Speakers
Description
Authors: Elena Gervilla Garcia (University of the Balearic Islands), Joan Pons Bauza (University of the Balearic Islands), Maria Wei Blanes Perez (University of the Balearic Islands), Victòria Romero Feliu (University of the Balearic Islands)
Chair: Elena Gervilla, Joan Pons
This symposium develops the applications of Social Network Analysis (SNA) in prevention research, integrating theoretical and methodological explanations, and a real application in the sports field. The four presentations address: (1) past and potential contributions of SNA to prevention research, (2) Some data measurement and depuration challenges in implementing SNA (3) How to implement SNA using R, and (4) a case study in sports teams to demonstrate practical utility. Collectively, the symposium provides a complete view about how SNA uncovers relational dynamics and its potential to be implemented in prevention research. Attendees will gain actionable insights into methodological implementation and practical advantages of SNA.
Abstract 1
What can Social Network Analysis add to Prevention Science?
Elena Gervilla Garcia (University of the Balearic Islands), Maria Wei Blanes Perez (University of the Balearic Islands), Victòria Romero Feliu (University of the Balearic Islands), Joan Pons Bauza (University of the Balearic Islands)
Introduction: Social Network Analysis (SNA) provides a powerful framework for mapping the relational structures that influence health behaviors. Despite its potential, SNA is still underutilized in prevention science. This presentation highlights the theoretical relevance of SNA for identifying structural leverage points-such as influential individuals and community clusters-to enhance the effectiveness and reach of prevention efforts.
Methods: We review and synthesize findings from case studies in public health and education, focusing on interventions that utilized SNA metrics, including centrality and community structure, to guide their strategies. Alongside established applications, we also discuss emerging and potential uses of SNA in the prevention field.
Results: Evidence demonstrates that both network position (e.g., influential nodes) and network topology (e.g., community structure) play critical roles in mediating the effectiveness of prevention strategies. For example, school-based interventions that identified and engaged highly central students as peer leaders achieved a 30% higher program adoption rate compared to random peer selection (Hunter et al., 2017). Smoking prevention messages disseminated through central nodes reached 65% of the student population within six weeks, compared to 38% in control schools. Anti-bullying campaigns targeting dense friendship groups resulted in a 24% reduction in reported incidents over one academic year (Paluck et al., 2016). Similarly, SNA-informed HIV prevention interventions engaging bridge individuals (high betweenness centrality) led to a 19% increase in condom use, versus 7% in standard outreach (Latkin et al., 2013).
Conclusion: Modeling social interdependencies through SNA enhances the precision and impact of prevention efforts, representing a shift from traditional individual-focused strategies. This presentation also explores new directions and future potential for SNA in advancing prevention science.
Abstract 2
Quality checks before applying SNA: Measurement and Depuration Strategies
Maria Wei Blanes Perez (University of the Balearic Islands), Victòria Romero Feliu (University of the Balearic Islands), Joan Pons Bauza (University of the Balearic Islands), Elena Gervilla Garcia (University of the Balearic Islands)
Introduction: Noise and missing data are examples of significant threats to the validity of social network analysis (SNA), particularly in prevention research. This presentation systematizes common data quality pitfalls and outlines essential preparatory steps for developing SNA in weighted networks.
Methods: Using sports team interaction data as an example, we demonstrate a comprehensive workflow for preparing adjacency matrices for SNA with the sna package in R.
Results: Key steps include data deduplication, matching and merging matrices, filtering data to address inconsistencies, modifying matrices to enable the calculation of specific SNA indices, and calculating inter-rater reliability to ensure coding accuracy. We further present practical guidelines for ethical data handling, including anonymization and secure storage, to protect participant confidentiality.
Conclusion: Rigorous data management and cleaning practices are foundational for valid and reliable SNA, especially in prevention science. By systematically preparing, matching, and filtering network matrices, and by calculating inter-rater reliability, researchers can minimize bias and maximize the impact of their analyses. This tutorial will equip attendees with practical strategies to enhance the quality and trustworthiness of SNA in prevention research.
Abstract 3
Social Network Analysis in R: A Step-by-Step Guide
Victòria Romero Feliu (University of the Balearic Islands), Maria Wei Blanes Perez (University of the Balearic Islands), Elena Gervilla Garcia (University of the Balearic Islands), Joan Pons Bauza (University of the Balearic Islands)
Introduction: Analyzing complex prevention ecosystems requires advanced tools capable of modeling multilayer networks, where diverse types of relationships and contexts interact. The R packages sna (Butts, 2008), statnet (Handcock et al., 2008), and xUCINET (Borgatti et al., 2002) offer robust frameworks for capturing and analyzing these intricate relational structures, supporting the development of effective prevention strategies.
Methods: This tutorial utilizes a social network dataset to illustrate a step-by-step guide for computing SNA indices using weighted networks. The workflow integrates the functionalities of sna, statnet, and xUCINET to provide a comprehensive and practical guide for prevention researchers.
Results: Our example includes the estimation of different centrality, density, reciprocity and fragmentation indexes. The session will include practical code snippets and demonstrations for essential tasks, illustrating how to uncover meaningful patterns.
Conclusion: Integrating sna, statnet, and xUCINET democratizes access to advanced social network analysis techniques, empowering prevention researchers to model and interpret dynamic, interacting relational layers. This approach facilitates more sophisticated insights into prevention systems, which extends beyond traditional analytical methods. Participants will gain hands-on experience in importing, visualizing, and analyzing social network data.
Abstract 4
Mapping Emotional Regulation Networks in Adolescent Sports Teams with SNA
Joan Pons Bauza (University of the Balearic Islands), Victòria Romero Feliu (University of the Balearic Islands), Elena Gervilla Garcia (University of the Balearic Islands), Maria Wei Blanes Perez (University of the Balearic Islands)
Introduction: Sports teams exemplify relational systems where social network analysis (SNA) can clarify complex dynamics among members. Emotional regulation is a key factor in adolescents’ well-being and mental health (Bird et al., 2021). Since adolescence is a period of heightened peer influence, understanding interpersonal emotional regulation within sports teams is particularly important (Tamminen et al., 2016).
Methods: We analyzed data from 218 adolescent sports teams, focusing on networks of interpersonal emotion regulation among teammates. SNA indices-including centrality, density, reciprocity, and fragmentation-were calculated to characterize the structure and quality of emotional support and regulation within teams.
Results: Athletes who engaged in more negative interpersonal emotion regulation efforts towards teammates (higher outdegree centrality) reported poorer mental health outcomes, including lower social functioning (β = -.241, p < .001), higher depression and anxiety (β = .122, p < .001), and greater loss of confidence and self-esteem (β = .114, p < .001). In contrast, those who received more positive regulation efforts (higher indegree centrality) showed improved social functioning (β = .210, p < .001). Teams with higher network density of positive efforts were associated with more task-oriented climates (r = .102, p < .001) and greater athlete satisfaction (r = .106, p < .001). Fragmented or low-reciprocity networks correlated with lower cohesion (r = -.079, p < .001) and more ego-oriented emotional climates (r = -.106, p < .001).
Conclusion: SNA offers a nuanced understanding of how emotional regulation and group belonging interact in adolescent sports teams. Promoting positive and cohesive emotion regulation networks appears beneficial for mental health, athlete satisfaction, and team climate, while fragmented networks are linked to poorer outcomes.
| Conflict of interest | We declare no conflicts of interest. |
|---|
