Homophily Effect Research Articles

Homophily Effects Among Oxford House Residents Utilizing Medication Assisted Treatment

ABSTRACT The effects of social support among persons utilizing medication assisted treatment (MAT) living in abstinence-based recovery homes is not clearly understood, and there is a need to identify social process components that engender social support for this population. The present investigation examined whether homophily (i.e., living with/without fellow MAT residents) among recovery home residents would moderate the mediating effects of social support on the relationships between stress and two recovery outcomes (quality of life, abstinence self-efficacy) among a sample of persons utilizing MAT living in recovery homes (Oxford Houses) in the U.S. Moderated mediational analyses were conducted to test whether potential mediating effects were moderated by homophily. Mediating effects were observed in both models, but they were significant only through the moderator (homophily) in that social support buffered the effects of stress among MAT residents who lived with other MAT residents in their recovery homes. Findings suggest that homophily effects are related to social networking for persons utilizing MAT living in abstinence-based recovery homes.

Homophily and behavior diffusion

There is continuing debate over the effect of homophily, which is the tendency for individuals to socialize with similar people, on behavior diffusion. We aimed to clarify this relationship from a social network perspective, using the agent-based modeling approach. The results demonstrate that homophily promoted the diffusion of behaviors that people had a strong propensity to adopt, but had a prohibitive effect when the adoption propensity was weak. When the adoption propensity was moderate, the effect was promotive at first and then became prohibitive. Moreover, we identified 3 types of homophily—status, value, and mixed (status–value)—and found that mixed homophily was most effective for behavior diffusion, followed by value homophily and then status homophily. These findings highlight the importance of behavior classes and homophily type in the relationship between homophily and behavior diffusion, and call for a serious consideration of both factors when empirically studying the related issues.

Consequences of homophily: does social status similarity enhance project performance?

This study examines the effect of status homophily between a team organizer and a project team on project performance. Due to the Matthew effect, high statuses of the team organizer and project team may lead to better project performance, respectively. However, similar-status association on a project can make it difficult to form an informal hierarchy and therefore cause internal conflict. Thus, status homophily may weaken project performance. We test this idea in the context of the Chinese film industry. The results show that both producer status and artistic team status enhance project performance. However, the joint effect (i.e., status homophily) negatively affects project performance.

The effect of age, environment and management on social contact patterns in sheep

Social structures of group-living farm animals can have important implications for animal welfare and productivity. Understanding which factors can have an effect on social behaviour is thus important in order to develop the best management strategies in livestock industries. Here, we studied the social network structure of a flock of 84 Poll Dorset ewes and collecting dyadic associations data through the use of proximity sensors during two study periods. First, we analysed the social structure of ewes at a group-level, by analysing the community structure, and at individual-level, by determining whether the ewes showed social differentiation in their association patterns. Second, we measured for the contribution of genetic relatedness, age, weight, reproductive status and previous management sub grouping on social associations to test for homophily effects. Lastly, we evaluated whether social clustering was influenced by the stocking density of individuals in a field, and by weather parameters, through the use of two climatic indices, the Temperature-Humidity Index (THI) and the Wind Chill Index (WCI). Our results showed that the pairwise associations between ewes are not-random and highly heterogeneous, both in total time spent in contact and in contacts duration. There was no evidence that ewes were subdivided into social communities, and at individual level, they showed markedly differentiated social relationships, demonstrating preferences in social ties. However, the factors that influenced the preferred social interactions between individuals changed over time. In the first study period ewes tended to maintain the social bonds formed in previous management sub grouping, most likely due to a social familiarization resulting from repeated interactions with the same individuals. In the second study period similarity in age influenced the strength of associations among ewes. We found no significant influence of reproductive status, weight (as an indicator of body size) and genetic relatedness on proximity associations in either study period. Moreover, our results showed the tendency of the ewes to form social clusters varied in relation to animals’ density, and Wind Chill Index (WCI). The identification of conditions that modify the social behaviour of sheep is critically important in order to implement management and productivity strategies and our results highlight how flock social structure can change depending on environmental and social contexts.

