Network Autocorrelation Models Research Articles

Depressive symptoms associations with online and in person networks in an online gaming community: a pilot study

Purpose Increases in video game use have led to mental health concerns, citing greater risk for depressive symptoms (DS) and reduced “in-real-life” (IRL) social involvement. However, recent studies have uncovered potential social benefits for online gaming. Many games provide avenues to extend real life relationships and make new online friendships. The purpose of this pilot study is to use social network analysis to determine associations between connections and DS in a gaming community. Design/methodology/approach As a pilot study, members of an online gaming site were asked to report demographic characteristics, DS, IRL social support, online social support and IRL people and members of the online community with whom they spoke to about important life matters. Multi-level modeling was used to parse variance described by demographic characteristics, IRL measures and online measures. Linear network autocorrelation modeling (LNAM) was used to determine relationships between network connections and DS. Findings Members (n = 37; µ = 24.76 years old, SD = 6.55; 100% male; 89.2% white) on average felt DS’ “not at all” to “several days” over the past two weeks. Multi-level modeling including online network measures explained 50% of variance (R2 = 0.50, F (9,27) = 2.98, p = 0.01); online connections were associated with DS (ß = 0.46). LNAM indicated DS were associated with IRL support (ß = −2.66), IRL connections (ß = 1.81), online support (ß = 2.40) and network effects (ß = 0.06), which indicates that a gamer’s DS were similar to those of their online connections. Originality/value Members with more DS may be seeking help via online channels. This may be important for future research to consider alternative forms of help-seeking behavior.

Imputation of attributes in networked data using Bayesian autocorrelation regression models

Misspecification in network autocorrelation models poses a challenge for parameter estimation, which is amplified by missing data. Model misspecification has been a focus of recent work in the statistics literature and new robust procedures have been developed, in particular cutting feedback. This paper shows how this helps in a misspecified network autocorrelation model. Where model misspecification is mild and the traits are fully observed, Bayesian imputation is routine. In settings with high missingness, Bayesian inference can fail, but a closely related cut model is robust. We illustrate this on a data set of graduate students using a Facebook-like messaging app.

Analysis of network interventions with an application to hospital-acquired infections.

Regional interventions to prevent the spread of hospital-acquired infections, vaccination campaigns, and information dissemination strategies are examples of treatment interventions applied to members of a network with the intent of effecting a network-wide change. In designing clinical trials or determining policy changes, it may not be cost effective or otherwise possible to treat all actors of a network. There is a notable lack of study designs and statistical frameworks with which to plan a network-wide intervention in this context and analyze the resulting data. This paper builds off of the network autocorrelation model in order to provide such a framework for a pre-post study design. We derive key quantitative measures of the network-wide treatment effect, exact formulas for power analyses of these measures, and extensions for the context in which the network is unknown. As the treatment assignation is part of the network-wide treatment, we provide methods for determining the assignation which optimizes the overall treatment effect over all members of the network subject to any arbitrary set of implementation costs and cost constraint. We implement these methods on Clostridioides difficile data for the state of California, where the hospitals are linked through patient sharing.

‘Burnout contagion’ among teachers: A social network approach

Previous studies have found that burnout is to some extent contagious and have argued it is a socially induced phenomenon. However, until now, actual social interactions and the long‐term effect of this contagion have remained largely unexplored. This study aimed to expand earlier findings on burnout contagion through the application of a social network approach. This approach assumes that some relationships provide more information on the feelings and attitudes of others. This study therefore not only identified interaction partners, but also examined how characteristics (multiplexity, frequency, and embeddedness) of the relationship with those partners relate to burnout contagion. Using (temporal) network autocorrelation models, burnout contagion was empirically investigated in the context of secondary school teams. Cross‐sectional analyses were performed on data obtained from 931 teachers working in 14 schools. Long‐term effects of burnout contagion were assessed among 578 teachers working in 12 schools. The results showed that interpersonal interactions act as conduits for burnout contagion, especially when relations are strong in terms of frequency, embeddedness, and multiplexity. The results also showed that features of relationships play a differential role in the contagion of different components of burnout. Moreover, long‐term effects were found for emotional exhaustion. This study thus provided evidence for the importance of interpersonal relationships in burnout contagion.Practitioner points Negative feelings are transmitted through personal interaction: As such, the importance of positive (social) experiences within the school team is stressed. Co‐rumination should be avoided as it may impact negatively on employees’ well‐being in both the short term and the long term. Given the contagious nature of burnout, interventions for preventing and reducing burnout should not be solely focused on increasing social support within the school team. External support might be necessary to disrupt a potential negative cycle within this team.

