Comments on: Savage Minds in American Anthropologist http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/ Notes and Queries in Anthropology — A Group Blog Thu, 09 Feb 2012 19:06:13 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: Rick http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-2/#comment-629148 Rick Thu, 11 Mar 2010 15:51:08 +0000 http://savageminds.org/?p=3260#comment-629148 Yes, those are some of the ways that you can use ego networks. Working with them is much more limiting than complete networks. Currently the only program you can use to analyze them is EgoNet and SPSS, which I was trying to figure out. SPSS, because on EgoNet you can only do one at a time, because that is the traditional way. The main things they are used for is to look at a person's personal support, and personal community. Because part of my field site is basically an open air drug market, and the whole area is just coming back from about 3 decades of crime and violence, I wanted to see if it has effected the networks set up for reciprocity. You can see this in Merry (1981), Stack (1978), and Achor (1974). Then there was also the issue with almost 40 years of complete dependence on Welfare, which changed relationships among people, but also relationships between people and the government. So, part of my question was what do people's networks actually look like after all this (compared to the literature), and what kinds of support do they turn to them for. From previous research, the violence created smaller and smaller pockets of community, that sometimes went down to a couple of blocks. This really impedes network growth. I found that many people simply did not use their networks to gain in dealing with government, which is problematic, because this city wants to empower people to take part in development. I also wanted to see if people considered the "community leaders," the self-appointed people that make noise and speak for the community, as a part of their personal networks. I found that it depends, but mostly not. The whole process is much more qualitative, than standard SNA. And, you are reduce to more of a traditional statistical methodology. There's such a lack of literature about it, that I had to write back and forth with Dr. Halgin, who wrote EgoNet. Oh, because those findings about the patterned differences between the networks of poor, black residents were pretty radical for the literature, I added a few more samples, and the effect disappeared. It was just sampling error. I guess I don't get famous today. There is a lot more variation within the group than between groups. Yes, those are some of the ways that you can use ego networks. Working with them is much more limiting than complete networks. Currently the only program you can use to analyze them is EgoNet and SPSS, which I was trying to figure out. SPSS, because on EgoNet you can only do one at a time, because that is the traditional way. The main things they are used for is to look at a person’s personal support, and personal community.

Because part of my field site is basically an open air drug market, and the whole area is just coming back from about 3 decades of crime and violence, I wanted to see if it has effected the networks set up for reciprocity. You can see this in Merry (1981), Stack (1978), and Achor (1974).
Then there was also the issue with almost 40 years of complete dependence on Welfare, which changed relationships among people, but also relationships between people and the government.
So, part of my question was what do people’s networks actually look like after all this (compared to the literature), and what kinds of support do they turn to them for. From previous research, the violence created smaller and smaller pockets of community, that sometimes went down to a couple of blocks. This really impedes network growth. I found that many people simply did not use their networks to gain in dealing with government, which is problematic, because this city wants to empower people to take part in development.
I also wanted to see if people considered the “community leaders,” the self-appointed people that make noise and speak for the community, as a part of their personal networks. I found that it depends, but mostly not.

The whole process is much more qualitative, than standard SNA. And, you are reduce to more of a traditional statistical methodology. There’s such a lack of literature about it, that I had to write back and forth with Dr. Halgin, who wrote EgoNet.

Oh, because those findings about the patterned differences between the networks of poor, black residents were pretty radical for the literature, I added a few more samples, and the effect disappeared. It was just sampling error. I guess I don’t get famous today. There is a lot more variation within the group than between groups.

Report this comment

]]> By: John McCreery http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-629133 John McCreery Thu, 11 Mar 2010 00:53:00 +0000 http://savageminds.org/?p=3260#comment-629133 Still curious about how you are using SNA. Partly arising from my own work, I'm imagining a couple of possibilities: (1) a branching tree pattern in which one person tells two or three others, who tell two or three others...... (2) a fireworks pattern, in which one individual tells a lot of people he knows and a few of those pass it on to others who tell a lot of people: a triggering sequence in which, in effect, you get a series of explosions of information, e.g., the word passing from one preacher to another, with each incorporating the message in their Sunday sermons. Just brainstorming, though. Still curious about how you are using SNA. Partly arising from my own work, I’m imagining a couple of possibilities: (1) a branching tree pattern in which one person tells two or three others, who tell two or three others…… (2) a fireworks pattern, in which one individual tells a lot of people he knows and a few of those pass it on to others who tell a lot of people: a triggering sequence in which, in effect, you get a series of explosions of information, e.g., the word passing from one preacher to another, with each incorporating the message in their Sunday sermons. Just brainstorming, though.