Multidimensional diffusion processes in dynamic online networks

We develop a dynamic matched sample estimation algorithm to distinguish peer influence and homophily effects on item adoption decisions in dynamic networks, with numerous items diffusing simultaneously. We infer preferences using a machine learning algorithm applied to previous adoption decisions, and we match agents using those inferred preferences. We show that ignoring previous adoption decisions leads to significantly overestimating the role of peer influence in the diffusion of information, mistakenly confounding influence-based contagion with diffusion driven by common preferences. Our matching-on-preferences algorithm with machine learning reduces the relative effect of peer influence on item adoption decisions in this network significantly more than matching on earlier adoption decisions, as well other observable characteristics. We also show significant and intuitive heterogeneity in the relative effect of peer influence.

Academic Achievement and Social Interactions: A Longitudinal Analysis of Peer Selection Processes in Inclusive Elementary Classrooms

This study analyzes the extent to which students in inclusive elementary classrooms select classmates for collaboration and play activities on the basis of academic achievement. To investigate this association, second and third graders (n = 506, mean age 7.6 years) were asked to nominate classmates with whom they collaborate and play at the beginning and middle of the school year. The academic achievement level of students was assessed with a standardized mathematical test at the first measurement point. Two longitudinal social network models were specified to analyze whether academic achievement level functions as a selection criterion for collaboration (academic context) and play activities (play context). Structural network effects, selection effects, and sex-related effects were included in the models. The selection effects on academic achievement revealed that high achieving students were more likely selected by peers for collaboration but not for play activities. Thus, academic achievement was only relevant for peer selection processes in the academic context. Further, a negative homophily effect on academic achievement was found, which indicates that students selected dissimilar peers in terms of academic achievement level for interactions in the academic and play context. Thus, on one hand there was no similarity in academic achievement among peers who liked to collaborate and play together, which is a positive result from an educational and inclusive perspective; on the other hand, there was a tendency to select high achieving rather than low achieving peers for collaborative activities. Teaching practices that can help prevent the formation of such a hierarchy in the classroom and implications for future studies on the link between academic achievement and peer selection processes are discussed.

Equilibrium homophily in networks

I look at the equilibrium effects of homophily in network formation games having positive externalities. I show that very weak preferences for network effects (such as a preference for being part of a clique) may result in very high levels of homophily at equilibrium. I present an empirical procedure that allows an evaluation of the role of network effects on the observed level of homophily in the data. I apply my method to the formation of friendship networks in the USA. I show that network effects are small and have a negligible contribution to the observed racial homophily in friendship networks.

Can Social Capital and Network Homophily Facilitate Community Participation?

Social networks and social capital have been a central discussion point for academicians and policy makers alike since last few decades. Intersecting literature on networks and applied areas stresses upon homophily effect operating through social categories of gender, caste and occupation play a crucial role in individual behaviour across health and related networks in the society. On the other hand, literature talks of social capital playing a vital role for the individual and collective society in various spheres of their lives. This empirical research triangulates the theories of social capital, social networks and community participation to study the role of networks in community participation in the context of local governance in rural society. We use indicators such as household network size as proxy to social capital, attendance and vocal participation as proxy to community participation while using the popular network concepts of peer-effect and homophily effect through logit models to assess the role of network-based social capital in community participation.

Comparing Gender Homophily among the Multilayer Media Social Networks of Face-to-Face, Instant Messenger and Social Networking Services: A Case Study of a High School Classroom

Abstract In which social worlds does gender homophily operate more strongly – offline or online? To address this question, the following two aspects must be considered. First, people currently use many types of online communication media. Second, to examine the homophily effects exclusively, it is necessary to control for other network formation mechanisms such as ‘foci’ and ‘triadic closure.’ For this study, I conducted a mixed-method research in a high school in rural Japan. I asked students about who they interacted with face-to-face (F2F), through instant messenger (IM), and social networking services (SNS) and then analyzed the social networks using exponential random graph models (ERGMs). Subsequently, I conducted semi-structured interviews to uncover the practices and social contexts of each communication media and explain the results of the quantitative analysis. The results showed that SNS was more gender heterogeneous than offline. In the IM network, a small gender homophily effect was initially observed. However, three months later, its strength decreased to almost the same as that in the SNS networks. From the qualitative research, some key mechanisms producing the difference in gender homophily are specified, such as precedence of F2F communication to IM interaction, independence of SNS communication from F2F, recommending functions, and hobby homophily. Overall, this study implies that considering offline or online alone may cause misunderstanding regarding homophily in organizations because the observed strength of homophily effects depends on whether the space is examined offline or online, what kind of media is examined, and when the online social network data are collected.

