Sociology 186: Advanced Social Network Analysis
Instructor: Jonathan F. Haynes Time: T/Th 12:50-2:05pm
Email: jhaynes@stanford.edu Location: 200-030
Office Hours: T/Th 2:05-3pm (120-243) Quarter: Winter 2008-2009
This online version of the syllabus has links to required articles. Last updated 2/18/09. Download the pdf version here.
Course Description
This course addresses how social scientists think about networks. We will cover a little bit of information across many disciplines - Soc, CS, Stats, MS&E, ICME - and the goal is to connect ideas together. Specifically, this course provides practical experience and an interdisciplinary perspective on the collection, management, exploration, and analysis of social network data. The emphasis is on developing technical skills for studying large-scale social networks. Topics include theories of social order, small worlds, scientific computing, network sampling, and network dynamics.
Course Goals
The primary goal is to encourage critical thinking about the role of social networks in our society. This will involve connecting theory from Sociology with other disciplines. A second goal is to learn technical skills for actually studying large social network datasets. This will include a gentle introduction to the Python programming language and learning NetworkX. NX is an open-source Python package for network analysis developed by mathematicians at Los Alamos National Laboratories. A third goal will be learning some basic data management skills and covering major trends in the study of social networks.
The Good
This course covers the cutting edge of one of the most exciting frontiers of social science. This is also a really good course for people who like learning by doing. The course material and assignments focus on conceptual understandings and practical knowledge.
The Bad
This course will not attempt to provide an in-depth literature survey of the field or an abstract discussion of networks. There is simply too much material to cover in just one course. We will jump around many topics where a common theme – networks – is what holds everything together.
The Ugly
This is a new course and the first of its kind in the country. The challenge is that interdisciplinary fields by their very nature are less well defined than intra-disciplinary fields, so we will need to connect together what may initially seem to be disparate topics.
At the end of the quarter, you will be knowledgeable about:
1) How social scientists think about networks,
2) The tools and techniques for working with large network data sets, and
3) How to incorporate social network measures into social science research.
Course Prerequisites
There are no prerequisites. While a background in any related discipline would be helpful, a sincere interest in a conceptual understanding of the material is all you need.
Course Sections (optional)
Wed. 1-2pm (basement Bldg 200)
Thurs. 2:15-3:05pm (basement Bldg 200)
Course Texts
REQUIRED:
(1) "Theories of Social Order: A Reader (Paperback)", eds: Michael Hechter, Christine Horne
(2) "Six Degrees: The Science of a Connected Age (Paperback)", Duncan Watts
(3) "Regional Advantage: Culture and Competition in
(4) "Learning Python, 3rd Edition (Paperback)", Mark Lutz
All required articles are available on the web. See citations below.
RECOMMENDED:
(5) "Network Analysis: Methodological Foundations (Paperback)", eds: Ulrik Brandes, Thomas Erlebach
Note: Previews for all the texts can be viewed at http://books.google.com/
Course Requirements
This course provides hands on experience working with large, real-world social network data. We will work through examples in class/section and problem sets will give students an opportunity to apply these techniques.
Problem Sets (30%) |
Three problem sets |
Midterm (30%) |
One in-class midterm exam; M/C, T/F, short answer, essays |
Project (40%) |
Individual or group paper/project and short presentation. |
Software
You may use any computer platform you wish, as the software we will use (Python, NetworkX, Graphviz) are all open-source. Personally, I recommend the Ubuntu Linux distribution.
Course Schedule
Note: (R) = required reading, (L) = optional reading, will be covered in lecture
Week 1: Introduction
1/6 (Tues) Course Overview
(R) Homans, George. “Explanation in the Social Sciences”. (Reader pp 9-14)
(L) Brandes, Ulrik and Thomas Erlebach. “Introduction”. (Network Analysis, pp 1-6)
1/8 (Thurs) Why study social networks?
(R)
(L) Brandes, Ulrik and Thomas Erlebach. “Fundamentals”. (Network Analysis, pp 7-15)
Week 2: Getting Started
1/13 (Tues) Small Worlds
(R)
(R) Travers, Jeffrey and
(L) Kleinberg, Jon. “The small-world phenomenon: An algorithmic perspective.” Proc. 32ndACM Symposium on Theory of Computing, 2000.
1/15 (Thurs) Introduction to Python; Search in Networks
(R)
(R) Lutz, Mark. Chapters 1 (Introduction), 4 (Python Object Types). (Python, pp 3-21, 65-92)
(R) Hagberg, Aric, Dan Schult, and Pieter Swart. “Tutorial”. NX Documentation. (pp 5-15)
(L) Koschützki, D. et al. “Centrality Indices”. (Network Analysis, pp 16-61)
[Section 1: Start IMDB project: basic python, lists, dictionaries, import, basic network measures.]
Week 3: Microstructure
1/20 (Tues) Strength of Weak Ties
(R) Granovetter, Mark. “The Strength of Weak Ties”, American Journal of Sociology, 78:6, 1973: 1360-1380. (in Reader, pp 299-309)
(R) Burt, Ronald. “Structural Holes and Good Ideas”. American Journal of Sociology. 110:2, 2004: 349-399.
(L) Hong, Harrison, Jeffrey Kubik, and Jeremy Stein. “Thy Neighbor’s Portfolio: Word-of-Mouth Effects in the Holdings and Trades of Money Managers.” Journal of Finance. December 2005. pp 2801-2824.
[Problem Set #1 OUT]
1/22 (Thurs) Peer Influence
(R) Moody, James. “Peer Influence Groups: Identifying Dense Clusters in Large Networks”. Social Networks. (2001) 23:261-283.
