Social Network and Pattern Analysis

Data Loading and Preparation

• Import the CSV data into Exploratory, ensuring correct delimiter settings.

• Create a new column by concatenating the students’ first and last names to form a "Full Name" column.

Filtering Data per Reading

• Branch the dataset for each reading to isolate the comments related to individual readings.

• Name each branch according to the corresponding reading.

Creating Nodes Files (Per Reading Branch)

• For each reading branch, create a summary of the data to generate the nodes file.

• Summarize the data by grouping it by "Full Name" and extracting the following metrics:

  • Number of messages posted

  • Number of replies received

  • Number of upvotes received

• These metrics will be used to create the nodes for the network graph.

Creating Links Files (Per Reading Branch)

• Generate links between authors who commented in the same section by performing a left join based on the “Range” variable (the section where the comment was made).

• Rename columns for clarity (e.g., "Source" and "Target").

• Filter out self-loops by removing rows where the "Source" is the same as the "Target."

• This creates the links between students who interacted in the same section of the reading.

Exporting Data for Gephi

• Export the nodes and links files for each reading in CSV format, ensuring correct naming (e.g., "nodes_Wise_2019.csv", "links_Brooks_2017.csv").

• Create a combined nodes and links file for the overall dataset (not filtered by reading).

Loading Data into Gephi

• Import the nodes and links files for each reading into Gephi.

• Set the graph type to "Undirected" and append the data to the existing workspace.

• Label the nodes by copying the "Id" column to the "Label" column.

Visualizing the Networks in Gephi

• Reading-specific networks: For each reading, visualize a network where the size of the nodes is determined by the number of messages posted, and the color is based on the number of replies received.

  • Node size: Based on message count (e.g., range from 10 to 40).

  • Node color: Based on the number of replies.

• Total network: Visualize a network that shows all the readings together, with nodes sized by the total number of messages and colored by the total number of replies across all readings.

Analysis

• Using the visualizations, analyze differences in conversation patterns between readings (e.g., comparing density, betweenness, or other network metrics).

• Explore how the interaction networks evolved week-to-week and what the visualizations reveal about student participation and engagement.