The structure and function of antagonistic ties in village social networks.

Negative or antagonistic relationships are common in human social networks, but they are less often studied than positive or friendly relationships. The existence of a capacity to have and to track antagonistic ties raises the possibility that they may serve a useful function in human groups. Here, we analyze empirical data gathered from 24,770 and 22,513 individuals in 176 rural villages in Honduras in two survey waves 2.5 y apart in order to evaluate the possible relevance of antagonistic relationships for broader network phenomena. We find that the small-world effect is more significant in a positive world with negative ties compared to an otherwise similar hypothetical positive world without them. Additionally, we observe that nodes with more negative ties tend to be located near network bridges, with lower clustering coefficients, higher betweenness centralities, and shorter average distances to other nodes in the network. Positive connections tend to have a more localized distribution, while negative connections are more globally dispersed within the networks. Analysis of the possible impact of such negative ties on dynamic processes reveals that, remarkably, negative connections can facilitate the dissemination of information (including novel information experimentally introduced into these villages) to the same degree as positive connections, and that they can also play a role in mitigating idea polarization within village networks. Antagonistic ties hold considerable importance in shaping the structure and function of social networks.

Population flow drives spatio-temporal distribution of COVID-19 in China.

Sudden, large-scale and diffuse human migration can amplify localized outbreaks of disease into widespread epidemics1-4. Rapid and accurate tracking of aggregate population flows may therefore be epidemiologically informative. Here we use 11,478,484 counts of mobile phone data from individuals leaving or transiting through the prefecture of Wuhan between 1 January and 24 January 2020 as they moved to 296 prefectures throughout mainland China. First, we document the efficacy of quarantine in ceasing movement. Second, we show that the distribution of population outflow from Wuhan accurately predicts the relative frequency and geographical distribution of infections with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) until 19 February 2020, across mainland China. Third, we develop a spatio-temporal 'risk source' model that leverages population flow data (which operationalize the risk that emanates from epidemic epicentres) not only to forecast the distribution of confirmed cases, but also to identify regions that have a high risk of transmission at an early stage. Fourth, we use this risk source model to statistically derive the geographical spread of COVID-19 and the growth pattern based on the population outflow from Wuhan; the model yields a benchmark trend and an index for assessing the risk of community transmission of COVID-19 over time for different locations. This approach can be used by policy-makers in any nation with available data to make rapid and accurate risk assessments and to plan the allocation of limited resources ahead of ongoing outbreaks.

Emergence and collapse of reciprocity in semiautomatic driving coordination experiments with humans.

Forms of both simple and complex machine intelligence are increasingly acting within human groups in order to affect collective outcomes. Considering the nature of collective action problems, however, such involvement could paradoxically and unintentionally suppress existing beneficial social norms in humans, such as those involving cooperation. Here, we test theoretical predictions about such an effect using a unique cyber-physical lab experiment where online participants (N = 300 in 150 dyads) drive robotic vehicles remotely in a coordination game. We show that autobraking assistance increases human altruism, such as giving way to others, and that communication helps people to make mutual concessions. On the other hand, autosteering assistance completely inhibits the emergence of reciprocity between people in favor of self-interest maximization. The negative social repercussions persist even after the assistance system is deactivated. Furthermore, adding communication capabilities does not relieve this inhibition of reciprocity because people rarely communicate in the presence of autosteering assistance. Our findings suggest that active safety assistance (a form of simple AI support) can alter the dynamics of social coordination between people, including by affecting the trade-off between individual safety and social reciprocity. The difference between autobraking and autosteering assistance appears to relate to whether the assistive technology supports or replaces human agency in social coordination dilemmas. Humans have developed norms of reciprocity to address collective challenges, but such tacit understandings could break down in situations where machine intelligence is involved in human decision-making without having any normative commitments.

Machine behaviour.

Machines powered by artificial intelligence increasingly mediate our social, cultural, economic and political interactions. Understanding the behaviour of artificial intelligence systems is essential to our ability to control their actions, reap their benefits and minimize their harms. Here we argue that this necessitates a broad scientific research agenda to study machine behaviour that incorporates and expands upon the discipline of computer science and includes insights from across the sciences. We first outline a set of questions that are fundamental to this emerging field and then explore the technical, legal and institutional constraints on the study of machine behaviour.

Algorithms for seeding social networks can enhance the adoption of a public health intervention in urban India.

