The Effects of Mutual Partner Violence on Women’s Mental Health

Abstract: 

This study examines the effects of five patterns of violence between domestic partners on women's mental health: (1) unilateral female-dominated, (2) unilateral male-dominated, (3) symmetrical mutual, (4) mutual female-dominated, and (5) mutual male-dominated. A sample of 835 low-income, ethnically diverse women in heterosexual relationships were interviewed. Violence was measured with the Severity of Violence Against Women/Men Scales, and mental health by the Hopkins Symptoms Check List. A majority of violent relationships were characterized by mutuality with most violence dominated by the male partner. Women in unilateral male-dominated relationships numbered 126, while 75 were in unilateral female-dominated relationships. Violence in mutually violent relationships was most predictive of women's mental health problems. Intimate partner violence is not simply unidirectional but characterized by mutual violence.

1The larger study was funded by grant R49/CCR610508 from the National Center for Injury Prevention and Control of the Centers for Disease Control and Prevention and the National Institute of Justice (NIJ) and grant 3691 from the Hogg Foundation for Mental Health, awarded to the third author. Additional funding was provided by grant 2001-WT-BX-0504 from NIJ, awarded to the third author. The results do not necessarily represent the official view of the funding agencies.

Table of Contents: 

    Introduction

    A growing body of research suggests that partner violence may not be gender biased, because both women and men are often found to be in the position of the victim and perpetrator. Indeed, over one-half of domestic violence perpetrators report also being the victim of domestic partner violence (Anderson, 2002). Thus, partner violence should be examined in terms of mutual violence.

    Although the stereotypical belief is that males are the perpetrators in almost all partner violence incidents, some studies have found that the rates of physical aggression from female to male partners are quite high and sometimes comparable to that of males (Kimmel, 2002; Harris, 1996) or higher (Archer, 2000; Brinkerhoff & Lupri, 1988). Using a nationally representative sample of the U.S. population, Vivian and Langhinrichsen-Rohling (1994) found that 12% of men and women reported aggressiveness toward their spouse during the prior year, and as many as 28% of the sample reported some form of aggression during the marriage. In a study conducted by Straus and Gelles (1986), 11% of women reported experiencing at least one act of male-to-female violence, while 12% of men reported experiencing at least one act of female-to-male violence. These statistics, however, do not take into account issues such as severity of the violence or the health outcomes of partner violence. Although the rates are similar, the physical and mental effects of partner violence affect women much more than it does men.

    Motivations of Partner Violence

    Current research shows that when partner abuse is examined in terms of motivation, the abusive behaviors that women commit are different in motivation from the abusive behaviors of men (Swan & Snow, 2002; Milardo, 1998; Vivian & Langhinrichsen-Rohling, 1994; Miller, 2001). The motivations behind partner violence can be influenced by both internal (e.g., anger or frustration) and external (e.g., self-defense or control) stressors. Self-defense, retaliation, preemptive strike, fear, and control are all frequently cited motivations behind violent acts.

    Self-defense and retaliation. Female-to-male violence as self-defense can range from retaliation for previous injuries and attacks to reactions from stress and psychological abuse. Women arrested for domestic violence frequently cite motivations related to retaliation, self-defense, wanting to compel communication, protecting other family members, and controlling the abusive situation (Dasgupta, 2001). Women are more likely to be motivated by self-defense than are men (Brinkerhoff & Lupri, 1988; Milardo, 1998; Vivian & Langhinrichsen-Rohling, 1994).

    Whereas men attribute aggression to situational factors and external causes, women tend to report aggression for retaliation or defense (Vivian & Langhinrichsen-Rohling, 1994). The primary focus of women during conflict is usually on self-protection and survival (Kerouac & Lescop, 1986). Women who use physical force are often victims themselves, trying to escape or stop an abusive partner. Swan and Snow’s (2002) study of women’s use of violence in intimate relationships found that approximately one-third of the women studied who used violence against their partners were also classified as victims. Women are also less likely than men to be aggressors in acts of partner violence. Only 12% of the women in Swan and Snow’s study were labeled as aggressors.

