This study examines the survival and evolution of 443K bidirectional mention ties on Twitter by merging datasets collected before 2015 and in the first few months of the COVID-19 pandemic (February to June, 2020). We hypothesize that strong pre-existing ties, marked by frequent communication and shared identities, endure and tolerate cognitive and stance differences over time. Our findings show that surviving ties are stronger than average pre-2015 ties but exhibit greater cognitive distance in COVID-19 discussions, suggesting that strong ties can tolerate different and even opposing opinions on contentious topics. This challenges traditional models of social influence and homophily, which predict increased cognitive similarity within strong ties. The findings imply the potential for old ties to function as network bridges, reducing political divides by connecting dissimilar social groups.

Despite extensive research on the relationship between international research collaboration (IRC) and research performance in science and technology (S&T), existing research has mostly examined single or comparative case studies, relatively small samples composed of developed countries, and uni-directional relations between empirical indicators. Although large scale network studies of IRC are becoming more common, 1) drivers of IRC network formation and 2) effects of the IRC network on policy-relevant performance outputs tend to be analyzed separately. Large scale analysis of the reciprocal dynamic relationship between IRC and national performance has yet to be conducted.

This research tests network effects on performance and vice versa simultaneously using a longitudinal co-evolution model on three decades of global S&T network and performance data. We employ the stochastic actor oriented model (SAOM) framework, also known as Siena models, to analyze data on 166 countries from 1993 to 2022. Yearly IRC networks are constructed from Web of Science's XML database. Corresponding national S&T performance data is gathered from Elsevier's fractional field-weighted citation index (FWCI), which disentangles national from internationally attributed citation impact. The models also account for geographic distance, national wealth, population metrics, political governance, and endogenous network processes.

The preliminary results support the hypotheses with positive and significant estimates for both effects. However, geographic distance appears to play a critical role in the transmission of the social effect of performance on the IRC network. Indeed, not controlling for geographic distance renders this effect insignificant in the face of the endogenous network dynamic of preferential attachment. Further analysis will be conducted incorporating different sensitivity tests in addition to tests for disciplinary and temporal heterogeneity.

Results of applying an Incremental Fuzzy Grouping algorithm on several real-world social media data sets will be presented. Fuzzy Grouping is a method of extracting group structures in which individual actors can be in multiple groups with different weights, at each point in time. Results of the developed Incremental Fuzzy Grouping algorithm will be compared with conventional grouping algorithms such as Leiden Grouping, K-Means Grouping and Spectral Clustering, applied on the same social media data set using various temporal window sizes. The ability of the proposed method to efficiently extract fuzzy groups over time, even for large scale data sets will be demonstrated through testing on data sets of different scale. In addition, the approximate computational cost and how that computational cost scales with data set size will also be compared between the proposed algorithm and conventional algorithms.

Note, the proposed fuzzy grouping algorithm does NOT operate on time windows pulled from the data set. Instead, it operates incrementally on the time stamped social media data set, overcoming one of the major challenges in analyzing over-time social media data using the conventional grouping algorithms. In addition, the proposed fuzzy group over-time tracking algorithm can employ an incremental algorithm for determining the “best” number of groups at any point in time. An advantage of the proposed algorithm is that it does not require the human user to select the time aggregation window or the desired # of groups.

Temporal changes in relationships are common in social networks. Do network structures change too or remain stable? Studying temporal change and stability in networks is especially important when actors engage in cross-sector collaboration over time to address complex public problems. This study argues that actors maintain regularities in temporal network structures and explores social processes underlying such regularities. Data constitute cross-sector collaboration ties, gathered through interview, that were created by 125 rural communities in Nepal with organizations in 2007 and in 2014 to help meet community needs. Data for 2007 captured such ties regarding the planning of drinking water projects for funding from the Rural Water Supply and Sanitation Program aimed at improving access to potable drinking water. Sixty-six communities received one-year funding whereas fifty-nine communities did not. Data for 2014 constituted post-funding collaboration ties between all 125 communities and organizations.

Network descriptives of collaboration networks for 2007 and 2014 are compared first. Next, bipartite exponential random graph models are estimated to determine network structures and nodal attributes affecting forming collaboration ties. Network structures include a tendency for popular organizations to become more popular and for communities and organizations to be part of network closure. Community attributes considered are size, remoteness, and future collaboration. Organization attributes include operative level (village, district, and central) and organization type. The estimates involve collaboration networks of all communities and of funded and unfunded communities with organizations separately for 2007 and 2014. The preliminary results indicate the existence of changes in collaboration ties along with stability of popularity structure. This adds to our knowledge that actors combine both change and structural stability in cross-sector collaboration networks.