Rate of intersexual interactions affects injury likelihood in Tasmanian devil contact networks

Abstract Identifying the types of contacts that result in disease transmission is important for accurately modeling and predicting transmission dynamics and disease spread in wild populations. We investigated contacts within a population of adult Tasmanian devils (Sarcophilus harrisii) over a 6-month period and tested whether individual-level contact patterns were correlated with accumulation of bite wounds. Bite wounds are important in the spread of devil facial tumor disease, a clonal cancer cell line transmitted through direct inoculation of tumor cells when susceptible and infected individuals bite each other. We used multimodel inference and network autocorrelation models to investigate the effects of individual-level contact patterns, identities of interacting partners, and position within the social network on the propensity to be involved in bite-inducing contacts. We found that males were more likely to receive potentially disease-transmitting bite wounds than females, particularly during the mating season when males spend extended periods mate-guarding females. The number of bite wounds individuals received during the mating season was unrelated to any of the network metrics examined. Our approach illustrates the necessity for understanding which contact types spread disease in different systems to assist the management of this and other infectious wildlife diseases.

Using Social Network Analysis to Understand Trust, Reciprocity, and Togetherness in Wildlife Tourism Microentrepreneurship

High levels of trust, reciprocity, and togetherness embedded within entrepreneurial networks are believed to facilitate cooperation that enables success among individual business owners. This study examines the effects of social influence, network characteristics, and entrepreneurial motivations on trust, reciprocity, and togetherness in a network of wildlife tourism microentrepreneurs. Thirty-seven wildlife tourism microentrepreneurs from North Carolina’s Pamlico Sound Region were recruited for in-person structured interviews. Data were analyzed using social network analysis, specifically a series of linear network autocorrelation models in conjunction with supportive qualitative assessment. Microentrepreneurs expressing a high level of trust were connected with microentrepreneurs expressing a low level of trust in their peers. Conversely, microentrepreneurs with strong feelings of reciprocity were connected with microentrepreneurs having similar feelings. These findings illustrate that the presence of equally reciprocal relationships is not an indication of equally trusting relationships. The findings also suggest that higher numbers of business ties tend to diminish the levels of trust, reciprocity, and togetherness toward connected peers.

Estimating the Attributable Disease Burden and Effects of Interhospital Patient Sharing on Clostridium difficile Infections.

To estimate the burden of Clostridium difficile infections (CDIs) due to interfacility patient sharing at regional and hospital levels. Retrospective observational study. We used data from the Healthcare Cost and Utilization Project California State Inpatient Database (2005-2011) to identify 26,878,498 admissions and 532,925 patient transfers. We constructed a weighted, directed network among the hospitals by defining an edge between 2 hospitals to be the monthly average number of patients discharged from one hospital and admitted to another on the same day. We then used a network autocorrelation model to study the effect of the patient sharing network on the monthly average number of CDI cases per hospital, and we estimated the proportion of CDI cases attributable to the network. We found that 13% (95% confidence interval [CI], 7.6%-18%) of CDI cases were due to diffusion through the patient-sharing network. The network autocorrelation parameter was estimated at 5.0 (95% CI, 3.0-6.9). An increase in the number of patients transferred into and/or an increased CDI rate at the hospitals from which those patients originated led to an increase in the number of CDIs in the receiving hospital. A minority but substantial burden of CDI infections are attributable to hospital transfers. A hospital's infection control may thus be nontrivially influenced by its neighboring hospitals. This work adds to the growing body of evidence that intervention strategies designed to minimize HAIs should be done at the regional rather than local level.

Relationships between social network characteristics, alcohol use, and alcohol-related consequences in a large network of first-year college students: How do peer drinking norms fit in?

