Effects of social interactions on individual AIDS prevention attitudes and behaviors in rural Malawi

This dissertation examines how social interactions affect AIDS-prevention attitudes and behaviors in rural Malawi. Analyses of AIDS conversation networks and attendance at local social events aim to determine (1) whether informal conversations about HIV/AIDS engender positive attitudes towards condoms and marital fidelity (and, if so, which characteristics of conversation networks matter most), and (2) whether participation in social events plays a significant role (and, if so, which types of events are most important).; This study uses network and social participation data from a household survey conducted in 1998 and 2001 by the Malawi Diffusion and Ideational Change project. Statistical associations are investigated using cross sectional analyses and causal relationships are explored using individual fixed effects models. The results show that AIDS awareness by itself is a necessary but insufficient condition for individual change in behaviors that put one at risk. Instead, AIDS conversation networks and social participation seem to be important forces conducive to positive attitudes toward condoms and marital fidelity.; The importance of the attitudinal characteristics of conversation partners in influencing a respondent's attitudes toward protective strategies and actual behavior is confirmed for both genders. The structural and compositional characteristics of conversation networks matter for individual change, but those effects are never as great as the direct effects of shared attitudes.; I find that the principal mechanism through which social interactions affect individual attitudes and behaviors is social influence. Social influence on both sexes works primarily through conversation networks; its strongest impact on individual change is reflected in attitudes toward protective strategies, worry about AIDS, and to a lesser extent condom ever-use. The most profound effects of AIDS conversation networks occur when a respondent has close, strong ties to his or her conversation partners but there is sufficient diversity in overall composition of the network.; Social learning, on the other hand, results from compositional diversity among conversation partners, and attendance at local social events. For men, greater attendance at dramatic performances or political meetings is associated with more favorable attitudes toward condom use outside marriage and greater reported ever-use of condoms, but at the expense of marital fidelity. Gerland, P. (2005). Effects of Social Interactions on Individual HIV/AIDS Attitudes and Behaviors in Rural Malawi. PhD dissertation. Princeton University.