[Air-l] re: network analysis programs

Willard Uncapher wuncapher at ucdavis.edu
Wed Sep 26 10:25:58 PDT 2001


At 01:08 PM 9/25/2001 -0500, you wrote:
>I'm trying to learn more about network analysis statistical programs. I 
>want to know which are used most often, which are Window-oriented, 
>limitations to various programs, etc.

There is a general, older page on social network analysis software: 
http://www.heinz.cmu.edu/project/INSNA/soft_inf.html (part of the 
International Network for Social Network Analysis- 
http://www.heinz.cmu.edu/project/INSNA/  ).  While I've known people who 
used such software (eg., Negopy), I am not particularly knowledgeable. 
Further the appropriateness, cost, and robustness of any particular 
software can vary between different tasks.

Barry Wellman's group and network would be knowledgeable (and he has a fine 
publications page 
http://www.chass.utoronto.ca/~wellman/publications/index.html ).

Nine theoretical wrinkles of your mathematical situation:  1. defining and 
locating 'actors/actants/nodes' for your data; 2. relating 'mere' 
connectedness and centrality within networks to power over and in a 
network; 3. considering networks in networks (from my perspective any 
'node' is itself a network); 4. relating links at the level of parts or at 
the level of super-sets to those which do not occur at the level of study, 
and vice versa; 5. clarifying the way the 'collection' of the data, 
particularly where it is 'easy' foregrounds certain questions over others; 
6. Deciding whether to distinguish networks based on content, media, 
frequency, etc. or to combine them as a unitary network; 7. Determining or 
accounting for what is missing from the network (temporary autonomous 
situations/zones, hidden links, sensitive links, etc.); 8. Conversely, 
setting a value for what might be in or constituting a link in a network 
based on play, mistake, or other indeterminancies (other than eliminating 
it as noise); and importantly, 9.  defining/deconstructing the boundary or 
level of the study, although boundary conditions can simply be asserted, 
often by someone evincing a perspective from 'outside' the network, even if 
they also 'belong' to it - the notion of 'deconstruction' is very much 
meant here.  How one answers those questions will have a great impact on 
where and how the statistics are collected, and on what the statistics 
produce, what they reveal, and how they are interpreted.

Good luck!   Willard


Willard Uncapher, Ph.D. / Ctr. for Technocultural Studies / UC, Davis / 
willard at well.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.aoir.org/pipermail/air-l-aoir.org/attachments/20010926/b12d53b4/attachment-0003.htm>


More information about the Air-L mailing list