Report this comment

]]> By: Rick http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-629124 Rick Wed, 10 Mar 2010 18:40:12 +0000 http://savageminds.org/?p=3260#comment-629124 This is based on geography. My field site contains two major ethic groups, Latinos and Blacks. These groups are refined into other groups, but my question was how does information flows from households in different sectors of the field site, and to other areas in the city outlined for development. Gender and ethnicity do not vary significantly from each other when it comes to overall numbers of alters, but the ego-networks are patterned differently by ethnic group. They are also used for different kinds of support and information flow depending on the area. This is mostly exploratory type research. It is just one part of other ethnographic techniques (PO, interviews, surveys), used to triangulate data for a rapid assessment. I agree that without alter-alter data SNA used by itself is kind of pointless. The same way you are using SNA to ask better questions, I'm using it right now to augment data so that I can make better recommendations. This is based on geography. My field site contains two major ethic groups, Latinos and Blacks. These groups are refined into other groups, but my question was how does information flows from households in different sectors of the field site, and to other areas in the city outlined for development. Gender and ethnicity do not vary significantly from each other when it comes to overall numbers of alters, but the ego-networks are patterned differently by ethnic group. They are also used for different kinds of support and information flow depending on the area.
This is mostly exploratory type research. It is just one part of other ethnographic techniques (PO, interviews, surveys), used to triangulate data for a rapid assessment.
I agree that without alter-alter data SNA used by itself is kind of pointless. The same way you are using SNA to ask better questions, I’m using it right now to augment data so that I can make better recommendations.

Report this comment

]]> By: John McCreery http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-629114 John McCreery Wed, 10 Mar 2010 07:37:54 +0000 http://savageminds.org/?p=3260#comment-629114 Without alter-to-alter relations, what's the point of the network analysis? To say one person has 103 alters all you need is a list. To classify them, all you need is categories. I could see, maybe, plotting them on a map of the city, to see if they are concentrated in certain places. Without alter-to-alter relations, what’s the point of the network analysis? To say one person has 103 alters all you need is a list. To classify them, all you need is categories. I could see, maybe, plotting them on a map of the city, to see if they are concentrated in certain places.

Report this comment

]]> By: Rick http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-629113 Rick Wed, 10 Mar 2010 07:02:48 +0000 http://savageminds.org/?p=3260#comment-629113 I'll do that the next time I get a gig requiring organizational anth. skills. Right now I'm trying to get a sense of how ego networks are patterned across geographic space in a very large city. Working with one person that has 103 significant alters, scattered over a city, and it would just be impossible to get the alter to alter data, beyond kinship and work ties. I've got to figure out a way for the city to increase dialogue with residents, so that we can plan inclusive development. We've actually got residents working with architects to preserve the unique character of communities. When all this is done, I'll probably present it at the AAA, and see how it goes. I'll have to join the AAA I guess. I’ll do that the next time I get a gig requiring organizational anth. skills. Right now I’m trying to get a sense of how ego networks are patterned across geographic space in a very large city. Working with one person that has 103 significant alters, scattered over a city, and it would just be impossible to get the alter to alter data, beyond kinship and work ties.

I’ve got to figure out a way for the city to increase dialogue with residents, so that we can plan inclusive development. We’ve actually got residents working with architects to preserve the unique character of communities. When all this is done, I’ll probably present it at the AAA, and see how it goes. I’ll have to join the AAA I guess.

Report this comment

]]> By: John McCreery http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-629112 John McCreery Wed, 10 Mar 2010 06:15:30 +0000 http://savageminds.org/?p=3260#comment-629112 P.P.S. If, at some future date, you decide to play with Pajek, you will save yourself a lot of grief converting data files if you use Jürgen Pfeffer's Text2Pajek utility. My current toolkit is Filemaker Pro for data storage and extraction, Text2Pajek for converting text files exported from Filemaker, and Pajek for SNA analysis and visualization. I'm working with two-mode networks, multiple attributes and multiple relations, and this set up makes it easy to extract and analyze interesting looking subnetworks. P.P.S. If, at some future date, you decide to play with Pajek, you will save yourself a lot of grief converting data files if you use Jürgen Pfeffer’s Text2Pajek utility. My current toolkit is Filemaker Pro for data storage and extraction, Text2Pajek for converting text files exported from Filemaker, and Pajek for SNA analysis and visualization. I’m working with two-mode networks, multiple attributes and multiple relations, and this set up makes it easy to extract and analyze interesting looking subnetworks.

