Wednesday, August 25, 2004

Learning UCINET

[I am reposting this with a correction: the data files I could not initially find are in fact provided with the UCINET installation. Thanks to Steve Borgatti and Emily Case for their help.]

I just started my own personal crash course in UCINET, one of the most popular tools for social network analysis. I have long known of UCINET and its creators Stephen Borgatti, Martin Everett, and Linton Freeman, but I never tried using it until now.

Before I even downloaded the software, I boldly predicted to a colleague that I would have UCINET mastered in a couple days. She cautioned me not to be so optimistic, and now that I'm well beyond a couple days into my crash course, I have to say she is right. I am finding the learning to be a challenging process of "connect the dots" where I can see most of the dots but have to guess the order to connect them in.

Here is a brief review of the helpfulness of different UCINET/SNA documents that I've found:

The official UCINET User's Guide covers the gritty details of UCINET in extended essay form. The User's Guide is technically well written; however, it's challenging to stay with its narrative flow and difficult to navigate as a reference. Nevertheless, for many UCINET questions, this document is the only place to turn.

Follow the Tutorial link on UCINET's main site to read a helpful introduction to social network analysis, written by Robert A. Hanneman, professor of sociology at UC Riverside. Unfortunately, this tutorial does not include any specific guidance to using UCINET.

There are a couple good sets of class notes out there to help the intrepid adventurer map his own course of UCINET exploration:

I recommend checking out the course schedule from Stephen Borgatti's July 2004 class on social network analysis, taught at the University of Essex. This includes the entire syllabus from the two-week course. There are many links, including a very promising document entitled "Getting Acquainted with UCINET and NetDraw" that uses data provided with the UCINET installation.

Another good set of class notes comes again from Robert A. Hanneman. His Tutorial on the UCINET site is a companion to his collaborative workshop called "An Introduction to Social Network Analysis using Pajek and UCINET." (Pajek is another tool similar to UCINET.) The course is a very quick introduction designed to get people up to speed and doing practical social network analysis.

There is also an active UCINET user's group on Yahoo! that I have just joined. And a wiki from Borgatti's recent course at Essex, which includes links to just about everything I've mentioned so far.

8 comments:

Anonymous said...

Hi Bruce, I attended Steve Borgatti's excellent 2004 course at the University of Essex and am responsible for initiating the course wiki you mention. You may also like to know that there is a course user group hosted at: http://groups.yahoo.com/group/SNAEssex2004/. We're all UCINET newbies and welcome anyone who is new to UCINET. Having tried myself to use UCINET before attending Steve's course I can empathise with you. I must admit having a few Eureka moments on the course! I've added a link to your site on the wiki, hope that's OK?

Anonymous said...

"No Charge Online Advertising Channels For Any Business"

Anonymous said...

Couldnt everyone USE FREE Advertising and Marketing Resources?

Sabrina said...

Hi! I'm an Italian student...my name is Sabrina.
I'd like to known if with ucinet is possible to have soem information about the equilibrium (nash equilibrium) between the agents on the graph (or network).
My contact is sabrina.costato@libero.it

Anonymous said...

Very helpful information. Thanks!

Pageboyz said...

Im rediscovering the pleasures of SNA software... this is truly helpful, esp with UCINET.

Hannah said...

Thanks! It's useful

adi said...

Hello, I am a post graduate student doing some research in social networks advertising. I need a tool which can visually represent social networks andcan also support applicationos some algorithms on the nodes and then giving visual resuls in the form of clusters.

Will UCINET be usefull for me?