    Women who use violence are more likely to do so toward a violent partner than toward a nonviolent partner (Brinkerhoff & Lupri, 1988). Violence toward an abusive partner is often a form of retaliation (Miller, 2001). Women’s violence as self-defense can also encompass retaliation. Retaliation can occur in response to physical or emotional abuse (Milardo, 1998) in the past or the present (Miller, 2001).

    Preemptive strike is a form of self-defense that may be used to prevent future abuse. For example, a mother may take action if she feels her child is in danger or a woman might prepare herself to take action against a partner who is coming home drunk or extremely angry. These forms of violence are all self-protection mechanisms frequently cited as motivations behind women’s violence.

    Fear and control. Control is a more common motivation for partner violence among males than among females (Vivian & Langhinrichsen-Rohling, 1994; Johnson, 1995; Swan & Snow, 2002). Control includes not only physical violence but threats, verbal abuse, isolation, and other tactics as well (Johnson, 1995). Sexual coercion is another form of control that is used by both men and women. Swan and Snow (2002) found that 28% of the women in their sample admitted that they had used sexual coercion. However, 45% of the women were also victims of sexual coercion, and 14% of the women’s partners used threats to force them to have sex.

    Fear is not only a motivation for partner violence, but also a result. Fear motivates violence more often in women than in men (Milardo, 1998). Women are more likely to report experiencing fear in a violent situation than men (Swan & Snow, 2002). Men in violent situations do not often report fear or perceive women as frightening (Morse, 1995; Swan & Snow, 2002). Thirty percent of female respondents in the National Youth Survey felt they were in danger of getting hurt during fights with their partner compared to only 14% of males (Morse, 1995; Swan & Snow, 2002). While men use fear as a control tactic, fear in women may motivate them to retaliate. Thus, the more male-to-female violence, the more fear, and the more fear, the greater the likelihood of retaliation, possibly resulting in the escalation of mutual violence.

    Methodological concerns. These motivations are all common reasons for partner violence. For these reasons, measures such as the Conflict Tactics Scale (CTS), which neglects to control for the motivations of partner violence, can produce inaccurate results. The CTS is frequently used in conjunction with studies on partner violence. The problem with the measure is that it only assesses the form of behavior and acts of violence without considering the context or impact of aggression (Vivian & Langhinrichsen-Rohling, 1994). This means that the CTS counts each act of violence in any context, even if the woman was striking back in self-defense. Women who strike their partners in self-defense are classified in the same categories as the husbands who beat them. In addition to not assessing motivation, the CTS also fails to take into account which partner initiated the violence and the consequences of the violence (Vivian & Langhinrichsen-Rohling, 1994). The CTS measures aggression and violence well, but other measures should be used to take into consideration specific motivations such as self-defense (Dasgupta, 2001) and outcomes or levels of injury (Morse, 1995) that differ between genders.

    Data can also be skewed by a subject’s failure to report information or to accurately report partner violence experiences. Research shows that males tend to underreport acts of physical violence (Harris, 1996; Milardo, 1998) or may be unwilling to report certain acts at all (Milardo, 1998), even on anonymous surveys. Females, on the other hand, tend to overreport acts of violence they have perpetrated on surveys (Harris, 1996; Marshall, 1994). Male injury rates could also be biased in some instances because males who are injured by their partners may be less likely to seek medical assistance due to social factors such as embarrassment, humiliation, or fear of accusation themselves. Errors in reporting such as these can have effects on the outcomes of research data.

    Severity of violence/injury. Despite having similar rates of partner violence, women and men are not equal in terms of the severity of the violence, as female-to-male violence tends to be less traumatic than male-to-female violence (Dasgupta, 2001). Males are generally bigger and stronger than women. The injuries suffered by males tend to be less severe overall when compared to the severity of violence experienced by women. Injury and hospitalization rates are much higher for women in partner violence situations than men (Morse, 1995). This reduced level of injuries as a result of female violence also contributes to lower hospitalization rates among abused males. Women, conversely, are likely to report injury and seek health-care services for abuse more often than males. Women are also more likely to seek the intervention of the judicial system, and to seek support and shelter from community-based intervention programs (Milardo, 1998).

