Blog of the Borg

Networks and Attributes

Steve Borgatti — Sat, 27 Aug 2011 18:52:31 +0000

In class yesterday, Ajay argued that the distinction between attributes and relations — so often made in intro-to-networks contexts — is not so clear. An attribute is a node-level variable, so centrality and any other node-level network concept is just another attribute. This is surely true, but it also a matter of semantics. Dan B prefers to use “attribute” in a more restricted way. An attribute is a characteristic of an actor themselves and not of their location in their social environment, such as their gender or attitude toward texting. In contrast, a node-level network property is a characteristic of a node’s position in the network, and can change if a tie is added or lost anywhere in the network. So gender and attitude are attributes while betweenness centrality is a node-level property. Joe said this distinction is problematic because network processes can cause attributes. For example, a person’s attitude is an attribute but it may have been formed as a result of contact with others with that attitude. But in my view the fact that attributes and network properties can cause each other is not relevant to the conceptual distinction between them; heat may cause fire but they are not the same thing.

The distinction here reminds me of the distinction between social networks and networks in general, which also came up in class. In my view, it is useful to restrict the term social network to just those cases where the nodes are entities, i.e., have agency. I can define a network in which the nodes are words and the ties indicate whether a given pair of words co-occurs in the same sentence in a given text. But I’d rather not call that a social network, and I certainly don’t want to apply social theory to it.

Political Networks Conference & Workshops

Steve Borgatti — Tue, 01 Feb 2011 02:27:20 +0000

4th Annual Political Networks Conference and Workshops

Call for Papers

The 4th Annual Political Networks Conference and Workshops will be held June 14-18, 2011 at the Gerald R. Ford School of Public Policy, University of Michigan, Ann Arbor. Approximately 150-200 political networks scholars from around the world are expected to be in attendance.

The study of political networks serves a key role in understanding governance, as politics is largely driven by relationships between actors, agencies, and institutions. A greater recognition of these relationships has begun to change the study of politics. We are soliciting papers that apply network ideas, from substantive insights to methodological innovations, to topics related to American politics, international relations, comparative politics, political theory, public administration, political methodology, or other areas of politics. We are particularly interested in proposals that are not only descriptive, but that also make causal claims with clear identification strategies. Submissions are encouraged from a wide range of disciplines and interdisciplinary fields including, but not limited to, political science, sociology, economics, public policy, anthropology, psychology, business, information systems, mathematics, physics, and complex systems.

The full conference will kick off on June 16 with an opening reception and keynote address by Garry Robins, University of Melbourne. A full schedule of panel presentations begins on June 17 and 18. Mark Newman of the University of Michigan will give the plenary address on June 17.

Due to the success of the workshop sessions held in prior years at Harvard and Duke, workshop offerings at this year?s event have been expanded. The first three days will be devoted to didactic sessions on network methodology by leading experts in the field, including Garry Robins, Steve Borgatti, Mark Newman, Carter Butts, David Siegel, Meredith Rolfe, and Michael Heaney.

Sessions are scheduled as follows:

June 14: Beginning Workshop on Network Analysis (assumes no prior training)
June 15: Computer Applications in Network Analysis
June 16: Specialized Workshops in Network Analysis and Keynote Address (5pm)
June 17: Conference Panels, Plenary Address, and Poster Session
June 18: Conference Panels, Plenary Panel, and Business Meeting

Key Deadlines:

March 15: Deadline for proposals
April 1: Invitations to conference announced
April 15: Deadline to confirm or decline invitations
May 1: Deadline for early registration

NSF-funded fellowships are available to support graduate students, postdoctoral fellows, and junior faculty to attend the conference and workshops. Scholars and students from outside the U.S. are also eligible to apply. The fellowship amount will be pro-rated to cover the number of workshop days. Conference attendance is required for all fellowship recipients.

To submit a proposal, register, apply for funding, find accommodations, or obtain other information, please visit

http://sitemaker.umich.edu/fordschool-pnc

Additionally, the 4th Annual Political Networks Conference and Workshops will be held immediately prior to the beginning of the ICPSR Summer Program at the University of Michigan.

There are two ICPSR opportunities that may be of interest to conference participants. First, ICPSR is offering a one-week course (June 20-24) on Advanced Network Analysis, which will provide an overview of how to handle and analyze very large-scale network data, such as Federal Elections Commission and Twitter data. Fellowships of up to $1,000 will be available to political networks conference attendees, who wish to attend this seminar as well. More information about this ICPSR course is available at

http://www.icpsr.umich.edu/icpsrweb/sumprog/courses/0135

Second, ICPSR is offering a four-week course (June 20-July 15) on Network Analysis. More information about this ICPSR course is available at

http://www.icpsr.umich.edu/icpsrweb/sumprog/courses/0131

Plan to come to the 4th Annual Political Networks Conference and Workshops, June 14-18, 2011, and stay in Ann Arbor for a summer of training in networks through the ICPSR Summer Program.

