Posted on September 25, 2009 by Eva Schiffer
Last Sunday I learnt this in a quite direct “slap in your face” kind of way. After returning from a nice outing to Baltimore with my family, we found that someone had entered the house through a (previously) blocked dog-flap, had had a good look around and left with an arm full of stuff (the amount of stuff he – she?? – took and left indicates one individual), including my laptop computer. No, I’m not good at making regular backups. But, just by the nature of my work (and my personality) I tend to collaborate, share, put stuff online before it’s a 150% polished product. And, guess what, you can’t steal the internet. At least our little neighborhood hoodlum can’t.
So everything I shared, I still have. Whatever I kept to myself, the thief got. I decided to see it philosophically – for what else can I do, after the fact.
Filed under: musings | 3 Comments »
Posted on September 10, 2009 by Eva Schiffer
Why do policy makers push for certain decisions? Together with a group of IFPRI researchers I work at better understanding how agricultural policy making processes in Malawi (and other African countries work) and how research results can play a bigger role in it. This question, how to bridge the gap between research and the rest of the world, has been bothering me for a while now. So instead of jumping into action, I actually sat down and brainstormed (in my own brain… what a storm) about the question what makes policy makers tick, what gets them to prefer a certain direction over another, what makes them push really hard for their solution. My very preliminary first list is this:
Why do actors push for specific directions?
* because of political negotiations, compromise
* because of voter pressure, public opinion
* because of media
* because of trusted advisor
* because of external shock: drought, market crash etc.
*because they can, because they personally benefit (abusive power)
* because of cultural norms
* because of past experience
* because of research findings
* because of general paradigm shift, “development fashion”
* because of donor priorities and funding.
(The order of the list signifies nothing) Any more ideas?
Filed under: exploring new ideas, musings, open questions | 6 Comments »
Posted on August 26, 2009 by Eva Schiffer
I’m still searching for a board game producer who can mass produce Net-Map boxes at the price and quality that makes sense. Today I found bluepanther, who look interesting and from their website I landed in a different universe that was unknown to me before: The world of Piecepack.
“Flexible. Portable. Affordable. Public domain. The piecepack is a set of boardgame parts that can be used to design and play a wide variety of games. Anyone may design and publish a piecepack rule set. Any manufacturer or individual may produce piecepacks.”
The piecepack website has an ever growing list of games that you can play with the boardgame parts. The sets of rules are developed and uploaded by passionate players. In ChoosySushi you are an overworked waiter in a sushi restaurant, in Interstellar Contest you are an alien who wants to build colonies in other aliens’ star systems, while Senat is the adaptation of an antique Egyptian forerunner of backgammon… By now there are 138 games you can play with this basic set of pieces, dices, pawns etc.
I’m fascinated by the idea and by the way the community of practice (or community of passion?) around it works. I would love to develop a similar approach with my colleagues who use Net-Map and adopt and adapt the rules in the field according to their needs, make it a process map, bundle or differentiate actors, invent new symbols for bottlenecks and future links, combine it with timelines, try out different ways of combining the knowledge of more than one interviewee etc.
Filed under: Other people's work, exploring new ideas | 6 Comments »
Posted on August 24, 2009 by Eva Schiffer
This is really simple and really helpful – at least for me. A structured to-do list that looks like this:
Try it out and you’ll see that it is a very easy way of being much more organized that with your standard to do list.
You think you don’t need the “not important, not urgent” square? But that’s where a lot of the fun stuff ends up. And by having your “to dos” organized like this you know that they are the cherry-on-the-top-stuff that you shouldn’t spend the majority of your time with. For example writing this post has been in the “not important, not urgent” box for quite some time. But maybe it’s still useful.
Filed under: exploring new ideas | 4 Comments »
Posted on August 16, 2009 by Eva Schiffer
Paolo makes an interesting point below about the research philosophical issues you would run into by developing what he calls Net-Map 0.2. The fact that by rating links following a mechanical rule (depending on their distance to the interview partner) one does not necessarily get closer to meaning and that it might lead us into a well known dead end to try to look for “the” reality by adding everyone’s story up quantitatively.
Also, when I thought about it more, I realized that we might run into a number of “internal” methodological problems.
1. I don’t know of any SNA program or algorithm that deals with weighted links. So finally, if we wanted to analyze anything quantitatively, we’d have to decide on a cut-off point (define that only links of a certain weight and above will be included) – how would we decide on that and would it make sense?
2. And more importantly, the logic of a network flow can be completely turned around by leaving out one link. If you take a circular flow involving 6 actors. One of these flows is three steps removed from our interview partner so we decide that’s the cut-off point and remove that link. All of a sudden we don’t have a circle but a line with a beginning and end point. And Paul, who used to be an equal partner in the circle is an isolate all of a sudden… This is a simple example and easy to spot, but in complex networks there will be a lot of flows that don’t make sense any more if we remove one or two links just because they are too far away from the interview partner…
Filed under: Uncategorized | 1 Comment »
Posted on August 14, 2009 by Eva Schiffer
This is a question that makes mapping networks tricky, if you go beyond everyone’s immediate links. When doing Net-Maps, we normally ask: “Who is involved in XY? How are they linked?” So our interview partners give us some information about their own links but also on the linkages between others.
Krackhard has coined the term “Cognitive Social Structures” for these kinds of perceived networks and proposes stacking the network perceptions of different interviewees on top of each other to get a more realistic view of the whole story.
My colleagues Noora Aberman and Klaus Droppelmann developed the idea of ranking the links with the following in mind: The more steps away from your immediate links something happens, the less sure can you be about it. So while adding the different networks up, links that involve the interview partner would have a heavier weight than those that are one step removed. The links one step removed would again have a greater weight than those two steps removed.
