Effectiveness of couples treatment for spouse abuse

Effectiveness of couples treatment for spouse abuse

Stith, Sandra M

Despite its controversy, carefully conceptualized and delivered couples treatment appears to be at least as effective as traditional treatment for domestic violence, and preliminary data suggests that it does not place women at greater risk for injury. However, the body of research on which these conclusions rest is sparse. Only six experimental studies have been done, each using different eligibility criteria, outcome measures, and treatment approaches. Thus, further study of this modality is warranted. Marriage and family therapists have an important part to play in continuing to develop and test innovative ways of helping couples end violence and improve their relationships-an endeavor that promises to improve the quality of the partners’ lives as well as those of their children.

Spouse abuse remains a pervasive problem in American society despite 20 years of widespread efforts to improve prevention, detection, and treatment as well as community and legal response. Spouse abuse has tremendous personal and social costs. According to a national survey, about 16% of married couples reported experiencing an incident of physical assault sometime during the year (Straus & Gelles, 1990). A consideration of cost must include medical, mental health, police, legal, and social services to victims and perpetrators. In 1994, females accounted for 39% of the hospital emergency room visits for all violence-related injuries but accounted for 84% of the persons treated for injuries inflicted by intimates (Bureau of Justice Statistics, 1998). American women are 4 times more likely to be injured by an intimate partner than in a motor vehicle accident (American Medical Association, 1992). Naturally, abusive relationships are related to significant distress, which decreases individual and family quality of life. Survey findings suggest that wife beating has significant adverse effects on women’s mental and physical health. One study indicated that severely assaulted women have twice as many headaches, 4 times the rate of depression, nearly 6 times more suicide attempts, and twice as many days in bed due to illness as women who were not abused (Gelles & Straus, 1988).

Undetected violence can also have significant effects for other members of the family not directly involved in abuse. Partner abuse has been linked to profound and long-lasting negative emotional and behavioral effects on children who witness assaultive behavior (Jaffe & Sudderman, 1995). Children who have observed violence between their parents have been found to assault their siblings and their parents, to commit violent crimes outside the family, and to assault their own intimate partners more often than children who have not witnessed violence between their parents (Straus & Gelles, 1990; O’Leary, 1988). Furthermore, children in families in which there is partner abuse are also more likely to be hit themselves (Jaffe, Wolfe, & Wilson, 1990).

Although historically spouse abuse has been viewed as a private family matter, it is now viewed as a societal problem as well as a crime subject to legal punishment. This paradigmatic shift in perspective has brought a great deal of change in how spouse abuse is handled in our communities. Whereas legal authorities were once reluctant to intervene in violence that was thought to be a private matter between partners, perpetrators now are often faced with jail time or mandatory treatment. Two clear messages emerge: the aggressor must be held accountable, and the victim must be protected. Beyond these two primary concerns, spouse abuse treatment goals vary according to the theoretical perspective of the treatment program. Currently most programs for offenders and all treatment programs that have been formally evaluated are designed for male offenders. Most treatment is administered to men in groups, while concurrent support services and treatment for women victims and their children are offered. This model of “parallel track” treatment is based on the belief that conjoint treatment will increase the danger to victims of abuse by forcing them to confront their abusers directly, will increase the emotional intensity of the couple relationship which may also lead to violence, and will suggest that the victim is at least partially responsible for her abuse because she is being asked to make changes in relationship patterns along with the perpetrator. Although a number of investigators have begun to explore ways to work with couples safely and productively, couples treatment remains controversial.

The primary purpose of this chapter is to review the research that has been conducted to determine the effectiveness of conjoint models of spouse abuse treatment. We will review the research on couples approaches to domestic violence treatment that has included some type of outcome assessment, and discuss some of the methodological concerns with the research that has been conducted to date. Because no programs addressing female violence have been assessed, this review will address only treatment programs focusing on male violence. Also, although the research on couples treatment is quite sparse, family therapists have been writing about their approaches for treating violent couples for some time; therefore we will also discuss several couples treatment programs that have not been empirically evaluated. First, however, because batterer group treatment remains the most widely used and most accepted treatment for domestic violence, we will begin by presenting a brief synopsis of the various kinds of batterers group treatment approaches, and what is known about the effectiveness of this approach.

BATTERERS’ GROUP TREATMENT APPROACHES

Most programs for offenders treat men in psychoeducational, gender-specific groups. Practice guidelines vary but most are based on an underlying conviction about what causes and maintains abuse. Thus, cognitive behavioral approaches are based on the belief that abuse-supporting beliefs, lack of behavioral self-control, and poor relationship skills result in abuse; feminist approaches are based on the belief that male socialization in the context of societally-sanctioned oppression of women results in abuse; and attachment-based approaches are based on the belief that abuse arises from the abusers’ attachment injuries from childhood being reactivated in contemporary relationships. In practice, most batterers’ treatment programs incorporate different theoretical approaches and interventions. However, although batterers’ treatment appears to produce a moderate reduction in recidivism of domestic violence when compared to control groups, no intervention has been shown to be differentially more effective than the other within the same sample (Babcock & La Taillade, 2000).

Batterer program evaluations suggest that from 50% to 80% of program completers are nonviolent at follow-up (see Table 1). However, this body of research is plagued with a variety of problems. Measures of effectiveness vary from study to study. Some studies consider reduction in violence as the measure of success whereas others use complete cessation of violent behavior as the criterion of success (Edleson & Tolman, 1992). Although simple re-assault rate does not reflect the severity or impact of physical abuse, nor whether emotional abuse is occurring, it does offer a gross indicator of behavior. Studies also vary in whose report is used to measure outcome. Some studies depend on male self-report, whereas others use the abused partner’s report, or police reports of re-arrest. Because men tend to report less violence than their partners, and only a small percentage of violence is reported to the police (Edleson & Brygger, 1986), partner report is considered the gold standard. Thus, the success rates reported in Table 1 are based on partner report and the measure of success is cessation of all physically violent acts, unless otherwise noted by an asterisk. It should also be noted that even when partners’ reports are the measure, only a small percent of the partners are located, especially at follow-up. Therefore, it should be taken into consideration that reported recidivism rates are based on the reports of female partners who were located and willing to respond.

Another issue that has been a factor in evaluating batterers’ group treatment approaches is problems with the methodological designs used. Studies that have no comparison group, or no pretest data, are limited in terms of what they tell us about effectiveness. Quasi-experimental designs, in which treatment completers are compared to some sort of control or contrast group, and true experimental designs, in which clients are randomly assigned to treatments and control groups, provide more robust evidence of what works. Table 1 separates quasi-experimental studies from true experimental designs.