Targeting structurally influential individuals within social networks can enhance adoption of health interventions within populations. We tested the effectiveness of two algorithms to improve social contagion that do not require knowledge of the whole network structure. We mapped the social interactions of 2,491 women in 50 residential buildings (chawls) in Mumbai, India. The buildings, which are social units, were randomized to (1) targeting 20% of the women at random, (2) targeting friends of such randomly chosen women, (3) targeting pairs of people composed of randomly chosen women and a friend, or (4) no targeting. Both targeting algorithms, friendship nomination and pair targeting, enhanced adoption of a public health intervention related to the use of iron-fortified salt for anemia. In particular, the targeting of pairs of friends, which is relatively easily implementable in field settings, enhanced adoption of novel practices through both social influence and social reinforcement.

A Prospective Sociocentric Study of 2 Entire Traditional Korean Villages: The Korean Social Life, Health, and Aging Project (KSHAP).

The Korean Social Life, Health, and Aging Project (KSHAP) was a multidisciplinary prospective study conducted in South Korea that measured various health biomarkers from blood, hair, and brain magnetic resonance imaging, and we examined their associations with sociocentric (global) social network data of older adults in 2 entire villages (or cohorts). Cohort K included participants aged 60 years or older, and cohort L included participants aged 65 years or older. We performed a baseline survey involving 814 of the 860 individuals (94.7% response rate) in cohort K in 2012 and 947 of the 1,043 individuals (90.8% response rate) in cohort L in 2017. We gathered longitudinal data for 5 waves in cohort K from 2011 to 2019 and 2 waves in cohort L from 2017 to 2022. Here, we describe for the first time the follow-up design of the KSHAP, the changes in social networks, and various biomarkers over a number of years. The data for cohort K are publicly available via the Korean Social Science Data Archive as well as the project website, and the data for cohort L will be shared soon.

Locally noisy autonomous agents improve global human coordination in network experiments.

Coordination in groups faces a sub-optimization problem and theory suggests that some randomness may help to achieve global optima. Here we performed experiments involving a networked colour coordination game in which groups of humans interacted with autonomous software agents (known as bots). Subjects (n = 4,000) were embedded in networks (n = 230) of 20 nodes, to which we sometimes added 3 bots. The bots were programmed with varying levels of behavioural randomness and different geodesic locations. We show that bots acting with small levels of random noise and placed in central locations meaningfully improve the collective performance of human groups, accelerating the median solution time by 55.6%. This is especially the case when the coordination problem is hard. Behavioural randomness worked not only by making the task of humans to whom the bots were connected easier, but also by affecting the gameplay of the humans among themselves and hence creating further cascades of benefit in global coordination in these heterogeneous systems.

Vulnerable robots positively shape human conversational dynamics in a human-robot team.

Social robots are becoming increasingly influential in shaping the behavior of humans with whom they interact. Here, we examine how the actions of a social robot can influence human-to-human communication, and not just robot-human communication, using groups of three humans and one robot playing 30 rounds of a collaborative game (n = 51 groups). We find that people in groups with a robot making vulnerable statements converse substantially more with each other, distribute their conversation somewhat more equally, and perceive their groups more positively compared to control groups with a robot that either makes neutral statements or no statements at the end of each round. Shifts in robot speech have the power not only to affect how people interact with robots, but also how people interact with each other, offering the prospect for modifying social interactions via the introduction of artificial agents into hybrid systems of humans and machines.

Social network correlates of free and purchased insecticide-treated bed nets in rural Uganda.

BACKGROUND: Malaria is a major cause of mortality and morbidity in Uganda. Despite Uganda's efforts to distribute bed nets, only half of households have achieved the World Health Organization (WHO) Universal Coverage Criteria (one bed net for every two household members). The role of peer influence on bed net ownership remains underexplored. Data on the complete social network of households were collected in a rural parish in southwestern Uganda to estimate the association between household bed net ownership and peer household bed net ownership. METHODS: Data on household sociodemographics, bed net ownership, and social networks were collected from all households across one parish in southwestern Uganda. Bed nets were categorized as either purchased or free. Purchased and free bed net ownership ratios were calculated based on the WHO Universal Coverage Criteria. Using network name generators and complete census of parish residents, the complete social network of households in the parish was generated. Linear regression models that account for network autocorrelation were fitted to estimate the association between households' bed net ownership ratios and bed net ownership ratios of network peer households, adjusting for sociodemographics and network centrality. RESULTS: One thousand seven hundred forty-seven respondents were interviewed, accounting for 716 households. The median number of peer households to which a household was directly connected was 7. Eighty-six percent of households owned at least one bed net, and 41% of households met the WHO Universal Coverage Criterion. The median bed net ownership ratios were 0.67 for all bed nets, 0.33 for free bed nets, and 0.20 for purchased bed nets. In adjusted multivariable models, purchased bed net ownership ratio was associated with average household wealth among peer households (b = 0.06, 95% CI 0.03, 0.10), but not associated with average purchased bed net ownership ratio of peer households. Free bed net ownership ratio was associated with the number of children under 5 (b = 0.08, 95% CI 0.05, 0.10) and average free bed net ownership ratios of peer households (b = 0.66, 95% CI 0.46, 0.85). CONCLUSIONS: Household bed net ownership was associated with bed net ownership of peer households for free bed nets, but not for purchased bed nets. The findings suggest that public health interventions may consider leveraging social networks as tools for dissemination, particularly for bed nets that are provided free of charge.

