What do centrality measures measure in psychological networks?

Area of Science:

• Psychology
• Network Science
• Data Analysis

Background:

• Centrality indices are common for analyzing psychological networks.
• These indices originated from social network analysis.
• Their suitability for psychological networks is questionable.

Purpose of the Study:

• Critically examine the application of popular centrality indices (degree, betweenness, closeness) in psychological networks.
• Assess the validity of underlying assumptions in the psychological context.
• Propose alternative approaches for measuring node importance.

Main Methods:

• Literature review of social network centrality indices.
• Critical examination of assumptions (e.g., flow, shortest paths, node distinctiveness, exchangeability).
• Application of these critiques to psychological network structures.

Main Results:

• Problems identified in social network literature also apply to psychological networks.
• Assumptions like 'flow' and 'shortest paths' may not align with psychological theories.
• Betweenness and closeness centrality appear particularly unsuitable for psychological networks.

Conclusions:

• Standard centrality indices may be inappropriate for psychological networks.
• Suggests developing tailored measures, reconsidering statistical importance metrics, or abandoning centrality concepts.
• Emphasizes the need to explicitly define centrality and its assumptions in psychological research.