to Syllabi List
545: Social Network Analysis
of Preventive Medicine
Keck School of Medicine
University of Southern California
Professor: Thomas W. Valente, PhD
1000 South Fremont Ave., Bldg. 8, Rm. 5133
phone: (626) 457-6678
fax: (626) 457-6699
Location: Alhambra campus Room 7059
Office Hours: Virtual office hours (24/7)
This course is
an introduction to the theory, methods and procedures of network analysis with
emphasis on applications to public health programs. The goal of the course is
to provide a working knowledge of the concepts and methods used to describe and
analyze networks so that professionals and researchers can understand the results
and implications of this body of research. The course also provides the training
necessary for scholars to conduct network analysis in their own research careers.
The course consists
of readings, class discussions, computer and data analysis assignments, and a
final paper. The data analysis assignments will be conducted using the UCINET
V network analysis software available to students in the class. Individual student
papers will use data that the student collects him/herself. The data collection
and entry process will be quite simple and consist of identifying a group (a class,
club, organization, etc.) that students can meet and then ask to complete a simple
one page questionnaire.
Students who complete
this course will be able to:
1. Read and comprehend
concepts presented in the social network literature
2. Use network
analysis as a research technique in their own research including knowledge of
what concepts are applicable and how to collect and analyze this type of data.
3. Explain how
network analysis contributes to theories or areas of study of interest to the
4. Develop a deeper
understanding of how interpersonal and mass communication contribute to the formation
of norms, social structure and decision-making, and hence an understanding of
the essential elements necessary to launch community development, communication
campaigns and/or health promotion projects.
1. Scott, J. (2000).
Social network analysis: A handbook (2nd Ed.). Newbury Park, CA: Sage.
2. Valente, T.
W. (1995). Network models of the diffusion of innovations. Cresskill, NJ: Hampton
student requesting academic accommodations based on a disability is required to
register with Disability Services and Programs (DSP) each semester. A letter of
verification for approved accommodations can be obtained from DSP. Please be certain
the letter is delivered to me as early in the semester as possible. DSP is located
in on the University Park campus in STU 301 and is open 8:30 a.m. – 5:00
p.m., Monday through Friday. The phone number is (213) 740-0776.
TOTALe (also known as BlackBoard) is the online learning portal through which
many USC professors provide electronic copies of their course materials, including
syllabuses, readings, and handouts. Students may obtain access TOTALe at learn.usc.edu
and use their USC computer user name and password to access the "MyUSC"
portal page. All courses that students are enrolled in that are using TOTALe will
appear on the page as a link. Simply follow the link to access online course materials
Proportion of Grade & due date
Week 3: Assignment
1: Matrix Calculation 5 % 2/xx
Week 4: Assignment 2: Krackplot/Pajek Graph 5 % 3/xx
Week 5: Assignment 3: UCINET Centrality 5 % 3/xx
Week 6: Assignment 4: UCINET Clique Detection 5 % 3/xx
Week 7: Assignment 5: UCINET Positional Equivalence 10 % 3/xx
Week 8: Assignment 6: Diffusion Network Modeling 1 10 % 4/xx
Week 9: Assignment 7: Diffusion Network Modeling 2 10 % 4/xx
Week 10: Assignment 8: Data Analysis of Class Project 1 10 % 4/xx
Week 11: Assignment 9: Data Analysis of Class Project 2 10 % 4/xx
Week 15: Final Paper 30 % 5/xx
Introduction & History: The first week is devoted to an overview
and history of the development of network analysis as a field of study. Students
will gain an understanding of the history of network research with discussions
of (1) which academic disciplines fostered its early growth, how and why; (2)
who are some of the major contributors to network analysis; and (3) where is network
analysis today and in the future. The first week is also devoted to introducing
the basic language of networks and providing an overview of the course.
1 & 2, pages 1-38
Valente, T. W.
(2000) Network analysis for public health programs. Unpublished paper. Department
of Preventive Medicine, School of Medicine, University of Southern California,
Los Angeles, CA
Knoke & Kuklinski:
Wellman, B. (1988).
Chapters 1 & 2. In B. Wellman & S. D. Berkowitz (1988) Social structures:
A network approach. Cambridge: Cambridge University Press.
Models: What is a network? What is network analysis? The second week
consists of an explanation of how a network is described. The lectures discuss
how to create a sociogram, how matrices are used to represent networks and how
network indices are computed from matrices.
3 & 4 , pages 39--84
Matrix Algebra. Newbury Park, CA: Sage. Chapter 1.
Marsden, P. V.
(1990). Network data and measurement. Annual Review of Sociology, 16, 435-463.
Burt, R. S. (1980).
