow, 2002). Five levels have frequently been identified in these frameworks: (1) words, (2) syntax, (3) the explicit textbase, (4) the situation model (sometimes called the mental model), and (5) the discourse genre and rhetorical structure (the type of discourse and its composition). The computational linguistic facility used in the correct study, Coh-Metrix (described more in the methods), allows us to capture these main levels of discourse. In the learning context, learners can experience communication misalignments and comprehension breakdowns at different levels. Such breakdowns and misalignments have important implications for the learning process.
Language is a primary means for expressing and exchanging content through a network. It is through language that participants are able to build connections and define social ties with other actors. With regard to analytical approaches, there has been extensive knowledge gleaned from manual content analyses of learners’ discourse during educational interactions. For instance, the early research of Bernstein (1971) highlighted that individuals with more complex social networks tend to demonstrate more formal and elaborated speech forms than those with more simple and densely connected personal networks. Milroy and Margrain (1980) reported that the variety of language in use is dependent on the density of the social network and the multiplexity of the ties. According to Granovetter (1973), the intensity of ties established between actors affords an opportunity to track the linguistic phenomenon of code-switching, whereby speakers change conversational styles as they converse with interlocutors from the different parts of their sub-networks. These earlier studies illustrate the relationship between social ties and language. However, the manual content analysis methods used in those studies are no longer a viable option with the increasing scale of educational data. Consequently, researchers have been incorporating automated linguistic analysis that range from shallow level word counts to deeper level discourse analysis.
Talk about social presence – affective, interactive and cohesive and indicating that interactive is typically always playing the role, while affective plays in betweenness, while cohesive plays in degree as well.
Also indicated that affective emerges as a separate construct within social presence scales.
Talk about social presence – affective, interactive and cohesive and indicating that interactive is typically always playing the role, while affective plays in betweenness, while cohesive plays in degree as well.
Also indicated that affective emerges as a separate construct within social presence scales.