The dynamics of personal relationships remain largely unexplored due to the inherent difficulties of the longitudinal data collection process. In this work, we analyze a dataset tracking the temporal evolution of a network of personal relationships among 900 people over the course of four years. We search for evidence that the network is in equilibrium, meaning that all macroscopic properties remain constant, fluctuating around stable values, while the internal microscopic dynamics are active. We find that the probabilities governing the network dynamics are stationary over time and that the degree distributions, as well as edge and triangle abundances match the theoretical equilibrium distributions expected under these dynamics. Furthermore, we verify that the system satisfies the detailed balance condition, with only minor point deviations, confirming that it is indeed in equilibrium. Remarkably, this equilibrium persists despite a high turnover in network composition, suggesting that it is an inherent characteristic of human social interactions rather than a trait of the individuals themselves. We argue that this equilibrium may be a general feature of human social networks arising from the competition between different dynamical mechanisms and also from the cognitive and material resources management of individuals. From a practical perspective, the fact that networks are in equilibrium could simplify data collection processes, validate the use of cross-sectional data-based methods like Exponential Random Graph Models, and inform the design of interventions. Our findings advance the understanding of collective human behavior predictability and our ability to describe it using simple mathematical models.

Relational event models (REMs) are statistical models for the analysis of sequences of relational observations. Estimating these models is straightforward (albeit potentially computationally demanding) when the exact sequence of events is known, such as in the case of phone call records. However, sequences of relational event data may be incomplete or difficult to gather, e.g., when online platforms only provide aggregated data on user behaviors or when networks derive from cognitive perceptions of relationships, like friendship perceptions. Those perceptions are typically collected through survey panels that do not gather instantaneous network changes.

We present an application of the Expectation-Maximization algorithm to estimate the parameters of a REM variant, the Dynamic Network Actor Model (DyNAM), when the sequences are partially observed. We consider the evolution of an incomplete sequence as the result of two conditionally independent DyNAM processes, one governing the event creation and the other the event deletion. The generalized Expectation-Maximization is considered for imputing missing data and generating a sequence of proposals that do not necessarily maximize the objective function. The Expectation step is computationally intractable and approximated by random samples of plausible event sequences aggregated following Multiple Importance Sampling schemes.

Additionally, we show that this implementation is not limited to the analysis of incomplete sequences of relational events but can also be applied to analyze the co-evolution of network panel data and relational event data. We illustrate the method by analyzing data on friendship networks collected in a panel design and online interactions.

Gender homophily is one of the most robust patterns in friendships, yet we lack clear insights into how structural factors shape this pattern. While extensive research has established that group composition influences patterns of homophilous associations, previous work has largely focused on ethnic contexts, where additional factors such as socioeconomic status and residential segregation complicate interpretation. Gender composition in schools provides a useful test case because, unlike ethnicity, it is usually unrelated to socioeconomic or residential barriers. We test two competing mechanisms for how gender composition shapes cross-gender friendship formation. The first suggests that increased mixing leads to more cross-gender friendships by providing greater opportunities for interaction. The second proposes a curvilinear relationship, where extremely imbalanced gender compositions may reduce cross-gender friendships due to heightened identity threat perceptions and stronger needs for group distinctiveness. Using random coefficient multilevel Siena models with sienaBayes, we analyze longitudinal friendship networks from two datasets: CILS4EU (794 classrooms across four countries) and Hungarian secondary schools (40 classrooms). This methodological approach allows us to examine how gender composition shapes the evolution of friendship ties while accounting for structural network dependencies. Our preliminary results reveal that the effect of gender composition may be gender-specific: while boys' preferences remain stable in response to the gender composition, girls' friendship choices vary. At moderate boy ratios, girls form more cross-gender ties, but in highly male-dominated classes, they cluster together and strengthen same-gender preferences, reinforcing gender-segregated structures.

The Stochastic Actor-Oriented Model (SAOM) is a model of network evolution, where estimated coefficients represent the timing of decisions and the propensity of actors to form or maintain relationships over time. One notorious challenge about these models is that the coefficients are on a latent rather than probability scale. As a consequence, they suffer from the same scalability issue as many generalized linear models, where the value of coefficients is affected by non-collapsibility due to unobserved heterogeneity. This makes them hard to interpret and to compare across models or datasets.

To address these interpretational challenges, we propose employing effect size measures based on empirical and counterfactual differences in predicted probabilities, thus calculating marginal effects, which avoid scaling issues. This is particularly challenging for network data however, as the interdependency of network statistics necessitates considering higher-order derivatives and discrete differences. We show that averaging over these marginal effects can differ in magnitude and sign compared to the original (latent scale) coefficients.

Since SAOM models unobserved decision sequences between network observations, researchers might either want to interpret effects at observation moments or take the simulated chains into account. We develop a flexible simulation-based method to allow combining either approach with different quantities of interest to be averaged over. An analysis of empirical data demonstrates differences between coefficient-based and marginal effect size interpretations for predicted tie and change probabilities.

These new tools will not only support mediation and meta-analysis for longitudinal network models but facilitate answering new research questions with SAOM.

Brokerage is a central concept in organizational network research, describing the influence actors exert by connecting or separating otherwise disconnected individuals. Traditionally, brokerage has been measured through structural positions in networks, often using Burt’s (1992) constraint measure. More recently, attention has shifted to brokerage orientations, reflecting semi-permanent behavioral tendencies such as tertius iungens (connecting others) and tertius gaudens (maintaining separation). Despite theoretical advancements, methodological approaches in network analysis have lagged. Most management studies still rely on non-network regression models that overlook tie interdependence, potentially leading to flawed interpretations. At the same time, while network models such as stochastic actor-oriented models (SAOMs) offer powerful tools for modeling network dynamics, they provide limited options for capturing complex brokerage behaviors beyond basic betweenness measures.