Still another issue to consider when interpreting batterers’ treatment outcomes is dropout rate. Although treatment seems to be successful for approximately half of those men who complete treatment, many studies report that only about one-third of batterers initially assigned to treatment, in fact, complete the program (Babcock & La Taillade, 2000). Men who choose to complete treatment tend to be more educated, more often employed, and less likely to have a criminal record than men who drop out (Babcock & Steiner, 1999). We know very little about what works for the dropout subpopulation. More rigorous studies must examine not only the outcome for those who complete treatment, but also must be clear about the dropout rate and compute at least some estimates of success rate based on the completers and noncompleters combined-a strategy becoming increasingly common in substance abuse treatment studies in which treatment attrition is also a significant problem (Liddle & Dakof, 1995). Table 1 shows the dropout rate for each study in which that information was provided.

Finally, the time period in which outcome is measured varies widely from study to study. Although more than 50% of the men who complete batterers’ treatment stop physically abusive behavior for some period of time following intervention (Tolman & Edleson, 1995), it is not clear how durable these changes are. Treatment providers and victims need to know if this behavior change lasts, and over what time period. In Table 1, the follow-up period of each study is reported.

Although group treatment programs for male offenders appear to be effective in eliminating physical violence for some men, there is no single approach that has been demonstrated to be the treatment of choice for all men under all circumstances. In fact, there may be negative effects for some men who participate in men’s groups (Edleson & Tolman, 1992). For example, group members may support each other’s negative attitudes about women or implicitly or explicitly support abusive behavior. Additionally, men may come home from group feeling like their wives have no reason to complain because they are not as abusive as other men in the group (Tolman, 1990). In one study, 10%-15% of female partners reported that their lives had worsened since their male partner began attending a batterer’s program (Gondolf, 2002). In another study a significant proportion of female partners reported an increase in verbal abuse since their partners attended the treatment program (Dutton, 1986). Given the lack of convincing evidence that traditional men’s treatment programs are widely effective, there is room to test other approaches, including conjoint treatment.

CONJOINT TREATMENT FOR DOMESTIC VIOLENCE

Although controversy exists about the appropriateness of treating violent couples together, a variety of reasons can be offered for providing couples therapy for domestic violence. First, a consistent research finding is that male batterers are a heterogeneous group (Gondolf, 1988; Saunders, 1992; Stuart & Holtzworth-Munroe, 1995). Holtzworth-Munroe and Stuart (1994) reviewed the batterer typology literature and reported that three descriptive dimensions (i.e., severity of marital violence, generality of violence [toward the wife or toward others], and presence of psychopathology/personality disorders) have consistently been found to distinguish subtypes of batterers. They suggest that three subtypes of batterers exist (i.e., family only, dysphoric/borderline, and generally violent/antisocial) and that tailoring treatment to each subtype of violent men might improve treatment outcome. From the growing domestic violence typology literature, it has become increasingly clear that all batterers do not need the same type of treatment. Most clinicians and researchers that advocate the use of conjoint approaches suggest that it should be limited to one subtype of batterer-the family-only batterer without apparent psychopathy, who is most likely to benefit from couple therapy (Stuart & Holtzworth-Munroe, 1995).

In addition to treating subgroups of batterers differently, there is also reason to include female partners in treatment. Both men and women are often violent in relationships. In fact, most research has found that women initiate and carry out physical assaults on their partners as often as do men (Stith & Straus, 1995). Despite the much lower probability of physical injury resulting from attacks by women, assaults by women are serious, just as it would be serious if men “only” slapped their wives or “only” slapped female fellow employees (Straus, 1993). If reciprocal violence is taking place in relationships, treating men without treating women is not likely to stop the violence. In fact, cessation of partner violence by one partner is highly dependent on whether the other partner also stops hitting (Feld & Straus, 1989; Gelles & Straus, 1988). Most importantly, when women use violence in relationships, they are at greater risk of being severely assaulted by their partners (Feld & Straus, 1989; Gondolf, 1998). Moreover, although men’s treatment groups address men’s role in intimate partner violence, they do not address any underlying relationship dynamics that may impact each partner’s decision to remain in the violent relationship despite the violence, or may play a part in maintaining the violence. Because 50% to 70% of battered wives remain with their abusive partners or return to them after leaving a women’s shelter or otherwise separating from them (Feazelle, Mayers, & Deschner, 1984), failing to provide services to both parties in an ongoing relationship may inadvertently disadvantage the female partner who chooses to stay. In a study involving the prediction of mild and severe husband-to-wife physical aggression with 1,870 randomly selected military personnel, Pan, Neidig, and O’Leary (1994) found that marital discord was the most accurate predictor of physical aggression against a partner. For every 20% increase in marital discord, the odds of mild spouse abuse increased by 102%, and the odds of severe spouse abuse increased by 183%. Because marital discord is a strong predictor of physical aggression toward a partner, it would seem that failure to address marital problems at some point in the treatment of men and/or women would make it likely that physical abuse would recur.

REVIEW OF LITERATURE ON CONJOINT TREATMENT FOR DOMESTIC VIOLENCE

Studies of typical family therapy client populations-those who come for regular outpatient marital therapy, not domestic violence treatment-show marital violence rates as high as 67% (O’Leary, Vivian, & Malone, 1992). Despite this staggering figure, only a few published studies have examined the effectiveness of couples treatment for husband-to-wife physical aggression. A number of issues limit the usefulness of this body of research in understanding the impact of couples counseling on treating domestic violence. Most of these studies are based on various types of cognitive behavioral treatment, leaving recent innovations in couples treatment such as Emotionally Focused Couples Therapy (Johnson, 1996) untested with domestic violence. Also, the therapeutic modality differs across studies with some studies testing individual couple therapy (one therapist working with one couple) while others examine multi-couple group therapy. Finally, as with studies of batterer treatment described above, methodological problems plague the study of conjoint treatment for domestic violence. We could find only six studies, for instance, that can be considered experimental. In this section we will review the existing outcome literature on couples treatment of domestic violence. The studies are summarized in Table 2.

