The assignment assumed experience with NodeXL, which we spent the first third of the semester learning during lab periods (about once a week). Though time consuming, it really helped speed up things in the second two units of the class. We spent a lot of time at the beginning of the unit in which this assignment was given going over how to transcribe social interactions while watching The Wire and driving home the point that careful attention to details were paramount.

We had students work in small groups with the idea that they could check each other's work and help resolve ambiguous interactions. However, we found that several groups had difficulty following instructions which lead to errors in the shared spreadsheet. For example, they would use the incorrect canonical name for one of the characters, or they would list unnamed characters despite explicit instructions not to. We were able to use another version of the data that Sam Alexander created to spot errors in the student-provided data and we then merged the two and students used that in their analysis for the paper. We also watched an episode of Law and Order in class; while students created their own network of the episode along the way, we provided them with a curated network for use in their paper.

Students did a great job on the analysis. These were honor students, so we had high expectations, and they did not disappoint. Students included well made network diagrams and supporting network statistics to make their arguments.

Things that might be changed are: more assignments in which students must uncover network recording errors and more police procedural data than from a single episode of Law and Order so that students can make a stronger argument. Perhaps annotating the latter could be a replacement for annotating The Wire since that data has already been collected.