A burgeoning area of research is using social network analysis to investigate college students' substance use behaviors. However, little research has incorporated students' perceived peer drinking norms into these analyses. The present study investigated the association between social network characteristics, alcohol use, and alcohol-related consequences among first-year college students (N = 1,342; 81% of the first-year class) at one university. The moderating role of descriptive norms was also examined. Network characteristics and descriptive norms were derived from participants' nominations of up to 10 other students who were important to them; individual network characteristics included popularity (indegree), network expansiveness (outdegree), relationship reciprocity, and network density. Descriptive norms were defined as participants' average perceived binge drinking frequency among their nominated peers. Network autocorrelation models revealed that indegree and descriptive norms were positively associated with participants' average number of drinks per week, binge drinking frequency, and alcohol-related consequences. Indegree and outdegree interacted with descriptive norms, such that when participants perceived less frequent binge drinking among their peers, outdegree was associated with less alcohol consumption but not consequences. When participants perceived more frequent binge drinking among their peers, indegree and outdegree were associated with more alcohol consumption but not consequences. The present results suggest that being popular and believing that heavy episodic drinking is normative among one's peers are associated with greater alcohol risk. Further, alcohol risks associated with nominating more peers may be enhanced or lessened depending on students' peer drinking norms. Implications for future research and interventions are discussed. (PsycINFO Database Record (c) 2018 APA, all rights reserved).

Relationship between network clustering in a therapeutic community and reincarceration following discharge

Recent qualitative work on Therapeutic Communities (TCs) suggests that they help residents change by creating an environment that is simultaneously challenging and supportive. There is evidence that social networks that feature numerous closed triads are both more supportive and more likely to influence individual behavior. This implies that TC residents whose peer social networks include more closed triads should have improved outcomes. The social network in this study consists of the affirmations exchanged between 1312 men who resided at a 90 bed TC in a Midwestern state over a period of eight years and includes a total of 34,667 weighted edges. The network was analyzed using the Temporal Network Autocorrelation Model (TNAM) based semiparametric Cox model, thereby using a statistical methodology that accounts for dependence between individuals in the network. Participants whose social networks of TC peers included a higher percentage of closed triads were at a decreased hazard of reincarceration following termination when controlling for age, length of stay and the number of peers who eventually graduated who affirmed the residents. These results support the longstanding TC contention that the community as a whole is the method of clinical treatment. Further quantitative research into TC processes and outcomes should ideally include social network surveys and statistics in order to avoid biases associated with violations of statistical independence assumptions.

A Network Autocorrelation Model to Predict Repeat Purchases in Multi-Relational Social Networks: Evidence from Online Games

It is well established in the literature that customer retention is at least as important as customer acquisition, especially in the freemium-based virtual economy, in which individuals are connected by multi-relational social networks, such as the “friendships” and “guilds” (virtual teams) functions in the context of online games. Yet, we do not know how these multi-relational social network effects simultaneously affect repeat purchases. In this study, we seek to examine the interdependent nature of repeat purchases of online game players embedded in multi-relational social networks. We examine two types of social network effects in the “friendship” function: a reinforcement (positive) effect and an attenuation (negative) effect. We also examine the herding effect (behavioral conformity within a virtual team) in the “guild” function. We develop a new hierarchical Bayesian model that supports a response variable for repeat purchases and that follows a Poisson distribution, and we simultaneously include both the two social network effects (reinforcement and attenuation) from the “friendship” function and also the moderating role of the herding effect from the “guild” function. Our study shows that both the reinforcement and attenuation social network effects have a significant effect on repeat purchases; however, the direction of the behavioral correlation of these two social network effects is starkly distinct: the reinforcement effect leads to repeat purchases among direct friends; in contrast, the attenuation effect leads to a divergence of repeat purchases among players who are indirect friends but share common friends. Moreover, we find that the size of the reinforcement effect is attenuated when the herding effect is accounted for. Besides theoretically developing and empirically validating a new method to predict repeat purchases in multi-relational social networks, this study has managerial implications for online game providers to develop effective targeting strategies to encourage players to engage in repeat purchases by leveraging multi-relational social network effects. Broader theoretical and practical implications for encouraging repeat purchases by leveraging social network effects in other online platforms are also discussed.