Experimental Design

Only six studies can be conceptualized as being experimental (i.e., couples are randomly assigned to two or more treatment conditions). One study compared individual treatment, multi-couple treatment, rigorous monitoring, and no treatment within a military setting (Dunford, 2000). Two studies compared individual couple therapy with multi-couple therapy for domestic violence (Harris, Savage, Jones, & Brooke, 1988; Stith, McCollum, Rosen, & Locke, 2002). Stith et al. (2002) also included a nonrandom comparison group in their analyses. Two studies compared a cognitive behavioral men’s group with a cognitive behavioral couples group (Braunen & Rubin, 1996; O’Leary, Heyman, & Neidig, 1999). Finally, one study compared individual treatment for substance abusers with a combination of individual treatment and behavioral couples therapy for substance abusers and looked at the impact of each treatment program on domestic violence (Fals-Stewart, Kashdan, O’Farrell, & Birchler, 2002).

Dunford (2000) conducted the only experimental study that included a conjoint treatment condition and a “no treatment” control group. He randomly assigned 861 Navy couples to one of four interventions: a 26-week cognitive behavioral therapy (CBT) men’s group followed by six monthly sessions, a 26-week CBT multi-couple group followed by six monthly sessions, a “rigorously monitored” group, and a control group. The control group did not receive any formal intervention. Victimized wives in the control group were contacted by the military agency responsible for preventing and responding to domestic violence-the Family Advocacy Center (FAC)-as soon as possible after the presenting incident to ensure that the women were not in immediate danger of continued abuse. Once their safety was assured, FAC provided wives with safety planning information. No other formal intervention was offered.

In the rigorously monitored group, a case manager at the FAC saw perpetrators monthly for 12 months and provided individual counseling. Every 6 weeks a record search was completed to determine if perpetrators had been arrested or referred to court anywhere in San Diego County. Wives were called monthly and asked about repeat abuse. They were told that they did not have to reveal anything about their husband’s behavior if doing so would place them in jeopardy. At the end of each treatment session, case managers sent progress reports to perpetrators and their commanding officers, specifying the presence or absence of instances of abuse. This process of rigorous monitoring was an attempt to create a “fishbowl” for perpetrators in which they felt that any instance of repeat abuse would be identified and dealt with by the military authorities.

The men’s group, which used a cognitive behavioral treatment approach, met weekly for 6 months and then monthly for another 6 months, for a total 1-year treatment period. Treatment was based on a curriculum developed by Saunders (1996) and Wexler (1999). Each session had a series of tasks that the group leader was obliged to complete including both didactic and process activities.

The multi-couple group was organized in much the same way as the men’s group, with 26 weekly sessions that included both didactic and process activities followed by six monthly sessions. As in the men’s group, the six monthly sessions that followed consisted of content review and process activities. The multi-couple group curriculum was also based on the cognitive behavioral model and was developed by Geffner and Mantooth (2000). The interventions were similar to those used in the men’s group, with the addition of wives to observe their partners being confronted about their abuse and to learn constructive ways of dealing with conflict (Dunford, 2000).

FAC records indicated that 71% of the cases were judged as having successfully completed treatment. Fifteen percent of the men were discharged from the Navy and therefore did not complete treatment. The remainder of the cases (14%) were considered as not having completed treatment. Thus, a conservative estimate of the dropout rate would be 29% if those leaving the Navy were defined as dropouts along with those labeled “not completing treatment.” Analysis of the data revealed that 83% of the men completing treatment (men’s, conjoint, and rigorous monitoring) did not re-injure their wives during a 1-year follow-up period. Because 79% of the men in the control group also did not re-injure their wives, there were no significant differences between groups on rate of re-injury. Findings also revealed no significant differences between groups on a variety of other outcome measures including “being pushed or hit,” sexual abuse, and control abuse. Thus, in the military setting, the no-treatment group seemed to do as well as the treatment groups.

Two issues limit the usefulness of this for understanding the effectiveness of conjoint treatment in the general population. First, this study was conducted with active duty military members. When repeat violence occurs with this group, commanders are notified and the recidivism can impact the individual’s career. Thus, findings from this study may not be generalizable to the civilian population. Also, a major problem with this study’s ability to compare conjoint couples treatment with the other interventions is that the average number of wives actually attending the conjoint group sessions was relatively low. The ratio of attendance of women to men was 2 women for every 5 men. Thus, although some review articles (e.g., Babcock & La Taillade, 2000) include Dunford’s (2000) study as a test of couples’ therapy for domestic violence, in actuality, very few couples participated in the conjoint treatment modality. Although couples were randomly assigned to groups, wives were not asked about their willingness to participate in treatment before the random assignment was made, nor were they required to attend, as were their active duty husbands. In addition, the multi-couple group was not conceptualized as a systemic intervention, but rather as an intervention that added wives to a men’s treatment program so that wives could “witness authority figures confronting the offensive and oppressive nature of spouse abuse” (Dunford, p. 469).

Although Dunford’s (2000) findings have limited usefulness in evaluating couples treatment, they did demonstrate that adding wives to a males’ treatment group did not increase the risk to the wives. Wives participating in the conjoint groups were no more likely to be assaulted or injured than wives whose husbands participated only in men’s treatment.

One early study (Harris et al., 1988) randomly assigned 81 couples that had contacted a family-service agency requesting relationship counseling to a multi-couple group treatment program or to couples counseling. Although the authors indicate that they also had a waiting-list control group, they do not compare the outcome of the treatment groups to the outcome of the control group. To be eligible for the program, a man had to use physical and/or sexual violence toward his partner or property and frighten or control her. The woman (when interviewed individually) had to indicate that she wished to remain in the relationship and report that she did not feel endangered. Furthermore, the man had to exhibit no psychotic symptoms, no evidence of serious brain injury, and no pervasive substance abuse that was not being treated concurrently.

The multi-couple group program consisted of 10 weekly 3-hour sessions. During the first 90 minutes of each session, the men and women met separately in same-sex peer groups. The women’s group focused on the process of ending victimization, and the men’s group concentrated on confronting violent behaviors and understanding attitudes that contribute to controlling behavior. Afterwards, the men and women met together with both group leaders for 1-hour teaching sessions on topics that included time-out procedures and the cycle of violence. Two review sessions were held at 1 month and 4 months after the program ended.

The individual couples counseling program was a “family systems-based form of treatment modified so that the therapist explicitly addressed the violence against the woman as the primary problem in the relationship using concepts developed in the group program” (Harris et al., 1988, p. 149). Treatment in this condition continued until the couple and therapist mutually agreed that all goals had been accomplished.

