Home Others Analysis of social networks

Analysis of social networks




Web pages




The analysis of social networks is the study of the structure and its influence on health and is based on theoretical constructs of sociology and mathematical foundations of graph theory. Structure refers to the regularities in the structuring of relationships between individuals, groups and / or organizations. When performing a social network analysis, the underlying assumption is that the network structure and the properties of that structure have a significant impact on the outcome of interest.

Because of the focus on network structure and not on individual characteristics and / or behaviors of network members, the data required for adequate analysis differ from those typically collected in non-relational epidemiological study designs. Typically, collect and analyze study designs that focus on individual traits / behavior and how those traits affect health. Attribute data is defined as data that reflects attitudes, opinions and behaviors of individuals or groups. Conversely, the analysis of social networks not only requires attribute data, but is based on the collection and analysis of relational data. Relational data refers to contacts, ties, and connections that relate one agent on a network to another. Relational data cannot be reduced to the properties of the individual agents themselves, but to a system / collection of agents.


The majority of studies on social networks use either whole (social-centered) networks or self-centered study designs. Whole network studies examine relationships between individuals or actors that are viewed as limited or closed for analytical purposes, when in reality the boundaries of the network are porous and / or ambiguous. When whole network studies are conducted, the focus of the study is on measuring the structural patterns of how individuals interact within the network and how these patterns explain specific health outcomes. The underlying assumption when performing an overall network analysis is that people making up a group or social network interact more than a randomly selected group of similar size.

In a sociocentric study, the members of the network are usually known or easy to identify, as the focus is usually on closed, a priori defined networks. For this reason, data collection for sociocentric network analysis includes enumerating all network members and managing saturation surveys for all network members. A saturation survey provides respondents with a list of all network members, and respondents are asked to identify members they are connected to. Actor-by-actor matrices can be created from this data and social network analyzes can be carried out.

If the network of interest does not have clearly defined boundaries, sociocentric studies lead to snowball- or respondent-driven samples to generate the network and collect data to identify structural patterns. In participant-oriented sampling, a small number of network members are interviewed and asked to name other network members, and these named members are also interviewed and asked to name other network members. This iterative process continues until all network members have been identified or for an a priori number of waves that were determined before the start of the study. When using subscriber-driven sampling, it is assumed that the sampling network is representative of all other segments of the network from which no data was collected. Subscriber-driven sampling uses name generator surveys to identify network members, followed by name interpreter questions to gather information about the named actors, their characteristics and relationships with the central actors.

Egocentric network designs, on the other hand, focus on a central actor, the ego, and the relationships between the ego and named actors or objects within their social networks. These types of designs collect data on the relationships between the ego and the objects, ages, with which they are associated. Self-centered study designs use either name generators or position generators to obtain both attribute and relational data that can be used to construct actor by actor from which egocentric data analysis can be made. Position generators are used to identify individuals who hold certain value roles, such as lawyers, while name generators, as discussed above, are questionnaires that ask the ego questions about people he or she is connected to in a certain way. Unlike in sociocentric studies, however, resource constraints exclude the subsequent survey of named ages, and therefore the ego serves not only as an informant for their own relationships with age, but also for age relationships with one another. Name generator questions, such as the sociocentric participant-oriented sample, are usually followed by name interpreter questionnaires.

Analysis of data from social networks

Network data is collected on an individual level, but analyzed on a structural level. The data is organized as an actor-to-actor matrix, as shown in Figure 1B. The data shown in Figure 1 shows the presence or absence of a tie. If the strength of a tie is also of interest, i. H. evaluated data, similarity or distance matrices could be used. Similarity matrices show stronger bonds with increasing numerical values, while increasing numerical values ​​in distance matrices reflect weaker bonds, since the bonds are weaker the greater the distance between two actors. Any actor-by-actor matrix can be converted into diagrams and analyzed using social network analysis software such as UCINET.
Graphics are visual representations of a network. Actors within a network are represented as nodes and the connecting lines of the nodes are representative of the connections between two actors. Charts can be directed to indicate that the relationship is being routed from one agent to another, or they can be rated, which indicates the strength of the bond. Although the visualization of the data is informative, the core of the analysis of social networks lies in the computation of descriptive measures that reveal important characteristics about 1) the position of network actors, 2) characteristics of network subsets, and 3) characteristics of complete networks.

