INSPECT-LB is a Lebanese research group that consists of 31 members contributing to a total of 674 publications.
1. Group Structure
INSPECT-LB is divided into 7 axes involved in separate fields of health research.
The following tree shows the size of these axes, the members involved, and the overall contribution of each member (node size). The chart below it shows the number of publications per axis.
2. Educational Attainment
The majority of members have a doctoral degree (n = 25), and the most common backgrounds are pharmacy (n = 19), nutrition (n = 3), medicine (n = 2), and nursing (n = 2).
3. Member Contribution
A contribution is defined as the authorship or co-authorship of an INSPECT-LB publication.
The average member contributed to 40.4 publications (SD = 84.9), while the median member contributed to 10 (IQR = 19).
4. Member Collaboration
A collaboration is defined as the involvement with another INSPECT-LB member in a publication. One publication typically entails multiple collaborations.
The graph below shows the number of collaborations (link thickness) between members (nodes), where more connected members are closer to each other. (Use your mouse to zoom in/out and change node positions)
The graph is connected (there are no subgroups who work separately), with a network density of 31.8% (the links that exist between members constitute 31.8% of the total possible links that could exist in the network if all members were connected to each other).
The average distance between members is 1.8 (the average member is 1.8 links away from another member), while the maximum possible distance between any pair of members is 4.
Although a core of 5 highly-connected members is apparent in the graph, the network is disassortative with degree assortativity1 of -0.37. Therefore, we can conclude that highly active members in the group tend to work, on average, with less active members more than each other. This provides evidence that INSPECT-LB is an inclusive environment where new entrants can easily reach and collaborate with senior members.
5. Communities
According to the Louvain algorithm, INSPECT-LB can be divided into 3 communities of members who collaborate more frequently than average2.
The following graph represents these 3 communities:
The first (in pink) includes 14 members.
The second (in purple) includes 14 members.
The third (in green) includes 3 members.
This division is not related to any known member attributes. For instance, there is no sign that members with the same educational background tend to collaborate more, since assortativity by educational background3 is -0.02.
6. Member Importance
In a social network, member importance can be evaluated in different ways. In the graph below, we report 3 measures of importance:
Betweenness centrality: Members ranking higher on this scale play a more important role in connecting others to the network. Therefore, their departure from the group causes a higher risk of connectivity disruption.
Degree centrality: Members ranking higher on this scale have a higher number of direct connections in the group.
Eigenvector centrality: Members ranking higher on this scale can be considered influencers or authorities.
Footnotes
Assortativity ranges between -1 and 1 and can be interpreted just like a correlation coefficient. Positive values indicate an association between members that have the same number of connections, and negative values indicate an association between members that have different number of connections.↩︎
Technically, communities are subgroups of members that are highly connected with each other, but weakly connected with other subgroups↩︎
Assortativity ranges between -1 and 1 and can be interpreted just like a correlation coefficient. Positive values indicate an association between members with the same educational background, and negative values indicate an association between members of different background.↩︎