Students will explore metrics of node centrality (degree, closeness, and betweenness) and develop an intuition for what these actually measure in the context of a social network. At first, students will compute these metrics by hand for specific graphs with a small number of nodes and edges.  These preliminary computations are designed to help the students warm up to the quantitative side of network analysis and help demystify the math from the definition of the metric.  We will look at some standard mathematical graphs (complete graph,  tree, “simple chain”, simple cycle) and then extend into small-scale graphs with less structure.  We will select these more generic graphs to specifically highlight the differences between degree centrality, closeness centrality, and betweenness centrality.    As the assignment progresses, we apply some of the observations from our warm-up to think about these metrics in the context of a social network.  The nodes now become people, and edges represent “interactions” between these people.  Students attempt to interpret these centrality metrics in this new context. 

For example,
●        In what ways is a person with a high degree centrality different from a person with a high closeness centrality?  Does one imply the other?   In the last component of the assignment, students begin working with large scale graphs and using NodeXL.  Given a provided social network, one familiar to the students, have them hypothesize how centrality measures for pairs or groups of characters will compare. 

For example,
●        Will Prof. McGonnagall have a higher closeness centrality than Dumbledore? Describe your reasoning. Then, using NodeXL, they can explore whether their intuition was correct.     The assignment will end with students putting together a summary of the different metrics, how they are defined, and what information they can provide about people in a social network.  We will ask students to use specific examples from social networks to support their explanations.