    Women receive significantly more serious injuries than do men (Dasgupta, 2001). Archer (2000) found that more than 60% of those who suffered an injury from an act of partner violence were women. Using data from the National Survey of Families and Households, Zlotnick, Kohn, Peterson, and Pearlstein (1998) found that 73% of those individuals reporting injuries from domestic violence were female. Even when the partner violence is mutual, women sustain higher levels of injury. Injury that results from partner violence can have a great effect on the victim’s mental health. It is, therefore, possible to conclude that male-to-female violence is much more serious in terms of severity of injury when compared to female-to-male violence.

    Some studies have reported that women perpetrate equal, if not greater, rates of violence than men (Straus & Gelles, 1986; Swan & Snow, 2002; Harris, 1996; Milardo, 1998). Using data from the National Youth Survey, Morse (1995) reported that among respondents between ages 27 and 33, 27.9% of women reported using violence against their partners compared to only 20.2% of men. Contrary to their hypothesis, Swan and Snow (2002) found that women committed significantly more moderate acts of partner violence than did their partners. The study also found that women’s partners experienced almost one and a half times more acts of severe physical violence than they perpetrated (Swan & Snow, 2002). Even among the women who do use violence, there were greater numbers of victims than aggressors within the group.

    Partner Violence and Mental Health

    Depression, posttraumatic stress disorder (PTSD), and suicidality have been found to be related to partner violence (Golding, 1999; Anderson, 2002). Golding’s meta-analysis on the effects of violence on women’s mental health (1999) found that in 18 studies on depression, the average rate of depression among women was 47.6%. Suicidality was found on average in 17.9% of women in 13 studies, and the rate of PTSD averaged 63.8% in 11 studies (Golding, 1999). The effect of partner violence on physical and mental health has been shown to be more severe in females than males (Coker, Smith, McKeowen, & King, 2000; Sharps & Campbell, 1999). Abused women experience higher rates of mental health problems and use more primary care and mental health-care services when compared to non-abused women (Sharps & Campbell, 1999).

    Hypotheses

    According to Johnson (1995), there are two distinct types of couple violence: couple violence and intimate terrorism. Common couple violence is not specifically gender related and consists of couple conflicts that “get out of hand” and result in minor violence. This type of violence occurs less frequently and rarely escalates to more severe forms of violence. Intimate terrorism, on the other hand, is a systematic, intentional form of violence that is perpetrated most often by males against their partners. It occurs more frequently, with much more severity, and it escalates in seriousness over time. Escalation in common couple violence and intimate terrorism is experienced in different rates as well. Escalation is thought to be more common in intimate terrorism because of the perpetrator’s need to exert and exercise control over the partner in varying degrees. Both partners are involved in conflict in common couple violence, and the issue can potentially be resolved between the partners without conflict, whereas in intimate terrorism the violence can occur without an issue to be resolved at all.

    The purpose of this study is to examine the relationship between mutual partner violence and women’s mental health. Women’s mental health will be examined using five distinct categories based on how partner violence is experienced: (1) women who experience unilateral male-to-female violence, (2) women who experience unilateral female-to-male violence, and women who experience either (3) male-dominated, (4) female-dominated, or (5) symmetrical mutual violence in the relationship. This research will examine the hypothesis that all women experiencing partner violence, regardless of the groups to which they belong, will be affected at least somewhat negatively in terms of mental health. Based on Johnson’s (1995) typology of partner violence, we expect that women in unilateral male-dominated partner violence relationships will have higher rates of mental health problems. Women in the mutual female-dominant group include women who experience more female-to-male violence, while women in the mutual male-dominant group include women who experience more male-to-female mutual violence. Women who experience mutual partner violence will have a significantly higher rate of mental health problems than women who experience either female-to-male violence or male-to-female violence because of the escalation that results from mutual abuse. The increased levels of abuse that result from continuing retaliation from acts of partner violence, whether the escalation is situational or over a period of time, will likely lead to higher levels of abuse and more severe outcomes. Females who experience unidirectional male-to-female abuse will have less severe mental health problems than those in the male-dominant mutual violence group because the levels of abuse will likely be lower, but will have more severe mental health problems than women in the female-to-male abuse group. Thus, mental health of women will be most strongly affected by membership in the following groups in descending order: (1) symmetrical mutual partner violence group, (2) male-dominant mutual violence group, (3) unidirectional male-to-female group, and (4) the two female-to-male groups, including both female-dominated mutual group and the unidirectional female-to-male group.

