Norms within Networks

Why do some mothers/caregivers immunize their children while others don't? What role do social norms within networks play in the inequitable uptake of childhood immunizations in northern Nigeria? A large body of research has been devoted to investigating the determinants of immunization acceptance in developing countries. Up to this point, however, most of this research has focused on individual-level explanations - such as maternal age, maternal education, household wealth, maternal knowledge about the benefits of immunizations, and previous health-seeking experiences - to account for variance in immunization use. Scant attention has been paid to the ways that social factors influence this outcome. This dissertation aims to fill this critical gap in the literature by investigating the impact of social network characteristics and norms on immunization use among a sample of mothers/caregivers in rural, northern Nigeria. More specifically, this study uses a formal social network analysis to examine how the structure of mothers/caregivers' relationships with opinion leaders and peers impact their decisions to immunize their children. This research makes a unique contribution to the literature by putting the social networks of mothers/caregivers at the center of its analyses, and examining how social relationships impact immunization decisions in a setting with some of the worst childhood immunization coverage rates in the world. This study draws on theories of social networks and social norms in order to explain the variance in childhood immunization coverage rates among a sample of mothers/caregivers in Zamfara State, northern Nigeria. The aims of this research are to 1) test whether injunctive and descriptive norms towards immunizations are linked to mothers/caregivers' immunization decisions; 2) examine whether the theorized constructs of social control and social learning underlie injunctive norms and descriptive norms around immunizations; 3) assess the degree to which injunctive and descriptive norms influence mothers/caregivers immunization behaviors; and 4) test whether the structural properties of closeness and frequency of communication about immunizations moderate the influence of injunctive and descriptive norms on mothers/caregivers' immunization use. This study employed a refined ego-centric network design that uses quantitative and geographic data collected in October-November 2011, from 550 mothers and 127 of their opinion leaders living across 22 paired villages in one local government area in Zamfara State, northern Nigeria. Validity tests were conducted to assess the accuracy of the injunctive norms measures and a latent variable model was utilized to generate more valid indicators of them. Mixed effect models, adjusting for clustering at the compound and village levels, were used to test all study hypotheses. Sensitivity analyses were conducted to confirm study results. The results indicate that injunctive and descriptive norms independently predict mothers/caregivers' immunization use in their peer networks, but not in their opinion leader networks. The results also confirm that social control underlies injunctive norms and that social learning underlies descriptive norms. Injunctive norms are more influential than descriptive norms in mothers/caregivers' peer networks, indicating that social control is operating as a stronger force on mothers/caregivers' immunization decisions than reflective observations of their peers' immunization behaviors. The results also show that that the influence of injunctive norms in opinion leader networks on mothers/caregivers' immunization use partly depends on whether or not they have a close relationship with their opinion leaders. Frequency of communication between mothers/caregivers and their opinion leaders and peers strengthens the influence of descriptive norms on immunization use. This indicates that both communication and observed immunization practices are necessary conditions for normative influences to operate. This study provides additional evidence to support the claim that health outcomes depend not just on individuals' own beliefs and actions, but also on the normative beliefs and actions of the people around them. To improve the effectiveness of interventions aimed at improving immunization acceptance, researchers and programmers alike should treat mothers/caregivers not only as individuals, but as members of meaningful and influential interpersonal networks. By embracing this approach, we will improve our understanding of the determinants of routine immunization use in developing countries, while accelerating the impact of immunization programs on child survival outcomes in contexts, like northern Nigeria, with some of the worst immunization coverage rates in the world.