Structure in personal networks: Constructing and comparing typologies

Area of Science:

• Social Network Analysis
• Sociology
• Computational Social Science

Background:

• Individual and social outcomes are influenced by personal network structure.
• Network structure reflects broader social variations, impacting social capital, support, and isolation.

Purpose of the Study:

• To propose a method for identifying typologies of personal network structures.
• To apply and compare this method with existing approaches using diverse datasets.

Main Methods:

• Utilized the Girvan-Newman algorithm to detect cohesive subgroups.
• Employed cluster analysis on three derived network measures to identify structural typologies.
• Compared findings with another recently introduced method for extracting structural typologies.

Main Results:

• Personal network structure can be effectively summarized using three measures from the Girvan-Newman algorithm.
• Cluster analysis successfully extracted structural typologies.
• Both methods captured significant variation, but showed disagreement and cross-classification.

Conclusions:

• The proposed method effectively categorizes personal network structures.
• The identified typologies have potential applications in understanding personal communities, social support, and social capital.
• Differences between typology detection methods warrant further investigation.