Background: Understanding parental communication with adolescents about HIV prevention is critical in HIV-endemic settings. Little is known about social network factors and perceived social norms as drivers of these conversations.

Methods: We collected sociocentric network data from all adult residents of eight villages in rural southwestern Uganda (2023–2024; response rate >92%). Parents with children aged 12–17 reported whether they had any HIV prevention discussion with their adolescent(s) in the last three months and their perceptions about whether other village parents had done so and thought parents should do so. We fitted a multivariable logistic regression model to estimate associations between any discussion and alter behavior, perceived injunctive norm, total degree, and other factors.

Results: Among 525 parents, 308 (59%) had discussed HIV prevention. Despite 366 (70%) having at least one non-spouse alter who talked with their adolescent and 309 (59%) having at least one non-spouse alter who supported such discussions, 156 (30%) parents incorrectly believed that most village parents had not talked with their adolescent and 239(47%) incorrectly believed that most village parents were unsupportive of such discussions. Parents who thought most parents supported having these discussions were themselves more likely to have engaged in HIV prevention discussions with their adolescents (adjusted odds ratio = 3.53; 95% CI 2.51–4.96; p < 0.001). Total out degree was also associated with having a discussion while alter discussion behavior was not.

Conclusion: Correcting misperceptions of supportive norms through network-targeted interventions may bolster efforts HIV prevention among youth in rural Uganda.

Background: Medical male circumcision is associated with decreased risk of HIV/STI transmission. Voluntary uptake in adulthood remains short of target goals in many countries despite increases over the past two decades. No studies have explored social network factors as drivers of voluntary medical male circumcision in HIV-endemic settings.

Methods: We conducted a sociocentric network study of all residents living in eight villages in southwest Uganda (response rate > 95%). We identified network ties using five locally-adapted name generator questions. We asked all men if they were circumcised. Explanatory variables included having circumcised male alters and male alters who believed men should get circumcised, perceptions of village-level norms about male circumcision, and total outdegree. We fitted a multivariable logistic regression model to estimate associations between circumcision and these explanatory variables of interest, adjusting for sociodemographic characteristics.

Results: Among 791 men, 263 (33%) were circumcised at 20 years old on average, with 25-46% circumcised across villages; 404 (51%) had at least one circumcised alter. Personal circumcision status was associated with having 2 circumcised alters (adjusted odds ratio [aOR]=1.71; 95% CI 1.08, 2.71, p=0.020) and the belief that most men in the villages were circumcised (vs. incorrectly thinking few to none were circumcised) (aOR=2.98; 95% CI 1.53-5.81, p<.001). Other explanatory variables were not associated with circumcision status.

Conclusion: Future research will explore perceived norms about circumcision within ego networks and other measures of network centrality as drivers of uptake. This research suggests peer- and norms-based interventions may help motivate voluntary male circumcision uptake.

Wife-beating (WB) —men physically abusing their wives or cohabiting partners —is a traumatic forms of intimate partner abuse. WB has been recognized as a long-term global health concern(1)and one of the indicators of the Sustainable Development Goals(2). The consequences for victims, families, and communities are complex and call for multiple levels of analysis (3). Communities with norms that support WB not only have increased perpetration risks but also become adverse environments for victims, discouraging reporting, prevention, and timely treatment (4,5). Victims of WB face numerous and diverse repercussions (ranging from detrimental mental health conditions, chronic diseases, and substance misuse, as well as higher risks of re-victimization and food insecurity)(6,7), including spillover effects on their children’s development and quality of life.(6,8) These unfortunate realities are prevalent in resource-limited settings and Low- and Middle-Income Countries (LMICs). Past research in these high-risk areas remains limited and has overwhelmingly centered on individual factors rather than broader social influences, such as social networks (4,8,9). Nevertheless, prior network studies of WB (4,10) have shown that WB support is shaped by the friends and family’s perceptions of WB. Between 2015 and 2019, we collected sociocentric data (including friendship, kin, and antagonistic ties) from 176 isolated villages in western Honduras through face-to-face interviews. WB attitudes were assessed via surveys. We estimated three mixed-effects logistic regression models by gender to assess individual, network, and village-level factors associated with WB support (n==9,891 women and n=5,323 men), measured by asking respondents' agreement on whether a husband is justified in hitting his wife for given situations (leaving without telling, neglecting children, arguing, burning food, refusing sex). Our results show that older age, higher education, and having a partner were protective factors, lowering the risk of WB support for men and women. Social networks were significantly associated with WB Support: having more antagonistic increased WB support, while more friendship ties had a protective effect (for both men and women). Family influence varied by gender—partnered women were more likely to accept WB if their father-in-law or siblings endorsed it. The WB endorsement of unpartnered women and men was associated to their father’s WB beliefs. Our study highlights the differential role of signed networks and family ties in shaping WB attitudes and beliefs. Our results emphasize the need to consider these cultural dynamics when designing studies or interventions to reduce physical partner violence in LMIC rural settings.

