Mapping Relational Closeness: A New Way to Understand Relationships in Learning Communities

• In Event: The Nuance of Patterns: Methods for Uncovering Intricate Quantitative Relationships

Abstract

This paper introduces the conditional joint network model (CJNM), a statistical framework for jointly modeling the existence and quality of social interactions in educational settings. By embedding individuals in a latent Euclidean space, CJNM captures relational closeness based on both binary connections and rated interaction quality. Applied to data from the 2023 mHealth Training Institute, the inclusion of the rating data revealed nuanced patterns that could have been overlooked (e.g., individuals who are socially peripheral despite frequent interaction). CJNM offers a more accurate and interpretable approach to understanding peer dynamics in learning environments, with implications for collaborative learning, group design, and educational interventions.

Authors

• Seewoo Li, University of California - Los Angeles

• Jinwen Luo, University of California - Los Angeles

• Minjeong Jeon, University of California - Los Angeles