CCK09 Benchmarking CCK08 social networks in the Moodle forums (Preview of the research)

I posted a comment in Prof. George Siemens blog that I intend to do a Social Network Analysis (SNA) of the Moodle forums in CCK09.

(ERRATA: edited July 26, 2009)
Prof. George Siemens has never given me any approval to conduct research on the Moodle CCK08 Forums. He merely stated that the forum posts are publicly available. This plan for a study or this blog is not connected to Prof. Siemens or Prof. Downes, it is not connected in any way to the University of Manitoba.

All errors or liability here are mine and I have done this as a private individual.
(END OF ERRATA)

His response, which I have taken as an approval (I will try to send a formal email later because I wanted to try first with sample data from CCK08 course. I did not want to be embarassed with a formal commitment when I could not actually do it.)

As the data is open (i.e. it is an open course - not password protected), analysis can be performed on the course forums and blogs. Conducting and SNA on blog conversations would be more difficult, but equally valuable. The course is not centralized as most courses are. This presents some difficulty in identifying contributors/participants and tracking how they relate/connect to others. (Siemens, 2009)

I decided to benchmark the forums in CCK08 course first. Here is the sample social network for Forum 1: What is connectivism?. I have not tried any analysis yet.

Download the Data set: http://www.mediafire.com/file/axwgbrijnim/forum1.0.zip (213.78 kb) - in Pajek format. I could not anonymize the Netdraw .vna file so this is the only format I can share.


Figure 1: Directed graph of CCK08 Moodle forum 1




Figure 2: Directed graph with total posts vector of CCK08 Moodle forum 1


Figure 3: Undirected graph CCK08 Moodle forum 1



Figure 4: Undirected graph with total posts vector of CCK08 Moodle forum 1


Figure 5: Netdraw graph (total posts are represented by the sizes of vertex labels)

My method is pretty tedious because I have to do a lot of manual data processing.

1. Use Aneesha Bakharia's script to extract th .vna data

• Copy and paste to .vna text files

2. Combine .vna of each discussion into one forum .vna

• open .vna in OOO Calc (use space separator)
• save as ODF spreadsheet
• copy node attributes to vector sheet
• copy tie data to net sheet
• sort ascending
• check for total posts
• edit nodes and net to sum redundant id posts
• combine in a complete sheet
• save as csv (space field delimeter and double quote text delimiter)

3. Clean up csv

• edit csv in text editor
• remove quotes in headers
• save as .vna

4. Convert to Pajek (I am more familiar with it than NetDraw)

• open .vna in NetDraw
• relations should be strength rather than talk (otherwise line values will not be exported to Pajek)
• export to Pajek data (.net & .vector)

ego networks are easier to study in netdraw but I could not anonymize the data set

5. Anonymize in Pajek

• Net->Transform->Add->Default Vertex Labels
• Analyze
• Create Graphs

Because of the manual steps in this method, it will take me probably a month to finish combining all the data. So stay tuned.

Hopefuly the final social network for the entire Moodle forum will be useful to the participants in the CCK09 course.

References:

Siemens, G. (2009, July 6). Comment posted in G. Siemens (2009, July 5), Connectivism & connective knowledge 2009. In Connectivism & Connective Knowledge [Blog]. Retrieved July 12, 2009, from http://ltc.umanitoba.ca/connectivism/?p=189.

Bakharia, A. & Dawson, Shane. Moodle SNA Analysis [Javascript program]. In Blackboard and WebCT - forum social network Analysis Tool. Random Syntax [Blog]. Retrieved July 12, 2009, from http://www.randomsyntax.com/blackboard-forum-social-network-analysis/.