***** To join INSNA, visit http://www.insna.org *****
From the example you gave, I think you may be interested in the whole
network among your participants, but I'd like to follow up with a few
clarification questions. So, what is the goal of your analysis? How do you
plan on using the network structure variables? What is your unit of
analysis, the 'whole' network among your participants, the ego network
(and/or the ties within) of each participant, or the participants?
If you are interested in investigating the structural characteristics of the
potential whole network among your participants, then Martina's response is
super helpful. The concern's about reliability would depends on how you
respond to her follow-up questions and how you set up the parameters for
If you are actually interested how ego-centric network's characteristics may
exhibit a relationship with other individual level variables (e.g.
isolation, depression, etc.), then the reliability regarding the structural
variables generated for each ego network depends mainly on the extent to
which participants reported their alters in a comprehensive fashion. You are
unlikely to have enough observations that allow you to construct the
structural variables like component or clique when your participants only
reported 3-5 alters on average. The structural variables would be invariant.
This is a known challenge for the egocentric network analysis.
That said, I think Martina's suggestion may still be useful even if network
characteristics may be a set of independent variables in your analysis. I am
not sure whether it would be theoretically and methodologically appropriate
to translate the simulation results back to the structural measures for each
participant. And I'd love to get additional feedback and thoughts on this.
Hope this helps.
Yuli Patrick Hsieh
Ph.D. in Media, Technology and Society
School of Communication
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.