Key Social Norms Take-aways This article highlights the critical need to rigorously validate quantitative proxy measures of social norms, particularly in IPV research. It shows that norms around IPV, such as male dominance and the acceptability of violence, are best captured using group-level aggregates of men’s attitudes. Men’s social norms were more strongly aligned with the attitudes of other men in their communities than with women’s, underscoring the importance of selecting appropriate, context-specific reference groups. This has significant implications for interventions targeting male behavior through social norms change.
Objectives The study aimed to demonstrate and validate techniques for exploring social norms proxies, focusing on IPV in Tanzania. It specifically sought to identify which reference group configurations best reflect normative influence and how such proxies relate to individuals’ perceptions of norms.
Methods Using data from the 2016 Tanzania International Men and Gender Equality Survey (IMAGES), which included 1,008 men and 1,008 women aged 15–49, the study constructed scores for IPV-related attitudes, descriptive norms (perceived behaviors), and injunctive norms (perceived approval). The researchers calculated intraclass correlation coefficients (ICCs) to assess attitudinal clustering across reference group configurations (e.g., gender, community, marital status, ethnicity), and used bivariate multilevel OLS regressions to test how group-level proxies predicted individual perceptions of norms.
Results The study found that IPV-supportive attitudes clustered most strongly among men within communities (ICC = 0.229) and among men within community-marital status groups (ICC = 0.236). Men’s individual-level descriptive and injunctive norms were significantly predicted by these male-group proxies, especially the male-only community aggregate. In contrast, women’s perceived norms were not positively associated with any group-level proxy and were sometimes negatively associated, suggesting differing mechanisms of social influence. The findings confirmed that attitudes among men are a valid proxy for men’s perceived norms, especially injunctive norms.
Conclusions The study concludes that using aggregate male attitudes at the community level is the most valid and appropriate proxy for IPV-related norms among men in Tanzania. It emphasizes the need to select reference groups carefully, ensure sufficient group sizes, and collect data from both men and women. For researchers and practitioners, the study provides a replicable framework to assess and apply norm proxies effectively, tailored to each new context, thereby improving the design and evaluation of social norms interventions.