Models of network structure. Annual Review of Sociology, 6, 79-141.
Diffusion: Network analysis has been a core methodology used to understand
the diffusion of innovations including the diffusion of health behaviors such
as smoking, family planning, substance abuse, and so on. These lectures provide
the student a basic understanding of how networks structure the diffusion of innovations
and how network analysis has contributed to the understanding of diffusion.
(1973). The strength of weak ties. American Journal of Sociology, 78:1360-1380
Klovdahl, A. S.
(1985). Social networks and the spread of infectious diseases: The AIDS example.
Social Science Medicine, 21(11), 1203-1216.
Valente, T. W.
(1996). The diffusion network game. Connections, 19(2), 30-37.
Valente, T. W (in
press). Models and methods for innovation diffusion. In P. Carrington, J. Scott
& S. Wasserman (Eds.) Models and Methods in Social Network Analysis. New York:
Cambridge University Press.
Centrality: Centrality is one of the most useful concepts in network
analysis. Week 4 is devoted to discussing various centrality measures and the
differences in their computation and application.
5, pages 85-99
Freeman, L. (1979).
Centrality in social networks: Conceptual clarification. Social Networks. 1, 215-239.
& Valente, T. W. (in press). The stability of centrality when networks are
sampled. Social Networks.
Valente, T. W.,
& Foreman, R. K., (1998) Integration and radiality: Measuring the extent of
an individual=s connectedness and reachability in a network. Social Networks,
& Faust, K. (1994). Chapter 6: Centrality and Prestige. In Social Network
Analysis: Methods and Applications. Cambridge, UK: Cambridge University Press.
Relational Models: Relational network models consist of the analysis
of direct ties between individuals. Relational models are concerned with who knows
whom and how the set of direct ties for an individual influences his/her behavior.
Relational variables consist of constructs such as number of nominations sent,
number of nominations received and personal network density.
6, pages 103-125
2 & 3
Valente, T. W.,
Watkins, S., Jato, M. N., Van der Straten, A., & Tsitsol, L. M. (1997). Social
network associations with contraceptive use among Cameroonian women in voluntary
associations. Social Science and Medicine, 45, 677-687.
Structural Models: Structural network models focus on how social networks
constitute sets of distinct positions or roles in the social space. Positional
(or role) analysis represents the dominant theme and (some consider) the major
insight of the network paradigm. The organizing principle is that an individual's
behavior is determined by his/her position rather than who he/she is connected
7, pages 126-148
Burt, R. (1987).
Social contagion and innovation: Cohesion versus structural equivalence. American
Journal of Sociology, 92, 1287-1335.
Graphical displays: Viewing networks represents a major attribute of
network analysis, just being able to see the structure. Tremendous advances in
network graphing have occurred and are occurring given improvements in computer
technology. This week we will learn how to use Pajek and Krackplot, two of the
more common display programs.
Freeman, L. C.
(2000). Visualizing Social Networks Journal of Social Structure.
Available at: http://www.heinz.cmu.edu/project/INSNA/joss/index1.html
and others (2002). Krackplot, 3.1.
McGrath, K. and
others (2003). Connections.
Mrvar, A., &
Batagelj, V. (No Date). PAJEK software for network visualization.
Thresholds: Thresholds models were first postulated by Granovetter in
1978 as a possible means to understand why collective behavior occurred in some
situations but not others. Granovetter encouraged application of threshold theory
to the diffusion of innovations and threshold theory provides a parsimonious explanation
of how networks structure diffusion of innovations.
Valente: Chapters 5, 7, & 8
(1978). Threshold models of collective behavior. American Journal of Sociology,
The Small World: One of the most popular uses of social network analysis
has been the study of the “small world” phenomenon. People are always
surprised when the know someone in common or having a connection with someone
from seemingly disconnected means. This week we discuss research conducted to
understand global connectivity and models used to explain the small world phenomenon.
1967. “The small world problem.” Psychology Today, 22:561-67.
Pool, Ithiel de
Sola and Manfred Kochen. 1978. “Contacts and influence.” Social Networks
1: 5-51. (only read pages 5 to 29 and 49-51).
and Stanley Milgram. 1969. “An experimental study of the small world problem.”
Watts, D. (2002).
The Small World, pages 1-45.
D. and H. Russell Bernard. 1978/79. “The reverse small-world experiment.”
Social Networks, 1:159-192.
Social Capital: The connections, norms and trust that bind individuals
and community have been referred to as social capital. Social capital has become
a hotly debated topic in social sciences for its seeming ubiquity and importance
to all aspects of life. Social network analysis provides the tools to measure
Lin, N. (1999).
Building a network theory of social capital. Connections, 22(1), 28-51.