Program Co-Chairs

Meredith Rolfe, University of Oxford (meredith.rolfe@sbs.ox.ac.uk)
Betsy Sinclair, University of Chicago (betsy@uchicago.edu)
University of Michigan Host Committee
Michael T. Heaney, Chair (mheaney@umich.edu)
Philip B. K. Potter, Co-Chair
Robert J. Franzese, Jr.
Elisabeth R. Gerber
William G. Jacoby

2011 SAGE Handbook of Social Network Analysis

Steve Borgatti — Tue, 25 Jan 2011 16:54:58 +0000

Edited by Carrington and Scott, coming out soon. Here’s the list of chapters:

1. Introduction Peter J. Carrington and John Scott

2. Social Network Analysis: An Introduction Alexandra Marin and Barry Wellman

3. The Development of Social Network Analysis-with an Emphasis on Recent Events Linton C Freeman

4. Network Theory Stephen P Borgatti and Virginie Lopez-Kidwell

5. Social Physics and Social Networks John Scott

6. Social Networks in Economics Sanjeev Goyal

7. Relational Sociology, Culture, and Agency Ann Mische

8. Personal Communities: The World According To Me Vincent Chua, Julia Madej and Barry Wellman

9. Social Support Lijun Song, Joonmo Son and Nan Lin

10. Kinship, Class, and Community Douglas R White

11. Animal Social Networks Katherine Faust

12. Networking Online: Cybercommunities Anatoliy Gruzd and Caroline Haythornthwaite

13. Corporate Elites and Intercorporate Networks William K Carroll and J P Sapinski

14. Political Dimensions of Corporate Connections Matthew Bond and Nicholas Harrigan

15. Policy Networks David Knoke

16. Social Movements and Collective Action Mario Diani

17. Crime and Social Network Analysis Peter J Carrington

18. Terrorist Networks: The Threat of Connectivity Renée C van der Hulst

19. Scientific and Scholarly Networks Howard D White

20. Cultural Networks Paul DiMaggio

21. Social Networks, Geography, and Neighbourhood Effects Ron Johnston and Charles Pattie

22. A Multiple-Network Analysis of the World System of Nations, 1995-1999 Edward L Kick, Laura A McKinney, Steve McDonald, Andrew Jorgenson

23. A Brief Introduction to Analyzing Social Network Data Robert A Hanneman and Mark Riddle

24. Concepts and Measures for Basic Network Analysis Robert A Hanneman and Mark Riddle

25. Survey Methods for Network Data Peter V Marsden

26. Survey Sampling in Networks Ove Frank

27. Qualitative Approaches Betina Hollstein

28. Analyzing Affiliation Networks Stephen P Borgatti and Daniel S Halgin

29. Positions and Roles Anuška Ferligoj, Patrick Doreian and Vladimir Batagelj

30. Relation Algebras and Social Networks Philippa Pattison

31. Statistical Models For Ties and Actors Marijtje A J van Duijn and Mark Huisman

32. Exponential Random Graph Models for Social Networks Garry Robins

33. Network Dynamics Tom A B Snijders

34. Models and Methods to Identify Peer Effects Weihua An

35. Kinship Network Analysis Klaus Hamberger, Michael Houseman and Douglas R White

36. Large-Scale Network Analysis Vladimir Batagelj

37. Network Visualization Lothar Krempel

38. A Reader’s Guide to SNA Software Mark Huisman and Marijtje A J van Duijn

Link to the advertisement:

Scott_HB of Social Network Analysis dis Compatibility Mode (2)

Great data visualization

Steve Borgatti — Fri, 10 Dec 2010 03:32:48 +0000

Take a look at these three sites.

Ethnographic rap

Steve Borgatti — Thu, 16 Sep 2010 00:14:50 +0000

Here’s a song that you can learn from:

http://www.youtube.com/watch?v=kyWTJWrH1aI

Here are the lyrics:

http://www.lyricsfreak.com/n/notorious+big/ten+crack+commandments_20101868.html

Administrative wrappers

Steve Borgatti — Mon, 06 Sep 2010 16:10:54 +0000

A recent study of our College of Business & Economics found that “the administrative and financial operations of the College are unduly bureaucratic and paper intensive.”

The job of university administration is to give professors the illusion that all they have to do is content – solve the mysteries of the universe, develop students, etc. When a professor has to exchange a dozen emails to be reimbursed for one trip, or several dozen to reserve multiple summer classrooms for a workshop, or literally hundreds of emails to administer his grant, the illusion is broken and the professor is not productive. He becomes part of administration and in fact a very incompetent part.

Note that the passive this-comes-from-the-University response doesn’t help (nor the University’s this-comes-from-the-State response). Computer programmers have a concept called a wrapper: when a set of software tools has a very complicated interface, the programmer figures it out once and writes a wrapper with a very simple interface so that in his own code he is shielded from that complexity. Administrators at the College level should not be passing along (and, worse, adding to) the University’s complexity, but rather shielding faculty from the stuff that faculty are really not good at.

