The problem of non-marital relationship violence has been of considerable interest to scholars for the past two decades. Although most studies have focused on dating violence among college populations, a growing body of scholarship has emerged that examines the problem among adolescents (Gray and Foshee, 1997; Jackson et al, 2000; Molidor and Tolman, 1998).

Within adolescent studies, in particular, fairly high rates of female-to-male violence have been documented (Morse, 1995; O'Keefe, 1997; O'Keefe and Treister, 1998). Debate continues concerning whether and how genders shapes dating and other partner violence. Some researchers suggest a gender neutral approach is warranted (Moffitt et al., 2000; Moffitt et al., 2001), whereas feminist scholars insist that gender — and male dominance specifically — must be at the foreground for a meaningful understanding of relationship violence to emerge (DeKeseredy and Schwartz, 1998; Dobash et al., 1992, 1998).

The case for sexual symmetry rests primarily on two bodies of scholarship — studies that find similar rates of relationship violence perpetration across gender and, more recently, research that suggests a common psychological profile for males and females who perpetrate relationship violence (Moffitt et al., 2000; Moffitt et al., 2001). Studies that produce similar rates of relationship violence across gender typically use a version of the Conflict Tactics Scale (CTS) (Straus, 1979).

In fact, CTS-based research has rather consistently found similar or higher rates of female-perpetrated than male-perpetrated physical violence among adolescents (Gray and Foshee, 1997; Morse, 1995; O'Keefe, 1997; O'Keefe and Treister, 1998). For example, using NYS data, Morse (1995) found consistently higher rates of female-to-male than male-to-female partner violence. She also reports that girls were more likely than boys to engage in violence labeled severe in the CTS (for example, kicking, hitting with a fist or object, use of weapons), and they were twice as likely to engage in violence that was nonreciprocal.

As significant for feminists, the CTS provide an impoverished understanding of partner violence because "its exclusive focus on 'acts' ignores actors' interpretations, motivations, and intentions. . . .The CTS omits the contexts of violence, the events precipitating it, and the sequence of events by which it progresses". (Dobash et al., 1992:76). Feminists suggest that although much of women's violence is used in response to men's violence, in self-defense or retaliation (Dobash et al., 1998:389), the CTS labels these women violent and as indistinguishable from the men from whom they are defending themselves. Moreover, given the CTS's acts-based operationalization of severity, severe assaults are indistinguishable from more inconsequential acts of violence (DeKeseredy and Schwartz, 1998; Dobash et al., 1992, 1998).

There is some evidence that adolescent dating violence may have features distinguishing it from partner violence in adulthood. Gray and Foshee's (1997:139; Kimmel, 2002; O'Keefe, 1997) survey of adolescents suggests that violent adolescent dating relationships are more often mutually violent than one-sided, and Morse (1995) suggests that the high rates of female-to-male partner violence she documented in the NYS may be a function of the age of her sample. Although national surveys have not documented race differences in rates of partner victimization for women (Bachman and Saltzman, 1995), surveys of adolescents suggest that African-American girls may have higher rates of partner violence perpetration than their counterparts in other racial groups (O'Keefe, 1997).