This article examine the influence of type of residence (X1), trust (X2), security conditions (X3), religious activities (D1), market operations (X4), number of village midwives (X5), household access to electricity (X6), the main crops of villagers (D2), and the number of poor households (X7) against physical violence at the household level in Indonesia using the logistic regression method. The source of research data comes from the 5th Wave of the Indonesian Family Life Survey (IFLS). The results of the bivariate analysis indicate that only X1, X6, D2, X7 could be included in multivariate analysis because it is significantly related to Y. Although it is not significantly related to Y, X4 and X5 still including in multivariate analysis because of p < 0.25. The other three independent variables, X2, X3, D1, cannot be included in multivariate analysis because they do not significantly contribute to Y and p > 0.25. The final logistic regression model only contained independent variables and can estimate Y by 6.74% significantly, X2 (4) = 28.79, p <0.01. The final model shows that only D2 and X6 have a negative relationship and explaining Y by -47.3% and 64%, respectively. Whereas X4 and X7, have a positive relationship and explain Y by 78.1% and 110.1%. This finding strengthens the influence of economic factors on physical phenomena on women at the household level. This article recommends to the all-party that concerns to physical violence at the household level, especially government institution at all level, to improve family food resilience, reduce poverty, and increase household electrification while providing media literacy to citizens.
physical violence, IFLS, electrification, household poverty, Indonesia