Both Straus (1990) and Johnson (1995) suggested that the contradiction can be resolved by taking a "dual population" approach. Straus argued that "ordinary" violence, such as slapping, shoving, and throwing things at a partner, is prevalent in the general population and is symmetrical; whereas "severe" violence such as choking, punching, and attacks with objects are rare in the general population but common in clinical populations and are male-predominant. Similarly, Johnson (1995) argued that "situational violence" is prevalent in the general population and symmetrical, whereas "intimate terrorism" is rare and is perpetrated almost exclusively by men. However, a review of 91 empirical comparisons found that symmetry and mutual violence perpetration is typical of relationships involving severe and injurious assaults and agency intervention, and of "intimate terrorists" as measured by Johnson's criteria. The discussion of these results suggests that much of the controversy arises because those who assert gender symmetry do so on the basis of perpetration rates, whereas those who deny gender symmetry do so on the basis of the effects of victimization, i.e. the greater harm experienced by women. Thus, the "different population" explanations of the controversy need to be replaced by a "perpetration versus effects" explanation. When prevention of perpetration is the focus, the predominance of symmetry and mutuality suggests that prevention could be enhanced by addressing programs to girls and women as well as boys and men. When offender treatment is the focus, the results suggest that effectiveness could be enhanced by changing treatment programs to address assaults by both partners when applicable. (Published Abstract)