What Data to Collect
Define Your Questions:
There are three elements to understand what data to collect for a specific social network analysis:
- Define the group(s) of people or institutions whose inter-relations you need to understand. This will define the nodes of the network
- Define what aspects of the relationships you need to know. This will define the meaning of the connections (edges) between the nodes in the network and the meaning of the results in your social network analysis.
- Define what properties of the network will change as a result of your program. This will define the focus of your evaluation.
The second point is a key issue to understand in social network analysis. All social networks, or information networks, can be defined only in relation to a specific context or subject of interest. The meaning, structure and other properties of the networks we study depend on the questions we ask; social network analysis is the same as any other technique in that respect.
To make this explicit, here is a simple example. Suppose we wanted to know how knowledge of IPM principles is distributed among a group of 10 school administrators. Either by survey or in a focus group we could start by asking our study participants the question "Name the other school administrators you have met and know". From this lists of names received we can represent the social network for who knows who as a set of 10 nodes (bubbles) connected by lines (or "edges" in network vocabulary) showing those people who report knowing each other (see the figure on the left).
Now suppose we ask the more specific question, "Name other school administrators you talk to about IPM". We can imagine that this question will result in a less densely connected network (i.e. one with fewer edges) since school administrators meet each other for a variety of reasons and so some of our group might know each other but not discuss IPM. Asking this question might result in the network shown in the second figure. The two networks are not the same because we asked different questions to elicit the data on which they are based.
The Question Defines the Network
To reiterate, social networks are defined in relation to specific issues or questions. For this reason it is crucial to invest effort deciding which questions to ask since these will define the networks you see.
The third point is the most difficult to address and should be the focus of planning a social network analysis. At a simplistic level there are only two things about a network that can be affected by an IPM program and hence be amenable to evaluation: the nodes of the network can change, or the edges can change. The extreme case is that a new network is formed as a result of the IPM program. For any of these possibilities your data will have to allow for comparison of "before and after" or "case and control" type studies.