    Method

    Participants

    Data were collected in Dallas County in the 6-wave longitudinal study Project HOW: Health Outcomes of Women. To be eligible to participate in the study, women had to either be married or in a dating or cohabitating heterosexual relationship for a minimum of 1 year, be between the ages of 21 and 48, and be living within 200% of the poverty level and/or receiving public aid. The sample consisted of 835 women who met all requirements out of the 998 women initially screened. The final sample included 303 African American, 272 European American, and 260 Mexican American women. Only Mexican American women who were born and/or educated in the United States were selected for the study in order to control for possible social/cultural differences that might confound the data.

    Procedure

    Women were recruited for the study from the south Dallas area by distributing fliers to businesses, health-care facilities, and community centers, as well as placing fliers on parked cars. Fliers were also mailed to women living in low-income census tracts in the target area. Public service announcements were made on the radio and in newspapers. Trained recruiters scouted potential candidates in stores, clinics, laundromats, and health fairs. Women who were interested in the study were allowed to provide names of family members and friends who also might be interested in the study.

    The research was conducted in one of two offices in the target area. The offices were located in rented office space in an area that was centrally located and in an ethnically diverse area. For participation in the study, women were compensated with a membership card, $15, a “Project HOW” canvas tote bag, and a t-shirt.

    The interviewers for this study consisted only of females due to the sensitive nature of the information collected. Both undergraduate and graduate students were trained to use the structured interview to collect data and to develop rapport among the participants. Three faculty-advised doctoral students were responsible for training the interviewers. Training involved dissecting each item in the interview and explaining techniques for asking questions. Standardization, confidentiality, and response bias were all stressed as issues. Trainees practiced by giving the interview to other trainees, friends, and/or family.

    When an interviewer felt prepared, the doctoral students assessed their performance by simulating the interview process in front of a video camera with a doctoral student who acted as a difficult participant. The potential interviewer was judged on her familiarity with the interview questions, consideration in asking conditional questions, ability to properly react to comments and questions by the participant, and the overall pacing of the interview.

    Interviewers were critiqued and given suggestions throughout the study to ensure accuracy and consistency. For the first wave of interviews, 62 students participated as interviewers, each interviewing between 1 and 57 participants.

    Confidentiality

    Strict confidentiality was maintained for all participants through specific procedures for interviewing. The Public Health Service provided a Certificate of Confidentiality ensuring the participant’s anonymity, maintaining that absolutely no outside party could obtain a woman’s name or her answers. Participants’ answers were not discussed with anyone not directly related to the study, including office workers not directly working with data. The interviewers themselves were uninformed of the participants’ last names, addresses, or the purposes and hypotheses of this study.

    When the participant arrived at the study, she completed an informed consent form and a registration form that matched the participant to her data. Informed consent was given to the participant in technical terms and signed by the principal investigator, then summarized in a way that was clear to the participant. Women also filled out forms that gave the researchers permission to contact them for later waves of the study.

    Measures

    Interviews were 3 hours long, consisting of both open-ended and closed-response questions in a one-on-one structured interview format. Interviewers recorded the response of participants on a carefully constructed data sheet. All items were constructed so they could be understood by someone with less than a high school education. The following specific measures used in the interview process were employed: the Severity of Violence Against Women Scales (SVAWS), the Severity of Violence Against Men Scales (SVAMS), and the Hopkins Symptoms Check List (SCL). The SVAWS and SVAMS consist of 46 items measuring violent behaviors of varying severity including shaking a finger at a person, throwing objects, hitting, biting, slapping, threatening with injury or death, choking, burning, and using a lethal weapon. The Hopkins SCL consists of 36 items measuring physical and psychological symptoms of anxiety and depression, including headaches, daydreaming, crying, feeling lonely, feeling blue, and feeling fearful.