Does anyone know of a study where this approach has been tried? What do you think?
Filed under: exploring new ideas, theoretical considerations | 1 Comment »
Posted on August 7, 2009 by Eva Schiffer
We learn about the Net-Map method as we go and define it as we use it. One problem we often encountered is: How do we consolidate different views of various interview partners to get to a common network map that describes the situation?
There are quantitative approaches to this, but I am wary about them, not only because they can be rather work intense and complicated but also because of the many assumptions you have to make and the kinds of distortions you get (e.g. one interviewee said A and B are linked, the other one said they are not – it’s likely that the second interviewees perspective is not represented in your combined map…).
For our research about fertilizer policy changes in Nigeria Noora Aberman and I used an approach that seems to be both pragmatic and quite reliable. At least for networks where there are no major conflicts or disagreements amongst interview partners. I’ve written about this when we were in the planning stage and because it worked out so well in the field, I’d like to recommend it again…:
First get a basic understanding of the issue you want to analyze and who might be knowledgeable about it.
Step 1: Interview a diverse mix of stakeholders and experts in individual interviews.
Step 2: Digest the information and draw a map of your new knowledge about this situation gathered through the interviews
Step 3: Invite interview partners and possibly others to a workshop where you introduce this consolidated map and discuss it with the participants as a way of validation.
Filed under: fine-tuning implementation | 1 Comment »
Posted on August 6, 2009 by Eva Schiffer
of mine are novels. One sister calls me a Buchtrinker (bookdrinker) because of the speed with which I gobble my books, the other one says: “Letter addict”, because she knows that I’ll also read the back of every corn flakes box or tomato tin in the house.
Most of the times I forget the story as fast as I read it though, and it’s more likely that I remember the color of the cover and the feel of the book than the title and author. Well, I have been thinking about keeping a reading list for a while and finally I’ve decided to wait no longer, here it is, the bookdrinker blog. I’d love to hear what you think about these books, if you’ve read them and maybe you can recommend others with a similar color of the cover or feel to it.
Filed under: exploring new ideas, musings | Leave a Comment »
Posted on July 28, 2009 by Eva Schiffer
The other day my colleague Regina Birner (IFPRI) asked me: “Do you think people generally overestimate their own influence (when drawing Net-Maps and setting up influence towers)?”
Hm, good question. It’s been shown by other social network analysts (plus, it’s somehow logical) that most people over-represent their own connectedness in a network: They know all of their own links but only some of the links other people have. So anytime you ask someone to draw a network that he or she is involved in, expect your artist to be in the midst of things and representing himself or herself as more central than other people would think they are (I have seen exceptions to this rule, though).
Now is the same true for influence towers? Do people in general think they are more influential than other people would think they are? From my experience I’d say it’s not that easy. There are those who think they are the center of the universe and consequently over-estimate their own influence. Then there are others who think it’s socially expected to be modest (or who actually are, for that matter) and they will underestimate their influence. One interviewee even refused to rate both, her own organization and that of the interviewer.
I have no quantitative analysis of this yet, so this is more of a gut feeling after doing a lot of interviews: Most interviewees seem to be rather spot on when rating their own influence and tend to agree with their network peers. I hope Regina and I find a way to launch a study big enough to do some more rigorous testing of this and other questions about reliability, replicability, interviewer bias and other aspects of Net-Map that make quantitatively oriented researchers nervous…
Filed under: open questions, technical details | Leave a Comment »
Posted on July 22, 2009 by Eva Schiffer
That’s the really interesting stuff, combining the analysis of social networks with the analysis of something else (XY): What kinds of networks are linked to greater innovations in companies (you need to collect indicators of innovativeness in addition to the network analysis), higher welfare in African villages (add welfare indicators) or being more effective terrorists (count successful acts of terrorism). That’s all about the question of how networks lead to something. And it get’s really mind boggling if you are able to link these results to specific structural issues of networks, such as centralization, flexibility over time, structural holes etc.
On the other hand you might also want to know what leads to people achieving certain positions within networks, what makes someone a hub in a hub-and-spoke network, why are some people so much better connected than others, act as boundary spanners or bottle-necks? Is it all about personality, income, geographic location, organizational structure, time spent in a specific field, cultural background or something completely different?
When people are exited about the new kind of data and understanding you can gather with social network analysis, they (and I am part of “they” here) often don’t realize that your understanding can move to the next level, if you combine network with non-network data.
Is there anyone out there who has combined social network and geographic data? I have this visual of a presentation in my head: First you show a social network map with the actors with highest centrality in the middle and the lesser connected nodes at the fringes (i.e. standard visualization of networks), on the next slide you see how these actors are distributed in space (e.g. where their farms are or where their offices are located) and then you put the social network on top of the geographic and the actors slowly move to their place in space while maintaining their network links. Not only would that look cool, but you would also be able to give your audience a very direct feel for whether or not the social networks are linked to the geographic position of the actors: Do they have stronger links with their immediate neighbors? Does the guy who lives in the middle of the village (or whose office is by the water cooler) really have the most dense network?
At the moment I neither have the data to play with nor would I know how to animate it, but maybe there is someone out there, who has already done that? Or, maybe you have the data and want to discuss how you could best visualize it? Or you are planning a research project and think: Wow, this is how I’m going to do it! But how am I going to do it? Drop me a line.
Filed under: exploring new ideas, open questions, theoretical considerations | 4 Comments »
Net-Map is an interview-based mapping tool that helps people understand, visualize, discuss, and improve situations in which many different actors influence outcomes. 