Report this comment

]]> By: Rick http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-629109 Rick Wed, 10 Mar 2010 05:09:46 +0000 http://savageminds.org/?p=3260#comment-629109 Wow, thanx. I've been using Ucinet 6, Netdraw, and Egonet. I've already used SPSS to analyze the ego networks statistically by aspects of certain variables. However, you can apparently analyze the entire composition of the networks by aggregating alter/tie data, with ego data from two separate data sets. You can do that for complete networks with Ucinet and others, but not with ego nets. I will check out Pajek. I avoided it when I first started to learn SNA because all the "how to" books said that other programs required more extensive knowledge of graph theory and matrix algebra. Unfortunately, I don't know if I'd be able to learn a new program by the time my report is due. I have a one page summary that I have to present to the client tomorrow. Wow, thanx. I’ve been using Ucinet 6, Netdraw, and Egonet. I’ve already used SPSS to analyze the ego networks statistically by aspects of certain variables. However, you can apparently analyze the entire composition of the networks by aggregating alter/tie data, with ego data from two separate data sets. You can do that for complete networks with Ucinet and others, but not with ego nets.

I will check out Pajek. I avoided it when I first started to learn SNA because all the “how to” books said that other programs required more extensive knowledge of graph theory and matrix algebra. Unfortunately, I don’t know if I’d be able to learn a new program by the time my report is due. I have a one page summary that I have to present to the client tomorrow.

Report this comment

]]> By: John McCreery http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-629103 John McCreery Wed, 10 Mar 2010 02:44:18 +0000 http://savageminds.org/?p=3260#comment-629103 P.S. I use Pajek for network analysis, one reason, besides the fact that the program is free and powerful, is that there is an excellent book called Exploratory Network Analysis with Pajek to refer to. Google "Pajek" and you will find it in a flash. P.S. I use Pajek for network analysis, one reason, besides the fact that the program is free and powerful, is that there is an excellent book called Exploratory Network Analysis with Pajek to refer to. Google “Pajek” and you will find it in a flash.

Report this comment

]]> By: John McCreery http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-629101 John McCreery Wed, 10 Mar 2010 02:41:21 +0000 http://savageminds.org/?p=3260#comment-629101 The short, sad answer is no, I haven't worked with SPSS. When I got into this, a fellow Mac user suggested DataDesk, which I have been using as my statistical package ever since. If I move one, it will probably be to R, but that's a pretty steep learning curve. The short, sad answer is no, I haven’t worked with SPSS. When I got into this, a fellow Mac user suggested DataDesk, which I have been using as my statistical package ever since. If I move one, it will probably be to R, but that’s a pretty steep learning curve.

Report this comment

]]> By: Rick http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-629097 Rick Wed, 10 Mar 2010 02:10:47 +0000 http://savageminds.org/?p=3260#comment-629097 @Mcreery: Have you ever worked with networks, specifically ego networks using SPSS, as outlined in a Wellman et al. article on the subject? If so, I have a few questions. This seems like a fantastic resource for what I'm doing, and there are some aspect of the process that are not clear to me. Thank you. @Mcreery:

Have you ever worked with networks, specifically ego networks using SPSS, as outlined in a Wellman et al. article on the subject?
If so, I have a few questions. This seems like a fantastic resource for what I’m doing, and there are some aspect of the process that are not clear to me.

Thank you.

Report this comment

]]> By: Digging Digitally » Wherefore are ye anthro/archaeo blogs? http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-628954 Digging Digitally » Wherefore are ye anthro/archaeo blogs? Sat, 06 Mar 2010 00:20:11 +0000 http://savageminds.org/?p=3260#comment-628954 [...] in the title of a recent American Anthropologist article by David H. Price. Savage Minds has a blog post on the AA article, with comments. AWBL contributor Michael E. Smith [...] [...] in the title of a recent American Anthropologist article by David H. Price. Savage Minds has a blog post on the AA article, with comments. AWBL contributor Michael E. Smith [...]

