Network Analysis of NASBS Membership

Background: The membership of the NASBS has grown considerably in recent years with diversity in subspecialty training, gender, and geography. The academic relationships and contributions of its membership have not been studied.

Objectives: The primary objectives of the study are: (1) Measure academic contributions of NASBS membership; (2) Identify influential nodes of academic collaboration; (3) Identify opportunities for future collaboration and mentorship.

Methods: Peer-reviewed publications of members of NASBS (2019 NASBS Web site) were identified using Scopus API author name search. Author and abstract records were collected in an SQL database for offline processing. Duplicate author profiles and alternate spellings were merged via automated similarity matching followed by manual review using institution affiliation and publication history. A commercial gender labeling database with scoring was applied to first names with manual intervention for lower confidence (<0.95) labels. Network structures were constructed and analyzed using the graph-tool python library to produce a weighted co-authorship network and compute centrality measures. Network maps were then produced using Gephi’s network visualization software.

Results: The coauthor network contained 952 members with found publications and 4,996 connections. A total of 846 (88.9%) members were contained in a single connected giant component, whereas, 102 members were unconnected and 64 members had a single connection. The maximum shortest-path distance between nodes was 8 with an average path length of 3.3. The mean number of connections without respect to weighting was 10.5 (SD = 12.2), median 7. The mean weighted number of connections was 9.7 (SD = 21.1), median 3.3. Member unweighted connection count range from 0 to 102 and weighted connections ranged from 0 to 260. A total of 333 members (35.0%) had unweighted connection above mean and 221 members (23.2%) had weighted connection count above mean. There were 578 (60.7%) members that were part of the shortest path (nonzero betweenness centrality). Girvan–Newman clustering identified 267 communities, where 13 contained at least 1% of the total membership each. The 3 largest communities contained 23.3, 8.4, and 6.9% of members. Weighted degree was correlated with publication count (r ² = 0.44) and weighted betweenness centrality was correlated with publication count (r ² = 0.40). There were 111 published members identified as women. 5.4% of women were unconnected versus 11.4% of men. Mean unweighted connection count for women was 8.42 (SD = 7.6), median 6 versus 10.7 (SD = 12.6), median 7 for men weighted connection count for women was 4.9 (6.1), median 2.5 versus 10.3 (22.3), median 3.4 for men. Mean betweenness centrality for women was 1.9 × 10–3 (3.7 × 10–3) versus 4.2 × 10–3 (0.0156) for men. Mean closeness centrality for women was 1.91 (0.53) versus 1.96 (0.59) for men. Average publication count for women was 32.3 (38.5) versus 70.5 (106) for men. Average citation count for women was 543 (1,012) versus 1,389 (2,893) for men.

Conclusion: Network mapping of membership of the NASBS helps to visualize the academic activities and relationships of the NASBS and reveals areas of concentration and influence within the specialty. Network analysis can help better understand demographic disparities and trends within the membership. These findings can be used to mentor and foster increased collaboration among the membership.

 

MissingLinkText

No conflict of interest has been declared by the author(s).