Research Methods for Identifying and Analysing Virtual Learning Communities
11 likes766 views
The document discusses methods for identifying and analyzing learning communities, emphasizing the importance of community characteristics such as trust, participation, and social protocols. It highlights various tools and models for analysis, including the use of scales like Rovai & Jordan’s Classroom Community Scale and interaction analysis metrics. The conclusions point to the value of mixed methods and the need for baseline data to inform the modeling and analysis of virtual learning environments.
More Related Content
Similar to Research Methods for Identifying and Analysing Virtual Learning Communities (20)
Research Methods for Identifying and Analysing Virtual Learning Communities
- 1. Methods for Identifying and Analysing Learning Communities Richard A. Schwier Virtual Community Research Laboratory Educa;onal Technology and Design University of Saskatchewan Higher Educa;on Development Centre University of Otago Dunedin, New Zealand February 7, 2011
- 2. Central Concerns • ShiNing focus of research • Atomized view of communi;es • Tools for analysis • Genera;on of models • Using research to inform development of online learning environments
- 3. Community Modeling Cons;tuents Comparison
- 6. Sense of Community • Chavis’ “Sense of Community Index” • Rovai & Jordan’s “Classroom Community Scale” (Chronbach’s alpha = .93) – Connectedness (.92) – Learning (.87) • Pre-‐post design (t-‐Test, p<.005)
- 7. Interac;on Analysis • Fahy, Crawford & Ally (TAT) • Intensity – “levels of participation," or the degree to which the number of postings observed in a group exceed the number of required postings – 858 actual/490 required = 1.75
- 8. Interac;on analysis • Density – Included only peripheral interac;ons – the ra;o of the actual number of connec;ons observed, to the total poten;al number of possible connec;ons 2a/N(N-‐1) = 2(122)/13(12) = .78
- 9. Reciprocity ra;o the parity of communication among participants
- 11. Characteris;cs of Community • Transcript analysis • Interviews • Focus groups
- 12. Characteris;cs • Awareness • Par;cipa;on • Social protocols • Trust • Historicity • Trajectory • Iden;ty • Technology • Mutuality • Learning • Plurality • Reflec;on • Autonomy • Intensity
- 13. Comparison of characteris;cs • Thurstone analysis
- 14. Thurstone Scale
- 15. Modeling Bayesian Belief Network Model of a Virtual Learning Community
- 16. BBN -‐ Query the network
- 17. BBN -‐ Query the network
- 18. Sense of Community Rovai & Jordan’s “Classroom Community Scale” (Chronbach’s alpha = .93) 90.0 67.5 45.0 Formal 22.5 Non-Formal 0
- 19. Intensity Fahy, Crawford & Ally (TAT) 2.0 1.5 1.0 Formal 0.5 0 Non-Formal
- 20. Density Fahy, Crawford & Ally (TAT) 0.8 0.6 0.4 Formal 0.2 Non-Formal 0
- 21. Reciprocity ra;o Instructors 15.0 11.3 7.5 3.8 Non-Formal 0 Formal
- 22. Reciprocity par;cipants 1.0 0.8 0.5 Mean Mean sd 0.3 sd 0 Formal Non-Formal 0.376276399
- 23. Order of importance -‐ elements Element Formal Non-‐formal Trust 1 7 Learning 2 3 Par;cipa;on 3 6 Mutuality 4 10 Intensity 5 7 Protocols 6 10 Reflec;on 7 2 Autonomy 8 10 Awareness 9 1 Iden;ty 10 4 Trajectory 11 13 Technology 12 4 Historicity 13 13 Plurality 14 7
- 24. And lately...
- 26. Interac;on analysis • Thread density and depth (Wiley, 2010) – Calcula;on of levels of replies in conversa;on threads – Data flawed, but useful Mean Reply Depth (MRD crude) = sum of reply depth for all messages/messages in the thread Mean Reply Depth (corrected)= MRD (crude) x ((n-‐b(childless messages)/n)
- 27. Do not akempt to read this!
- 28. Do not akempt to read this! Mulitlogue/discussion Simple Q&A/chit-‐chat Monologue/no discussion
- 30. Keep an eye on... Technology Enhanced Knowledge Research Ins;tute (TEKRI)-‐ hkps://tekri.athabascau.ca/ George Siemens & data analy;cs
- 31. Conclusions • Cycle of analysis is more important than specific tools used • Mixed methods seems reasonable, and worked well in prac;ce • Baseline data are needed to situate findings • Modeling is an act of systema;c specula;on influenced by data (not limited by data) • Most enjoyable part: the hunt