Sixty-seven percent of the 35 couples assigned to individual couples counseling dropped out before completing treatment, whereas only 16% of the 23 couples assigned to the multi-couple group dropped out. Only a small number of the initial couples completed all pre and post measures (5 couples who had completed couples counseling; 16 couples who participated in the multi-couple group). Repeated-measures analysis of variance indicated that scores on all aspects of psychological well-being assessed (i.e., levels of violence, mood states, self-confidence, and social support) did not vary by treatment group or by sex, hut that participants’ mean pre- and posttests scores for all the measures were significantly different; that is, participants showed positive changes over time, regardless of sex or treatment group. Follow-up results revealed that the goal of stopping the physical violence was achieved for over 80% of the couples based on reports by women who participated in the follow-up interview, and that these results did not differ by treatment group. However, because couples in the individual couples treatment program were much less likely to complete treatment, “all couple counseling for this population has ceased in this setting” (Harris et al., 1988, p. 154). Despite this apparent condemnation of couples therapy, the group intervention included a strong couples component.

Stith et al. (2002) conducted another study comparing individual couples treatment for domestic in sustaining gains. Approximately 92% of the respondents who were able to be located (62% of the participants) indicated that there had been no further incidents of violence after treatment.

In the final study that we were able to identify, which used experimental methods to assess the impact of couples treatment on domestic violence, Fals-Stewart et al. (2002) compared individual treatment for substance abusers with behavioral couples therapy (BCT). In BCT the spouse is included in treatment with the substance-abusing patient to build support for sobriety. Behavioral techniques such as communication training and “Caring Day” assignments are used to reduce risk of relapse. In this study, 80 married or cohabiting male patients with a primary drug abuse diagnosis (with the primary drug of abuse not being alcohol) in a substance abuse outpatient clinic were randomly assigned to one of two treatments. The individual treatment was a behavioral coping skills program. BCT was the other treatment. Both treatments comprised 56 therapy sessions over a 6-month period. Individual treatment had all sessions with the patient alone; BCT had 12 of the sessions with the patient and female partner together. In addition to having fewer substance abuse relapses, couples in BCT also reported more positive relationship adjustment on multiple measures (Fals-Stewart, Birchler, & O’Farrell, 1996). In addition, Fals-Stewart et al. (2002) reported that although nearly half of the couples in both groups reported male-to-female violence in the year before treatment, the number reporting violence in the year after treatment was significantly lower for BCT (18%) than for individual treatment (43%). Thus, although BCT did not specifically focus on domestic violence, conjoint treatment for drug abuse had a significant impact on reducing domestic violence recidivism, while individual treatment did not seem to impact domestic violence recidivism.

Results from each of the experimental studies indicate that violent men who are treated with their female partners (either in individual couple therapy or as part of a multi-couple group) reduced their violence. The study by Fals-Stewart et al. (2002) found that couples therapy for substance abuse was more effective than individual therapy for substance abuse in reducing domestic violence. The study by Brannen and Rubin (1996) found that the multi-couple therapy was more effective than gender-specific therapy in reducing domestic violence for male abusers with a history of alcohol problems. Significant differences were not found between individual couple treatment and multi-couple treatment or between gender-specific treatment and multi-couple treatment in reducing or eliminating violence in the other four studies. There is no evidence from the six experimental studies reviewed here that women are more likely to be endangered in these carefully screened, domestic violence-focused conjoint treatment programs than they are in programs that treat men individually. In addition, adding conjoint treatment sessions to individual treatment for substance abuse appears to reduce domestic violence recidivism more than does treating male substance abusers individually.

Quasi-Experimental Design

In an interesting study of the natural history of domestic violence before and after alcoholism treatment, O’Farrell, Van Hutton, and Murphy (1999) followed couples receiving behavioral marital therapy (BMT) for two years. Comparison rates of domestic violence for a demographically matched nonalcoholic sample were derived from a nationally representative survey of violence in American families (Straus & Gelles, 1990). In the year before BMT, the alcoholics had a significantly higher prevalence of violence and frequency of violent acts (4 to 6 times more prevalent and substantially more frequent) than did their counterparts in the nonalcoholic comparison sample. Of the original 88 couples completing treatment and 1-year follow-up assessments, 75 provided 2-year follow-up data on violence. The percentage of couples experiencing any violent act decreased from 61.3% in the year before BMT to 22.7% in the first year after BMT and to 18.7% in the second year after BMT. The frequency of violent behaviors also decreased significantly. Additionally, during both the first and the second year after BMT the prevalence and frequency of violence by alcoholics were no longer significantly higher than among their counterparts in the nonalcoholic comparison sample. However, violence prevalence and frequency remained at least 1.5 times higher than in the comparison sample.

Another interesting finding from this study was that the extent of violence after BMT was associated with the extent of the alcoholics’ drinking. Overall, relapsed alcoholics had a greater level of violence than did remitted alcoholics. Remitted alcoholics no longer had elevated domestic violence levels compared with matched controls whereas relapsed alcoholics did, and the frequency of violence was correlated with the number of days the alcoholic drank. The authors emphasized that “reduced violence among the remitted alcoholics and persistent violence among the relapsed alcoholics does not permit unequivocal interpretation that the continuing drinking caused the continued violence, since other variables may account for the variations in both drinking and violence” (O’Farrell et al., 1999, p. 321). They also emphasized that, like the other nonexperimental studies reviewed here, we cannot conclude that the changes in violence were caused by BMT, because a control group without BMT was not included.

Pre-Posttest Designs

Outcome has been reported for a number of other programs developed to treat couples in battering relationships. Eight studies are reported in Table 2 and two other studies are briefly described in this section. However, studies regarding the effectiveness of these programs are plagued by substantial methodological shortcomings including limited sample sizes, lack of standardized assessment instruments, outcome reported by only one partner, and one-sample designs. Although these studies do not help us understand how conjoint treatment compares with no treatment or with individual treatment, they do help us understand whether offenders receiving these types of treatment reduce their level of violence. Of course we cannot conclude that the treatment is responsible for the change, because a variety of other factors including time, threats by the partner to divorce, arrest, and so forth, may be responsible for noted changes.

Only one study found that violence continued for all couples treated in a conjoint treatment program. Lindquist, Telch, and Taylor (1983) pilot tested a 9-week, 2-hour per week multi-couple group with 8 couples. The program focused on positive communication skills, stress management, anger control, and problem solving. Of the 10 individuals who responded at the 6-week follow up, all indicated that the group continued to be of some benefit, and 50% reported no recidivism. However, at the 6-month follow up, all of the 4 couples responding reported between one and four violent incidents.

