A survey experiment of women's attitudes about intimate partner violence against women in rural Bangladesh.

According to the Demographic and Health Surveys (DHS) in poorer countries, 50 % of women of reproductive age report that wife hitting or beating is justified. Such high rates may result from structural pressures to adopt such views or to report the perceived socially desirable response. In a survey experiment of 496 ever-married women aged 18-49 years in rural Bangladesh, we compared responses to attitudinal questions that (1) replicated the 2007 Bangladesh DHS wording and portrayed the wife as transgressive for unstated reasons with elaborations depicting her as (2) unintentionally and (3) willfully transgressive. The probabilities of justifying wife hitting or beating were consistently low for unintended transgressions (.01-.08). Willful transgressions yielded higher probabilities (.40-.70), which resembled those based on the DHS wording (.38-.57). Cognitive interviews illustrated that village women held diverse views, which were attributed to social change. Also, ambiguity in the DHS questions may have led some women to interpret them according to perceived gender norms and to give the socially desirable response of justified. Results inform modifications to these DHS questions and identify women for ideational-change interventions.