Participants will be introduced to the underlying theory of network analysis and how to utilize social network analysis to analyze relational data in educational settings. The development arc traverses all stages of this quantitative methodology, including cleaning and reshaping data, addressing missing data, developing intuition for inference, and data visualization in R. Participants will be able to:

• import data into R and manage data files;
• clean data and combine data sets;
• investigate the potential impacts of missing data;
• visualize data to explore the data and generate publication quality figures; and
• analyze networked data at multiple levels (node-level, graph-level, and model-level).

Networked data can take many forms, and many forms of data can be treated as networks.

• The most common form of network data is adjacency data, which is a set of nodes and the connections between these nodes (stored most often as a square matrix or as an edgelist). In educational settings this might include a group of students and the connections could be homework partners.
• Co-occurrence data is fairly typical, this is like a situation where Persons A, B, and D all attended a meeting together, so by virtue of the probability that they talked at this meeting, they all get a connection to each other, or citation networks are examples of co-occurrence data.
• Text can also be treated as a network, with words serving as nodes and placement in the document serving as the edges. Actions such as steps in a problem solution, or instructional moves, can be treated as a network.

Network methods exist for dealing with various types of data.

Participants will leave the workshop having completed initial data cleaning, analysis, and visualization on their own data set.