There is scant research on the extent to which, like victims, perpetrators may have experienced victimisation before and/or, are still subject to ongoing victimisation. However, where this has been explored, the percentage of persons affected by prior or ongoing victimisation is significant. For instance, in national surveys, around half of domestic violence perpetrators reported that they were also victims of partner assaults e.g., 49% of respondents to the 1985 National Family Violence Resurvey who reported perpetrating domestic violence also stated that they were victimised by their partners (Straus & Gelles, 1990 cited in Anderson, 2002). Furthermore, analyses done on the National Survey of Families and Households (NSFH), in addition to interviewing both partners in heterosexual marital and cohabiting relationships, found that 64% of respondents who reported perpetrating domestic violence also reported being victimised by violence (Umberson, Anderson, Glick, & Shapiro, 1998 cit

Whereas this finding may reflect a context in western countries where considerable improvements have been made with regard to gender equality, in Uganda male dominance, characterised by a patriarchal family system results in females often being controlled by men and hence, women are more likely to be victims than perpetrators. This is not to say however, that males cannot be victims or females do not perpetrate violence in Uganda (a finding of this study) but to raise questions about the importance of local context in studies of domestic violence.

Although this current study in Uganda did not collect data on the causes of domestic violence, Andersons (2002) analysis gives an insight into the association that exists between psychosocial factors and domestic violence. For example, there was a significant positive association between violence perpetration and mental health/drug and alcohol problems for men and women respectively, thus suggesting that mental health and substance abuse are associated with an increased risk of domestic violence. However, when violence is not controlled, the odds of violence perpetration are increased by about 2% (exp [0.017] = 1.016) for men and 3% for women (exp [0.025] = 1.025) for each unit increase in depression (Anderson, 2002). This finding suggests that depression and substance abuse are associated with a risk of domestic violence perpetration not only for men but also for women. This finding gives an insight into some of the psychosocial factors that influence perpetrating domestic violenc

However, researchers who examine partner violence within national survey data normally focus on either perpetration or victimisation (e.g. Straus and Gelles, 1990; Straus, 1991; Anderson, 2002). Anderson (2002) argued that this separation creates problems for researching partner violence, such as problems identifying causal order, masking the ways in which experiences of intimate violence may differ by gender and other social locations. Andersons (2002) study on perpetrators and victims was carried out among a subsample of 7395 married and cohabiting heterosexual couples drawn from Wave 1 of the National Survey of Families and Households (NSFH-1), a nationally representative sample of the US (Anderson, 2002 p.855). Andersons (2002) findings revealed gender symmetry in intimate partner violence victimisation and also perpetration (see Table 2). For this review, attention is paid to gender symmetry in intimate partner violence victimisation since this was an important finding from a