Media, social capital, and health information diffusion: A multilevel approach

In this dissertation, I aim to clarify the mechanisms through which social capital and media exposure interact and jointly influence health-related outcomes. Moreover, I explore how the media's routine transmission of health information and media health campaigns exert their effects through audience's social capital. First, using the National Survey of Parents and Youth (NSPY) dataset, I examined the interactive effects of parents' campaign exposure and antidrug-specific social capital at both individual- and geographically-aggregated levels on parents' drug-related talk with their children and other antidrug ads-related talk. I found main effects of parents' campaign exposure and parents' participation in antidrug-specific community activities on their talk about drugs with their children and other antidrug ads-related talk. More interestingly, there was a negative interactive effect between campaign exposure and participation in antidrug-specific community activities on these two talking behaviors, which supports the substitution model. In contrast, there was neither a contextual effect of aggregate-level antidrug-specific social capital nor a cross-level interaction involving aggregate-level social capital. Extending these results to natural coverage effects, I tested whether media exposure and social capital at both individual- and community levels interact and jointly influence healthy lifestyle behaviors using the Annenberg National Health Communication Survey (ANHCS) dataset. Then, I tested whether these interactive effects are mediated by interpersonal health communication. I found that the associations between media use and healthy lifestyle behaviors were stronger among respondents with high individual-level social capital, which supports the substitution model. Moreover, interpersonal health communication was found to mediate these interactive effects. Unlike my hypothesis, however, community-level social capital (i.e., state- and county-level) did not bear any significant relationship with healthy lifestyle behaviors. Neither were there interactive effects of community-level social capital and individual-level explanatory variables on health-related behaviors. Overall, these results suggest that media exposure and individual-level social capital serve as substitutive health information sources rather than complementary ones.