Social Networks and Health: Models, Methods, and Applications

Social Networks and Health: Models, Methods, and Applications

Thomas W. Valente

Language: English

Pages: 292

ISBN: B0052XUFP0

Format: PDF / Kindle (mobi) / ePub


Relationships and the pattern of relationships have a large and varied influence on both individual and group action. The fundamental distinction of social network analysis research is that relationships are of paramount importance in explaining behavior. Because of this, social network analysis offers many exciting tools and techniques for research and practice in a wide variety of medical and public health situations including organizational improvements, understanding risk behaviors, coordinating coalitions, and the delivery of health care services.

This book provides an introduction to the major theories, methods, models, and findings of social network analysis research and application. In three sections, it presents a comprehensive overview of the topic; first in a survey of its historical and theoretical foundations, then in practical descriptions of the variety of methods currently in use, and finally in a discussion of its specific applications for behavior change in a public health context. Throughout, the text has been kept clear, concise, and comprehensible, with short mathematical formulas for some key indicators or concepts. Researchers and students alike will find it an invaluable resource for understanding and implementing social network analysis in their own practice.

Prepper's Natural Medicine: Life-Saving Herbs, Essential Oils and Natural Remedies for When There is No Doctor

Essentials of Blood Banking: A Handbook for Students of Blood Banking and Clinical Residents (2nd Edition)

La cura para la fatiga: Descubre cómo combatir la fatiga y recuperar tu vitalidad

The Fountain of Truth: Outsmart Hype, False Hope, and Heredity to Recalibrate How You Age

Practical Radiological Anatomy

100 Perfect Hair Days: Step-by-Steps for Pretty Waves, Braids, Curls, Buns, and More!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

that models that included both selection and peer influence explained initiation into marijuana use more fully than did either factor alone. Other studies have also attempted to disentangle influence from selection (Engels et al., 1997; Fisher & Bauman, 1988). Fisher and Bauman (1988) looked at stable dyadic friendships and showed that the two persons became similar in their smoking behavior, suggesting influence. On the other hand, dynamic friendship dyads also showed similarity, indicating an

personal networks can be collapsed into compositional and variance measures. For example, the number or percentage of females and the IQV of the gender variable can be calculated. Each individual in the data can be characterized as the degree his or her personal network is female and the sample characterized as the extent to which the percent female varies. For relation type, the proportion or number of the personal networks who are family members and the diversity of that proportion can be

contacts—our networks—are an important part of our identity. A third reason for the growth in network analysis has been that computing and graphical display software has become available in the last decade. Analysis of social networks in the past was cumbersome and lacked the convenience of graphical displays to highlight network features and properties. The advent of sophisticated computing technologies created the platforms for major analytic advances to proceed. Today, there are dozens of

The network has been constructed from information only about who belonged to which department. A matrix (network) of department-by-department affiliations can also be constructed from this same data. The cells of the department-by-department matrix indicate the number of students who share departments. The department-by-department matrix is created by multiplying the transposed matrix by the original one (reversing the order of the multiplication) so the first matrix is 4 × 27 and the second is

each network is created. This sample is then used to calculate network parameters of interest. In this way, a large distribution of network parameters can be obtained based on randomly generated networks in an efficient manner. One issue is how to generate the random networks. In the ERGM framework, the simulated networks are created by randomly generating networks based on the structural parameters of interest (density, reciprocity, transitivity). These simulated networks are referred to as

Download sample

Download