Application of social network analysis to communicable disease research

1
Application of social network analysis to
communicable disease research

• Sheila McCarthy, MSc
• PhD student Healthcare and Epi UBC
• Jan. 21, 2004

2
Social Network Analysis

• SNA is a research perspective that focuses on
  relationships between and among social entities,
  and on the patterns and implications of these
  relationships
• (Wasserman and Faust, 1994)
• Both theory and methods

3
Theory

• Relational ties are primary
• Attributes of actors are secondary
• Actors and their actions are viewed as
  interdependent
• Relational ties are channels for transfer
• Different types of relations identify different
  networks even when imposed on identical people
• The structure of relations among individuals and
  the location of individuals in the network have
  important consequences both for the individuals
  and for the system as a whole.

4
Methods

• Distinct from the methods of traditional
  statistics and data analysis
• Data are the function of two interdependent units
• Must incorporate social context
• i.e. analyze ties and the structure of ties
• Software Pajek, UCInet, Multinet, and more
• Multiple levels of analysis individual, small
  groups, whole network

5
Social Network Analysis

• SNA is particularly attractive to epidemiologists
  interested in STIs, bloodborne infections, and TB
  because these diseases spread directly and
  indirectly through relational ties
• SNA offers new approaches to research, but also
  to intervention and prevention strategies

6
Example 1

• Illustrates a sociogram (visual representation of
  a network)
• Components, sub-groups, individuals
• Network structures related to disease dynamics
• Intervention, prevention

7
Example 1 from Rothenberg et al.Sociogram
HIV/AIDS Colorado Springs, 1982-1999 325 nodes
(actors) and their ties (relations, links)
8
Example 2

• Illustrates matrix algebra
• Illustrates the use of social events (in epi,
  epidemiologically-relevant places
• Intervention, prevention

9
Example 2 from Dynamic Social Network Modeling
and Analysis Workshop Summary and Papers (2003)
http//books.nap.edu/books/0309089522/html/index.h
tml
10
Example 2 from Dynamic Social Network Modeling
and Analysis Workshop Summary and Papers (2003)
http//books.nap.edu/books/0309089522/html/index.h
tml
11
Example 2 from Dynamic Social Network Modeling
and Analysis Workshop Summary and Papers (2003)
http//books.nap.edu/books/0309089522/html/index.h
tml
12
Example 2 from Dynamic Social Network Modeling
and Analysis Workshop Summary and Papers (2003)
http//books.nap.edu/books/0309089522/html/index.h
tml

• Goals
• Study patterns of interactions
• Concern with issues related to similarity/dissimil
  arity
• Results
• Two more or less distinct groups
• Positions core/periphery could also be
  determined

13
Conclusions

• SNA can be used for
• Research into disease dynamics
• Intervention and prevention
• Based on an assumption of importance of
  relationships among interacting units
• Offers new methods of analysis

14
References

• Wasserman and Faust. 1994. Social Network
  Analysis Methods and Applications.
• Breigner, Carley, and Pattison. 2003. Dynamic
  Social Network Modeling and Analysis.
  (http//books.nap.edu/books/0309089522/html/index.
  html)