Homophily as a Process Generating Social Networks: Insights from Social Distance Attachment Model

Real-world social networks often exhibit high levels of clustering, positive degree assortativity, short average path lengths (small-world property) and right-skewed but rarely power law degree distributions. On the other hand homophily, defined as the propensity of similar agents to connect to each other, is one of the most fundamental social processes observed in many human and animal societies. In this paper we examine the extent to which homophily is sufficient to produce the typical structural properties of social networks. To do so, we conduct a simulation study based on the Social Distance Attachment (SDA) model, a particular kind of Random Geometric Graph (RGG), in which nodes are embedded in a social space and connection probabilities depend functionally on distances between nodes. We derive the form of the model from first principles based on existing analytical results and argue that the mathematical construction of RGGs corresponds directly to the homophily principle, so they provide a good model for it. We find that homophily, especially when combined with a random edge rewiring, is sufficient to reproduce many of the characteristic features of social networks. Additionally, we devise a hybrid model combining SDA with the configuration model that allows generating homophilic networks with arbitrary degree sequences and we use it to study interactions of homophily with processes imposing constraints on degree distributions. We show that the effects of homophily on clustering are robust with respect to distribution constraints, while degree assortativity can be highly dependent on the particular kind of enforced degree sequence.

Specification and Estimation of Network Formation and Network Interaction Models with the Exponential Probability Distribution

In this paper, we model network formation and network interactions under a unified framework. The key feature of our model is to allow individuals to respond to incentives stemming from interaction benefits on certain activities when they choose friends (network links), while capturing homophily in terms of unobserved characteristic variables in network formation and activities. There are two advantages of this modeling approach: first, one can evaluate whether incentives from certain interactions are important factors for friendship formation or not. Second, in addition to homophily effects in terms of unobserved characteristics, inclusion of incentive effects in the network formulation also corrects possible friendship selection bias on activity outcomes under network interactions. A theoretical foundation of this unified model is based on a complete information cooperative game. A tractable Bayesian MCMC approach is proposed for the estimation of the model. We apply the model to empirically study American high school students' friendship networks with the Add Health data. We consider two activity variables, GPA and smoking frequency, and find a significant incentive effect from GPA, but not from smoking, on friendship formation. These results suggest that the benefit of interactions in academic learning is an important factor for friendship formation, while the interaction benefit in smoking is not, even though homophily in smoking behavior is important for a smoker to link to other smokers. On the other hand, from the perspective of network interactions, both GPA and smoking frequency are subject to significant positive interaction (peer) effects.

A computational study of homophily and diffusion of common knowledge on social networks based on a model of Facebook

In this paper, we introduce homophily to a game-theoretic model of collective action (e.g., protests) on Facebook and study the effect of homophily in individuals’ willingness to participate in collective action, i.e., their thresholds, on the emergence and spread of collective action. We use three different networks (a real Facebook network, an Erdős–Renyi random graph, and a scale-free network) and conduct computational experiments to study contagion dynamics (the size and the speed of diffusion) with respect to the level of homophily. We provide a series of parametric results on the time to achieve a specified contagion spread, on the spread of contagion at different times, and the probability of cascades. We demonstrate that these behaviors are highly nonlinear and nonmonotonic in homophily. Networks with randomly assigned thresholds result in both smaller and slower diffusion compared to the networks characterized by homophily and heterophily.

