Talk:Social network analysis of an aid agency

From KM4Dev Wiki
Revision as of 14:49, 15 September 2010 by Davide Piga (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

See the original thread of this E-Discussion on D-Groups

Ben Ramalingam, 2009/10/08

Hi all,

Does anyone know of examples of social network analysis (SNA) being applied within an aid agency?

I have a number of examples of SNA being used in cross-organisational multi-stakeholder processes (disaster response in Mozambique, civil society networks in Ghana, INEE etc), and work I did myself on actor-focused network analysis in SDC, but I have seen nothing which formally uses SNA in a particular agency... I am particularly interested in getting my hands on a social network visualisation.

Any tips or suggestions very warmly recieved.

Rick Davies, 2009/10/08

Hi Ben

A few years ago, as part of a piece of work with Oxfam Australia (with Rosy Eyben) I was given access to internal data on staff membership of the many and varied committees, working groups and internal mailing lists, within Oxfam International. I used Netdraw to produce a network diagram of the staff relationships, as expressed in the form of shared membership of these different bodies. But for a number of reasons I did not take it further, e.g. by getting staff to predict what the aggregate network structure would look like, or to comment on the aggregate structure that was actually produced by the data available.

Subject to Oxfam's permission, I may be able to share a version of the network diagram where the names of individual staff have been anonomysed.

It would be interesting to discuss how such data could best be analysed, and tracked over time

Bakhtiar Ali, 2009/10/08

Hi Davies

Regarding SNA, You developed diargramme based on "shared membership". What i learned is that SNA is done based of many kind relationship like, emotional, personal, professional, bla bla.., that indicate relationship has many dimensions. It would be nice to exactly point out first these relationship types.

Anybody would like to suggest types of relations?

However, my intention is not to deviate discussion from the Ben issue. It is just to add something more.

Thanks to Ben for starting such nice and important discussion.

Rick Davies, 2009/10/08

Hi Dr Bakhitar

You are exactly right, there are many kinds of relationships between members of an organisation that could be the focus of a social network analysis.

In the Oxfam example, my choice of focus was opportunistic. I found that data was available, and analysis of this data had the potential to shed light on aspects of organisational structures that typically do not appear in traditional organograms ( which focus on more formal and stable structures within organisations). Further out on an imagined continuum of formality/informality you could find, and possibly map, friendship relationships. This would be more challenging, not only to map, but also to interpret.

A lot has been written about the relationship between organisational structure and strategy. Being able to map structures is one way of starting an analysis of this relationship.Once a method of mapping has been identified one of the important choices to then make is whether to go for an inductive/interpretative approach versus a more predictive/ theory testing approach. Or in simpler words, show the results and get participants interpretations, or elicit participants predictions of what the structure will look like, then share and discuss the actual findings (the mapped structure)

Giulio Quaggiotto, 2009/10/08

Hi Ben,

A couple of years ago when I was at the IFC I conducted a social network analysis working with Rob Cross at the University of Virginia to analyse knowledge flows within the department, focusing in particular on the issue of onboarding: how long does it take new staff to integrate in the informal networks within the department. We then developed an induction programme focusing in particular on this aspect and conducted a follow up analysis to measure impact.

In terms of visualisation, we used the Network Roundtable's software On the site, you should be able to find examples of visuals. I believe Valdis Krebs also has a software for this. If you are interested in visualising networks as they form in social media within an organisation, Trampoline is an interesting piece of software to use. See this fun example :-)

Valerie Brown, 2009/10/09

Not quite on Ben's question, but to add to the uses of social network analysis, which I hav efoun dvery interesting: I have been involved in 2 uses. One is a 3-day national conference on knowledge management,where a friendship network was created from participants inputs into a survey and was ready at the start of the conference -a great contact-oling process. Another was it use in a research design on the concept of the bazaar as a social hub for a community or region, identifying the bazaar's social networks for commerce, support, friendship, knowledge flows, conviviality etc.and theit interconnections.

Reinier Battenberg, 2009/10/09


Out of curiousity I surfed around a bit. Just to see if there are opensource solutions for this issue as well. (after all, most of us are spending public money)

And yes there is:

Looks neat, just an Excell plugin that can pull data from your mailbox, twitter and other sources and graph it all out. Perhaps someone running windows could try this out on an archived folder of this mailinglist to show what it can do.

Social Networks Visualizer SocNetV also offers a built-in web crawler, allowing you to automatically create networks

SONIVIS:Tool aims at analysing social (virtual) information spaces like Wikis. These spaces are investigated by using different network definitions (collaboration/information networks). Clustering algorithms and statistiscal analyses are provided.

For the real powerusers there is a statistical language/tool called R. It is a huge collection of libraries that can do magic with big datasets. It has a very steep learning-curve, but for real (profesional) number-crunches it is supposed to be totally worth it. And it has social networking libraries

Barbara Fillip, 2009/10/09

A few links on Social Network Analysis

One lesson I learned: whatever tool you pick, make sure to ask if all the functions will work on different systems (PC. vs. mac) and different browsers and that there are no organizational constraints (as in... the IT department can't participate because they're barred from using IE... or something like that).

Rick Davies, 2009/10/09

Hi all

Re Renier's list below

There is also Netdraw, which is free, and Louisa Clark's manual on how to use it (if you are not an SNA expert)

Elizabeth Corley, 2009/10/09


Some of you might be interested in the publication by Ulrich Ernst on the value of social linkages.

Patti Anklam, 2009/10/09

Hello, Ben,

I have been working with IFAD and IDRC using social network analysis to determine the impact of knowledge network development on the overall connectivity of the network. We have baseline data from last fall and will be doing a follow-up comparison next spring.

As part of this work, we are also doing county-level network analyses in 2 countries.

We will be writing up results when we have completed the work in the spring.

Nancy White, 2009/10/10

I do hope someone will summarize this thread on the wiki. I'm writing offline at the moment, but I believe there is already a social network page... so you might want to search first

Roxanna Samii, 2009/10/11

Hi Ben,

Patti Anklam did an SNA exercise for IFAD-funded ENRAP network. Ciao Roxy