Comment

Comment by Johannes Castner on October 21, 2011 at 3:58pm

UCINET is only useful for very basic static (one slice in time) analysis, but for that it is pretty good. However, for my statistical network work I'm likely to stick with R, which is incidentally free and has ever improving packages (whatever some innovative statistician or network methodologist might invent will surely be available as an R package soon thereafter).

Indeed, our projects are similar, particularly in terms of methodology. I went to a network analysis seminar yesterday and they said that the next methodological hurtles that are being tackled in this field are dynamic and spatial ones, which is good news for us as consumers of these people's output. They all reported major discomforts with ERGM models, by the way, which have been simply imported from physics without good theoretical motivations.  

Comment by John McCreery on October 21, 2011 at 9:20am

That's an amazing project you've got underway. What software are you using for the network analysis? I stumbled around for a while with UCINET but settled on Payek. The combination of price (it's freeware) and the book Exploratory Social Network Analysis with Pajek proved irresistible to an independent scholar getting started on his own.

I note a lot of similarities between our two projects. We are both dealing with archival data, historical depth, and a wealth of documentary information whose relation to the networks we'd like to explore.

In my networks, the edges connect creatives to project teams that produced winning ads. The line values aren't distances. They are a nominal scale in which integers denote roles (1=copywriter, 2=planner, 3=creative director.....), in an industry where several individuals may be credited with the same role and the same individual may play multiple roles.

Comment by Johannes Castner on October 21, 2011 at 6:06am

Oh wow ...I might be able to get the first two chapters scanned in once I have the book (they have this service here at my university, for such cases).

For my part, I am currently working on mapping the entire road network of Zambia and all of the changes thereof that occurred between 1990 and now, using high resolution satellite images that have been taken on a yearly basis. The roads are the edges of the graph with the distance property and the cities, towns and villages are the nodes, with various population attributes. There also exists data on every new road project between 1990 and now with the name of the contractor, the amount of money of the contract and these data can be linked with the roads on the map. 

For my work, I am particularly interested in the Chinese and Indian contractors and in where the Chinese and Indians live in Zambia, as I want to later (next year) spend a year in Zambia to map changes in Zambians' personal and business networks due to the increasing Chinese involvement in Zambia (using mainly ethnography, but also other methods, such as looking into business files of current and past employees). I also want to use text mining techniques on massive amounts of news papers and magazines in Zambia (which have been arduously collected by the National Archives of Zambia) to construct a sentiment network, so as to see how people's systems of attitudes toward other Zambian tribes, Indians and the Chinese, related to all kinds of aspects of life, have changed over time. In particular, I am interested in seeing if the various native tribes and groups of Zambia have come to feel closer to each other, in terms of identity, now that a great number of Chinese and Indian people are present.     

Comment by John McCreery on October 21, 2011 at 5:16am

We may have hit a snag. When I did my one-click order the screen said that the book was in stock and could be delivered within a few days. The confirmation message changed that to 2-4 weeks.

Comment by John McCreery on October 21, 2011 at 3:57am

OK, I'm persuaded. Just ordered the book from Amazon.jp. Should be here in a few days. In the meantime, could you tell  us a bit about how you came to have these particular interests?  

In my own case, I have been dabbling with network analysis for, it must already be four or five years, as part of project exploring the world of Tokyo advertising creatives, of which I have been a small part for nearly three decades. As a spin-off from that, I am now working on a Pajek cookbook for researchers who analyze 2-mode networks. If you'd like to learn more, I'd be happy to run on about it.

Comment by Johannes Castner on October 20, 2011 at 5:20pm

Dear John, while it is true that this book is a bit pricey, note that it is not just another network analysis book, but it is about the dynamics on networks. It was highly recommended to me by a great Harvard Sociologist and Statistician. But the real reason why I would love to discuss this book here, in an Anthropology forum, is that I would love to see Anthropologists getting into the business of dynamic theory construction. Anthropologists, as a group, have by far the greatest amount of field experience in the most diverse human settings among all of the social scientists and I believe that this experience is invaluable. However, if this experience never gets translated into a precisely stated theory that can be tested, then this knowledge will be lost to us, much like an ancient language that is no longer spoken and that was never written down. Writing ethnographic books is not enough, as it most often amounts to little more than a particular observer's interpretation of states of affairs (often incongruent, with incongruences that would be cleared away if the theory had been mathematically stated), that can not be tested in a meaningful matter. Anthropology is too important to thus fall to the wayside and this is why I call for a theoretical Anthropology that is technically able and mathematically and statistically literate (only in those subfields of mathematics and statistics that are culturally and socially relevant, of course). This also has the potential to finally give marginalized populations a voice in economic policy discussions as their views could be more precisely stated.

Comment by John McCreery on October 20, 2011 at 4:01pm

Johannes, to participate in this group will require a serious commitment in money (it is not a cheap book) and time. It look stone on the scale of other only partially read network analysis books in my library. Could you tell us a bit more about why you think these invest,ends are a good idea?

• ‹ Previous
• 1
• …
• 11
• 12
• 13
• Next ›
• Page