All of the other studies reported that couples treatment had some positive impact on domestic violence. For example, Geffner, Mantooth, Franks, and Rao (1989) developed a psychoeducational conjoint therapy based on Rational Emotive Therapy principles. This program includes both individual couple treatment and couple group treatments. Riza, Stacey, and Shupe (1985) evaluated the program and reported that one year after counseling, 78% of treated couples reported a reduction or cessation of violence and 79% remained together. All of the couples surveyed recommended the program to others.

Taylor (1984) described an eight-stage individual couples treatment program based on a cognitive behavioral approach. His stages included observing stressful cues, changing internal cognitive dialogues, learning to express one’s self more assertively, and more effectively resolving conflicts. He reported that at 6-month follow-up 65% of the 50 couples he treated were violence-free. He did not report how these data were gathered nor on whose report they were based.

Deschner, McNeil, and Moore (1986) reported on a 10-session anger control group for violence-prone spouses and parents. In their groups, they treated both child and spouse abusers. The first few weeks were gender specific and addressed anger management skills and a commitment to nonviolence; the remaining weeks were conjoint and addressed anger management skills interactionally. Research on the program (Deschner et al., 1986) found that 8 of the 15 couples were violence free at 8-month follow-up. In a further study of this approach, Deschner and McNeil (1986) reported on 134 persons that had come to at least one session of treatment. Of these, 45 had been involved with child abuse alone, 7 with both child and spouse abuse, and the rest with partner abuse only. All had experienced at least two episodes of violence in their family. Eighty-two attended at least four sessions, a dropout rate of 39%. During the follow-up survey 8 months after treatment ended, researchers were able to contact 54% of the group members. According to their self-reports, only 15% had reverted to physical violence at any time since leaving treatment. Most were still with the same partner and were managing to avoid violence.

Harris (1986) conducted a study of 40 couples randomly selected from over 200 couples seen between 1978 and 1983 using Walker’s (1979) model of conjoint counseling for battering couples. The model was an early version of cognitive behavioral treatment in which both partners were taught anger management skills. Initially, individual sessions were held with each partner and then conjoint sessions were interspersed with individual sessions. Although the authors reported a 73% success rate, they did not explain how they defined success.

Perez and Rasmussen (1997) described a program that targeted couples at risk for battering. The prevention model is based on a Bowenian approach that focuses on partner differentiation, reducing emotionality, and promoting rational thinking. Couples in the program had experienced one or more episodes of the man using highly coercive behavior, “mild” violence to control the woman, and/or both partners had used coercive behavior. Couples typically participated in weekly, 1 hour-and-15 minute sessions. The number of sessions ranged from 1 to as high as 50, although the average was 8 to 10. Outcome was measured by therapist reports. Of the 118 couples that participated in the program between 1992 and 1995, according to therapist reports, 62% were improving or improved, 27% had little improvement or no improvement, and the remaining 11% were in crisis or referred. The authors did not report rates of violence after treatment or exactly how therapists arrived at improvement rates.

Johannson and Tutty (1998) reported results from exploratory research that evaluated two 12-week groups for couples who had previously completed 24-week separate gender family violence groups. Eligibility criteria for the group included no ongoing severe violence and willingness of both partners to sign a no violence contract. The main focus of the couples group was to help couples integrate the communication, problem solving, and conflict resolution skills learned in the gender-specific groups. Each week the facilitators introduced a theme, such as the effects of childhood abuse or personal values, and the couples practiced communicating their personal experiences with respect to the themes while other couples and the facilitators observed and intervened if communication became difficult. Fifteen couples began the group, but 2 dropped out. Overall, posttest scores on psychological and physical abuse declined. However, because participants were asked to report on violence during the past 12-months each time, considerable overlap occurred in the time period assessed. During treatment, no violence was reported by either partner at the weekly group check-ins or the monthly private screenings. One year after the group had ended, the group facilitators contacted 11 of the 13 couples. Seven remained in their relationship with 4 reporting no further incidents of abuse. Two of the 3 couples who experienced reccurence of violence came back to the agency seeking further relationship counseling.

What Have We Learned From Research on Conjoint Approaches to Domestic Violence?

As indicated earlier, results from the experimental and quasi-experimental studies reviewed here indicate that domestic violence-focused conjoint approaches with carefully screened couples appear to be at least as effective as gender-specific treatment approaches. All but one of the 1-sample studies indicates that after completing conjoint treatment, some couples are able to end the violence in their relationships. No studies have demonstrated that conjoint treatment is more dangerous than gender-specific treatment.

EMPIRICALLY UNTESTED APPROACHES TO COUPLES TREATMENT OF DOMESTIC VIOLENCE

Several models of couples therapy for domestic violence are being used clinically but have not been empirically tested. We have chosen to discuss three of these because we believe they address important issues in treatment and are rich ground for future research. These models are the Ackerman Institute Model, the Cultural Context Model, and Solution-Focused Brief Therapy.

The Ackerman Institute Model

The work of Goldner, Penn, Sheinberg, and Walker (1990) at the Ackerman Institute for the Family represents an effort to escape the polarization of feminist and systemic ideas in the treatment of domestic violence. The model that has arisen from this work is grounded in the belief that neither the feminist nor the psychological/systemic perspectives by themselves are adequate to explain the complexity of violence in attached relationships. The feminist perspective’s focus on power inequity and the culturally sanctioned domination of women by men does not leave room for the ambivalence felt by both men and women in attached relationships who experience both conflict and love for one another. Similarly, exclusive reliance on the psychological/systemic perspective runs the risk of holding victims and perpetrators equally accountable for the violence in the relationship and ignoring the social forces that support violence against women.

The treatment model (Greenspun, 2000) that has grown from this attempt to combine the feminist and psychological/systemic views conceptualizes violence as multiply-determined, the outgrowth of male power and control, couple relational dynamics, and individual factors (e.g., past histories of abuse and trauma, and neurobiological predispositions to violence). Treatment is provided on an outpatient basis by one or two therapists working directly with the clients while a team observes. Greenspun reports that most couples treated with this model so far are self-referred for general marital therapy and violence is identified as a problem during agency intake. A small percentage is court-ordered.

Treatment progresses through three stages-evaluation, individual couple work, and group therapy. During the evaluation stage, the treatment team determines whether or not the couple is appropriate for conjoint work. Some factors arc clearly exclusionary (e.g., a history of sociopathy) whereas others may require additional treatment before couples therapy begins (substance abuse, neurobiological factors such as attention deficit disorder, or neurological impairment affecting impulse control, for instance). The primary criteria for inclusion, however, are: “the man’s ability to take full responsibility for his use of violence, his capacity to tolerate hearing the woman’s description of being victimized by him, and his willingness to work toward stopping his abusive behavior” (Greenspun, 2000, p. 160).