A simple example of the current mentality: I hired outside instructors to teach some of our workshops. They were asked to fill out several forms in order to get paid. I noticed that two of the forms were nearly identical. In fact, it was clear that one was a newer version of the other. When I asked about this, I was told “We know. But when we give both forms to Accounting it seems to work so we don’t mess with it.”

Harvard professor fakes data?

Steve Borgatti — Mon, 30 Aug 2010 14:14:44 +0000

Harvard professor Marc Hauser was found guilty of “scientific misconduct”. Harvard doesn’t say exactly what that entails, but it is clear is that he and his assistants had the task of coding videotaped monkey behavior. This is a task which should always be given to someone who does not know what the study’s hypotheses are. It should also be noted that the scientific method is designed to protect a researcher from his/her unconscious biases. It does nothing to prevent people from simply faking it.

http://arstechnica.com/science/news/2010/08/harvard-professor-found-guilty-of-scientific-misconduct.ars

BTW, the guy’s field of study is morality.

Ritualistic Science

Steve Borgatti — Sun, 18 Apr 2010 13:40:34 +0000

Much of “science” today is not so much scientific as scientistic. That means that the forms and behaviors of science are used, but the underlying logic is missing. Scientism is very much like the phenomenon of cargo cults, in which Pacific islanders built airports and control towers out of wood in order to bring back the planes that came with World War II. They didn’t realize that air traffic control panels actually did something, so they simply built facsimiles out of wood.

We can see scientism in every aspect of published research today. We see it in papers that think that theory is a set of interconnected hypotheses, rather than as the reason why X leads to Y. We see it in work in which each hypothesis is justified by a collection of independent, even mutually contradictory, reasons why the hypothesis must be true. Instead of the hypotheses being tests of an underlying theory, the hypotheses are the principal claims of the theory, and the justifications for these hypothesis can be a smorgasbord of ideas that embody entirely different theories.

Another aspect of scientism is the belief that there are universal best practices in research that are independent of the research question and the research setting. As a result, there is a great deal of argumentation from authority. E.g., we use this measure because so and so did. We even see it used to justify hypothesis: we expect X to lead to Y because so-and-so said it would.

A tiny example of this ritualistic belief in best practices can be found in network analysis where people argue that certain variables *must* be collected. For example, it has been claimed that in organizational network research, every study must measure the “workflow” network, which indicates who is required to interact with whom because of the nature of their jobs. The idea is that this set of ties can determine many other ties, such as interaction, and so must be taken account of, regardless of the study objectives.

To borrow a page from the scientistic consider the following counter-argument from authority. Ron Burt is a major luminary in the field who has published more than 100 papers on dozens of research projects. How many of these included the workflow network? None. Mark Granovetter is also a major luminary. Does his work take account of the workflow network? No. What about Jim Coleman? Brian Uzzi? Jim Moody? Woody Powell? Anybody?

Graduate students need to be inoculated to generate antibodies against any prescription of this kind. The moment someone says “you can’t do meaningful research unless you include < insert favorite variable >” your scientific spider sense should start tingling and you should become wary. Such rules serve to replace and obviate the need for thought.

Apple’s negative ties

Steve Borgatti — Sat, 17 Apr 2010 16:16:33 +0000

Dan Halgin pointed this out to me. I wish they had included Adobe and Amazon as well:

Duke Political Networks Conference May 19-21, 2010

Steve Borgatti — Sat, 17 Apr 2010 14:20:52 +0000

Dear Networkers,

The program for the upcoming Duke Political Networks Conference has been posted online at

http://www-personal.umich.edu/~mheaney/Duke_Conference_Program.pdf

Please check your role at the conference and bring any concerns to my attention as soon as possible.

Our guaranteed rate of $129 per night at the Washington Duke Inn and Golf Club is ending on Sunday, April 18. For more information about hotel and travel, go to

http://www.poli.duke.edu/politicalnetworks/travel.html

We are still working out the details for the online paper archive where the papers will be posted. We are asking that papers be posted by Wednesday, May 12, one week before the start of the conference. I hope to send out the instructions for uploading the papers shortly.

If you have not yet registered, please do so ASAP at http://www.poli.duke.edu/politicalnetworks/index.php

Please do not hesitate to contact me if you have any questions about program or Mike Ward (mw160@duke.edu) if you have any questions about logistics or other hosting arrangements.

We very much look forward to seeing you soon at the 3rd Annual Political Networks Conference.

Sincerely,

Michael T. Heaney, Ph.D.

Assistant Professor of Organizational Studies & Political Science

University of Michigan

722 Dennison Building

500 Church Street

Ann Arbor, MI 48109-1042

Cell phone: 202-236-3369

E-mail: mheaney@umich.edu

http://www.umich.edu/~mheaney/