Effects of homophily and academic reputation in the nomination and selection of Nobel laureates

In collective decision-making, a group of independent experts propose individual choices to reach a common decision. This is the case of competitive events such as Olympics, international Prizes or grant evaluation, where groups of experts evaluate individual performances to assign resources, e.g. scores, recognitions, or funding. However, there are systems where evaluating individual’s performance is difficult: in those cases, other factors play a relevant role, leading to unexpected emergent phenomena from micro-scale interactions. The Nobel assignment procedure, rooted on recommendations, is one of these systems. Here we unveil its network, reconstructed from official data and metadata about nominators, nominees and awardees between 1901 and 1965, consisting of almost 12,000 individuals and 17,000 nominations. We quantify the role of homophily, academic reputation of nominators and their prestige neighborhood, showing that nominees endorsed by central actors – who are part of the system’s core because of their prestigious reputation – are more likely to become laureate within a finite time scale than nominees endorsed by nominators in the periphery of the network. We propose a mechanistic model which reproduces all the salient observations and allows to design possible countermeasures to mitigate observed effects.

Achievement in student peer networks: A study of the selection process, peer effects and student centrality

Isolating the homophily effect from the socialization process is crucial in understanding the links between student achievement and peer performance. Then, the combined effects on student achievement of peer performance and of student centrality within a network have been scarcely studied. While friendship links have been addressed, researches rarely investigated instrumental ties. In this paper, 120 college students were asked about their friends and about their advisors. Results show a selection effect (i.e., students reassemble with friends sharing same levels of performance), and a socialization process (i.e., peer performance is significantly associated with later student achievement). While the centrality univariate effects are significant for student performance, only being a prestigious friend remains important when peer performance is taken into account.

Credibility of climate change denial in social media

Public perception about the reality of climate change has remained polarized and propagation of fake information on social media can be a potential cause. Homophily in communication, the tendency of people to communicate with others having similar beliefs, is understood to lead to the formation of echo chambers which reinforce individual beliefs and fuel further increase in polarization. Quite surprisingly, in an empirical analysis of the effect of homophily in communication on the level of polarization using evidence from Twitter conversations on the climate change topic during 2007–2017, we find that evolution of homophily over time negatively affects the evolution of polarization in the long run. Among various information about climate change to which people are exposed to, they are more likely to be influenced by information that have higher credibility. Therefore, we study a model of polarization of beliefs in social networks that accounts for credibility of propagating information in addition to homophily in communication. We find that polarization can not increase with increase in homophily in communication unless information propagating fake beliefs has minimal credibility. We therefore infer from the empirical results that anti-climate change tweets are largely not credible.

Nonparametric identification in index models of link formation

We consider a dyadic link formation model with a homophily effect index and a degree heterogeneity index. We provide nonparametric identification results for the potentially nonparametric homophily effect function, the realizations of unobserved individual fixed effects and the unknown distribution of idiosyncratic shocks, up to normalization. We propose a novel form of scale normalization on an arbitrary interquantile range, which is not only theoretically general but also proves particularly convenient for the identification analysis. We then use an inductive “in-fill and out-expansion” algorithm to establish our main results.

An Empirical Comparison of Statistical Models for Longitudinal Elementary Teachers’ Advice Networks

이 연구의 목적은 초등교원의 종단 조언 네트워크에 관한 통계 모형을 경험적으로 비교하는 것이다. 두 가지 연구 문제는 첫째, 종단 지수임의 그래프 모형 (TERGM) 과 확률론적 행위자 기반 모형 (SAOM) 사이의 결과의 유사점과 차이점은 무엇인가? 둘째, 교원의 종단 조언 네트워크에 적합한 통계 모형은 무엇인가? 이를 위해 2018학년도 1학기 초등 교원의 조언 네트워크 (n = 54)와 교원 소진을 4회 반복 측정하였고, 이러한 자료를 종단 지수임의 그래프 모형과 확률론적 행위자 기반 모형으로 분석하였다. 그 분석결과, 종단 지수임의 그래프 모형과 확률론적 행위자 기반 모형 사이의 유사성은 교원의 조언 네트워크 변화에 전이성 효과, 동일 학년 유사성 효과, 공식 지도자의 대중성 효과 및 개인 성취감 결여의 활동성 효과이었다. 반면에 두 모형의 차이점은 교원의 조언 네트워크 변화에 호혜성 효과와 교원 총 경력의 대중성 효과이었다. 이러한 차이는 주로 시간에 대한 관점 (이산 대 연속성)과 시간 종속성에 대한 처리방법 (장기종속 대 1차 마르코프 과정)에 기인한다고 볼 수 있다. 이러한 결과를 토대로, 교원의 종단 조언 네트워크 자료에 적합한 통계 모형을 제안했다.The purpose of this study was to empirically compare statistical models for longitudinal elementary teachers’ advice networks. The two research questions were: first, what are the similarities and differences between the results of the temporal exponential random graph model and the stochastic actor-oriented model? Second, which statistical models are appropriate for longitudinal teachers’ advice networks? To this end, we measured teachers’ advice networks (n=54) and teachers’ burnout four times in one elementary school in 2018, which we analyzed by temporal exponential random graph models and stochastic actor-oriented models. Results showed that similarities between two types of models included the transitivity effect, the grade-level homophily effect, the alter effect of formal leader on the change of teachers’ advice networks, and the reduced feelings of personal accomplishment ego effect on the change of teachers’ advice network. Meanwhile, differences between the two models were the reciprocity effect and the total teaching experience effect on the change of teachers’ advice networks. These differences are mainly due to the viewpoint about time (discrete versus continuous) and the treatment of temporal dependencies (longer dependence versus a first-order Markov process). Based on these results, we suggest appropriate statistical models for longitudinal teachers’ advice network data.

