Welcome to Week 3 of the Social Networks course at the University of Maine at Augusta! This week, we consider the point of distinction that social network analysis draws between itself and more traditional research in social science. In the United States of America, our culture favors explanations of events that are based in the traits of individual people. As a consequence of thinking of individuals as the objects to study, most social scientists tend to work with research tools that ignore relations. In contrast, social network analysts think of the world in terms of relations. As we think of society in different terms, so our research tools must be different. The key objects for studying networks are matrices, graphs and two kinds of lists.
Before you work through this lecture, please be sure to read “Individuals Versus Networks” first. It contains all the necessary conceptual tools you’ll need to build networks in four ways.
Last Week’s Course Survey
I’d like to thank the 23 students who provided feedback on the course so far via the course survey. I promise you that I read each one of the comments and have thought about them individually and put together.
Put together, I’m glad that the question to which most students (22 out of 23) responded “very comfortable” or “some was about the syllabus. This is the one document with which you ought to feel most comfortable in the whole course — it ought to be obvious and clear, and I’m glad most of you find it to be so. Attention to the 2 of 23 students who only answered “somewhat comfortable” regarding the syllabus: if you have a question, please send it to me. I’d love to have whatever conversation we need to have to move you into that “very comfortable” column about the syllabus.
Students reported less comfort with the lectures. Only a bare majority (12 of 23) reported feeling “very comfortable” with lectures, while 11 of 23 reported feeling “somewhat comfortable.” Regarding the readings, 3 of 23 reported feeling “very comfortable,” while 18 of 23 reported feeling “somewhat comfortable.”
I feel like we’re mostly on target with these answers. Course lectures should be making you feel a bit less than completely comfortable, because they should be requiring you to stretch and work at the edges of your capability. I know the readings are placed at a high level, but that’s a good thing for the 300-level. I know that getting through the readings can be a bit tiring and certainly require more attention than readings at the 100-level, but trust me — if you keep at it, you’ll find that your comprehension ability for this kind of reading will increase and what you thought was difficult will turn out to be not so hard by the end of the semester.
If you are the 1 student who answered that the lecture made you feel “somewhat uncomfortable,” or one of the 2 students who answered that way regarding the readings, please get in touch with me for help and perhaps some in-person review, because while I want this course to stretch you and maybe make your thinking muscles a bit sore from time to time, the course should never hurt. If we meet each other over the phone, over video chat or in person, we can work together to ease the pain and make the course a fun kind of challenge again.
Again, thanks for letting me know how it’s been going, and do get in touch one-on-one if we need to resolve something straightaway. Although we’re separated by a pair of screens for most of the course, that doesn’t mean we have to be strangers!
Multiplexity is Everywhere… Even in Organized Crime
One of the most exciting aspects of social network analysis is the fact that it’s a really hot topic in research right now. In the American Sociological Review (the top-rated and top-read journal in all of sociology), the second article from the top is Chris Smith and Andrew Papachristos’ “Trust Thy Crooked Neighbor: Multiplexity in Chicago Organized Crime Networks.” Using archival documents, Smith and Papachristos looked at the classic era of organized crime in the early 20th Century when Al Capone ruled the roost and his political buddies kept handing him get-out-of-jail-free cards. They didn’t just want to tell a pretty story, though: they worked to trace the networks of organized crime activity at the time.
Out of more than 5,000 pages of old organized crime records including Chicago Crime Commission investigatory reports, FBI documents, tax investigations and newspaper clippings, Smith and Papachristos identified 1,030 individuals involved in the Capone network of organized crime, and they identified 3,496 ties of criminal conduct connecting them. Here’s a picture of the Capone network, taken from p. 656 of Smith and Papachristos’ article. Ties in this graph represent the relation of shared criminal activity, measured as “co-arrests, co-offenses, criminal associations, illegitimate business associations, political corruption, and union corruption”:
Al Capone is the person with the most ties in this criminal network (316 — nearly 10% of all ties in the network — involve Capone). You can see just by looking at this network graph that there are some individuals who are central to the network — involved in many relations — but that there are many people on the fringe of the network, too. Many people involved in Chicago organized crime were associated with just a few others.
