Chasing the turning points

Most processes, be it for example the implementation of a new policy or a development project, usually have a few crucial turning points, which decide over the success or failures of these processes. These moments are of special interest because they allow us to learn and to improve future activities. Until very recently Net-Map was used to explore the situation of one point in time. Then we started to think about ways of using Net-Map to investigate a process over a certain time period. The solution we found was the combination of Net-Map with the time line tool. The time line tool is a participatory rural appraisal (PRA) method to capture key events in the past.

This exercise can help to understand who were the drivers of change at the key events, how they were linked, which goals they had and how much they really influenced the process according to the interview partner.

The following steps describe only one possible option how to combine the two tools to draw networks and influences at different points in time:

  1. Select a process that you want to analyze. This could be for example the successful implementation of an agricultural reform or the development of a project, which did not go the way expected.
  2. If possible, do a literature review of the process if it is analyzed already. Reading in the newspapers or progress reports can also help to gain knowledge in advance.
  3. Identify people who have a sufficient knowledge about the case and who were involved in the process the whole time or at different stages. Try to include people from different levels, such as policymakers, extension agent of government agencies or NGOs, farmers, scientists, etc.
  4. Develop time lines with the people identified which highlight all events in relation to the process. You can do this as a group exercise which has the advantage of saving time, since you have to deal with only one time line. You can also do this as individual interviews, which has the advantage of giving every interview partner the opportunity to express her or his view, without fearing negative consequences. The disadvantage is that you have to collapse the time line afterwards which can be difficult in cases where they are not alike.
  5. The next step is a very important one: Ask the interview partner about the goals of the different actors, e.g. whether they were in favor or against the implementation of the new policy.
  6. After you have established the time line you select interview partners who seem to have a sound understanding of the whole process. You ask them to select the two to four most important turning points that led to the success or failure of the whole process.
  7. For each of the points in time you ask who was involved at that point, how those involved were linked and how much they influenced the process at this particular point in the process. The different opportunities on how to draw a network map are described on this homepage under “About”.

As usual some creativity and careful thoughts are required to keep the exercise as short as possible. Also consider to split up the interviews into two or three sessions. (This innovative use of Net-Map was conceived in discussions between Jennifer Hauck, Eva Schiffer, Roukaya Zimmermann and Regina Birner)

How much money? How much trust?

After my Net-Map presentation in Beijing, a colleague remarked: “But it does make a difference, how much money I get from someone, not only if I do or don’t.”

My colleague is right: it is interesting to know how strong the links in my network are: Who is my biggest supporter, who is my most trusted source of advice, who is responsible for the most powerful disturbance?

In social network analysis these questions are approached by assessing the weight of links. Basically you pose the above questions to your interviewee. There are different ways how you can pose the question and note down the answers on your Net-Map.

  • You can ask your interview-partners to rate the links according to importance and note a “1” next to the most important link, a “2” next to the second most important link and so on. This is especially useful if the links describe a flow that is not easily quantified (e.g. advice) and you will get most useful data if you look at ego-networks, asking: “Which link is most important for you?” And not: “Which link is most important for the network?”
  • If something easily quantifiable flows through the links (e.g. money), you can ask the interview-partner to quantify how much flows between the actors in a specified period of time (e.g. US$/year) and write the amounts next to the links. The same is possible if the link refers to regular interactions and you want to know e.g. “How often do the actors meet to exchange information?” This however requires an interviewee that is very knowledgeable about the network and might otherwise lead to a lot of guesswork.
  • If you want to get a general understanding about primary and secondary links, without going into too much detail, you can choose different colours, e.g. dark green for most important/most trusted flows of advice and bright green for less important/less trusted ones. Or you can vary the thickness of lines.

When visualizing the results with the available computer programs, you can either add the weight as a figure next to the links or vary the thickness of lines according to weight.

However, let me issue a word of warning: Rating is a rather time-consuming endeavour, especially in complex networks. Another colleague told me that in her group exercises the participants enjoyed drawing networks and rating links but were exhausted and lost interest, when she asked them to put actors on influence towers. This might be due to the fact that both rating of links and putting actors on influence towers asks for very similar information about the importance of actors in the network.

So as a rule of thumb I would recommend always keeping the exercise as short as possible and either rating links or putting actors on towers. If you, however, decide that you need both kinds of data, look for other ways of condensing the mapping: E.g. limit the map to one or two crucial kinds of links. Or limit – and pre-define – the actors you want to include.

Kinds of links

Think of everyone you know. Try to make a list of all different ways you interact with these people. What do you give to and get from them?

Both lists (who you know and what kinds of links you maintain) will barely ever be absolutely complete, so this is more of an exercise to start and get your mind going, to get an expansive understanding of what we mean by “kinds of links”. Thinking of my own personal networks, there are people that I

• Cook for/provide food to
• Give or get money from
• Get formal instructions from
• Am related to (as family members)
• Exchange information with
• Greet on the street
• Argue with
• Give or get advice from
• etc.

Some of these links go into two directions (e.g. exchange information), some of them mostly go into one direction (e.g. mostly I am the one who cooks), others go in one direction by definition (e.g. giving instructions), or don’t have a direction at all (e.g. being relatives).

Also, these links may differ in the way they can change my behavior (or that of members of the network): Some are rather direct and clear (e.g. formal instructions), while others (e.g. providing food) might have a far more subtle and less clearly marked impact. My relation to people I greet on the street is far less stable and important to me than to those that I seek out for advice.

And finally, some of these links will matter in some situations, while others will be more important in other contexts. When it comes to making decisions concerning the house I live in, other people and links will influence me, than when I talk about my work life.

What does all this mean for using Net-Map?
There is an endless number of possible kinds of links you can draw. Drawing all possible links would be as futile as trying to draw a map of the world in 1:1 scale. Focus on those links that are relevant in your field and concerning the decisions you want to understand better.

In your pre-test you will get a better feeling for your links, especially by looking out for those that somehow “don’t work”. Some links will occur basically between everyone. Collecting them is time consuming and the resulting data tends to be boring. So rather cut them out or define them more specifically (e.g. instead of “exchange of information” ask about “exchange of information concerning project implementation”). Some kinds of links will nearly always occur together (in some setting giving command and giving funding are tied to each other). Think about bundling them or only choosing one of them. Some links will be very rare. Maybe your definition is too narrow. Or this link is not really central to this kind of network.

Discuss your links with people who are part of or knowledgeable about the network you analyze, to make sure they are relevant, understandable and meaningful to them. It might be that you get very different pictures of the links in your different pre-test interviews. Try to find out whether this is really due to different views of the network of your interview-partners, or whether your definitions of links are unclear and your interview partners talk about different things, when they discuss who is e.g. giving advice to whom.

Some examples for kinds of links that we have used in our studies are given in the “Case Study” section.