The peer effect on pain tolerance.

Background and aims Twin studies have found that approximately half of the variance in pain tolerance can be explained by genetic factors, while shared family environment has a negligible effect. Hence, a large proportion of the variance in pain tolerance is explained by the (non-shared) unique environment. The social environment beyond the family is a potential candidate for explaining some of the variance in pain tolerance. Numerous individual traits have previously shown to be associated with friendship ties. In this study, we investigate whether pain tolerance is associated with friendship ties. Methods We study the friendship effect on pain tolerance by considering data from the Tromsø Study: Fit Futures I, which contains pain tolerance measurements and social network information for adolescents attending first year of upper secondary school in the Tromsø area in Northern Norway. Pain tolerance was measured with the cold-pressor test (primary outcome), contact heat and pressure algometry. We analyse the data by using statistical methods from social network analysis. Specifically, we compute pairwise correlations in pain tolerance among friends. We also fit network autocorrelation models to the data, where the pain tolerance of an individual is explained by (among other factors) the average pain tolerance of the individual's friends. Results We find a significant and positive relationship between the pain tolerance of an individual and the pain tolerance of their friends. The estimated effect is that for every 1 s increase in friends' average cold-pressor tolerance time, the expected cold-pressor pain tolerance of the individual increases by 0.21 s (p-value: 0.0049, sample size n=997). This estimated effect is controlled for sex. The friendship effect remains significant when controlling for potential confounders such as lifestyle factors and test sequence among the students. Further investigating the role of sex on this friendship effect, we only find a significant peer effect of male friends on males, while there is no significant effect of friends' average pain tolerance on females in stratified analyses. Similar, but somewhat lower estimates were obtained for the other pain modalities. Conclusions We find a positive and significant peer effect in pain tolerance. Hence, there is a significant tendency for students to be friends with others with similar pain tolerance. Sex-stratified analyses show that the only significant effect is the effect of male friends on males. Implications Two different processes can explain the friendship effect in pain tolerance, selection and social transmission. Individuals might select friends directly due to similarity in pain tolerance, or indirectly through similarity in other confounding variables that affect pain tolerance. Alternatively, there is an influence effect among friends either directly in pain tolerance, or indirectly through other variables that affect pain tolerance. If there is indeed a social influence effect in pain tolerance, then the social environment can account for some of the unique environmental variance in pain tolerance. If so, it is possible to therapeutically affect pain tolerance through alteration of the social environment.

Analyzing big data in social media: Text and network analyses of an eating disorder forum.

Social media plays an important role in everyday life of young people. Numerous studies claim negative effects of social media and media in general on eating disorder risk factors. Despite the availability of big data, only few studies have exploited the possibilities so far in the field of eating disorders. Methods for data extraction, computerized content analysis, and network analysis will be introduced. Strategies and methods will be exemplified for an ad-hoc dataset of 4,247 posts and 34,118 comments by 3,029 users of the proed forum on Reddit. Text analysis with latent Dirichlet allocation identified nine topics related to social support and eating disorder specific content. Social network analysis describes the overall communication patterns, and could identify community structures and most influential users. A linear network autocorrelation model was applied to estimate associations in language among network neighbors. The supplement contains R code for data extraction and analyses. This paper provides an introduction to investigating social media data, and will hopefully stimulate big data social media research in eating disorders. When applied in real-time, the methods presented in this manuscript could contribute to improving the safety of ED-related online communication.

