[Air-L] About statistical analysis for networks that share some of their actors

[Xanat Meza]

Hello everyone!
I have been asking this question in other more related groups, but so far, there are no answers, so I am giving it a shot here, as some of the members of this group might be conducting network analyses.
I have a stakeholder network that is segmented in three different categories. However, some of the actors in the network are included in two or even the three sub-networks. 
My question is, which type of statistical network analysis (or any statistical test for that matter) can be used to test differences in such a network? Usually, a node level T-test or a network ANOVA, for example, can be conducted, with the assumption that the groups have totally different sets of actors. 
If anyone has an idea of what can proceed in this case, or has research literature or case study based papers, I would be grateful.
Xanat V. Meza

Ph.D. Kansei, Behavioral and Brain SciencesUniversity of Tsukuba
M.A. Media and Communication
Yeungnam University
B.D. Graphic Communication Design
Universidad Autonoma Metropolitana