Thursday, October 27, 2011

Ushering peace: Using SNA to nip violent crisis in the bud


A lot has been written about analyzing social networks to understand life cycle of revolutions and how they can be coordinated better using social network technologies. However, another potential utility for SNA can be found on the other end of the spectrum. Analyzing communities that show early signs of unrest to address issues before they snowball in to a crisis.

The groundwork:

The first step for such a study would involve identifying the community that is likely to have the greatest degree of resentment, the potential to act on it and to identify some of their key demands. Experience shows that certain communities like labor, youth, the unemployed, traders etc. have a more significant role to play in conflicts than say the retirees.

Once a preliminary analysis has been done to identify the target community (which is considered outside the scope of this topic), a demographic analysis would be crucial in framing the questions and identifying the geographical areas to target. 

The challenge:

At the outset, the author understands that collecting data under certain circumstances would be a far greater challenge than within an organization where employees are mandated to provide it. The data collecting agency, whether government or NGO, may also be received with hostility and uncooperativeness. The reliability of the data can be a point of concern. These issues, while important and relevant, will have to be the subject of another blog post.  

The first step:

Once a target community and the geographical areas have been identified, the next step would be identifying the most efficient way of reaching the member of those communities. Often time this will come in the form of established groups and associations, whether political or non-political, who represent that community. Assuming that these organizations would be open to talking about their issues, the next step would be the actual survey.

The author envisages this to be a four-step, potentially iterative, process (who said crisis management was easy!):

1.     Data collection for issue identification.
2.     Data analysis for identifying key focus area and brainstorming possible solutions.
3.     Follow-up survey to identify mass appeal of possible solutions.
4.     Data analysis of follow-up survey. Iterate step 2-4, if necessary.

The survey:

The following would be the broader segments in which such a survey can be divided:

Identifying information:

·      Name
·      Age
·      Gender
·      Position in the organization
·      Location
·      Academic qualification
·      Profession (including ‘unemployed’ as an option, if relevant)
·      Religious affiliation (if relevant)

Issue identification information:

·      Three most significant issues of concern, in order of priority. (This question would not limit the options to pre-defined choices. The analyses of the answer will help determine the key areas of concern.)


The network:

·      How frequently have you communicated in the last two months?
o   Never
o   Occasionally (at least once a week)
o   Frequently (more than once a week)


The analysis:

Once the data is available, several important interpretations can be drawn:

·      The first and the obvious one would be to do an issue based analysis to identify key focus areas i.e. what is the most pressing issue that needs to be addressed?
·      Once the key issues have been identified, the key players will need to be determined. Here in conjunction with an individual’s ‘Position’ in an organization we will need to look at the following attributes of the nodes:

o   Eigen vector centrality: Who is the most important node in the network? Does it correspond to the position?
o   In-degree: Whom are the most people communicating with?
o   Out-degree: Who is reaching out to the most people?
o   Any clusters, particularly based on issues or locations?
o   Closeness at an organizational level to determine the right people to engage with.

Taking it to the next level:

The survey and analysis can be further extended to look at inter-organizational relations. This will be a key analysis as crises are seldom the results of an effort of one organization alone. The desired information can be extracted from the important position holders in each organization to determine what kind of contacts they share with other peer organizations and then analyze if they are focused on the same issues.

The benefit:

The result of this analysis will be invaluable for agencies to not only determine the issues of substance but to also identify the stated and un-stated decision makers in each of the relevant organizations. Pro-actively collecting this information and doing the analysis can act not only as a control but also a prevention mechanism.

2 comments:

Christopher Tunnard said...

OK, but the problem is that the reader really needs to have a hypothetical in order to evaluate how useful the information you outline would be. Simply saying that "certain communities like labor, youth, the unemployed, traders etc. have a more significant role to play in conflicts than say the retirees" is a bit simplistic. Which youth? Traders of what? Which organizations? What agencies?

I'm not saying that the methodology you have might not be appropriate, but without any context at all, it is really difficult to assess.

Shashank said...

Agreed. I was using the youth unrest in Kashmir over the summer of 2010 as a guiding beacon for my thought process. However I decided to exclude the specific mention of the convoluted issue so as to not deviate from my core focal point. But I can see that the process would be easier to understand when exemplified.