*****  To join INSNA, visit http://www.sfu.ca/~insna/  *****

Dear Socnetters,

I'm working with a survey team who will be getting data on the ego-centric
networks of a large number of people. We are using a name generator and a
network structure matrix to get most of the ties (assuming that the respondent
can tell if person 1 knows person 2).

After we've completed the section on each of the people, we were interested in
events that people do together. This data will be used to model the movement of
people around the city. The network structure will be in a computer program
during the interview, so we can compute a selection criteria however we like. We
don't want to take a random subset, but would rather select the likely people
who would not be associating with each other.

So with the following data on each person:
Frequency of socializing (ordinal from daily to yearly)
Type of relation (family, friend, coworker)
Strength of tie (1-10)
and data on transitive relations,

we want to select up to 8 different alters who would regularly see the
respondent, but are from different parts of the network.

Does anyone have a suggestion for an algolrithm that would select the most
different clusters or most diverse alters for interaction. Currently we are
considering the 8 most/least embedded people, or the strongest people within
subgroups (network minus respondent).

Thanks,
BERNiE

|- Bernie Hogan                             -|
|- Ph.D Student                             -|
|- Department of Sociology                  -|
|- NetLab, Knowledge Media Design Institute -|
|- University of Toronto                    -|
|-                                          -|
|-        [log in to unmask]          -|

_____________________________________________________________________
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.sfu.ca/~insna/). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.