The position of network actors or the networking of network actors is often referred to as a measure of cohesion. There are two common measures of cohesion

Distance = the length of the shortest path that connects two actors

The distance between points 15 and 11 is 5

Density = total number of relational ties divided by the total possible number of relational ties

Components and cliques measure properties of network subsets

A component is a part of the network in which all actors are connected either directly or indirectly.

(Howe et al.)

Nodes 1, 6, and & 7 form a clique

A clique is a subgroup of actors who are all directly connected to each other and no other member of the network is connected to all members of the subgroup. Clique analysis is the most common method to identify dense subgroups within a network.

  • Node number 19 has a degree centrality of 9, which is the highest in the sociograph. The general measure of centralization refers to how tightly a diagram is organized around its most central point. The network structure measures discussed above can then be used to parameterize predictive regression models that relate relational data to attribute data. For example, Lee et al. After the generation of measures of the network structure with methods of social network analysis, a multivariable regression to evaluate the connections between centrality measures and hospital characteristics.


Textbooks & Chapters

Scott J. Social Network Analysis: A Handbook. Newbury Park: Sage, 2000.
This book provides an introduction to social network analysis. It gives a brief overview of the theoretical foundations of social network analysis and discusses the key techniques required to carry out this type of analysis. Specifically, questions of study design, data collection and measures for the structure of social networks are discussed.

Carrington PJ, Scott J, Wasserman S. Models and Methods in Social Network Analysis Cambridge: Cambridge University Press, 2005.
This book offers a more detailed methodological approach to analyzing social networks. Chapter 2 provides a brief discussion of study designs, while Chapter 3 focuses on methods of data collection and model fitting.

Wasserman S, Faust K. Social Network Analysis: Methods and Applications. Cambridge: Cambridge University Press, 1994.

M. E. J. Newman. Networks. An introduction. First edition Oxford University Press, 2010
This book is an introductory text that deals with social networks and the analysis of social networks.

Methodical articles

Analysis of Social Networks: A Methodical Introduction
Journal: Asian Journal of Social Psychology
Release year: 2008

Collection methods for network data

Author: PV Marsden
Year of publication: 2011

The art and science of dynamic network visualization

Author: S Bender-deMoll, DA McFarland
Release year: 2006

simcity buildit cheats iphone

Dyad dynamics in social networks: associative, relational and proximity mechanisms

Author: MT Rivera, SB Soderstrom, B Uzzi
Year of publication: 2010

A glossary of terms for navigation in the field of social network analysis

Author (s): P. Hawe, C. Webster, A. Shiell
Release year: 2004

Public Health Network Analysis: History, Methods, and Applications

Author: DA Luke, JK Harris
Year of publication: 2007

Application item

A (very) brief introduction to R

Author (s): P Torfs, C Brauer
Year of publication: 2012

A comparative study of social network analytics tools

Autor (en): Combe et al
Year of publication: 2010

Software for analyzing social networks

Author(s): M Huisman, MAJ van Duijn
Year of publication: 2005

The prevalence of obesity on a large social network over 32 years

Author: NA Christakis, JH Fowler
Year of publication: 2007

Is Obesity Contagious? Social networks vs. environmental factors in the obesity epidemic

Author: E Cohen-Cole, JM Fletcher
Release year: 2008

Detection of implausible social network effects in acne, height and headache: longitudinal analysis

Author: E Cohen-Cole, JM Fletcher
Release year: 2008

Structural features of social networks and their connection with social support for older people: Who provides support?

