How to Collect the Data
Where to Find the People
The data for Social Network Ananlysis (social network analysis) can be collected using several of the other approaches discussed in other modules; for example surveys, questionnaires, or focus groups can all be used to gather data for social network analysis. In these cases the data will be elicited from the participants (as in the example on the "What to collect" page). In some circumstances it might be possible to obtain suitable data from existing sources. For example:
- Attendance lists from meetings and conferences
- Email discussion threads
- Working group participant lists
Irrespective of whether your data will be collected from participants in your program, or existing sources, the important issue is to collect data that will give you the information you need. To make this more concrete we'll continue with the example from the previous page. Suppose our project is intended to increase the level of discussion among school administrators about IPM. Clearly, an evaluation of the project will need to assess whether that goal aim has been achieved. Asking ourselves how we should collect the data to demonstrate an increased level of discussion is the same as asking the question how we ourselves would know that the project has been successful. What we need are:
- Baseline data on the level of discussion about IPM among a case group of administrators, and a control group. (Cases are administrators who will participate in the IPM project, and controls are administrators who will not)
- Repeat data at one or more subsequent time point(s) for the case group and control group about level of discussion amongst themselves about IPM.
This basic case/control cross-sectional study has the least number of sets of observations required to answer our question, and by the same token, allow for effective evaluation. The figure below tries to capture these concepts. In the case group, in the baseline there are 9 two-way conversations about IPM among 10 administrators and the same number in the control group. In the repeat observation there are 13 conversations in the case group and 11 in the control group.