We use sociological social psychology to compare the potential of two forms of dyadic exchange studied in the social exchange literature, reciprocal and negotiated exchange, in their ability to produce trust and cooperation outside of the exchange dyad. Reciprocal exchange has been shown to produce stronger feelings of solidarity and trust in a dyad (Molm, Collett, and Schaefer 2007; Molm, Melamed, and Whitham 2013; Molm, Schaefer, and Collett 2009; Molm, Whitham, and Melamed 2012; Molm, Schaefer, and Collett 2007), in part because networks of reciprocal exchange are more likely to be perceived by participants as groups to which they belong (Savage and Sommer 2016). While this can reliably produce trust and cooperation within a dyad, it is unclear what effect this has on cooperation towards those outside of that dyad, or those not involved in exchange. There is little evidence that the trust and solidarity cultivated in reciprocal exchange persists into new forms of exchange, much less into new relations (Cheshire, Gerbasi, and Cook 2010). Even worse, the production of feelings of group belonging can actively harm extra-dyadic relations, as the affective attachments that support in- and out-group boundaries can be a barrier to cooperation and solidarity outside the perceived group (Aksoy 2020; Balliet, Wu, and De Dreu 2014; Chen and Li 2009; Tarrant et al. 2012; Triplett 2012). However, it could also be the case that perceptions of interpersonal conflict that are adaptive for negotiated exchange (Kuwabara 2011) are carried into other interactions in a way that hampers trust and cooperation (Peysakhovich and Rand 2016).

Using a behavioral experiment, we compare the potential of negotiated and reciprocal exchange in their potential to produce cooperation towards people outside of the exchange dyad, what will be called extra-dyadic cooperation. Participants will be randomly assigned to engage in either reciprocal or negotiated exchange with a fixed partner. After performing social exchanges, participants will then be randomly paired with a new partner and play a one-shot prisoner's dilemma, an experimental game where players have the opportunity to cooperate and maximize their mutual benefit, despite the incentive to maximize their own individual gains at the expense of their partner, i.e. to defect in the game. Unlike social exchange, the prisoner's dilemma is a situation in which, regardless of one's beliefs about the other player's likely actions, it is always better from a selfish point of view to not benefit the other player, making this a more stringent test of which exchange form can more reliably produce trust and cooperation beyond the dyad.

Sociologists have recognized the power of reciprocity in social processes for decades, with Alvin Goulder in 1960, for example, proposing a “norm of reciprocity.” According to this norm, people are expected to respond in kind to actions directed towards them by another, typically referring to positive, prosocial actions. In recognition of the foundational role of reciprocity in social processes, statistical analyses of social networks routinely control for this process. Even though the effect of reciprocity is sizeable in almost all social networks of interest, the overabundance of mutual dyads is continuously taken-for-granted. The purpose of this paper is to examine systematically the degree to which the presence of reciprocity shapes social network configurations. We accomplish this by estimating Exponential Random Graph Models (ERGMs) on a sample of over 40 social networks. The networks in our sample represent diverse genres of social interaction (e.g., friendship among adolescents, advice seeking between coworkers, cyberbullying on social media) and include relationships defined by both amity and conflict. Then, we rely on our ERGM estimates to simulate a series of empirically-informed networks and quantify the extent to which local tendencies toward reciprocity inform the broader macro-level structures of social networks (e.g., centralization, clustering, connectedness). Our results suggest that the social effects of reciprocity can extend beyond the immediate dyad to influence broader, network-level topologies that often improve population-level outcomes.