The stability of centrality measures when networks are sampled

By Elizabeth Costenbader, Thomas W. Valente. 2003

799 citations recorded by the source corpus.

Subfields: Sociology and Political Science, Statistical and Nonlinear Physics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Sociology and Political Science

In Statistical and Nonlinear Physics

Citations in the corpus

Is the Sample Good Enough? Comparing Data from Twitter's Streaming API with Twitter's Firehose (2013, background): “More recently, (Costenbader and Valente 2003) and (Borgatti, Car-ley, and Krackhardt 2006) have studied the affect of data error on common network centrality measures by randomly deleting and adding nodes and edges.”
A Critical Review of Centrality Measures in Social Networks (2010, background): “Furthermore, Costenbader and Valente (2003) also analyze the stability of different CM in presence of incorrect or incomplete information on the structure of a network (e.g., for the analysis of a sample of the network).”
Identifying Opinion Leaders to Promote Behavior Change (2007, methodology): “Simulations on empirical data have shown that a sample of 50% is 80% reliable at identifying opinion leaders (Costenbader & Valente, 2003).”
The Alignment of the Informal and Formal Organizational Supports for Reform: Implications for Improving Teaching in Schools (2010, background): “These response rates are well above the threshold of 80% considered generally acceptable for social network analysis to minimize the effects of missing data (Costenbader and Valente, 2003).”
Social Anxiety Disorder as a Densely Interconnected Network of Fear and Avoidance for Social Situations (2017, methodology): “We then evaluated the stability of the centrality metrics by implementing a subset bootstrap procedure (Costenbader and Valente 2003).”
Opinion Leadership and Social Contagion in New Product Diffusion (2011, background): “An important study by Costenbader and Valente (2003) of 59 different social networks of sizes similar to ours indicates that indegree in human networks is a robust metric as long as response rates are higher than 10%–20%.”
Assessing the bias in samples of large online networks (2014, background): “This finding relates to previous work on the effects f snowballing in network sampling (for instance, Granovetter, 976; Frank, 1977, 1978; Costenbader and Valente, 2003) but it s not directly comparable because of the way in which data is etrieved from online sources: instead of asking people to poi”
Mapping Change in Large Networks (2008, background): “distribution [16], but working with sets and assessing the accuracy of the clusters requires a different approach.”
Urban Polycentricity and the Costs of Commuting: Evidence from Italian Metropolitan Areas (2010, background): “a given network’s set of nodes ( Costembader and Valente, 2003 – p. 291; Zemljic and”
Prosocial signaling and cooperation among Martu hunters (2015, background): “This network has greater coverage than most social networks attempting comprehensiveness, and more than is necessary to ensure the robustness of centrality measures (Borgatti, Carley, & Krackhardt, 2006; Costenbader & Valente, 2003).”
How Correlated Are Network Centrality Measures? (2008, methodology): “Further information on these datasets are available in Costenbader & Valente, 2003;Valente, 1995.”
Identification of influencers - Measuring influence in customer networks (2008, background): “Costenbader and Valente [23] evaluated the stability of centrality measures when networks are sampled in the face of inaccurate or incomplete network data.”
Study on centrality measures in social networks: a survey (2018, methodology): “To get over the limitations of Costenbader and Valente (2003) and Borgatti (2006), a computational approach was taken and robustness was examined for large sample of graphs.”
Social network analysis of animal behaviour : a promising tool for the study of sociality (2008, background): “Missing data can significantly alter network measures (Kossinets 2006), although centrality measures calculated from subsets of the data may retain their relative relationships to each other despite sometimes large absolute differences (Costenbader & Valente 2003).”
What do centrality measures measure in psychological networks? (2019, background): “…measures that take into Running head: WHAT DO CENTRALITY MEASURES MEASURE? account shortest paths (e.g., betweenness centrality) or distances (e.g., closeness centrality) can change drastically when a node is removed from or added to the network (Costenbader & Valente, 2003, Epskamp et al., 2017).”
Dynamic Properties of Foreign Exchange Complex Network (2019, background): “Closeness centrality (ca ) measures the mean of the distance from node a to all others [70].”