AOM Session #809: Summary

I attended a symposium on Monday afternoon of AOM entitled, Email Data in Social Network Analysis (full details from the program are below the break). It was co-sponsored by the Organizational Management Theory (OMT) and Organizational Communication and Information System (OCIS) divisions. The panel members were: Adam Kleinbaum (Harvard), Marshall Van Alstyne (Boston U.), Nathaniel Bulkley (U. of Michigan) and Eric Quintane (U. of Melbourne); Ron Burt was also in the audience.

Paraphrasing Adam Kleinbaum, here’s the motivation for the symposium:

There have been few innovation in data collection methods in social network analysis field since it began 70 years ago. This symposium is intended to crystallize a conversation that is emerging in different corners of the field about how and when to use email data for social network analysis.

The benefits including comprehensiveness (including longitudinal), ability to observe content of ties (via content analysis) and analysis at a new scale (far more people in a common network). The challenges include getting access to email, the ability to get and use data (permission), and being able to manage and analyze data (e.g., the need for data manipulation skills).

There are also major theoretical challenges including:
– what does an email tie really mean in an analysis.
– what kind of validity does the tie have?
– when and under what circumstances is email analysis a helpful complement to other analysis?

Marshall Van Alstyne and Nat Buckley both presented details of studies on executive recruiting. You can find some of this research at Van Alstyne’s SSRN site. Kleinbaum talked about his study of a very large data set that includes: 30,000 employees; 46 million email messages, and 5.3 million calendar entries. His interest is on unit-to-unit level communication. Eric Quintane provided a methodological paper on similarities and differences between perceived and observed networks.

In all, I found the session to be quite informative and highly relevant to the work that I do on large-scale communication networks. I’m not convinced that the symposium met its lofty goals (“crystallize how and when to use email analysis”) but it keep an important dialog going. The researchers on the symposium panel are forming a Google group on using email data for SNA analysis. I’ll post the name once I track it down again!

Program Session #: 809 | Submission: 11746 | Sponsor(s): (OMT, OCIS)
Scheduled: Monday, Aug 6 2007 12:20PM – 2:10PM at Philadelphia Marriott in Franklin 7
E-Mail Data in Social Network Analysis
Chair: Adam M. Kleinbaum; Harvard U.;
Chair: Eric Quintane; U. of Melbourne;

Although the field of social network analysis dates back at least 70 years, there have been relatively few major innovations in the sourcing of network data; the dominant source of data throughout the modern history of social network analysis has been the questionnaire. In recent years, however, a new source of network data has emerged: electronic mail. In this symposium, we aim to explore the implications of e-mail data for substantive social network research in organizations. Specifically, we have three objectives: First, we hope to present and discuss a series of novel, interesting papers that use e-mail for substantive network analysis. Second, we hope to step back from the particular papers presented and explore important theoretical questions about the use of e-mail for network analysis. Finally, and more generally, we hope that this symposium will serve as a focal point for the emerging conversation among the Academy of Management’s members about e-mail data in social network analysis.

Formal and Informal Structures in Inter-Unit Coordination
Presenter: Adam M. Kleinbaum; Harvard U.;

Information, Technology and Information Worker Productivity: Task Level Evidence
Presenter: Sinan Aral; MIT Sloan School of Management;
Presenter: Marshall Van Alstyne; Boston U.;

Using Email to Disambiguate Relationships Between Social Capital and Performance
Presenter: Nathaniel Bulkley; U. of Michigan;
Presenter: Marshall Van Alstyne; Boston U.;

Email and Survey Data. The Correspondence Between Perceived and Observed Networks
Presenter: Eric Quintane; U. of Melbourne;
Presenter: Adam M. Kleinbaum; Harvard U.;


2 Responses

  1. Steve,
    Thanks for sharing this summary. (And it was nice to meet you at the conference by the way.) What was the gist of the talk on The Correspondence Between Perceived and Observed Networks?

  2. Yeliz — It was a real pleasure meeting you a well. I enjoyed talking with you about your research interests.

    The gist of talk was they did a survey of folks in one organization looking at media choice. I don’t remember the exact findings but it was basically along the lines that people in the organization had fairly consistent patterns of usage for concluding if email was a “substitute” or “complement” to other communication channels. [The paper was perhaps misnamed as it wasn’t really about perceived and observed networks in the same way as other work with similar titles.]

    The problem I had with that paper wasn’t so much the study itself (which was done just fine) as it was the implicit implication of how that study was going to be used in light of the other work being done by the panelists. [They were all part of the same inter-related research teams.]

    The stated purpose of the symposium was figuring out when it is appropriate to use email in SNA; e.g., when can it be as a proxy for other relationships. When you start to get into really large email data sets, I think you have to be really careful about the inferences you make around what kinds of social ties email communication represents.

    I think the choice an individual makes about using email as a substitute or complement to other communication channels (and, therefore, to what kind of relationship it represents) varies over time, by geographic location, and within organizational unit (e.g., by structure).

    There are strong organizational norms that come into play in those decisions–how your boss and how your teammates communicate will shape how you communicate.

    If any of the same determinants for email usage are related to other variables of interest in a study, that introduces all sorts of potential bias that will confound study results.

    I asked a question raising these issues at the symposium. I wasn’t entirely sure from the response if I’d made my point very clearly or not.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: