Pub Date : 2025-11-20DOI: 10.1016/j.socnet.2025.11.001
Minheng Chen , Yang-chih Fu , Xin Guo , Qiang Fu
The estimation and measurement of the size of egocentric networks have sparked vigorous discussion and debate. Drawing on datasets from the Taiwan Social Change Survey, this study explores methodological issues pertaining to the change of core networks in Taiwan from 1997 to 2017 via a modified Poisson mixture approach, assesses the efficiency of name generators as a survey instrument via Fisher Information Maximizer, and investigates the role of social desirability in reporting core networks. Net of other effects, the study finds that individuals expressing a strong sense of social desirability report significantly fewer close contacts and face a higher risk of social isolation. Name generators in this study are associated with trivial design errors and can yield estimates comparable to those produced by exact enumeration. These findings are situated in the drastic changes in face-to-face survey interviews as well as the cultural context of Taiwan and, more broadly, East Asia. They call for further research inquiries into methodological issues regarding measuring and estimating egocentric networks in a transnational and modern setting.
对自我中心网络大小的估计和测量引发了激烈的讨论和争论。本研究以台湾社会变迁调查资料为基础，运用修正的泊松混合方法，探讨1997 - 2017年台湾核心网路变迁的方法问题，运用Fisher Information Maximizer评估名称生成器作为调查工具的效率，并探讨社会期望度在核心网路报告中的作用。考虑到其他影响，研究发现，表现出强烈的社会渴望感的个人报告的亲密接触明显减少，面临更高的社会孤立风险。本研究中的名称生成器与微不足道的设计错误有关，并且可以产生与精确枚举产生的结果相当的估计。这些发现是基于面对面调查访谈的剧烈变化，以及台湾乃至更广泛的东亚的文化背景。他们呼吁对在跨国和现代环境中测量和估计自我中心网络的方法问题进行进一步的研究和调查。
{"title":"Social isolation by design: Bias in measuring core networks in Taiwan?","authors":"Minheng Chen , Yang-chih Fu , Xin Guo , Qiang Fu","doi":"10.1016/j.socnet.2025.11.001","DOIUrl":"10.1016/j.socnet.2025.11.001","url":null,"abstract":"<div><div>The estimation and measurement of the size of egocentric networks have sparked vigorous discussion and debate. Drawing on datasets from the Taiwan Social Change Survey, this study explores methodological issues pertaining to the change of core networks in Taiwan from 1997 to 2017 via a modified Poisson mixture approach, assesses the efficiency of name generators as a survey instrument via Fisher Information Maximizer, and investigates the role of social desirability in reporting core networks. Net of other effects, the study finds that individuals expressing a strong sense of social desirability report significantly fewer close contacts and face a higher risk of social isolation. Name generators in this study are associated with trivial design errors and can yield estimates comparable to those produced by exact enumeration. These findings are situated in the drastic changes in face-to-face survey interviews as well as the cultural context of Taiwan and, more broadly, East Asia. They call for further research inquiries into methodological issues regarding measuring and estimating egocentric networks in a transnational and modern setting.</div></div>","PeriodicalId":48353,"journal":{"name":"Social Networks","volume":"85 ","pages":"Pages 1-12"},"PeriodicalIF":2.4,"publicationDate":"2025-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145555347","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-06DOI: 10.1016/j.socnet.2025.10.003
Julie Riddell , Srebrenka Letina , Kathryn Skivington , Daniel Archambault , Valerie Wells , Emily Long , Ruth Hunter , Mark McCann , the MINI team
The methods used to study and intervene in networks have evolved over the past 90 years. Network research has gained strength from being an interdisciplinary field, but this has also meant that the innovations occurring within some disciplinary communities may not have disseminated to others. Disciplines focussing on health improvement may not have adopted innovations coming from other disciplines relevant to network science. Through a review of the literature, we focus on the key aspects which can inform community-based networks and health improvement (NHI) projects. The review aims to draw these innovations together to understand the range of methods currently available for NHI research and practice. We conducted a narrative synthesis of published literature that may be relevant for NHI projects, synthesising existing work on methods of network data collection, visualisation and intervention approaches.
Searches were conducted between the 8–11th January 2021, within the following databases; ACM Digital Library, EconLIt, ERIC, IEEE Explore, Medline, PsycInfo, Scopus, Social Sciences Citation Index and Sociological Abstracts. The expert community was also consulted to identify relevant research. Searches focused English language papers relating to methods for data collection, visualisation, and implementing interventions, but not statistical analysis. The search was not restricted to studies applied to health.
We used a systemic review methodology to identify peer-reviewed articles that met pre-defined inclusion criteria. Data extraction was restricted to 84 papers published since 2018, of which some were included within more than one category (Network Data collection N = 41, Network Visualisation N = 32, Network Intervention N = 30).
Analysis uncovered a diverse range of approaches to collecting, visualising and interventions using network data, and based on the included studies (rather than existing typologies) we developed a preliminary threefold typology of network and health improvement methods. We found nine types of network visualisation, eight types of data collection, and six types of network intervention approaches. Data visualisations commonly used node-link (circle-line) diagrams to visualise networks whilst key player interventions mostly used whole network data collection, and interventions using personal reflections of networks mostly used egonets. Visualisation was a feature of eight out of 30 intervention papers. Evaluation of the network methods was highly variable.
Our findings suggest potential areas for future methodological research around the use of network methods in community interventions, we propose further integration of data collection and visualisation approaches as part of intervention design, and encourage the network intervention community to integrate methods testing as part of their project to improve the evidence base for network methodology.
{"title":"Methods for interventions using networks to improve health: A narrative synthesis of methodological research on network data collection, visualisation and intervention","authors":"Julie Riddell , Srebrenka Letina , Kathryn Skivington , Daniel Archambault , Valerie Wells , Emily Long , Ruth Hunter , Mark McCann , the MINI team","doi":"10.1016/j.socnet.2025.10.003","DOIUrl":"10.1016/j.socnet.2025.10.003","url":null,"abstract":"<div><div>The methods used to study and intervene in networks have evolved over the past 90 years. Network research has gained strength from being an interdisciplinary field, but this has also meant that the innovations occurring within some disciplinary communities may not have disseminated to others. Disciplines focussing on health improvement may not have adopted innovations coming from other disciplines relevant to network science. Through a review of the literature, we focus on the key aspects which can inform community-based networks and health improvement (NHI) projects. The review aims to draw these innovations together to understand the range of methods currently available for NHI research and practice. We conducted a narrative synthesis of published literature that may be relevant for NHI projects, synthesising existing work on methods of network data collection, visualisation and intervention approaches.</div><div>Searches were conducted between the 8–11th January 2021, within the following databases; ACM Digital Library, EconLIt, ERIC, IEEE Explore, Medline, PsycInfo, Scopus, Social Sciences Citation Index and Sociological Abstracts. The expert community was also consulted to identify relevant research. Searches focused English language papers relating to methods for data collection, visualisation, and implementing interventions, but not statistical analysis. The search was not restricted to studies applied to health.</div><div>We used a systemic review methodology to identify peer-reviewed articles that met pre-defined inclusion criteria. Data extraction was restricted to 84 papers published since 2018, of which some were included within more than one category (Network Data collection N = 41, Network Visualisation N = 32, Network Intervention N = 30).</div><div>Analysis uncovered a diverse range of approaches to collecting, visualising and interventions using network data, and based on the included studies (rather than existing typologies) we developed a preliminary threefold typology of network and health improvement methods. We found nine types of network visualisation, eight types of data collection, and six types of network intervention approaches. Data visualisations commonly used node-link (circle-line) diagrams to visualise networks whilst key player interventions mostly used whole network data collection, and interventions using personal reflections of networks mostly used egonets. Visualisation was a feature of eight out of 30 intervention papers. Evaluation of the network methods was highly variable.</div><div>Our findings suggest potential areas for future methodological research around the use of network methods in community interventions, we propose further integration of data collection and visualisation approaches as part of intervention design, and encourage the network intervention community to integrate methods testing as part of their project to improve the evidence base for network methodology.</div></div>","PeriodicalId":48353,"journal":{"name":"Social Networks","volume":"84 ","pages":"Pages 202-219"},"PeriodicalIF":2.4,"publicationDate":"2025-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145465537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-27DOI: 10.1016/j.socnet.2025.10.002
Sean Everton, Seth Gray, Chad Machiela, Rob Schroeder
This paper aims to demonstrate the potential value of latent space network models (LSNMs) in supporting crisis managers during early community network intervention when information and resources are typically limited and the community is most vulnerable. Community network intervention requires crisis managers to identify existing ties between response network members and to foster and develop relations to eliminate (or minimize) bottlenecks and bridge gaps in human service coverage. We argue that in early crisis intervention, when responders have an incomplete understanding of the situation and limited resources, crisis managers may employ LSNMs to model relationships between actors who share parallel objectives (such as shelter volunteers or substance use disorder treatment specialists) and roles (such as law enforcement and emergency medical responders). Doing so would allow managers to compare the impact of proposed courses of action based on limited existing data and to guide the development of subnetworks to implement intervention initiatives. To demonstrate the utility of LSNMs, we examine the crisis response network that emerged following Oregon Governor Tina Kotek’s January 2024 state of emergency declaration concerning the fentanyl abuse in Portland, the state’s largest city. We find that LSNMs can assist early emergency responders with limited initial network data to model networks and identify critical limitations, assess risks associated with intervention strategies, and prioritize network development efforts to address shortcomings within available resources.
{"title":"Leveraging latent space network models for community intervention","authors":"Sean Everton, Seth Gray, Chad Machiela, Rob Schroeder","doi":"10.1016/j.socnet.2025.10.002","DOIUrl":"10.1016/j.socnet.2025.10.002","url":null,"abstract":"<div><div>This paper aims to demonstrate the potential value of latent space network models (LSNMs) in supporting crisis managers during early community network intervention when information and resources are typically limited and the community is most vulnerable. Community network intervention requires crisis managers to identify existing ties between response network members and to foster and develop relations to eliminate (or minimize) bottlenecks and bridge gaps in human service coverage. We argue that in early crisis intervention, when responders have an incomplete understanding of the situation and limited resources, crisis managers may employ LSNMs to model relationships between actors who share parallel objectives (such as shelter volunteers or substance use disorder treatment specialists) and roles (such as law enforcement and emergency medical responders). Doing so would allow managers to compare the impact of proposed courses of action based on limited existing data and to guide the development of subnetworks to implement intervention initiatives. To demonstrate the utility of LSNMs, we examine the crisis response network that emerged following Oregon Governor Tina Kotek’s January 2024 state of emergency declaration concerning the fentanyl abuse in Portland, the state’s largest city. We find that LSNMs can assist early emergency responders with limited initial network data to model networks and identify critical limitations, assess risks associated with intervention strategies, and prioritize network development efforts to address shortcomings within available resources.</div></div>","PeriodicalId":48353,"journal":{"name":"Social Networks","volume":"84 ","pages":"Pages 191-201"},"PeriodicalIF":2.4,"publicationDate":"2025-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145415548","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-15DOI: 10.1016/j.socnet.2025.10.001
Daniel Cowen , Tobias Stark , Vincenz Frey , Andreas Flache
People become friends with one another primarily due to things they have in common, like shared demographic characteristics or shared interests. But on what similarities are people becoming friends at different stages of knowing one another? To study this, we use a longitudinal dataset that followed a cohort of students of one study programme at a Swiss technical university. We model how demographic traits of nationality and gender, and less observable interest-related traits of being social, a partygoer, a smart and hard-working student, contribute to friendships. Using Stochastic Actor- Oriented Models, we find a baseline level of both demographic and interest-related friendship homophily, indicating that there are differing reasons behind friendships. We also find that homophily based on demographic traits diminishes over time when students get to know those in their study cohort better. This suggest that homophily on observable traits is mainly relevant when people first meet but becomes less important over time.
{"title":"Trends of Friends – Time dynamics of Surface- and Deep- level traits in friendship formation and maintenance","authors":"Daniel Cowen , Tobias Stark , Vincenz Frey , Andreas Flache","doi":"10.1016/j.socnet.2025.10.001","DOIUrl":"10.1016/j.socnet.2025.10.001","url":null,"abstract":"<div><div>People become friends with one another primarily due to things they have in common, like shared demographic characteristics or shared interests. But on what similarities are people becoming friends at different stages of knowing one another? To study this, we use a longitudinal dataset that followed a cohort of students of one study programme at a Swiss technical university. We model how demographic traits of nationality and gender, and less observable interest-related traits of being social, a partygoer, a smart and hard-working student, contribute to friendships. Using Stochastic Actor- Oriented Models, we find a baseline level of both demographic and interest-related friendship homophily, indicating that there are differing reasons behind friendships. We also find that homophily based on demographic traits diminishes over time when students get to know those in their study cohort better. This suggest that homophily on observable traits is mainly relevant when people first meet but becomes less important over time.</div></div>","PeriodicalId":48353,"journal":{"name":"Social Networks","volume":"84 ","pages":"Pages 180-190"},"PeriodicalIF":2.4,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145323688","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-04DOI: 10.1016/j.socnet.2025.09.004
Yunsub Lee , Xinwei Xu
Aggregated relational data (ARD) provides valuable information for inferring structural features of personal social networks at scale. Following recent ARD studies, we suggest a formal parameter for agent-based modeling (ABM) that helps reflect multiple structural features of extended social networks (e.g., size; variation; distribution) and apply it to a widely known classic ABM—Axelrod’s cultural dynamic model. Results show that when incorporating realistic network features estimated from ARD, the model generates outcomes substantially different from its original results. Our study highlights ARD's potential to enrich ABM in reflecting more realistic networks that better connect micro-processes with macro-phenomena.
{"title":"Use of aggregated relational data in agent-based modeling","authors":"Yunsub Lee , Xinwei Xu","doi":"10.1016/j.socnet.2025.09.004","DOIUrl":"10.1016/j.socnet.2025.09.004","url":null,"abstract":"<div><div>Aggregated relational data (ARD) provides valuable information for inferring structural features of personal social networks at scale. Following recent ARD studies, we suggest a formal parameter for agent-based modeling (ABM) that helps reflect multiple structural features of extended social networks (e.g., size; variation; distribution) and apply it to a widely known classic ABM—Axelrod’s cultural dynamic model. Results show that when incorporating realistic network features estimated from ARD, the model generates outcomes substantially different from its original results. Our study highlights ARD's potential to enrich ABM in reflecting more realistic networks that better connect micro-processes with macro-phenomena.</div></div>","PeriodicalId":48353,"journal":{"name":"Social Networks","volume":"84 ","pages":"Pages 164-179"},"PeriodicalIF":2.4,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145220009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-27DOI: 10.1016/j.socnet.2025.09.003
Jonathan Januar , H. Colin Gallagher , Johan Koskinen
The clandestine nature of covert networks makes reliable data difficult to obtain and leads to concerns with missing data. We explore the use of network models to represent missingness mechanisms. Exponential random graph models provide a flexible way of parameterising departures from conventional missingness assumptions and data management practices. We demonstrate the effects of model specification, true network structure, and different not-at-random missingness mechanisms across six empirical covert networks. Our framework for modelling realistic missingness mechanisms investigates potential inferential pitfalls, evaluates decisions in collecting data, and offers the opportunity to incorporate non-random missingness into the estimation of network generating mechanisms.
{"title":"In the shadow of silence: Modelling missing data in the dark networks of crime and terrorists","authors":"Jonathan Januar , H. Colin Gallagher , Johan Koskinen","doi":"10.1016/j.socnet.2025.09.003","DOIUrl":"10.1016/j.socnet.2025.09.003","url":null,"abstract":"<div><div>The clandestine nature of covert networks makes reliable data difficult to obtain and leads to concerns with missing data. We explore the use of network models to represent missingness mechanisms. Exponential random graph models provide a flexible way of parameterising departures from conventional missingness assumptions and data management practices. We demonstrate the effects of model specification, true network structure, and different not-at-random missingness mechanisms across six empirical covert networks. Our framework for modelling realistic missingness mechanisms investigates potential inferential pitfalls, evaluates decisions in collecting data, and offers the opportunity to incorporate non-random missingness into the estimation of network generating mechanisms.</div></div>","PeriodicalId":48353,"journal":{"name":"Social Networks","volume":"84 ","pages":"Pages 147-163"},"PeriodicalIF":2.4,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145157542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-17DOI: 10.1016/j.socnet.2025.08.003
Shira Offer , Claude S. Fischer , Keunbok Lee
Egocentric networks are dynamic. Prior research has typically measured change in broad network characteristics or simply in membership turnover but given relatively little attention to the history of alter-ego relationships. Using rich information about alters over three waves in the UCNets study, we develop a novel approach that delineates a “trajectory” over time for each of the alters in the network and then uses all these observed trajectories to identify types of ego networks. Results from Multilevel Latent Growth Models reveal six distinct trajectories for alters: continuously active, awakened, dormant, dropped, transitory, and new. The distribution of those six trajectories coalesces at the ego-level into three network types – anchored, shifting, and regenerative – each with unique dynamics and compositional features. To illustrate the contribution of this approach, we examine the associations between life events and the three types of network dynamics among UCNets' young adults. Findings reveal subtle patterns of change: some events shape networks by reinforcing their cores, while others expand or reconfigure networks’ near and distant peripheries.
{"title":"The gears in network dynamics: The alter-trajectory approach","authors":"Shira Offer , Claude S. Fischer , Keunbok Lee","doi":"10.1016/j.socnet.2025.08.003","DOIUrl":"10.1016/j.socnet.2025.08.003","url":null,"abstract":"<div><div>Egocentric networks are dynamic. Prior research has typically measured change in broad network characteristics or simply in membership turnover but given relatively little attention to the history of alter-ego relationships. Using rich information about alters over three waves in the UCNets study, we develop a novel approach that delineates a “trajectory” over time for each of the alters in the network and then uses all these observed trajectories to identify types of ego networks. Results from Multilevel Latent Growth Models reveal six distinct trajectories for alters: continuously active, awakened, dormant, dropped, transitory, and new. The distribution of those six trajectories coalesces at the ego-level into three network types – anchored, shifting, and regenerative – each with unique dynamics and compositional features. To illustrate the contribution of this approach, we examine the associations between life events and the three types of network dynamics among UCNets' young adults. Findings reveal subtle patterns of change: some events shape networks by reinforcing their cores, while others expand or reconfigure networks’ near and distant peripheries.</div></div>","PeriodicalId":48353,"journal":{"name":"Social Networks","volume":"84 ","pages":"Pages 131-146"},"PeriodicalIF":2.4,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145095566","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-15DOI: 10.1016/j.socnet.2025.09.001
Alessandro Lomi, Philippa E. Pattison
{"title":"Duality: The first fifty years and beyond","authors":"Alessandro Lomi, Philippa E. Pattison","doi":"10.1016/j.socnet.2025.09.001","DOIUrl":"10.1016/j.socnet.2025.09.001","url":null,"abstract":"","PeriodicalId":48353,"journal":{"name":"Social Networks","volume":"84 ","pages":"Pages 123-126"},"PeriodicalIF":2.4,"publicationDate":"2025-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145060458","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-12DOI: 10.1016/j.socnet.2025.08.004
