*****  To join INSNA, visit  *****

On Hana's question,

Unless I misunderstand, it sounds like more timing information is needed before centrality analysis can be done properly.  If I read it right, the data include the periods in which alliances were formed but not necessarily the periods in which they are dissolved or otherwise changed.  To get centrality by firm by period, therefore, the periods of entry and exit for both nodes and links must be added so that centrality is computed over the alliance relationships that really existed for each period.

With many periods and many measures to calculate, it is worth considering using available libraries (JUNG or NetStat, for example) to put together some simple batch programs to handle what otherwise starts to be a lot of pointing and clicking -- work that becomes error-prone in a task like this.  Graphical user interfaces are very nice for exploratory analysis, but the old command line batch program is still a great way to compute multiple measures for multiple periods.  (This is what I do, and the NetStat routines from the folks at CASOS were hassle-free to integrate into the home-grown software I use to compute measures on panels of networks.)

Happy hunting,

Mark T. Kennedy, Ph.D.
Department of Management & Organization
Marshall School Business | University of Southern California
[log in to unmask]

SOCNET is a service of INSNA, the professional association for social
network researchers ( To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.