Article type
Year
Abstract
Background: Knowledge translation is a key element of evidence-informed healthcare decision making. Social media like Facebook and Twitter provide a rapid and easily accessible platform for wider dissemination of systematic reviews and are an integral mode of communication for a global organization like Cochrane with diverse stakeholders.
Objectives: To evaluate the formation of network structures and to analyze the clusters and key influential person within clusters.
Methods: We downloaded data from the Cochrane Facebook Group (https://www.facebook.com/groups/63721740498/) by using Netvizz application (https://apps.facebook.com/netvizz/) on 23 March 2015. We imported the data file to visualize it via a software called Gephi. We applied the Mathieu Jacomy’s Force Atlas 2 algorithm layout to evaluate network patterns and to identify social network clusters.
Results: Overall there were 8577 users in the Cochrane Group. However we could download data for only 4762 users(nodes) since there is a restriction in the app for the amount of data that can be downloaded. The gender distribution was 51.3 % males and 47.5% females and the remaining 1.13% were unspecified . Members of these groups speak 55 different languages.
One user in this network has been identified as a highly influential user with more than 75 connections within the group; 2504 (52.2%) users who are members of the group, do not have any network relations with each other (Fig1). We have identified nine clusters based on a modularity algorithm using the connections (edges) and in each cluster there are one or two influential users who have a relationship with almost everyone within that cluster (Fig2). We also identified 7 users who act as conduits between different clusters but most of them are not the key-influencers (Fig3).
Conclusions: The vast majority of users within the group are not connected to each other and are probably passive consumers of information or represent members who are not actively involved in Cochrane activity. There is a need to reach out to them as well as encourage more interaction within clusters to promote diversity of opinions and increase engagement and outreach.
Objectives: To evaluate the formation of network structures and to analyze the clusters and key influential person within clusters.
Methods: We downloaded data from the Cochrane Facebook Group (https://www.facebook.com/groups/63721740498/) by using Netvizz application (https://apps.facebook.com/netvizz/) on 23 March 2015. We imported the data file to visualize it via a software called Gephi. We applied the Mathieu Jacomy’s Force Atlas 2 algorithm layout to evaluate network patterns and to identify social network clusters.
Results: Overall there were 8577 users in the Cochrane Group. However we could download data for only 4762 users(nodes) since there is a restriction in the app for the amount of data that can be downloaded. The gender distribution was 51.3 % males and 47.5% females and the remaining 1.13% were unspecified . Members of these groups speak 55 different languages.
One user in this network has been identified as a highly influential user with more than 75 connections within the group; 2504 (52.2%) users who are members of the group, do not have any network relations with each other (Fig1). We have identified nine clusters based on a modularity algorithm using the connections (edges) and in each cluster there are one or two influential users who have a relationship with almost everyone within that cluster (Fig2). We also identified 7 users who act as conduits between different clusters but most of them are not the key-influencers (Fig3).
Conclusions: The vast majority of users within the group are not connected to each other and are probably passive consumers of information or represent members who are not actively involved in Cochrane activity. There is a need to reach out to them as well as encourage more interaction within clusters to promote diversity of opinions and increase engagement and outreach.