Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Not All Bonds Are Created Equal: Dyadic Latent Class Models for Relational Event Data

Rumana Lakdawala, Roger Leenders, Joris Mulder

TL;DR

The paper introduces the Dyadic Latent Class Relational Event Model (DLC-REM) to address unobserved heterogeneity in dynamic relational data by assigning each dyad to latent classes with class-specific rate parameters and a concomitant model linking class membership to covariates. It demonstrates that DLC-REM generalizes stochastic block models by enabling dyadic-level heterogeneity with flexible latent structure, estimated via an EM algorithm and implemented in R using flexmix. Through simulations, it shows improved recovery of latent structure and predictive performance compared to SB-REM, while remaining parsimonious. The MID application reveals four distinct dyadic interaction patterns driven by factors such as major-power status, alliances, democracy differences, and contiguity, highlighting the model’s capacity to uncover nuanced, context-dependent conflict dynamics. Overall, DLC-REM provides a versatile tool for modeling heterogeneous relational event processes with practical implications for forecasting and policy analysis.

Abstract

Dynamic social networks can be conceptualized as sequences of dyadic interactions between individuals over time. The relational event model has been the workhorse to analyze such interaction sequences in empirical social network research. When addressing possible unobserved heterogeneity in the interaction mechanisms, standard approaches, such as the stochastic block model, aim to cluster the variation at the actor level. Though useful, the implied latent structure of the adjacency matrix is restrictive which may lead to biased interpretations and insights. To address this shortcoming, we introduce a more flexible dyadic latent class relational event model (DLC-REM) that captures the unobserved heterogeneity at the dyadic level. Through numerical simulations, we provide a proof of concept demonstrating that this approach is more general than latent actor-level approaches. To illustrate the applicability of the model, we apply it to a dataset of militarized interstate conflicts between countries.

Not All Bonds Are Created Equal: Dyadic Latent Class Models for Relational Event Data

TL;DR

The paper introduces the Dyadic Latent Class Relational Event Model (DLC-REM) to address unobserved heterogeneity in dynamic relational data by assigning each dyad to latent classes with class-specific rate parameters and a concomitant model linking class membership to covariates. It demonstrates that DLC-REM generalizes stochastic block models by enabling dyadic-level heterogeneity with flexible latent structure, estimated via an EM algorithm and implemented in R using flexmix. Through simulations, it shows improved recovery of latent structure and predictive performance compared to SB-REM, while remaining parsimonious. The MID application reveals four distinct dyadic interaction patterns driven by factors such as major-power status, alliances, democracy differences, and contiguity, highlighting the model’s capacity to uncover nuanced, context-dependent conflict dynamics. Overall, DLC-REM provides a versatile tool for modeling heterogeneous relational event processes with practical implications for forecasting and policy analysis.

Abstract

Dynamic social networks can be conceptualized as sequences of dyadic interactions between individuals over time. The relational event model has been the workhorse to analyze such interaction sequences in empirical social network research. When addressing possible unobserved heterogeneity in the interaction mechanisms, standard approaches, such as the stochastic block model, aim to cluster the variation at the actor level. Though useful, the implied latent structure of the adjacency matrix is restrictive which may lead to biased interpretations and insights. To address this shortcoming, we introduce a more flexible dyadic latent class relational event model (DLC-REM) that captures the unobserved heterogeneity at the dyadic level. Through numerical simulations, we provide a proof of concept demonstrating that this approach is more general than latent actor-level approaches. To illustrate the applicability of the model, we apply it to a dataset of militarized interstate conflicts between countries.
Paper Structure (18 sections, 10 equations, 11 figures, 4 tables)

This paper contains 18 sections, 10 equations, 11 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Dyadic Latent Class Relational Event Model
  3. Model Overview
  4. Concomitant Model
  5. Relation to Stochastic Block-Models
  6. Technical Implementation of the DLC-REM
  7. Model Fitting
  8. Model assessment
  9. Implementation in R
  10. Numerical Simulations as a Proof of Concept
  11. Analyzing interaction dynamics from militarized disputes data
  12. Description of the datasets
  13. Model specification
  14. DLC-REM Fitting I: Identifying the Optimal Number of Classes
  15. DLC-REM Fitting II: Model Interpretation
  16. ...and 3 more sections

Figures (11)

  • Figure 1: SB-REM class representation.
  • Figure 2: DLC-REM class representation
  • Figure 4: Data Generating Structure
  • Figure 5: Fitted dyadic LC, K=2
  • Figure 6: Fitted dyadic LC, K=3
  • ...and 6 more figures