The University of Southampton
University of Southampton Institutional Repository

Developing a methodology for social network sampling (abstract)

Developing a methodology for social network sampling (abstract)
Developing a methodology for social network sampling (abstract)
Researchers are increasingly turning to network theory to describe and understand the social nature of animal populations. To make use of the statistical tools of network theory, ecologists need to gather relational data, typically by sampling the social relations of a population of animals over a given time-period. Due to effort constraints and the practical difficulty involved in tracking animals, these sampled relational data are almost always a subset of the actual network. Measurements of the sample – such as average path length, clustering, and assortativity – are assumed to be informative as to the structure of the real-world social network. However, this assumption is problematic. Due to artefacts of the sampling process, the various network measures taken on the sample may be biased estimators of the true values. For example, just as we would get a biased estimate of mean human height by selecting for a sample those people who stood out in a crowd, we will get a biased estimate of a measure like mean connectivity if we sample individuals who are socially prominent. This problem can only be solved by developing a qualitative theory of network sampling, answering questions such as what proportion of the whole network needs to be sampled before a given level of accuracy is achieved, and what sampling procedures are least biased? To develop such a theory, we need to be able to generate networks from which to sample. Ideally, we need to perform a systematic study of sampling protocols on different known network structures. But currently available data on animal social networks are unsuitable as these networks were themselves sampled. The simulation methods of artificial life provide the way forward. We have developed a computational tool for generating artificial social networks that have user-defined distributions for network properties (such as the number of nodes, and the density) and for key the measures of interest to ecologists (such as the average degree, average path length, clustering, betweenness, and assortativity). This tool allows us to perform the required systematic analyses of the biases inherent in different sampling regimes (e.g., snowball sampling) applied to different network structures. We will present details of this system, and show we are using it to develop robust sampling methods for social network data. We see the system as the first in a series of works that will allow us to develop a qualitative theory of social network sampling to aid ecologists, and eventually social scientists, in their social network data collection.
765
MIT Press, Cambridge, MA
Franks, D.
ebf40669-d4e2-4f58-b228-e57cd191b526
James, R.
b2bf9dc8-4cb2-41cf-b77c-6b042856f60f
Noble, J.
440f07ba-dbb8-4d66-b969-36cde4e3b764
Ruxton, G.
4b83d7ab-d477-4489-8750-90cd357488a1
Bullock, S.
Noble, J.
Watson, R.
Bedau, M. A.
Franks, D.
ebf40669-d4e2-4f58-b228-e57cd191b526
James, R.
b2bf9dc8-4cb2-41cf-b77c-6b042856f60f
Noble, J.
440f07ba-dbb8-4d66-b969-36cde4e3b764
Ruxton, G.
4b83d7ab-d477-4489-8750-90cd357488a1
Bullock, S.
Noble, J.
Watson, R.
Bedau, M. A.

Franks, D., James, R., Noble, J. and Ruxton, G. (2008) Developing a methodology for social network sampling (abstract). Bullock, S., Noble, J., Watson, R. and Bedau, M. A. (eds.) In Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems. MIT Press, Cambridge, MA. p. 765 .

Record type: Conference or Workshop Item (Paper)

Abstract

Researchers are increasingly turning to network theory to describe and understand the social nature of animal populations. To make use of the statistical tools of network theory, ecologists need to gather relational data, typically by sampling the social relations of a population of animals over a given time-period. Due to effort constraints and the practical difficulty involved in tracking animals, these sampled relational data are almost always a subset of the actual network. Measurements of the sample – such as average path length, clustering, and assortativity – are assumed to be informative as to the structure of the real-world social network. However, this assumption is problematic. Due to artefacts of the sampling process, the various network measures taken on the sample may be biased estimators of the true values. For example, just as we would get a biased estimate of mean human height by selecting for a sample those people who stood out in a crowd, we will get a biased estimate of a measure like mean connectivity if we sample individuals who are socially prominent. This problem can only be solved by developing a qualitative theory of network sampling, answering questions such as what proportion of the whole network needs to be sampled before a given level of accuracy is achieved, and what sampling procedures are least biased? To develop such a theory, we need to be able to generate networks from which to sample. Ideally, we need to perform a systematic study of sampling protocols on different known network structures. But currently available data on animal social networks are unsuitable as these networks were themselves sampled. The simulation methods of artificial life provide the way forward. We have developed a computational tool for generating artificial social networks that have user-defined distributions for network properties (such as the number of nodes, and the density) and for key the measures of interest to ecologists (such as the average degree, average path length, clustering, betweenness, and assortativity). This tool allows us to perform the required systematic analyses of the biases inherent in different sampling regimes (e.g., snowball sampling) applied to different network structures. We will present details of this system, and show we are using it to develop robust sampling methods for social network data. We see the system as the first in a series of works that will allow us to develop a qualitative theory of social network sampling to aid ecologists, and eventually social scientists, in their social network data collection.

Text
ALIFExi-abstracts-0025.pdf - Other
Download (13kB)

More information

Published date: 2008
Venue - Dates: Artificial Life XI: Proceedings of the Eleventh International Conference on the Simulation and Synthesis of Living Systems, 2008-01-01
Organisations: Agents, Interactions & Complexity

Identifiers

Local EPrints ID: 266736
URI: http://eprints.soton.ac.uk/id/eprint/266736
PURE UUID: 22485156-4870-41bd-9e03-20749d3772be

Catalogue record

Date deposited: 01 Oct 2008 16:39
Last modified: 26 Jun 2020 16:43

Export record

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×