Poor mental health, peer drinking norms, and alcohol risk in a social network of first-year college students

ObjectiveCollege students with anxiety and depressive symptomatology face escalated risk for alcohol-related negative consequences. While it is well-established that normative perceptions of proximal peers' drinking behaviors influence students' own drinking behaviors, it is not clear how mental health status impacts this association. In the current study, we examined cross-sectional relationships between anxiety and depressed mood, perceived drinking behaviors and attitudes of important peers, and past month alcohol consumption and related problems in a first-semester college student social network. MethodParticipants (N = 1254, 55% female, 47% non-Hispanic White) were first-year students residing on campus at a single university who completed a web-based survey assessing alcohol use, mental health, and social connections among first-year student peers. Network autocorrelation models were used to examine the independent and interactive associations between mental health and perceptions of close peers' drinking on drinking outcomes, controlling for important variables. ResultsMental health interacted with perceptions to predict past-month drinking outcomes, such that higher anxiety and higher perceptions that peers drink heavily was associated with more drinks consumed and consequences, and higher depression and perceptions was associated with more drinks consumed, heavy drinking frequency, and consequences. Attitudes that peers approve of heavy drinking were associated with more drinks consumed and heavy drinking frequency among students with lower (vs. higher) depressed mood. ConclusionsThis study provides strong evidence that perceiving that close peers drink heavily is particularly risk-enhancing for anxious and depressed college students, and offers implications about alcohol intervention targeted at these subgroups.

An event- and network-level analysis of college students' maximum drinking day

BackgroundHeavy episodic drinking is common among college students and remains a serious public health issue. Previous event-level research among college students has examined behaviors and individual-level characteristics that drive consumption and related consequences but often ignores the social network of people with whom these heavy drinking episodes occur. The main aim of the current study was to investigate the network of social connections between drinkers on their heaviest drinking occasions. MethodsSociocentric network methods were used to collect information from individuals in the first-year class (N=1342) at one university. Past-month drinkers (N=972) reported on the characteristics of their heaviest drinking occasion in the past month and indicated who else among their network connections was present during this occasion. ResultsAverage max drinking day indegree, or the total number of times a participant was nominated as being present on another students' heaviest drinking occasion, was 2.50 (SD=2.05). Network autocorrelation models indicated that max drinking day indegree (e.g., popularity on heaviest drinking occassions) and peers' number of drinks on their own maximum drinking occasions were significantly associated with participant maximum number of drinks, after controlling for demographic variables, pregaming, and global network indegree (e.g., popularity in the entire first-year class). ConclusionBeing present at other peers' heaviest drinking occasions is associated with greater drinking quantities on one's own heaviest drinking occasion. These findings suggest the potential for interventions that target peer influences within close social networks of drinkers.

Resistance to peer influence moderates the relationship between perceived (but not actual) peer norms and binge drinking in a college student social network

IntroductionAdolescent and young adult binge drinking is strongly associated with perceived social norms and the drinking behavior that occurs within peer networks. The extent to which an individual is influenced by the behavior of others may depend upon that individual's resistance to peer influence (RPI). MethodsStudents in their first semester of college (N=1323; 54.7% female, 57% White, 15.1% Hispanic) reported on their own binge drinking, and the perceived binge drinking of up to 10 important peers in the first-year class. Using network autocorrelation models, we investigated cross-sectional relationships between participant's binge drinking frequency and the perceived and actual binge drinking frequency of important peers. We then tested the moderating role of RPI, expecting that greater RPI would weaken the relationship between perceived and actual peer binge drinking on participant binge drinking. ResultsPerceived and actual peer binge drinking were statistically significant predictors of participant binge drinking frequency in the past month, after controlling for covariates. RPI significantly moderated the association between perceptions of peer binge drinking and participant's own binge drinking; this association was weaker among participants with higher RPI compared to those with lower RPI. RPI did not interact with the actual binge drinking behavior of network peers. ConclusionsRPI may function to protect individuals from the effect of their perceptions about the binge drinking of peers, but not from the effect of the actual binge drinking of peers.

Network Autocorrelation Modeling: A Bayes Factor Approach for Testing (Multiple) Precise and Interval Hypotheses

Currently available (classical) testing procedures for the network autocorrelation can only be used for falsifying a precise null hypothesis of no network effect. Classical methods can be neither used for quantifying evidence for the null nor for testing multiple hypotheses simultaneously. This article presents flexible Bayes factor testing procedures that do not have these limitations. We propose Bayes factors based on an empirical and a uniform prior for the network effect, respectively, first. Next, we develop a fractional Bayes factor where a default prior is automatically constructed. Simulation results suggest that the first two Bayes factors show superior performance and are the Bayes factors we recommend. We apply the recommended Bayes factors to three data sets from the literature and compare the results to those coming from classical analyses using p values. R code for efficient computation of the Bayes factors is provided.