(R) Lutz, Mark. Chapter 8 (Lists and Dictionaries). (Python, pp 152-171)
(L) Hill, Shawndra, Foster Provost, and Chris Volinsky. “Network-Based Marketing: Identifying Likely Adopters via Consumer Networks”. Statistical Science. (2006) 22:2, pp 256–276.
(L) Thompson, Clive. “If You Liked This, Sure to Love That”. The
[Section 2: Continue IMDB project, more network measures, methods in NX reference.]
Week 4: Macrostructure
1/27 (Tues) Communication networks: email, mobile
(R) Diesner, Jana, Terrill L. Frantz, and Kathleen M. Carley. “Communication Networks from the Enron Email Corpus ‘It’s Always about the People. Enron is no Different’”. Computational & Mathematical Organization Theory. Vol. 11:3, October 2005.
(R) Onnela, J.-P., J. Saramaki, et. al. “Structure and tie strengths in mobile communication networks.” Applied Physical Sciences. 10.1073, April 2007.
1/28 (Wed) Optional Special Section on Data-Preprocessing (6-7pm, 120-215) (how the raw IMDB data became the data set for PS#1)
1/29 (Thurs) Communication networks: instant messenger
(R) Leskovec, Jure and Eric Horvitz. “Worldwide Buzz: Planetary-Scale Views on an Instant-Messaging Network”, Microsoft Research, Technical Pub. MSR-TR-2006-186, June 2007.
(R) Lutz, Mark. Chapters 12 (if Tests), 13 (while and for Loops). (Python, pp 236-277)
[Course Project Proposal DUE IN CLASS]
[Section 3: Optional Wed. session]
[Problem Set #1 DUE FRIDAY
Week 5: Culture, Community, and Social Order
2/3 (Tues) Culture & Community - Silicon Valley
(R) Saxenian, Annalee. Chapters 1-2. (Regional Advantage, pp 1-58)
2/5 (Thurs) Culture & Community - Route 128 Boston
(R) Saxenian, Annalee. Chapters 3-4. (Regional Advantage, pp 59-104)
[Section 4: Start SiVNAP project]
Week 6: Review/Midterm Week
2/10 (Tues) Path Dependence, Unintended Consequences
(R) Saxenian, Annalee. Chapters 5-7. (Regional Advantage, pp 105-168)
[IN CLASS REVIEW SESSION]
2/12 (Thurs) MIDTERM EXAM
[No Section]
Week 7: Network Sampling
2/17 (Tues) Sampling
(R) Granovetter, Mark. “Network Sampling: Some First Steps”, American Journal of Sociology, 81:6. (May 1976), pp 1287-1303.
(R) Salganik, Matthew and Douglas Heckathorn. “Sampling and Estimation in Hidden Populations Using Respondent-Driven Sampling”. Sociological Methodology. 34:1. (Nov 2004), pp 193-240.
2/19 (Thurs) Measure Robustness
(R) Borgatti, Steve, Kathleen Carley, and David Krackhardt. “Robustness of Centrality Measures under Conditions of Imperfect Data”. Social Networks. (2006) 28:124–136.
[Problem Set #2 DUE THURSDAY
[Section 5: questions relating to PS2. Problem Set #3 OUT 2/19/09]
Week 8: Network Dynamics; Spontaneous Order
2/24 (Tues) Norms and Affiliations
(R) Horne, Christine. “Explaining the Emergence of Norms”. (Reader, pp 129-139)
(R) Simmel, Georg. “The Web of Group Affiliations”. (Reader, pp 291-298)
2/26 (Thurs) Spontaneous Order
(R) Carley, Kathleen. “A Theory of Group Stability.” American Sociological Review, Vol. 56, No. 3 (1991), pp 331-354.
(R) Hayek, Friedrich. “Cosmos and Taxis”. (Reader, pp 221-236)
[Section 6: Continue PS3 project]
Week 9: Applications
[Problem Set #3 DUE TUESDAY
3/3 (Tues) Finding central actors
(R) Borgatti, Steve. “Identifying sets of key players in a network”. Computational, Mathematical and Organizational Theory. 12:1. 2006. pp 21-34.
3/5 (Thurs) Special Topics
[No section]
Week 10: Conclusion
3/10 (Tues) Student presentations of projects
3/12 (Thurs) Student presentations of projects
[COURSE PROJECT PAPER DUE THURSDAY
End of Quarter
The Fine Print – Course Policies, Honor Code
Independent work: Problem sets and the take-home final exam for this course must be your own work, with assistance only from the course staff. Please use your common sense and remember that the primary goal is to learn the course material. At all times, you must abide by the Stanford Honor Code (http://honorcode.stanford.edu).
Participation: Attendance and participation are neither required nor graded; however, you should plan on attending regularly. It may be difficult to do well in the course if you miss a lot of class.
Auditing: In order to provide the best experience for enrolled students, attendance by auditors and outside visitors must be approved in advanced by the instructor. In general, the answer is “no”. If the course topic interests you, you should enroll.
Credit/No Credit: Students who opt for pass/fail must complete all course requirements to pass the course.
Late Policy: A problem set will be marked down to the next lower grade (e.g., A to A-, A- to B+, and so on) for each day (part or full) that it is late. However, each student is allotted five days for the entire quarter for any needed extensions, for any reason without penalty. Please use these days wisely, as additional extensions will only be granted in advance by the instructor and only if deemed absolutely necessary. Weekends and University holidays do not count as late days. Finally, the midterm, project, and final exam dates are fixed.
Revision 021809. J. Haynes © 2009.