network analysis
Definition
Network analysis refers to a set of methods for modeling relationships among variables as systems of interconnected nodes and edges, rather than treating variables in isolation. In psychological research, these methods can be applied to time-series data to uncover the structural organization of co-occurring phenomena. Dynamic Exploratory Graph Analysis, a network psychometric approach, was used to map the covariation of six facial expressions across 220 videos of global political leaders, revealing a two-dimensional structure in which positive and negative emotions form distinct patterns and anger exhibits relative autonomy compared to other emotions. Network analysis methods have also been proposed as a means of mapping social media users to neighboring communities outside their ideological echo chambers, enabling crowdsourcing interventions to reach the extreme partisans most likely to share partisan misinformation.
Sources: Tomašević & Major (2024), Pretus et al. (2024)
Related Terms
- misinformation (1 shared article)
- crowdsourcing (1 shared article)
- fact-checking (1 shared article)
- trust (1 shared article)
- emotions (1 shared article)
- exploratory graph analysis (1 shared article)
- populism (1 shared article)
- affective dynamics (1 shared article)
- computational social science (1 shared article)
Applications
Network Analysis and Emotion Dynamics
Emotion dynamics research examines how emotional states covary, fluctuate, and change over time, and network analysis provides a formal framework for representing these relationships. Dynamic Exploratory Graph Analysis applied to facial expression recognition data from political leaders' videos identified a two-dimensional network structure, with the first derivative model showing that anger was negatively correlated with most other emotions and more autonomous in highly populist speakers. As populist rhetoric increased, happiness became more contingent on co-occurring positive emotions, while anger grew less connected to the broader emotional network.
Sources: Tomašević & Major (2024)
Network Analysis and Misinformation Interventions
Crowdsourcing interventions designed to combat partisan misinformation depend on exposing users to fact-checks from sources outside their immediate ideological community, and network analysis methods offer a mechanism for identifying those neighboring communities. The approach addresses a core tension in polarized social media environments: ingroup sources inspire trust but rarely provide dissonant corrections, while distant outgroup sources generate dissonance but are often dismissed. By mapping community connections, network analysis can help calibrate the ideological distance between fact-checkers and recipients to a range that is dissonant enough to prompt belief updating without triggering wholesale rejection.
Sources: Pretus et al. (2024)