Characterizing super-spreaders using population-level weighted social networks in rural communities.

Sociocentric network maps of entire populations, when combined with data on the nature of constituent dyadic relationships, offer the dual promise of advancing understanding of the relevance of networks for disease transmission and of improving epidemic forecasts. Here, using detailed sociocentric data collected over 4 years in a population of 24 702 people in 176 villages in Honduras, along with diarrhoeal and respiratory disease prevalence, we create a social-network-powered transmission model and identify super-spreading nodes as well as the nodes most vulnerable to infection, using agent-based Monte Carlo network simulations. We predict the extent of outbreaks for communicable diseases based on detailed social interaction patterns. Evidence from three waves of population-level surveys of diarrhoeal and respiratory illness indicates a meaningful positive correlation with the computed super-spreading capability and relative vulnerability of individual nodes. Previous research has identified super-spreaders through retrospective contact tracing or simulated networks. By contrast, our simulations predict that a node's super-spreading capability and its vulnerability in real communities are significantly affected by their connections, the nature of the interaction across these connections, individual characteristics (e.g. age and sex) that affect a person's ability to disperse a pathogen, and also the intrinsic characteristics of the pathogen (e.g. infectious period and latency). This article is part of the theme issue 'Data science approach to infectious disease surveillance'.

Assortative mixing and resource inequality enhance collective welfare in sharing networks.

Resource sharing can impose an economic trade-off: One person acquiring resources may mean that another cannot. However, if individuals value the social process itself that is a feature of economic exchanges, socio-structural manipulations might improve collective welfare. Using a series of online experiments with 600 subjects arrayed into 40 groups, we explore the welfare impact of 2 network interventions. We manipulated the degree assortativity of the groups (who were engaged in resource sharing) while keeping the number of people and connections fixed. Distinctly, we also manipulated the distribution of sharable resources by basing endowments on network degree. We show that structural manipulation (implementing degree assortativity) can facilitate the reciprocity that is achievable in exchanges and consequently affect group-level satisfaction. We also show that individuals are more satisfied with exchanges when each node is unequally endowed with resources that are proportional to the number of potential recipients, which again facilitates reciprocity. Collective welfare in settings involving resource sharing can be enhanced without the need for extra resources.

Inequality and visibility of wealth in experimental social networks.

Humans prefer relatively equal distributions of resources, yet societies have varying degrees of economic inequality. To investigate some of the possible determinants and consequences of inequality, here we perform experiments involving a networked public goods game in which subjects interact and gain or lose wealth. Subjects (n = 1,462) were randomly assigned to have higher or lower initial endowments, and were embedded within social networks with three levels of economic inequality (Gini coefficient = 0.0, 0.2, and 0.4). In addition, we manipulated the visibility of the wealth of network neighbours. We show that wealth visibility facilitates the downstream consequences of initial inequality-in initially more unequal situations, wealth visibility leads to greater inequality than when wealth is invisible. This result reflects a heterogeneous response to visibility in richer versus poorer subjects. We also find that making wealth visible has adverse welfare consequences, yielding lower levels of overall cooperation, inter-connectedness, and wealth. High initial levels of economic inequality alone, however, have relatively few deleterious welfare effects.

Using Trellis software to enhance high-quality large-scale network data collection in the field.

Trellis is a mobile platform created by the Human Nature Lab at the Yale Institute for Network Science to collect high-quality, location-aware, off-line/online, multi-lingual, multi-relational social network and behavior data in hard-to-reach communities. Respondents use Trellis to identify their social contacts by name and photograph, a procedure especially useful in low-literacy populations or in contexts where names may be similar or confusing. We use social network data collected from 1,969 adult respondents in two villages in Kenya to demonstrate Trellis' ability to provide unprecedented metadata to monitor and report on the data collection process including artifactual variability based on surveyors, time of day, or location.

