| node type: | woman, event |
| edge type: | participated in |
| number of nodes: | 18 women, 14 events |
| number of edges: | 89 |
| source: | A. Davis, B. B. Gardner, M. R. Gardner: Deep South. The University of Chicago Press, 1941 |
| data: | SouthernWomen1.txt |
The edge-repulsion LinLog drawing shows two main groups of women participating in different social events. A few women, in particular Pearl, are not clearly assigned to one of the groups. Indeed, a meta-analysis of the data by Freeman [1] showed considerable disagreement among 21 studies about the assignment of Pearl: Some studies assign her to the left group, some to the right group, others to no or both groups.
A comparison of the edge-repulsion LinLog drawing with the node-repulsion LinLog drawing illustrates the different interpretations of the identified clusters: While the earlier drawing assigns Olivia and Flora to the second group because they exclusively attend to events of this group, the latter drawing separates them from the second group because they attend only to few such events.
[1] Linton C. Freeman: Finding social groups: A meta-analysis of the southern women data. In Ronald Breiger, Kathleen Carley and Philippa Pattison (eds.): Dynamic Social Network Modeling and Analysis. Washington, D.C.: The National Academies Press, 2003.
|
| Edge-repulsion LinLog model |
 |
 |
| Fruchterman-Reingold model (for comparision) | Node-repulsion LinLog model (for comparison) |
| back |