The individual couple therapy stage is brief-five sessions-and aims at addressing the violence directly and helping the perpetrator stop it. Interventions during this phase include helping the perpetrator identify exceptions to his use of violence, connecting impulse toward violence to past abusive experiences, acknowledging and validating the woman’s experience of the man’s violence, and addressing safety.

The last stage of treatment-group therapy-is used to augment the changes that have begun in the individual couple stage and to generalize treatment into a wider community dedicated to nonviolence. Both multi-couple and gender-specific groups are used. Goals of this phase of treatment include reducing the social isolation of couples, and providing peer support for nonviolence along with peer confrontation of abusive acts or attitudes.

To date, no outcome data are available for this model.

The Cultural Context Model

Almeida and Durkin (1999) described a couples treatment model for domestic violence-the Cultural Context Model-that has grown from over 15 years of clinical experience. This model is based in a feminist analysis of the important part that culture plays in determining many levels of behavior including violence in attached relationships. Culture in this model is a broad category that can serve as both a source of significant strength and connection as well as the medium for attitudes and beliefs that serve to oppress women and children and cede power to men. Thus, any change effort must raise consciousness about the power of men and the role of women and place these issues in their cultural, historical, and sociopolitical contexts.

Domestic violence treatment in the Cultural Context Model is conceptualized in three phases. In the first phase, male perpetrators are seen in male-only groups. The three goals of this phase are to help men become accountable for their violence; help them understand the interplay of gender, race, class, culture, sexual orientation, and domestic violence; and expose them to nonviolent male sponsors. As individuals, sponsors serve a variety of functions in treatment-models, mentors, and confronters-while collectively they bring the voice of the community into the treatment setting and challenge the patriarchal view that violence inside the family is a private matter, of no concern to “outsiders.”

Phase 2 of therapy involves group treatment for men and women in gender-specific groups. The goals of these groups are to build on the work of phase 1 and to begin to address relationship issues. Men and women still meet separately in this phase because the emphasis of treatment is different for each. For men, the focus is on learning to value personal accountability to others over concern with self while the focus in women’s treatment is on empowerment and stepping out of the role of bearing sole responsibility for the well-being of their families.

Phases 1 and 2 of therapy last for at least 9 months. Throughout this time, therapists look for evidence that men are working to change patterns of dominance. Both men’s reports of their own actions and attitudes and, more importantly, women’s reports concerning their partners are used to judge whether or not change is occurring. Only when the therapists are convinced that significant change has occurred do the couples begin conjoint therapy.

In phase 3, both men and women remain connected to their gender-specific groups but begin to meet together for couples counseling as well. Almeida and Durkin (1999) did not describe specific goals of couples treatment but did note that groups remain a primary focus of treatment during this phase. Information gained by therapists in couples sessions may be taken back to group to preserve the emphasis on community as the arena for accountability. Thus, a man who reverts to abusive or domineering actions will be challenged in his gender-specific group as well as, presumably, during the couple session.

Almeida and Durkin (1999) reported that in 15 years of using the treatment format described above, “no woman participating in our program has ever been physically hurt” (p. 321). No data on dropout or other outcome data are reported for this model.

Solution-Focused Domestic Violence Couples Treatment

Lipchik and Kubicki (1996), a family therapist and a batterer intervention specialist respectively, described an approach to couples treatment based on the Solution-Focused Brief Therapy model (see, for example de Shazer, 1982, 1985, 1988). This model most resembles typical couples therapy. Couples are seen in an outpatient setting by two therapists, one of who conducts the session while the other observes from behind a one-way mirror. Following an initial conjoint session, partners are seen individually to further assess violence and to formulate safety plans for both the man and woman. Ongoing couples treatment is undertaken only if the following conditions are met:

1. The man says he really wants to stop being abusive in any way his partner experiences it.

2. The man takes responsibility for the abuse.

3. The man takes responsibility for contributing to the quality of the relationship.

4. [The woman] expresses the desire for emotional and physical violence to cease.

5. [The woman] takes responsibility for contributing to the quality of the relationship (Lipchik & Kubicki, 1996, p. 70).

If these conditions are not met, couples therapy is not indicated and one or both partners are referred to whichever noncouples treatments appear to be appropriate-individual therapy, batterer’s intervention, shelters, and so forth. Furthermore, couples are evaluated for substance abuse, psychiatric impairment, and the presence of some positive feeling about the relationship before couples treatment begins.

Couples treatment consists of applying the solution-focused model to the issue of violence and whatever other concerns the couples present. Lipchik and Kubicki (1996) made clear that this approach represents a radical departure from more traditional approaches to domestic violence treatment. They eschewed the stance of separating the couple, advocating for the victim, and confronting and resocializing the man in favor of taking a collaborative stance with both partners, motivating the man to take responsibility for his violent acts by not increasing his resistance through confrontation, identifying positive aspects of the relationship and amplifying them, and helping the couple set mutually satisfying goals for treatment. The authors addressed issues of safety in treatment by careful selection of appropriate couples and discontinuing couples treatment if there was a recurrence of abuse. Although Gingerich and Eisengart’s (2000) review of outcome studies of Solution-Focused Brief Therapy provides some support for the efficacy of this approach with a variety of types of problems, and Stith et al. (2002) reported on the effectiveness of an integrated version of this model, no studies have tested its usefulness as presented by Lipchik and Kubicki for couples treatment of domestic violence.

PRIORITIES FOR FUTURE RESEARCH

As is clear from the foregoing review, much work remains to be done to fully evaluate the efficacy of couples treatment for domestic violence and to specify with which couples it is appropriate. Both small- and large-scale studies can be useful in this regard. At present, cognitive behavioral approaches are the most common models that have been tested. Other models, including the three we have described in this chapter, are being used clinically with violent couples and should be tested empirically. In addition, recent innovations in marital therapy, including the work of Johnson (1996) and Gottman (1999) may lead to even more useful models for working with couples between whom there has been violence. The development and testing of such models is a further step in the process of carefully evaluating couples therapy for domestic violence.

In addition to simply doing more outcome research in this area, the literature could be strengthened by attention to some of the methodological problems in existing studies. As noted earlier, few studies to date meet the criteria for sound outcome research. It is also important for researchers to use sound and clearly specified outcome criteria. Issues to consider include whose report of outcome (victim, abuser, or a third party like police reports) should be used, what criteria for outcome (e.g., absence of violence, reduction in violence, increase in couple satisfaction, or decrease in patriarchal values) will be measured, and what measurement tool will be used.

