***** To join INSNA, visit http://www.insna.org *****
IF you have the mediated networks as 1 mode (person x person) matrices where ties represent something like "communicate with this person using e-mail" (etc.), you can just correlate and/or regress (QAP) each against the advice network to see how similar they are. Perhaps one ICT type is more related to advice sharing than another.
If you have them as two mode people by ICT matrices, you could run affiliations to convert to 1 mode matrices as above (but note these will be "use the same technology" matrices not "communicate with this person using e-mail" matrices) then QAP. This does change the theoy somewhat as its more of a girders-similarity approach than a pipes-flows argument.
(| 1. Comparing ICT-mediated networks | >
| > Dear socnetters:
| > I have a dataset consisting of 17 peoples advice-network in a company
| > that is highly distributed. I have also asked the informants how the
| > different relations is supported by their available ICT (e-mail, mobile
| > phones and IM/SMS) and I would like to analyse closer the difference
| > between the media use within the group. (For instance if some mediated
| > networks are highly similar to the advice network, towards the manager
| > in the groups, etc). Does any of you have suggestions of what kind of
| > analysis would be appropriate for a comparison of mediated networks and
| > advice-networks in such a small group? Or do you know about articles
| > where similar analysis has been conducted?
| Did you see the recent thread(s) on homophily? It seems like you could
| arrange the media data as attribute-data; of course, this then ignores
| any network data you have regarding the media (that is, 'Alice talks
| to Bob on her cellphone, and Bob talks to Xen with email'). I don't
| know if you have that sort of data.
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.