Politics of the precautionary principle: assessing actors’ preferences in water protection policy

This paper analyzes the prospects for introducing the precautionary principle in water protection policy. In situations where a problem enters the political agenda and scientific uncertainties remain about causes or effects, political actors can justify state intervention based on the precautionary principle. It allows for public action even if risks related to the problem remain unclear. While the precautionary principle is widely applied in health and environmental policy fields all over the world, the mechanisms leading to its adoption are not fully understood. To close this gap, the paper investigates decision-makers preferences for the precautionary principle and further asks: Which factors promote political actors’ preferences for precautionary policy measures? In order to answer this question we study the case of emerging micropollutants—a water quality issue that recently entered political agendas, where many uncertainties remain about sources and effects. We rely on data gathered through a standardized survey among the political elite in Switzerland, which represents one of the first countries that adopted policy measures to reduce micropollutants in water bodies, despite the uncertainties that remain. Results analyzed through a temporal network autocorrelation model reveal that actors embedded in collaborative governance arrangements have the tendency to prefer precautionary action. Certain aspects of policy design, such as problem prioritization and target group membership, also impact the prospects for introducing the precautionary principle.

Heterogeneous susceptibilities in social influence models

Network autocorrelation models are widely used to evaluate the impact of social influence on some variable of interest. This is a large class of models that parsimoniously accounts for how one's neighbors influence one's own behaviors or opinions by incorporating the network adjacency matrix into the joint distribution of the data. These models assume homogeneous susceptibility to social influence, however, which may be a strong assumption in many contexts. This paper proposes a hierarchical model that allows the influence parameter to be a function of individual attributes and/or of local network topological features. We derive an approximation of the posterior distribution in a general framework that is applicable to the Durbin, network effects, network disturbances, or network moving average autocorrelation models. The proposed approach can also be applied to investigating determinants of social influence in the context of egocentric network data. We apply our method to a data set collected via mobile phones in which we determine the effect of social influence on physical activity levels, as well as classroom data in which we investigate peer influence on student defiance. With this last data set, we also investigate the performance of the proposed egocentric network model.

The Power of Peers: How Transnational Advocacy Networks Shape NGO Strategies on Climate Change

What explains variation in the tactical choices of non-governmental organizations (NGOs)? This article uses network autocorrelation models to establish how the tactical choices of climate change NGOs are shaped by their embeddedness in transnational advocacy networks. Specifically, it finds that NGOs are more likely to adopt protest tactics when adjacent organizations – those with whom they have direct ties – have already done so. The choices of equivalent organizations – those that occupy similar relational roles in the network – do not appear to be influential. Qualitative evidence also shows that NGOs are affected by relational pressure from their peers, which alters their perception of costs and benefits. These findings enhance understanding of how networks influence actors’ behavior and offer insights into the relational processes that generate protest in global politics.

Network autocorrelation models with egocentric data

Network autocorrelation models have been widely used for decades to model the joint distribution of the attributes of a network's actors. This class of models can estimate both the effect of individual characteristics as well as the network effect, or social influence, on some actor attribute of interest. Collecting data on the entire network, however, is very often infeasible or impossible if the network boundary is unknown or difficult to define. Obtaining egocentric network data overcomes these obstacles, but as of yet there has been no clear way to model this type of data and still appropriately capture the network effect on the actor attributes in a way that is compatible with a joint distribution on the full network data. This paper adapts the class of network autocorrelation models to handle egocentric data. The proposed methods thus incorporate the complex dependence structure of the data induced by the network rather than simply using ad hoc measures of the egos’ networks to model the mean structure, and can estimate the network effect on the actor attribute of interest. The vast quantities of unknown information about the network can be succinctly represented in such a way that only depends on the number of alters in the egocentric network data and not on the total number of actors in the network. Estimation is done within a Bayesian framework. A simulation study is performed to evaluate the estimation performance, and an egocentric data set is analyzed where the aim is to determine if there is a network effect on environmental mastery, an important aspect of psychological well-being.