Smith and Papachristos’ article really gets interesting when they examine multiplexity. The authors don’t just graph shared involvement in crime; they track the relation of “personal connection” too (measured family, friends, and funeral attendance). Ties of personal connection generate this network between the same nodes:
Faint gray dots represent nodes who aren’t part of the personal network. Finally, Smith and Papachristos look at a network generated by the relation of “legitimate activities”: perfectly legal “associations in business, formal organizations, politics, and unions”:
It’s immediately apparent that most of the far-flung, fringe nodes in the criminal network do not tend to have personal or legitimate network ties to other members of the organized crime scene. Fringe participants in organized crime have a one-dimensional connection to the crime network. The minority of nodes in the Capone-Era Chicago organized crime network who do have multiplex ties reaching beyond crime are the most central figures in the crime network. Smith and Papachristos conclude from their social network research that personal and legitimate network ties help create a backbone structure around which the center of the network of criminal activity was based. “Even when multiplex ties are rare,” the authors write (p. 663), “they are still structurally relevant. Overlappying ties tend to be thick ties where trust and reciprocity [the tendency for favors to be returned] are likely to reside.”
With the tools and skills you’ll be gaining in this course (plus a fair amount of legwork), you could complete a study like theirs. The sky’s the limit. In her follow-up work on the subject, Chris Smith is looking at how women in these organized crime networks experienced “relational inequality,” connecting the field of gender stratification to network analysis. Imagine what you could come up with.
Thinking About Health, Stuck at the Individual Level
As you’ve read this week, the usual focus on data that describes individual leaves out a broad range of possible explanations for the outcomes we care about. One outcome most everybody cares about is health. Why do some people remain healthy throughout their lives, while others are plagued with with chronic troubles? Ross and Wu (1995) rejected the notion that health is all about the body; their research examined a number of ways in which people with higher education end up with better health. Their list of possible explanatory factors included better health practices, to be sure, but also included variables that were related to social context: quality of job, control of one’s own life, and most crucially for our consideration in this class, social support. Sounds like a network’s in that last explanation, doesn’t it? As this video notes, Ross and Wu describe networks but don’t quite measure them:
We don’t have to be stuck in the practice of shoving network ideas into individual data. Network data sets us free.
Practice, Practice! Transform a Network from One Form to Another
In this week’s reading, you learned how to depict a network as an adjacency list, an adjacency matrix, a graph, or an edge list. Now practice that skill! In this week’s Padlet exercise, I want you to make up a fictional network. Include node labels so I know which node is which. Then I want you to depict that one, single network using two of the four forms for depicting a network. Ready? Go to it!
(P.S. Does the Padlet not work for you? That’s OK — you can also answer the question by posting a comment in the comment box at the end of this lecture. If you’re going to be posting a graph as an image and can’t get the Padlet to work, send me your your work by e-mail.)
Don’t Forget Homework
Don’t forget that the syllabus lists homework to be completed each week. Homework #2 is due by 11:59 pm Eastern time on Saturday, September 17:
- Read “Individuals versus Networks.” Click here to read, download, or print.
- Read and Participate in Lecture #3 by September 17 (that’s this lecture right here!)
- Homework #2, due by 11:59 pm Eastern time on Saturday, September 17:
1. Define a set of people who you consider to be your “family.” 2. Describe your family using the individual-level approach. 3. Describe your family by drawing a social network graph, by writing out a social network edge list, and by creating a social network matrix. 4. Submit your work by Saturday, September 17 at 11:59 pm in the area titled “Homework #2” in the “Homework Assignments” section on the left-hand side of our Blackboard page.
Do you have any questions about this lecture or the week’s readings? If so, post them in the comments box below!
Ross, Catherine E., and Chia-ling Wu. 1995. “The Links Between Education and Health.” American Sociological Review 60(5): 719-745.
Smith, Chris M. and Andrew V. Papachristos. 2016. “Trust Thy Crooked Neighbor: Multiplexity in Chicago Organized Crime Networks.” American Sociological Review 81(4): 644-667.