ONA Project

The see syllabus for details on the presentation and final paper that present the results of the project.

This is is your main assignment in this course. It is a team assignment, although you can subdivide into smaller teams if you prefer. The idea is to do a complete social network analysis of an organization (or some part of it). This includes data collection, analysis, and interpretation of results. The results will be presented in class, and also written up as a group term paper. This is a complex assignment and you should feel free to contact the instructor and TA as often as necessary to ask for help.

In addition, we have written the following online guide:

As I see it, a project like this has the following steps:

    • Design. This is where you figure out what your research questions are, what part of your organization you are going to study, how you are going collect the data, and how you will ensure confidentiality and privacy. Beginners have a strong tendency to give the design step short shrift, which is a very bad idea. It almost results in a lot pain later.

    • Collect. This is where you actually collect the data. For most of you, this will be via an online survey. For others, this will come from an internal database, such as a billings database in which each employee has recorded the amount time they have spent on each project over a given period. Note you will need to get permission from the organization to do this. In all likelihood, this will require a quid pro quo in which you provide the organization with some of the results. It is very important that you not over-promise here and compromise the privacy of your respondents.

    • Enter data. This is where you convert the data from, say, an online survey tool, into something UCINET can use. You should feel free to seek help on this from your IT department or tech-savvy teenager. You don't want to get bogged down here. Our TA Wookje can also help.

    • Visualize. This is where you draw diagrams of your network(s).

    • Measure. This is where you calculate metrics, such as node centrality.

    • Analyze. Here's where you look for relationships between measured variables. For example, you might use network position to predict performance.

    • Interpret. The interpretation typically has two aspects, diagnosis and prescription. The diagnosis part is where you figure out what the results mean for the organization and/or individuals studied. The prescription part is where you make suggestions, based on the diagnosis, of what steps should be taken to correct problems, improve performance, etc.