This project has three specific aims. First, to test cross-culturally leading, current hypotheses about the cause of family violence (wifebeating, child abuse, physical punishment of children, sibling violence). These hypotheses have been derived from the psychiatric, psychological and sociological literature on family violence. Second, to test cross-culturally hypotheses about the prevention or control of family violence. These hypotheses have been developed through a review of anthropological reports about societies where family violence is rare or absent. Third, to develop a list of specific psychosocial factors that are highly predictive of low rates of family violence and may have some applicability to the family violence problem in the United States. The research design for this study is a holocultural study. The sample is a probability sample of one-hundred small-scale and present societics from all geographical regions of the world. The data will be collected from ethnographic reports included in the Human Relations Area Files data archive. Methodologically, the research will use bivariate and multivariate statistical analyses and control for potential sources of bias such as regional variation, systematic data errors, and interdependence of sample units.