Author: TE Seeman, LF Berkman
Year of publication: 1988

Social network analysis of patient exchanges between hospitals in Orange County, California

Autor(en): BY Lee, SM McGlone, Y Song, TR Avery, S Eubank, CC Chang, RR Bailey, DK Wagener, DS Burke, R Platt, SS Huang
Year of publication: 2011

Transmission network analysis for tuberculosis contact examinations

Autor(en): VJ Cook, SJ Sun, J Tapia, SQ Muth, DF Argüello, BL Lewis, RB Rothenberg, PD McElroy
Year of publication: 2007



Description: R contains several packages relevant for social network analysis: igraph is a generic network analysis package; sna carries out sociometric analyzes of networks; Network manipulates and displays network objects; PAFit can analyze the development of complex networks by estimating preferred binding and knot fitness; tnet carries out analyzes of weighted networks, two-mode networks and longitudinal networks; ergm is a set of tools for analyzing and simulating networks based on exponential random graph models exponential random graph models; Bergm provides tools for Bayesian analysis for exponential random graph models, Hergm implements hierarchical exponential random graph models; 'RSiena' enables the analysis of the development of social networks with the help of dynamic actor-oriented models; latentnet has functions for network latent position and cluster models; degreenet offers tools for the statistical modeling of network degree distributions; and networksis offers tools for simulating two-part networks with fixed edges.


Description: statnet is a suite of software packages that implement a variety of network modeling tools.

International Network for Social Network Analysis (INSNA) is a professional association for researchers interested in network analysis. The website contains descriptions of the SNA software, news, scientific articles, technical columns, abstracts, and book reviews. The site offers graduate programs, courses, discussion forums, I-Connect, bibliographies, and publications on SNA. INSNA also provides a Journal of Social Networks and hosts an annual international conference for social networks and other SNA events.


This website accompanies the chapter A Readers' Guide to SNA Software. In J. Scott and P. J. Carrington (Eds.) The SAGE Handbook of Social Network Analysis (pp. 578-600). The website offers several links and references related to social network analysis software.

Combe et al. (2010). A comparative study of social network analytics tools. France: Web Intelligence & Virtual Enterprises, Saint-Etienne

The aim of this article is to describe the functionalities of analyzing social networks. In addition, the article explains and compares some of the widely used software tools devoted to social network analysis. Software packages discussed in detail include Pajek, Gephi, NetworkX, and igraph.

Huisman, M. und van Duijn, M.A.J. (2005). Software for analyzing social networks . In P. J. Carrington, J. Scott und S. Wasserman (Hrsg.) Models and Methods in Social Network Analysis (S. 270-316). New York: Cambridge University Press.

This book is intended to complement the book Social Network Analysis: Methods and Applications by Wasserman and Faust. Huisman's book is a collection of articles dealing with specific network methods and methods for analyzing social network data. In addition to reviewing network measurements and analysis, this book provides information on software for analyzing social networks.

Web pages

International Network for Social Network Analysis (INSNA)
Site Map: International Network for Social Network Analysis (INSNA) is a professional association for researchers interested in network analysis. The website contains descriptions of the SNA software, news, scientific articles, technical columns, abstracts, and book reviews. The site offers graduate programs, courses, discussion forums, I-Connect, bibliographies, and publications on SNA. INSNA also provides a Journal of Social Networks and hosts an annual international conference for social networks and other SNA events.


Website Overview: Statnet is a suite of network analysis software packages that implements the latest advances in statistical modeling of networks. The analytical framework is based on Random Graph Models (ergm) of the exponential family. statnet offers a comprehensive framework for ergm-based network modeling, including tools for model estimation, model evaluation, model-based network simulation and network visualization. This broad functionality is driven by a central Markov Chain Monte Carlo Algorithm (MCMC).
statnet has a different purpose than the excellent UCINET or Pajek packages; the focus is on the statistical modeling of network data. Statnet's statistical modeling functions include ERGMs, latent space and latent cluster models. The packages are written in a combination of (the open source statistics language) R and (ANSI standard) C and are called via the R command line. And because it runs in the R package (www.r-project.org), you also have access to the full functionality of R, including the 'network' and 'sna' packages from Carter Butts. statnet has a command line interface, not a GUI, with syntax similar to R.