Report this comment

]]> By: Wednesday Round Up #105 « Neuroanthropology http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-628933 Wednesday Round Up #105 « Neuroanthropology Thu, 04 Mar 2010 13:01:31 +0000 http://savageminds.org/?p=3260#comment-628933 [...] Savage Minds in American Anthropologist Kudos to SM! What public anthropology could and should be based on the author’s experience – go [...] [...] Savage Minds in American Anthropologist Kudos to SM! What public anthropology could and should be based on the author’s experience – go [...]

Report this comment

]]> By: Rick http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-628925 Rick Thu, 04 Mar 2010 06:06:11 +0000 http://savageminds.org/?p=3260#comment-628925 I see what you mean now. There are very new and creative ways of using SNA with what people call web 2.0. I've heard of it, but don't know much about it. I see what you mean now. There are very new and creative ways of using SNA with what people call web 2.0. I’ve heard of it, but don’t know much about it.

Report this comment

]]> By: John McCreery http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-628924 John McCreery Thu, 04 Mar 2010 03:06:01 +0000 http://savageminds.org/?p=3260#comment-628924 A couple of comments. First, the basic idea that studying the actual networks through which transactions or information flows goes back a long way. But the current acceleration of interest in the field begins in the 1990s, when computers and software improve to the point that processing and visualizing network data became something that individuals could do using PCs. In just the last few years, the rise of Facebook, Twitter, and other forms of online social networking and the creation of mammoth citation databases for those more academically inclined, has added fuel to the fire. Second, I see where you are coming from. That said, the hard slog of contacting people, gathering structured information, coding, cross-tabulating and analyzing the results has always been a regular feature of social survey research. That is what makes folks who can now have Web crawlers gather information for them and do data mining with off-the-shelf tools get so excited about this stuff. It is also why, in my particular case, I have become an advocate of piggy-back ethnography using found data. In my current research, I am using credits data from ad contest annuals, combined with books by and published interviews with the central figures in the winner networks. Other folks may have an assistant or two in the field. I have a trade press that generates hundreds of pages of "field notes" per month. When I get around to the depth interviews with key informants I will know why I picked them and be able to ask smarter questions based on the SNA and desk research. A couple of comments. First, the basic idea that studying the actual networks through which transactions or information flows goes back a long way. But the current acceleration of interest in the field begins in the 1990s, when computers and software improve to the point that processing and visualizing network data became something that individuals could do using PCs. In just the last few years, the rise of Facebook, Twitter, and other forms of online social networking and the creation of mammoth citation databases for those more academically inclined, has added fuel to the fire.

Second, I see where you are coming from. That said, the hard slog of contacting people, gathering structured information, coding, cross-tabulating and analyzing the results has always been a regular feature of social survey research. That is what makes folks who can now have Web crawlers gather information for them and do data mining with off-the-shelf tools get so excited about this stuff.

It is also why, in my particular case, I have become an advocate of piggy-back ethnography using found data. In my current research, I am using credits data from ad contest annuals, combined with books by and published interviews with the central figures in the winner networks. Other folks may have an assistant or two in the field. I have a trade press that generates hundreds of pages of “field notes” per month. When I get around to the depth interviews with key informants I will know why I picked them and be able to ask smarter questions based on the SNA and desk research.

Report this comment

]]> By: Rick http://savageminds.org/2010/02/24/savage-minds-in-american-anthropologist/comment-page-1/#comment-628917 Rick Wed, 03 Mar 2010 17:45:06 +0000 http://savageminds.org/?p=3260#comment-628917 "I think it’s used more as a tool in practicing anthro. than in academic anthro. though, which is why I don’t think it’s taught in very many depts." I'm gonna have to eat those words a bit. I think this might also be a difference between British and American anth. It seems like this method is used and taught more among the British than us. I can't say that for sure, but it seems like much of the literature is coming from the British. “I think it’s used more as a tool in practicing anthro. than in academic anthro. though, which is why I don’t think it’s taught in very many depts.”

I’m gonna have to eat those words a bit. I think this might also be a difference between British and American anth. It seems like this method is used and taught more among the British than us. I can’t say that for sure, but it seems like much of the literature is coming from the British.

Report this comment

]]>