    Results

    Table 1 illustrates the prevalence of violence within the sample of women divided into six different categories: nonviolent relationships, symmetrically violent relationships, mutually violent male-dominant and female-dominant relationships, and unidirectional male-dominant and female-dominant relationships.

    Of the entire sample of women (n = 835), 77.4% had experienced some form of partner violence. As hypothesized, 53.5% (n = 445) were in relationships that involved mutual partner violence; 25.4% (n = 212) had experienced mutual male-dominated violence, which was the most common form of mutual violence; 16.6% (n = 139) were in symmetrical mutual relationships; and 11.3% (n = 94) were in mutual female-dominated relationships. One-hundred and twenty-six women (15.1%) reported experiencing unidirectional male-to-female violence only. The lowest number of women was in the unidirectional female-to-male violence only category, with 9% of the women having reported it.

    Regression analysis was used to measure the effects of the different types of violence on the mental health of women. The results of the regression analyses are presented in Table 2. There was no effect of being in the nonviolent group (n = 189, 22%) on mental health ( = -.039, F (1, 187) = .292, ns. Violence in mutual male-dominated relationships was most predictive of mental health problems (B = .363, p < .001), accounting for 13% of the variance of these women’s psychological distress, F (1, 239) = 36.348, p <.001. Women in the male-dominated unidirectional violence group tended to experience mental health problems as well (B = .172), F (1, 124) = 3.787, p = .054, but membership in this group explained only 3% of the variance in mental health.

    Violence in the symmetrical group (B = .009), F (1, 154) = .012, ns, was not predictive of women’s mental health problems. Membership in the female-dominated mutual violence group (B = -.080), F (1, 46) = .297, ns, as well as female-dominated unilateral violence group (B = -.084), F (1, 73) = .514, ns, were also not predictive of women’s mental health problems.

    Thus, contrary to our hypothesis, the symmetrical mutual relationships did not have the strongest negative effect on women’s mental health. Being in a male-dominated relationship had the strongest effect, followed by being in a unidirectional male-dominated relationship. None of the other types of relationships had a significant effect on women’s mental health.

    Discussion

    This study examined the role of mutual partner violence on women’s mental health. By expanding on Johnson’s (1995) dichotomy of partner violence as either mutual (common couple violence) or unidirectional (intimate terrorism), this study was able to illustrate the complexity of mutual violence in violent relationships. As these findings illustrate, relationships characterized by mutual partner violence can result in several different outcomes based on the type of mutual violence alone. A majority of the relationships in this study were characterized by mutual partner violence (n = 646, 77%), and of these relationships, only 17% were labeled as mostly symmetrical mutual partner violence. Thus, the remaining 83% of the mutually violent relationships were not necessarily “mutual” in terms of frequency and severity. Indeed, a majority of the mutual violence relationships were dominated by the male partner (n = 212, 25%). By simply exploring partner violence as either mutual or unidirectional, a majority of the victims of partner violence become labeled as “mutually violent.” These results clearly illustrate that most relationships involve mutual violence in some form, but also that in most cases of mutual violence one partner (typically the female) faces more frequent and severe violence than the other. It is difficult to truly understand the health outcomes of the victims of unequally mutual violent relationships when they are classified in the same category as the violence perpetrators.

    These findings also suggest that recent literature suggesting that common couple violence is a less severe form of violence (Johnson, 1995) may not be entirely accurate. Indeed, we found that women who experience mutual partner violence are the most likely to have mental health problems. Not only is mutual violence the most prevalent form of violence in violent relationships, it can be as or more dangerous to women’s mental health as unidirectional violence.

    Implications

    The findings from this study, first and foremost, illustrate the point that partner violence as a whole cannot be examined as male-to-female violence or female-to-male violence. This study clearly illustrates that male-to-female violence is more consequential to women’s health outcomes. The major finding from this study is that low-income women are very likely to be victims of partner violence, and face detrimental mental health problems as a result. A second major finding is that a substantial number of the relationships characterized by violence involve mutual violence; in our study two-thirds of the violent relationships involve mutual violence. Recent literature has come to recognize the significant role that mutual violence plays in violent relationships, but researchers continue to downplay it. The results of this study show that this is not warranted. Mutual violence is common in violent relationships and relevant to understanding the outcomes of women’s mental health.

    Moreover, mutual violence is divisible based on which partner perpetrates the majority of the violence and which sustains the most severe injuries. Recognizing partner violence as mutual is not enough to accurately predict health outcomes of women, as males or females can be the dominant perpetrators within the mutual partner violence. Couples characterized by mutual violence can be completely different based on factors such as frequencies of perpetration (such as male-to-female violence), the use of self-defense, and the physical inequalities of the partners (such as the fact that males tend to be naturally stronger and larger than women).

    Being in a mutually violent, male-dominated relationship had the strongest effect on mental health outcomes. From this finding, we can conclude that it is important to differentiate among the various categories of domestic violence. The findings of this research show that mutuality cannot stand as a category on its own. Even when a couple is mutually violent, it is much more common for the male to be dominant in the relationship, and this form of violence is likely different from a mutually violent relationship where the woman is the primary perpetrator or where there is symmetrical perpetration between the male and female partners. Just as unidirectional violence is unacceptable as categorizing male-to-female and female-to-male violence, mutual violence is an unacceptable category when trying to understand the role that different forms of mutual violence play on women’s mental and physical health.

    Based on the findings of this study, dual arrest laws should be reexamined. Dual arrest refers to the arrest of a couple engaged in partner violence based on the assumption that both partners were engaged in the violence or when law enforcement officers are unable to discern a primary perpetrator of the violence. These laws can be unfair to women. Women who are arrested under this policy may not be guilty of the violence but may be charged because they made an attempt to stop or prevent violence from an abusive partner. A woman acting in self-defense may find herself in police custody because she could not explain the situation or her partner made an attempt to put blame back on her for the violence to avoid prosecution. Situations like these are unfair to the victim, who may undergo more severe forms of violence than the perpetrator and still be convicted of the same crime. Dual arrest laws need to be reevaluated because the women who are victimized and need attention can easily be arrested and incarcerated along with the violent partner just for acting in self-defense and not being able to justify it.

    Limitations

    The research data from this study were collected from a very large sample of low-income women from the Dallas area participating in a longitudinal study that consisted of a comprehensive interview of women’s relationship history and health outcomes. Some of the limitations for this study, however, are that men are not represented in the sample data. Information about men’s role in the violence and relationship is attained from interviewing women about their partner’s behaviors. Thus, the results are biased toward the women’s memory and perceptions. Women tend to overreport their own use of violence and underreport their partners’ use of violence (Harris, 1996; Marshall, 1994). It is possible that these findings would change if a more accurate measurement method were used. Therefore, future research may benefit from conducting surveys that include both men and women in equal proportions in order to collect data fairly on violence that includes both men and women.

    Conclusion

    Partner violence takes on many forms. It is a topic that is often either oversimplified in research or overcomplicated by too many different categories of violence. Partner violence is clearly not a black and white issue. It is important to understand that one partner can be violent as well as the other in any given situation based on a person’s own motivations, drives, and abilities. The important point to discern in partner violence is the difference between victim and perpetrator. In unidirectional violence this division is clear—for example, the male is the perpetrator and the female is the victim. However, as this study illustrates, the violence equation is rarely that simple. Indeed, most violent relationships are characterized by mutual violence. Moreover, mutual violence relationships are rarely symmetrical and are often dominated by men. These types of violent relationships can have significant effects on women’s mental health outcomes.

    References

    • Anderson, K. L. (2002). Perpetrator or victim? Relationships between intimate partner violence and well-being. Journal of Marriage and the Family64, 851–863.
    • Archer, J. (2000). Sex differences in aggression between heterosexual partners: A meta-analytic review. Psychological Bulletin, 126, 651–680.
    • Brinkerhoff, M., & Lupri, E. (1988). Interspousal violence. Canadian Journal of Sociology,13, 407–435.
    • Coker, A., Smith, P. H., McKeowen, R., & King, M. (2000). Frequency and correlates of intimate partner violence by type: Physical, sexual, and psychological battering. American Journal of Public Health, 90, 553–560.
    • Dasgupta, S. D. (1999). Towards an understanding of women’s use of non-lethal violence in intimate heterosexual relationships. Harrisburg, PA: National Electronic Network on Violence Against Women.
    • Golding, J. (1999). Intimate partner violence as a risk factor for mental disorders: A meta-analysis. Journal of Family Violence, 14, 99–132.
    • Harris, M. (1996). Aggression, gender, and ethnicity. Aggression and Violent Behavior,1, 123–146.
    • Johnson, M. (1995). Patriarchal terrorism and common couple violence: Two forms of violence against women. Journal of Marriage and the Family, 57, 283–294.
    • Kerouac S., & Lescop, J. (1986). Dimensions of health in violent families. Health Care for Women International,7, 413–426.
    • Marshall, L. (1994). Physical and psychological abuse. In W. R. Cupach & B. H. Spitzberg (Eds.), The dark side of interpersonal communication. Hillsdale , NJ : Lawrence Erlbaum.
    • Milardo, R. (1998). Gender asymmetry in common couple violence. Personal Relationships, 5, 423–443.
    • Miller, S. (2001). The paradox of women arrested for domestic violence. Violence Against Women, 7, 1339–1376.
    • Morse, B. (1995). Beyond the conflict tactics scale: Assessing gender differences in partner violence. Violence and Victims, 10, 251–272.
    • Sharps, P., & Campbell, J. (1999). Health consequences for victims of violence in intimate relationships. Consequences of Violence, VII, 163–178.
    • Straus, M. A., & Gelles, R. J. (1986). Societal change and change in family violence from 1975 to 1985 as revealed by two national surveys. Journal of Marriage and the Family, 48, 465 – 479.
    • Swan, S., & Snow, D. (2002). A typology of women’s use of violence in intimate relationships. Violence Against Women, 8, 286–319.
    • Vivian, D., & Langhinrichsen-Rohling, J. (1994). Are bi-directionally violent couples mutually victimized? A gender sensitive comparison. Violence and Victims, 9, 107–124.
    • Zlotnick, C., Kohn, R., Peterson, J., & Pearlstein, T. (1998). Partner physical victimization in a national sample of American families: Relationship to psychological functioning, psychosocial factors, and gender. Journal of Interpersonal Violence, 3, 156–166.

    Table 1: Frequencies for Types of Violence Relationships (n = 835)

     

    Frequency

    Percent

    Valid Percent

    Cumulative Percent

    Violence type

     

     

     

     

    No violence

    189

    22.6

    22.6

    22.6

    Symmetrical mutual

    139

    16.6

    16.6

    39.3

    Mutual female-dominated

    94

    11.3

    11.3

    50.5

    Mutual male-dominated

    212

    25.4

    25.4

    75.9

    Unidirectional female-dominated

    75

    9.0

    9.0

    84.9

    Unidirectional male-dominated

    126

    15.1

    15.1

    100.0

    Total

    835

    100.0

    100.0

     

    Table 2: Simple Regression of Mental Health Problems on Types of Violence (n = 835)

     

    Beta

    R 2

    F chg

    p

    Violence Group

     

     

     

     

    No violence

    -.039

    .002

    .292

    .590

    Symmetrical mutual

    .009

    .000

    .012

    .913

    Mutual female-dominated

    -.080

    .006

    .297

    .588

    Mutual male-dominated

    .363

    .132

    36.348

    .000

    Unidirectional female-dominated

    -.084

    .007

    .514

    .476

    Unidirectional male-dominated

    .172

    .030

    3.787

    .054