Identity, trust and altruism: An experiment on preferences and microfinance lending

We conduct laboratory and online experiment treatments to study willingness to lend, potential crowd out of philanthropic motivation by financial returns, and preference over borrowers in micro-finance lending. We distinguish between perceptions of transaction-related factors, such as neediness and trustworthiness, and identity-related factors such as ethnicity and gender. By looking at both channels together we are able to assess the extent to which identity-related differences in lending can be attributed to differences in how transaction-related factors are perceived among people with similar and different identities. We find that (1) both financial return and philanthropic motivation affect the amount lent, with little evidence that the former crowds out the latter; (2) lenders prefer borrowers with whom they share gender and ethnic similarity even after controlling for perceived riskiness and neediness; and (3) lenders are willing to trade greater risk in order to help more needy borrowers, but at a rate sensitive to their own degree of financial exposure. Lack of crowding out, homophily effects in some subgroups, and robustness to decision order are found in additional treatments with three subject pools, including one of non-student participants.

Spatial Modeling Approach for Dynamic Network Formation and Interactions

This study primarily seeks to answer the following question: How do social networks evolve over time and affect individual economic activity? To provide an adequate empirical tool to answer this question, we propose a new modeling approach for longitudinal data of networks and activity outcomes. The key features of our model are the inclusion of dynamic effects and the use of time-varying latent variables to determine unobserved individual traits in network formation and activity interactions. The proposed model combines two well-known models in the field: latent space model for dynamic network formation and spatial dynamic panel data model for network interactions. This combination reflects real situations, where network links and activity outcomes are interdependent and jointly influenced by unobserved individual traits. Moreover, this combination enables us to (1) manage the endogenous selection issue inherited in network interaction studies, and (2) investigate the effect of homophily and individual heterogeneity in network formation. We develop a Bayesian Markov chain Monte Carlo sampling approach to estimate the model. We also provide a Monte Carlo experiment to analyze the performance of our estimation method and apply the model to a longitudinal student network data in Taiwan to study the friendship network formation and peer effect on academic performance. Supplementary materials for this article are available online.

Students of the Same Opinion Flock Together. The Interplay of Homophily and Influence in Forming Homogenous Opinion Networks.

Many studies report homogeneity regarding political opinions within friendship networks. Two mechanisms, homophily and influence, can lead to this empirical finding. Homophily describes a selection mechanism whereby individuals with similar opinions are more likely to befriend each other. Influence means that individuals are directly influenced by their friends. Furthermore, it is assumed that political discussions are a transmission mode for political opinions and thus increase the effects of homophily and influence. Theoretically, it can be assumed that both mechanisms are at work, however, empirically it is difficult to separate the effects from each other. To investigate both mechanisms and their interaction with political discussion, stochastic actor-based co-evolution models are estimated. The data for the analysis was collected in a three-wave survey panel among students during their first semester at university. The results indicate that both mechanisms can impact network homogeneity, however, which mechanism is at work depends on the contemporary nature of the topic. The analyses also show that political discussions do not increase the effect of homophily and influence.