For the past four years, I have been experimenting with a social learning platform, which I call Pace Commons, which is based on Elgg social networking software.
Pedagogically, my major goal was to flatten the typical instructor centered hierarchies endemic to most learning management systems (such as Blackboard). As a teacher educator, it seemed philosophically inconsistent to be working with teachers and teacher candidates to promote student-centered, autonomy supportive classrooms while using a tool that reinforced traditional power dynamics. This project was greatly influenced by the work of Dron & Anderson in Teaching Crowds.
I also developed Pace Commons to be a research platform. I have been investigating these research questions:
- What types of patterns of interactions can we see in teachers and teacher candidates using a social learning platform?
- What types of learning networks emerge in teachers and teacher candidates using a social learning platform?
What was supposed to happen, of course, was that my students would be so empowered by this type of learning environment that they would throw off their shackles, so to speak, and embrace the opportunity of autonomy supportive learning. After all, that is exactly what happened to two cohorts of eighth graders when I did something similar in their biology class.
At you might guess, for these teachers and teacher candidates, not so much. Instead some other interesting findings emerged.
This is the first of a series (of currently undetermined length) where I will lay out my methods and the types of data I am looking at.
What Does This Type of Learning Environment Look Like?
Pace Commons works like other social networks (that means you, Facebook) where users can make create pages (think wikis), blog posts, comment on the postings of others, share photos and files, etc. There is even a Pinterest-y type feature. For my classes, I assign students to groups (one group per class typically), and delineate the work of the course via pages. Here is an example of one such course. In these two images, you can see how the work is laid out into units. This online course is called “Computer Science for Teachers,” and gives teachers and teacher candidates a solid background in the history of CS in Education, along with experiences in using block-based coding, text-based coding, and robotics in K-12 schools. Each of the units listed links to sub-pages with the details and requirements for each project and learning activity. Throughout the course, students share their work via blog posts, pages, and/or photos.
What Kind of Data Have I Been Collecting
The system provides all the typical data collected for each post — type of user activity (e.g., blog post, comment, photo, etc.), date and time stamps, the actual posts themselves, and system-generated identifiers for each post.
How Am I Analyzing These Data
For the time being, I am working to do two things with these data. First, I want to see the patterns of interactions and the networks these interactions form. The image below illustrates this type of analysis. In it, we can see the clusters of interactions between students in a particular course. Then, I am investigating a small set of questions. What kind of interactions and learning networks are emerging? What are the types of activities that seem to generate larger or smaller networks? What is the role of influencers within this type of learning network?
The second thing I am working on is to using textual analysis tools to explore the content of these interactions as well. For example, what are the types of things around which the students are interacting? Are there some topics and/or students who are more or less “attractive” to the others?
Let’s look at some preliminary data
Here is a screenshot from a typical blog post. In it, we can see the end of the original blog itself, allow with a small set of responses. I responded to the blog post, and the blog author responded to my comment. Also, one other student responded to the original post.
This is what a network graph of that set of interactions would look like. The original author is indicated as a blue circle, and the two respondents are indicated as red circles. The double arrows between nodes 1 and 2 indicate the interaction between me and the author. The single arrow from node 3 indicates that student’s response to the original post. Not too complex, but it makes the point.
And this is what a textual analysis of this set of interactions look like. I created a corpus of the original blog post and its responses, looking at both word frequencies (depicted as a word cloud) and networks of words (depicts as links between words). This word was done in Voyant-Tools. You can play with an interactive version here.
Conclusions (for now)
Obviously, none of this is very Earth shattering for now. People interact, their interactions can be graphed, and their interactions demonstrate sets of keywords.
For now, this is nothing more than a proof of concept.
Next steps: to do this type of analysis across an entire course. Stay tuned.