Measuring Hidden Support for Physical Intimate Partner Violence: A List Randomization Experiment in South-Central...

Area of Science:

• Sociology
• Public Health
• Gender Studies

Background:

• Physical intimate partner violence (IPV) persists globally, yet understanding its drivers is challenging due to sensitive and illegal aspects.
• Traditional survey methods, like direct questioning, may yield unreliable data on attitudes towards IPV due to social desirability bias and fear of reporting.

Purpose of the Study:

• To assess the reliability of direct questioning survey methods for measuring attitudes supporting IPV.
• To identify the true predictors of continued support for wife-beating in rural Ethiopia.
• To compare direct versus indirect questioning techniques for sensitive topics.

Main Methods:

• A survey of 809 adult Arsi Oromo men and women in rural Ethiopia.
• Embedded a list randomization experiment (LRE) to increase respondent privacy.
• Randomly assigned participants to direct or indirect questioning about wife-beating acceptability.

Main Results:

• Indirect questioning revealed that approximately one-third of the sample found wife-beating acceptable.
• Lower education (<3 years schooling) and lack of female participation in household economic decision-making predicted higher endorsement of wife-beating.
• Individuals with these characteristics were more likely to underreport their approval when asked directly.

Conclusions:

• Direct questioning significantly underreports support for wife-beating, questioning its accuracy for IPV attitude measurement.
• Addressing IPV requires open dialogue and interventions targeting less educated individuals and households with unequal decision-making power.
• Higher education and shared couple decision-making were associated with the absence of support for wife-beating.