Lack of sexual behavior disclosure may distort STI testing outcomes.

BACKGROUND: Men who have sex with men (MSM) globally have a high burden of curable sexually transmitted infections (STIs). MSM do not frequently receive rectal STI testing because of several barriers, such as not being out (disclosure of sexual behavior). We evaluate whether Chinese MSM select an STI test (rectal vs urethral) appropriate for their sexual behavior (insertive and/or receptive), and the interactions with being out. METHODS: This was a secondary analysis of data from a cross sectional MSM survey conducted at a multisite randomized controlled trial (RCT) (December 2018 to January 2019) around uptake of gonorrhea and chlamydia testing among Chinese MSM (N = 431). We collected socio demographics, relevant medical and sexual history, and disclosure of sexual behavior (outness). We estimated the decision to test and test choice, and the extent to which disclosure plays a role in decision making. RESULTS: Among 431 MSM, mean age was 28 years (SD = 7.10) and 65% were out to someone. MSM who indicated versatile sexual behavior and were out to someone had a 26.8% (95%CI = 6.1, 47.5) increased likelihood for selecting the rectal test vs the ure thral test, compared to those versatile and not out. Versatile MSM out to their health provider outside of the study context had a 29.4% (95%CI = 6.3, 52.6) greater likelihood for selecting the rectal STI test vs the urethral test, compared to versatile MSM not out to their health provider. CONCLUSIONS: Sexual behavior and outness may affect gonorrhea and chlamydia testing provision. Apart from clinicians, community based efforts may reduce stigma based barriers to testing.

Association Between Degrees of Separation in Physician Networks and Surgeons' Use of Perioperative Breast Magnetic Resonance Imaging.

BACKGROUND: Perioperative magnetic resonance imaging (MRI) is frequently used in breast cancer despite unproven benefits. It is unclear whether surgeons' use of breast MRI is associated with the practices of other surgeons to whom they are connected through shared patients. METHODS: We conducted a retrospective study using Medicare data to identify physicians providing breast cancer care during 2007-2009 and grouped them into patient-sharing networks. Physician pairs were classified according to their "degree of separation" based on patient-sharing (eg, physician pairs that care for the same patients were separated by 1 degree; pairs that both share patients with another physician but not with each other were separated by 2 degrees). We assessed the association between the MRI use of a surgeon and the practice patterns of surgical colleagues by comparing MRI use in the observed networks with networks with randomly shuffled rates of MRI utilization. RESULTS: Of the 15,273 patients who underwent surgery during the study period, 28.8% received perioperative MRI. These patients received care from 1806 surgeons in 60 patient-sharing networks; 55.1% of surgeons used MRI. A surgeon was 24.5% more likely to use MRI if they were directly connected to a surgeon who used MRI. This effect decreased to 16.3% for pairs of surgeons separated by 2 degrees, and 0.8% at the third degree of separation. CONCLUSIONS: Surgeons' use of perioperative breast MRI is associated with the practice of surgeons connected to them through patient-sharing; the strength of this association attenuates as the degree of separation increases.

Social norms, misperceptions, and mosquito net use: a population-based, cross-sectional study in rural Uganda.

BACKGROUND: Mosquito net use is an essential part of malaria prevention. Although previous research has shown that many people sleep under a mosquito net in endemic areas, it is unknown whether people underestimate how common it is to sleep under a net every night. Furthermore, perceived social norms about whether most others sleep under a mosquito net every night may contribute to personally sleeping under a net, given decades of research showing that people often mimic others' behaviours. METHODS: Population-based data were collected from 1669 adults across eight villages in one rural parish in southwestern Uganda. Individuals' perception about whether most adults in their community sleep under a mosquito net every night was compared with whether daily mosquito net use was the actual norm in their community to identify the extent of norm misperception. The association between whether an individual perceived daily mosquito net use to be the norm and personal mosquito net use was assessed while adjusting for the ratio of nets:people in the household and other factors. RESULTS: Although the majority (65%) of participants reported sleeping under a mosquito net every night (and 75% did so among the 86% of people with at least one net), one-quarter of participants thought that most adults in their community did not sleep under a mosquito net every night. Another 8% were unsure how many nights per week most adults in their community sleep under a mosquito net. Participants who perceived that daily mosquito net use was the norm were 2.94 times more likely to report personally sleeping under a mosquito net every night (95% CI 2.09-4.14, p < 0.001) compared to participants who thought doing so was not normative, adjusting for other factors. CONCLUSIONS: Results suggest an opportunity for anti-malarial interventions to reduce misperceptions about mosquito net use norms and emphasize the commonness of daily mosquito net use in malaria-endemic regions. If people correctly perceive most others to sleep under a net every night, then they may personally do so when possible and support others to do so too.