Our study addresses these gaps by conceptualizing brokerage as a dynamic process based on personal preferences and introducing novel SAOM effects that capture different brokerage orientations and the impact of brokerage positions on actor behavior. Using a unique dataset, we illustrate the application of these effects and offer practical guidance for interpretation. Our contributions include (1) integrating brokerage orientations into a widely-used network modeling framework, enabling researchers to study brokerage as behavioral trajectories without relying on trait-based data, as is necessary when modeling interorganizational relations; (2) providing a method to estimate the impact of brokerage positions while accounting for network interdependencies; and (3) fostering a coevolutionary perspective on brokerage, aligning network dynamics with brokerage outcomes and advancing research on organizational networks.

Social network researchers often assume that mutual nominations between actors are more likely to occur than one-sided nominations. However, this assumption does not always hold, particularly in large social networks. Actors who nominate many others often do not achieve complete reciprocation from all their nominees. In this presentation, we introduce a new parameter termed “overchoosing,” which integrates reciprocity and outdegree popularity effects as their product. This parameter captures an endogenous tie-formation process not adequately represented by existing network configurations. We hypothesize a negative estimate for the overchoosing parameter, indicating that the greater the number of outgoing nominations by an actor, the lower the likelihood that each nomination will be reciprocated. To test this hypothesis, we implemented the overchoosing parameter within stochastic actor-oriented models (SAOM). Analyses of secondary school friendship networks (N > 1000, two waves) demonstrated that the model including the overchoosing parameter provided a better fit to observed data compared to a model without it, as evaluated by the triad census goodness-of-fit indices—particularly for the triadic out-star without reciprocation (021D). We also present and discuss findings from additional analyses of online social networks and bitcoin trust networks (Ns > 500, number of waves > 100) by Bayesian SAOM optimized for parallel computation on supercomputers (Chan et al., 2022). These insights help refine our understanding of how reciprocal relationships form or dissolve in larger social networks.

The cryptocurrency market has grown exponentially to become a significant part of the global financial system, with a market capitalisation of $3.05 trillion as of 26 February 2025. Designed to maintain a stable value pegged to a reserve asset, stablecoins have gained momentum among traders and investors as a bridge between traditional fiat currencies and the decentralized world of cryptocurrencies.

Despite the perceived stability provided by built-in decentralised algorithms and their relative autonomy from fundamentals, stablecoin networks are not immune to critical events, with adaptive responses not yet fully understood. This study examines how the transaction networks of two Ethereum-based stablecoins adapt in response to the Terra-Luna crisis, one of the most tragic panic crises in the cryptocurrency world. By using Relational Event Models (REMs), we study the evolution of these stablecoin network structures and the mechanisms driving their post-shock adaptation. In particular, our analysis explores the role of triadic closure as a potential stabilising mechanism and investigates whether the formation of new triadic connections has shaped the overall structure during the crisis period and helped to restore network cohesion after disruptions.

Our REM analysis reveals a shift in network dynamics after the crisis. Under stable conditions, tie formation is mainly driven by reciprocation and transitive closure, while after the shock these mechanisms weaken significantly, suggesting a disruption of established relational patterns. Cyclical closure becomes more prominent, suggesting a shift towards alternative reconfiguration strategies to maintain network cohesion. Using REMs, this study provides one of the first researches on how stablecoin networks adapt in response to critical events and demonstrates the importance of these models for studying network dynamics and formation mechanisms, with potential implications for research on the microstructure of financial markets.

Empirical contact networks remain underexploited in revealing fundamental mechanisms driving social behaviours. We propose an original modeling framework for generating temporal networks, designed to reproduce key observables from empirical data [1]. Our hypothesis is that some of these observed features are intrinsically linked to the spatial constraints of face-to-face interactions. Unlike conventional network-based approaches, our models incorporate the critical role of spatial constraints in shaping interaction patterns. This approach builds on the core idea of Starnini et al. [2], but shifts to a more fundamental framework by considering social homogeneity in agents behaviour.

Starting from a pedestrian model with continuous space and discrete time, a ”contact” occurs when two agents face each other within a certain radius. A temporal network is constructed with nodes representing agents and links defined by these contacts. Our simulations explore various dynamics, interaction mechanisms, boundary conditions, and spatial configurations to assess the geometry's impact. The resulted temporal networks are then compared against empirical time-varying networks collected during four conferences [3].

One key result concerns the inter-contact duration distribution, which is power-law distributed with an exponent −3/2 in the empirical data. We reproduce this result using three different pedestrian models: two-dimensional random walk; active Brownian particles; and the Vicsek model [1]. This suggests that this property can be recovered by any pedestrian dynamics as soon as it has a random underlying mechanism.

We believe that this novel approach to network generation offers a framework to link observations on temporal networks to sociological interpretations.