Analysis of social networks
Next offer: 6.-10. June 2016 8: 30-12: 30
Software used: R

Analysis of social networks

premed post bac programs

Host / Program: University of Michigan / Courses
Software used: Gephia, Netlogo, R

Summer workshop to analyze social networks

Next offer: 6.-10. June 2016
Software used: UCINET

Network modeling for epidemics

Next offer: August 9 to September 2, 2016 (Seattle, WA)
Software used: R, Statnet package

Interesting Articles

Editor'S Choice

The science of the flu shot
The science of the flu shot
Once in the air with the autumn cold, flu vaccinations are being given out in clinics and pharmacies across the country. Vaccination, while imperfect, is the most reliable way to avoid a potentially fatal infection. While many accept it as a seasonal inconvenience, the flu kills about 19,000 Americans in an average year. After the pioneering work of Hilary Koprowski,
Alumni publish 'Semper Fi' with Jai Courtney in the lead role
Alumni publish 'Semper Fi' with Jai Courtney in the lead role
Three Columbia alumni released Semper Fi earlier this month, a feature film starring Jai Courtney and distributed by Lionsgate. The film was produced by Alumna Karina Miller '04, co-written by Alumna Sean Mullin '06, and co-written and directed by Alumna Henry-Alex Rubin '95.
Columbia filmmakers make a splash at the Nashville Film Festival
Columbia filmmakers make a splash at the Nashville Film Festival
Current student Asad Farooqui and alumna Fany de la Chica '18 are represented at this year's Nashville Film Festival.
Review: 'The Emperor of All Diseases
Review: 'The Emperor of All Diseases'
It's difficult, if not impossible, to reach middle age without seeing the ravages of cancer up close.
Alexandra Carter
Alexandra Carter
As director of the Law School's Mediation Clinic, Alexandra Carter ’03 has been training students in various forms of alternative dispute resolution since 2008. Under her guidance, students learn negotiation strategies and advise clients in federal, state, and New York courts; Cases range from family business disputes to complaints filed with the U.S. Equal Employment Opportunity Commission. In 2016, Carter partnered with the United Nations Institute for Education and Research and their students are the exclusive providers of alternative dispute resolution classes for the United Nations Diplomatic Corps in New York. She is currently training judicial and administrative directors in New York state courts which will soon require most civil disputes to be resolved through alleged mediation rather than in public courts. In 2019, Columbia University honored Carter with the Presidential Award for Teaching for its innovative pedagogy and commitment to its students. Carter developed her passion for mediation and teaching as a student at the Law School's Mediation Clinic, led by Professor Carol Liebman, who became her mentor and role model. As a student, Carter won the Jane Marks Murphy Prize for Clinical Advocacy and the Lawrence S. Greenbaum Prize for Best Oral Argument in the 2002 Harlan Fiske Stone Moot Court Competition. Prior to enrolling in Law School, Carter was a private equity analyst with Goldman Sachs and Fulbright Fellow in Taiwan, where she researched contemporary literature to assess cross-strait political tensions. After Carter got her J.D. received, she worked on the U.S. District Court for the District of Massachusetts and then joined Cravath, Swaine & Moore as a litigator. She was retired to the academy by Liebman and other mentors from Columbia Law School. Carter's new take on negotiation is the subject of her upcoming general interest book Ask for More: 10 Questions to Negotiate Anything, which will be the main title published by Simon & Schuster in May 2020.
Ars Nova announces Melis Aker '18 as a resident of Play Group 2019
Ars Nova announces Melis Aker '18 as a resident of Play Group 2019
Alumna Melis Aker '18 joins Alumna Julia May Jonas '12 as a new member of the Ars Nova Plays Group 2019. The Play Group is a two-year residency where members become part of the Ars Nova resident artist community.
Clinic for Rehabilitation and Regenerative Medicine
Clinic for Rehabilitation and Regenerative Medicine
What are tendonitis and tendinitis? Tendons are strong strands of tissue that connect muscles to bones. Tendonitis is when a tendon is inflamed. It can hit any tendon in the body. When a tendon is inflamed, it can cause swelling, pain, and discomfort. Another problem called tenosynovitis is linked to tendonitis. This is the inflammation of the lining of the tendon sheath around a tendon. Usually the vagina is inflamed by itself, but both the vagina and tendon can be inflamed at the same time.