Online social integration is associated with reduced mortality risk.

Social interactions increasingly take place online. Friendships and other offline social ties have been repeatedly associated with human longevity, but online interactions might have different properties. Here, we reference 12 million social media profiles against California Department of Public Health vital records and use longitudinal statistical models to assess whether social media use is associated with longer life. The results show that receiving requests to connect as friends online is associated with reduced mortality but initiating friendships is not. Additionally, online behaviors that indicate face-to-face social activity (like posting photos) are associated with reduced mortality, but online-only behaviors (like sending messages) have a nonlinear relationship, where moderate use is associated with the lowest mortality. These results suggest that online social integration is linked to lower risk for a wide variety of critical health problems. Although this is an associational study, it may be an important step in understanding how, on a global scale, online social networks might be adapted to improve modern populations' social and physical health.

Formation of raiding parties for intergroup violence is mediated by social network structure.

Intergroup violence is common among humans worldwide. To assess how within-group social dynamics contribute to risky, between-group conflict, we conducted a 3-y longitudinal study of the formation of raiding parties among the Nyangatom, a group of East African nomadic pastoralists currently engaged in small-scale warfare. We also mapped the social network structure of potential male raiders. Here, we show that the initiation of raids depends on the presence of specific leaders who tend to participate in many raids, to have more friends, and to occupy more central positions in the network. However, despite the different structural position of raid leaders, raid participants are recruited from the whole population, not just from the direct friends of leaders. An individual's decision to participate in a raid is strongly associated with the individual's social network position in relation to other participants. Moreover, nonleaders have a larger total impact on raid participation than leaders, despite leaders' greater connectivity. Thus, we find that leaders matter more for raid initiation than participant mobilization. Social networks may play a role in supporting risky collective action, amplify the emergence of raiding parties, and hence facilitate intergroup violence in small-scale societies.

Exposure, hazard, and survival analysis of diffusion on social networks.

Sociologists, economists, epidemiologists, and others recognize the importance of social networks in the diffusion of ideas and behaviors through human societies. To measure the flow of information on real-world networks, researchers often conduct comprehensive sociometric mapping of social links between individuals and then follow the spread of an "innovation" from reports of adoption or change in behavior over time. The innovation is introduced to a small number of individuals who may also be encouraged to spread it to their network contacts. In conjunction with the known social network, the pattern of adoptions gives researchers insight into the spread of the innovation in the population and factors associated with successful diffusion. Researchers have used widely varying statistical tools to estimate these quantities, and there is disagreement about how to analyze diffusion on fully observed networks. Here, we describe a framework for measuring features of diffusion processes on social networks using the epidemiological concepts of exposure and competing risks. Given a realization of a diffusion process on a fully observed network, we show that classical survival regression models can be adapted to estimate the rate of diffusion, and actor/edge attributes associated with successful transmission or adoption, while accounting for the topology of the social network. We illustrate these tools by applying them to a randomized network intervention trial conducted in Honduras to estimate the rate of adoption of 2 health-related interventions-multivitamins and chlorine bleach for water purification-and determine factors associated with successful social transmission.

Actual Versus Perceived HIV Testing Norms, and Personal HIV Testing Uptake: A Cross-Sectional, Population-Based Study in Rural Uganda.

HIV testing is an essential part of treatment and prevention. Using population-based data from 1664 adults across eight villages in rural Uganda, we assessed individuals' perception of the norm for HIV testing uptake in their village and compared it to the actual uptake norm. In addition, we examined how perception of the norm was associated with personal testing while adjusting for other factors. Although the majority of people had been tested for HIV across all villages, slightly more than half of men and women erroneously thought that the majority in their village had never been tested. They underestimated the prevalence of HIV testing uptake by 42 percentage points (s.d. = 17 percentage points), on average. Among men, perceiving that HIV testing was not normative was associated with never testing for HIV (AOR = 2.6; 95% CI 1.7-4.0, p < 0.001). Results suggest an opportunity for interventions to emphasize the commonness of HIV testing uptake.