Quantitatively measured outcome is important to the evaluation process for couples treatment, but qualitative methods also have a place. Although there is no empirical support for the notion that women who participate in conjoint domestic violence-focused therapy are in more danger of being abused than women who are left out of the treatment process, little is known about how women experience domestic violence-focused couples therapy. It is important to know what conditions help them feel safe to participate fully in couples sessions and what conditions lead them to hold back. In addition, philosophical concerns must mesh with empirical ones. If couples treatment decreases violence and improves relationships but leaves women feeling more to blame and in danger, it would clearly be a mistake to judge such a treatment successful. Qualitative methods can certainly make a contribution in understanding women’s experience.

Treatment dropout is a major problem in both domestic violence treatment and in domestic violence outcome research. Most studies report that 30% or more of the participants who begin treatment do not complete it. However, little is known about what differentiates those who drop out from those who complete treatment. Perhaps completers are a clear subpopulation of those who are referred for domestic violence treatment and outcomes are therefore only generalizable to that specific group. A better understanding of why clients drop out of treatment would set the stage for adjusting our treatment models to better meet clients’ needs and therefore reduce dropout rates. In addition to understanding what leads men to drop out of treatment, it is important to also understand more about the experience of women whose partners drop out compared with women whose partners complete. If partners of dropouts are at more risk, outreach services might be made available to them. Answering such questions about both men and couples who drop out of treatment will give the most complete view of the efficacy of using couples treatment for domestic violence, and its differential effectiveness over other intervention modalities. Process and qualitative research designs could help us begin to understand the factors that lead clients to leave treatment without completing it and the experiences of those who drop out.

Other Research Directions

Unfortunately, we could find no studies that have examined the cost effectiveness of couples treatment nor any that examine who is best able to deliver it and what “dose” of treatment is most effective. Also, no one has conducted research examining the effectiveness of any of these conjoint models with diverse populations. Each of these issues deserves further investigation.

Domestic violence research has increasingly shown a strong link with various other mental health disorders, especially substance abuse and depression. Conjoint treatment models that target individuals with dual diagnoses of both domestic violence and substance abuse or domestic violence and depression could be developed and compared with conjoint treatment models developed to treat these issues in isolation.

We also need research efforts to develop and test programs that treat the entire family, including witnessing children. We know that domestic violence has a profound impact on children living in violent families, and family therapists are especially appropriate mental health providers to develop and test these models. Finally, with communities increasingly mandating treatment for female domestic violence offenders, we need to develop and test treatment programs that specifically address the needs of aggressive female partners.

Implications for Therapists

Domestic violence is common in couples that come for general marital counseling. As noted earlier, O’Leary et al. (1992) found violence rates as high as 67% for couples in a regular outpatient population. Thus, family therapists are already working with violent couples. Unfortunately, they are often not aware that they are doing so. One major implication from the existing body of literature is that family therapists need to become more aware of domestic violence as an issue and make assessment for violence a routine part of treatment regardless of the presenting problem.

Assessment for domestic violence should include both a written assessment instrument and a thorough interview conducted independently with each partner. The most widely used assessment measure is the Conflict Tactics Scale (CTS; Straus, 1979). This is an 18-item self-report inventory of conflict resolution tactics. Both partners indicate whether they or their partner engaged in any of a number of physically aggressive behaviors in the year prior to the assessment.

In addition to a written assessment, each partner should be interviewed separately. The therapists should ask questions in a nonjudgmental, routine way. Because many couples do not label the slapping, hitting, or shoving that takes place in their relationship as violence, it is important to ask about specific acts and behaviors rather than ask general questions about “violence.”

When assessment reveals that there has been violence between client partners, regular couples treatment protocols need to be modified. Although much that family therapists do in couples treatment of nonviolent couples is also useful when there has been violence, all of the couples treatment approaches reviewed in this chapter-tested and untested-have made substantial changes from regular outpatient couples treatment practice. The common core ingredients that appear in most of the successful programs include:

1. Clients are carefully screened into the program.

* Clients who have seriously injured their partners are excluded.

* Both clients (in separate interviews) must report that they want to participate in couples treatment and that they are not afraid to express their concerns to their partners.

2. The primary focus of treatment is on eliminating all forms of partner abuse (physical, emotional, verbal), not on saving marriages.

3. Most programs emphasize taking responsibility for one’s own violence and include a skill-building component including teaching such skills as recognizing when anger is escalating, de-escalating, and taking time-outs.

4. Effectiveness in all the successful programs reviewed here is measured by reduction or elimination of violence.

The therapist’s primary responsibility when working with violent couples is to assure the safety of all family members, and this issue must remain at the forefront of treatment until safety is assured. At times, assuring safety means not agreeing to work conjointly with a couple if the risk of repeated violence is too high, if the perpetrator is not motivated to change his violent behavior, or if the victim is being coerced to come to couples therapy as another controlling tactic by the perpetrator. When couples are judged appropriate for conjoint work, specific modifications that therapists should make include having regular individual meetings with partners-especially victims-to develop safety plans, assess for recurrences of violence, and ascertain whether or not the victim feels safe to continue to conjoint treatment. During the assessment phase, individual meetings allow the therapist to determine whether or not the victim is a willing participant in treatment or is being coerced to come. Finally, family therapists must be familiar with, and use, the safety net of community resources for victims and perpetrators of domestic violence including police and judicial resources, women’s shelters, and group treatments available through batterers’ programs. It is unlikely that couples treatment alone will be sufficient for any but the mildest cases of domestic violence.

As family therapists become more knowledgeable about domestic violence and couples treatment, it is also important that they become more active in the growing regulation of the field. As of 1996, coalitions from more than half of the states and the District of Columbia had adopted standards governing programs or individuals that provide batterer intervention, and 13 other states were developing standards (Healey & Smith, 1998). Although most experts would agree that not enough is known about the effectiveness of current interventions, couples counseling is expressly prohibited in 20 state standards for batterer interventions. Family therapists must educate other domestic violence treatment providers and regulatory bodies about the potential benefits of careful couples treatment in some cases of domestic violence.