Burt, R. S. (in
press) The network structure of social capital. In R. I. Sutton & B. Staw
(Eds.) Research in organizational Behavior. Greenwich, CT: JAI Press.
Van Meter, Karl
M. 1999. "Social Capital Research Literature: Analysis of
Keyword Content Structure and the Comparative Contribution of Author Names."
Lomas, J. (1998).
Social capital and health: Implications for public health and epidemiology. Social
Science and Medicine, 47, 1181-1188.
Critical mass and Geographic models: Critical mass models are the population
level analog to thresholds. The critical mass has been defined as the point at
which diffusion becomes self-sustaining. Many behavior change programs attempt
to reach critical mass. Geographers have long made major contributions to diffusion
theory and recently, with the advent of GIS, have made contributions to epidemiology.
This week we will review some readings on spatial networks.
6 & 9
Klovdahl, A. S.,
Potterat, J. J., Woodhouse, D. E., Muth, J. B., Muth, S. Q., and Darrow W. W.
(1994). Social networks and infectious disease: The Colorado Springs study. Social
Science Medicine, 38(1), 79-88.
Ego-centric Networks: How do you measure ego-centric networks? What are
some common instruments used and common measures created from ego-centric data
such that one gets a sense of structure generalizable from sampled units?
Marsden, P. V.
(1987). Core discussion networks of Americans. American Sociological Review, 52,
Campbell, K. E.
and Lee, B. A. (1991). Name generators in surveys of personal networks. Social
Networks, 13, 203-221.
Brewer, D. D. (1991).
Forgetting as a Cause of Incomplete Reporting of Sexual and Drug Injection Partners.
Sexaully Transmitted Diseases, 166-176.
Burt, R. (1984).
Network items and the general social survey. Social Networks, 6, 293-339.
Ego-centric Networks (cont.):
Valente, T. W.,
& Vlahov, D. (2001). Selective risk taking among needle exchange participants
in Baltimore: Implications for supplemental interventions. American Journal of
Public Health, 91, 406-411.
& Saba, W. (1998). Mass media and interpersonal influence in a reproductive
health communication campaign in Bolivia. Communication Research, 25, 96-124.
Valente, T. W.,
& Saba, W. (2001). Campaign recognition and interpersonal communication as
factors in contraceptive use in Bolivia. Journal of Health Communication. 6(4),
Paper presentations (cont.)
for Group Identification/Selection
In this class you
are expected to collect some data from a small group in order to complete network
analysis computations for the final paper. The group can be any collection of
people as long as they seem like a group. Groups should have at least 15 members
and no more than 50. Groups of about 30 are ideal from a practical standpoint.
You will only need to interview the group once and the questionnaire will take
less than 3-4 minutes to complete.
Examples of groups
that have been used in the past:
1. A seminar class at the school
2. A dance or exercise class
3. A religious group
4. A school cooperative group
5. A dormitory hall
6. A school-based interest group Students are encouraged to consult with the professor
or TA for advice about their group identification/selection process. Students
are required to submit a draft of their questionnaire to the professor before
solicit the group’s approval to interview them.
Guidelines for the Final Project
1. The paper you
are expected to write should be no more than 20 pages including tables, figures,
references and appendices (should you have any such as a copy of your questionnaire).
The paper should be double-spaced throughout with font sizes no smaller than 10
2. The paper should
consist of a scaled-down version of a journal submission. In other words, the
paper should have the following sections: introduction, theory or literature review
(this can be brief and lightly referenced); methodology and data; results; discussion;
and conclusion; and references, tables, figures and appendices.
3. Your questionnaires
should be short - no more than five or six questions which basically gather information
on a couple of different networks such as friendship and advice and perhaps a
demographic question or two to measure gender, or departmental affiliation. You
can also collect the same network data with both nominations and roster techniques
so that you can compare networks.
paper will consist of descriptive analysis of the networks you have measured and
then an analysis of the centrality, group structure and positional structure of
the networks. Optionally you may compare different networks that you have measured
(advice versus friendship for example or roster versus nominations). You may then
analyze your networks in terms of demographic data that you have collected: For
example, does the network breakdown into groups based on gender or department
affiliation? Finally, you may write on the potential for diffusion or behavior
change to occur within the network or from outside the network.
For your questionnaires
a. Include a disclaimer clause such as the following:
is completely voluntary and your participation is optional. Your responses will
be kept strictly confidential and the responses will be converted to numeric form
and not individually analyzed by anyone in this group. You may have the results
presented back to youin one month in a manner which will not permit the identification
of individual respondents in any way.
b. Remember to
include a space for the respondents to write their own names on the questionnaire
so that you can identify them.