Wednesday, August 17, 2011

Is there a way to compare these sociograms?

I am much taken with the potential of social media to support the creative enterprise through commentary, supportive or not. So, in that spirit, I am hoping for comments on how my co-author and I might analyze the diagrams reproduced below.
A word or two of background. These are sociograms, a fairly venerable way to display relations between individuals, or indeed any entities, such as cities or corporations. They can take many forms, as the literature indicates.

In our work, Sonya Glavac and I examined neighboring behaviors in seven neighborhoods in Glendale, Arizona, using measures of behavior found in the urban and social work literature. The goal was to see if residents living in Home Owner Associations behaved differently than residents living in traditional (organic) neighborhoods. We found some differences in how residents viewed their neighborhoods but few of these turned out to be statistically significant.

We also asked residents to mark on maps the neighbors with whom they interacted, lending things and so forth. We counted ties and whether reciprocal relations exist (i.e. neighbor A interacts with neighbor B, but does B interact with A with the same strength?). We created the following sociograms from the data that we collected.

Sociogram of neighboring behaviors in an HOA

Sociogram of neighboring links in a non/HOA neighborhood

We can see that although there are similar numbers of interactions, these take different forms: the HOA shows more concentration on certain individuals while the second neighborhood shows more randomness. The question is: is there a way to test these interpretations?

No comments:

Post a Comment