CONCLUSION

Violence between partners in intimate relationships is a complex and troubling phenomenon, one that challenges our conceptions of love and intimate relationships and exacts an immense toll in human suffering and social costs. In order to bring domestic violence out from behind the wall of “private family matters,” advocates for battered women some 30 years ago took a strong and uncompromising stance that all violence against women is best dealt with as a crime, that it reflects the patriarchal nature of society, and that any attempt to examine couple dynamics in domestic violence adds to battered women’s victimization. At the same time, some aspects of domestic violence were, of necessity, left out of the advocacy movement’s analysis. In order to advance an important social change agenda, advocates downplayed the prevalence of female assaults on male partners; the wish many victims have to stay with abusive partners, albeit without continued violence; the different subgroups of batterers, each of which might need a different intervention approach; and the possibility that conjoint treatment could help end violence and empower women in their relationship without suggesting they are responsible for being abused. Over the past 10 years, researchers and treatment professionals have begun to consider these issues in more depth, with a growing interest in conjoint couples therapy being one result.

It is important to note that all of the experimental and quasi-experimental studies reported in this chapter have been published since 1995. The topic of marital therapy for domestic violence was not reviewed in the 1995 Special Issue of the Journal of Marital and Family Therapy because no one was conducting experimental studies on this topic at that time. Because domestic violence is such a common occurrence in couples, it is interesting that relative to other issues in this volume, family approaches to domestic violence are not as advanced methodologically. One reason for the underdevelopment of research in this area is that historically, prejudices have existed against conjoint work with couples when violence has occurred.

However, at present, clinical use of couples treatment for domestic violence is increasingly common although the outcome research on this approach remains underdeveloped. It appears as though carefully conceptualized and delivered couples treatment is at least as effective as gender-specific treatment and that couples treatment does not place women at greater risk for injury. However, the body of research on which these conclusions rest is sparse and often unsophisticated. Only six experimental studies have been done and they use different eligibility criteria, outcome measures, and treatment approaches. On balance then, couples treatment certainly appears to hold promise as an effective approach to selected cases of domestic violence, and merits further study. Marriage and family therapists have an important part to play in continuing to develop and test innovative ways of helping couples end violence and improve the quality of their relationships-an endeavor that promises not only to improve the quality of their own lives but also to improve the lives of their children and society as well.

REFERENCES

References marked with a double asterisk are recommended for clinicians.

**Almeida, R. V., & Durkin, T. (1999). The cultural context model: Therapy for couples with domestic violence. Journal of Marital & Family Therapy, 25(3), 313-324.

American Medical Association. (1992). Violence against women. Journal of The American Medical Association, 267(23), 107-112.

**Babcock, J. C., & La Taillade, J. J. (2000). Evaluating interventions for men who batter. In J. P. Vincent & E. N. Jouriles (Eds.), Domestic violence: Guidelines for research-informed practice (pp. 37-77). London: Jessica Kingsley.

Babcock, J. C., & Steiner, R. (1999). The relationship between treatment, incarceration, and recidivism of battering: A program evaluation of Seattle’s coordinated community response to domestic violence. Journal of Family Psychology, 13(1), 46-59.

Brennen, S. J., & Rubin, A. (1996). Comparing the effectiveness of gender-specific and couples groups in a court-mandated spouse abuse treatment program. Research on Social Work Practice, 6(4), 405-424.

Bureau of Justice Statistics. (1998). Violence by intimates: Analysis of data on crimes by current or former spouses, boyfriends, and girlfriends. Retrieved June 3, 2002, from http://www. ojp.usdoj.gov/bjs/abstract/vi.htm

Chen, H., Bersani, C., Myers, S. C., & Denton, R. (1989). Evaluating the effectiveness of a court sponsored abuser treatment program. Journal of Family Violence, 4, 309-322.

Davis, R. L. (1998). Domestic violence: Facts and fallacies. Westport, CT: Praeger.

de Shazer, S. (1982). Patterns of brief family therapy. New York: Guilford.

de Shazer. S. (1985). Keys to solution in brief therapy. New York: W. W. Norton.

de Shazer, S. (1988). Clues: Investigating solutions in brief therapy. New York: W. W. Norton.

Deschner, J. P., & McNeil, J. (1986). Results of anger control training for battering couples. Journal of Family Violence, 1(2), 111-120.

Deschner, J. P., McNeil, J., & Moore, M. (1986). A treatment model for batterers. Social Casework, 67, 55-60.

Dobash, R., Dobash, R. E., Cavanagh, K., & Lewis, R. (1996). Re-education programs for violent men-an evaluation. Research Findings, 46, 1-4.

Dunford, F. W. (2000). The San Diego Navy experiment: An assessment of interventions for men who assault their wives. Journal of Consulting and Clinical Psychology, 68(3), 468-76.

Dutton, D. G. (1986). The outcome of court-mandated treatment of wife assault: A quasi-experimental evaluation. Violence & Victims, 1(3), 163-175.

Dutton, G. D., Bodnarchuk, M., Kropp, R., Hart, S. D., & Ogloff, J. R. P. (1997). Wife assault treatment and criminal recidivism: An 11-year follow-up. International Journal of Offender Therapy and Comparative Criminology, 41(1), 9-23.

Edleson, J. L., & Brygger, M. P. (1986). Gender differences in reporting of battering incidences. Family Relations: Journal of Applied Family & Child Studies, 35(3), 377-382.

Edleson, J. L., & Grusznski, R. J. (1988). Treating men who batter: Four years of outcome data from the Domestic Abuse Project. Journal of Social Service Research, 12, 3-22.

Edleson, J. L., & Syers, M. (1991). The effects of group treatment for men who batter: An 18-month follow-up study. Research on Social Work Practice, 1, 227-243.

Edleson, J. L., & Tolman, R. M. (1992). Intervention for men who batter: An ecological approach. Thousand Oaks, CA: Sage.

Fals-Stewart, W, Birchler, G. R., & O’Farrell, T. J. (1996). Behavioral couples therapy for male substance-abusing patients: Effects on relationship adjustment and drug-using behavior. Journal of Consulting and Clinical Psychology, 64, 959-972.

Fals-Stewart, W., Kashdan, M., O’Farrell, T. J., & Birchler, G. R. (2002). Behavioral couples therapy for drug-abusing patients: Effects on partner violence. Journal of Substance Abuse Treatment, 22, 87-96.

Feazelle, C. S., Mayers, R. S., & Deschner, J. P. (1984). Services for men who batter: Implications for programs and policies. Family Relations, 33(April), 217-223.

Feder (1999). Women and domestic violence: An interdisciplinary approach. Binghampton, NY: Haworth Press.

Feld, S. L., & Straus, M. A. (1989). Escalation and desistance of wife assault in marriage. Criminology, 27, 141-161.

Ford, D. A., & Regoli, M. J. (1993). The criminal prosecution of wife batterers: Process, problems, and effects. In N. Z. