Posted by ljshaffer
It is amazing what you can get accomplished at work when you all is quiet and distractions are at a minimum. This past summer I collected some networked type data. I asked folks at my field site in southern Mozambique to describe the social-ecological system in which they live and how it all connects together. The interviews are part of a longer-term (I hope) project to map/describe the savanna social-ecological system (SES) of southern Mozambique from the perspective of the people who live and work there. Maps can then be used as a focal point for discussions about key elements of SES sustainability, building long-term adaptive capacity, locating best intervention sites, identifying risk and uncertainty, potential tipping points, etc.. I got the idea from a paper I read about a Turkish team of environmental scientists who used interviews, cognitive mapping, and graph theory to construct maps of a local lake ecosystem from the perspective of the stakeholders. They used their mapping method to run policy simulations and facilitate the creation of a participatory environmental management plan.
- Özesmi, U., & Özesmi, S. L. (2004). Ecological models based on people’s knowledge: a multi-step fuzzy cognitive mapping approach. Ecological Modelling, 176(1), 43-64.
Özesmi, U., & OeZESMI, S. (2003). A participatory approach to ecosystem conservation: fuzzy cognitive maps and stakeholder group analysis in Uluabat Lake, Turkey. Environmental Management, 31(4), 0518-0531.
This afternoon I finally got a chance to play with my data. It has taken me this long to get to it because I first had to learn how to use the software analysis program, Gephi. Learning new software, at least for me, takes time, solitude, and a lot of button pushing. I make a mess, delete, start over, delete, repeat et nauseum. Basically, having interruptions (student or otherwise), or at least the potential for them, does not make for a good learning environment for me.
Gephi is relatively easy, particularly if you want to download large datasets or use the datasets they give you – which are aimed at social network analysis. However, I chose Gephi because it allows you to look at other sorts of networked data. Including data like mine, linked social-ecological system elements drawn from TEK interviews. But to do this I had to figure out how to configure a data set for importing into the software. Surprisingly, Gephi doesn’t have a tutorial on how to put together a basic .csv (comma delimited) file for importing into their program. I guess they assume everyone who uses this open source software is in the know.Thank goodness for Literature Geek (a.k.a. Amanda Visconti) and History Blogger. Both researchers have provided detailed instructions about setting up basic .csv data files in Excel for use in Gephi.
- Literature Geek. 9 Sept 2013. Get Your Data into Gephi: A Quick and Basic Tutorial. Accessed 8 Jan 2015.
- History Blogger. 17 Aug 2013. Getting Started With Gephi. Accessed 8 Jan 2015.
So here are the initial results from a single mapping interview.
First some translation and the legend. A machamba is the local Ronga word for agricultural field and esteiras are floor mats handmade from Cyperus papyrus stalks. The colors denote different types of capital: green for natural capital, red for human capital, and blue for infrastructure. The purple denotes a process rather than a type of capital (I haven’t yet figured out what to do with this). I enlarged the nodes for trees and bees because my informant told me at the end of the interview that these elements were “super important for the life of the community.” The enlarged nodes of hospital, electricity and machamba show that they are sustainable elements, while fire and charcoal are unsustainable elements in this SES. I provided weights, based on the perceived strength of connections, for each of the edges but they didn’t come out so well – particularly for elements that are connected but there isn’t really any perceived connection strength like charcoal and esteiras (both household money-making activities). There were also negative weights given to connections between elements that were perceived as bad. I’d like to figure out how to show good versus bad connections too.
At this point, I am still playing. While I plan to create individual maps for each informant, my goal is to eventually link all the maps together. The interviews were long but people provided a whole lot of very detailed information. I’m looking forward to seeing more results from my playdays.