David Kretschmer , Lars Leszczensky
Friendship segregation between Muslim and non-Muslim youth in Europe is well documented. However, previous network studies provide only snapshots, thus ignoring whether interreligious friendship-making changes throughout adolescence. A recent non-network study suggests increasing in-group friendships among Muslim girls and stability among Muslim boys, but it could not explain these differences and did not consider interdependence with non-Muslims’ friendship-making. To overcome these limitations, we study the trajectories of friendship-making among Muslim and non-Muslim boys and girls and assess the explanatory power of three key determinants of interreligious friendship-making dynamics: interreligious attitudes, religious norms that constrain out-group friendships, and reactions to friendship-making behavior of other groups. Addressing the methodological limitations of non-network research, we study friendship trajectories with stochastic actor-oriented models for network evolution applied to five waves of longitudinal friendship network data among 1122 Muslim and non-Muslim youth in German schools. We find that Muslim girls start out with at least as many interreligious friends as Muslim boys but that their tendency to have non-Muslim friends decreases substantially throughout adolescence. By contrast, the religious friendship-making of both Muslim boys and non-Muslims of either gender remains stable over time. We show that the increase in in-group friendships only applies to Muslim girls with high religiosity and that it is particularly strong for cross-gender friendships, suggesting that gendered religious norms can explain differences in the dynamics of Muslim boys’ and girls’ friendship-making. By contrast, interreligious attitudes and reactions to shifts in other groups’ friendship-making do not contribute to the observed friendship-making trajectories.
{"title":"Stable or dynamic? Explaining the development of Muslim and non-Muslim boys’ and girls’ friendship-making across adolescence","authors":"David Kretschmer , Lars Leszczensky","doi":"10.1016/j.socnet.2025.08.004","DOIUrl":"10.1016/j.socnet.2025.08.004","url":null,"abstract":"<div><div>Friendship segregation between Muslim and non-Muslim youth in Europe is well documented. However, previous network studies provide only snapshots, thus ignoring whether interreligious friendship-making changes throughout adolescence. A recent non-network study suggests increasing in-group friendships among Muslim girls and stability among Muslim boys, but it could not explain these differences and did not consider interdependence with non-Muslims’ friendship-making. To overcome these limitations, we study the trajectories of friendship-making among Muslim and non-Muslim boys and girls and assess the explanatory power of three key determinants of interreligious friendship-making dynamics: interreligious attitudes, religious norms that constrain out-group friendships, and reactions to friendship-making behavior of other groups. Addressing the methodological limitations of non-network research, we study friendship trajectories with stochastic actor-oriented models for network evolution applied to five waves of longitudinal friendship network data among 1122 Muslim and non-Muslim youth in German schools. We find that Muslim girls start out with at least as many interreligious friends as Muslim boys but that their tendency to have non-Muslim friends decreases substantially throughout adolescence. By contrast, the religious friendship-making of both Muslim boys and non-Muslims of either gender remains stable over time. We show that the increase in in-group friendships only applies to Muslim girls with high religiosity and that it is particularly strong for cross-gender friendships, suggesting that gendered religious norms can explain differences in the dynamics of Muslim boys’ and girls’ friendship-making. By contrast, interreligious attitudes and reactions to shifts in other groups’ friendship-making do not contribute to the observed friendship-making trajectories.</div></div>","PeriodicalId":48353,"journal":{"name":"Social Networks","volume":"84 ","pages":"Pages 110-122"},"PeriodicalIF":2.4,"publicationDate":"2025-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145046129","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号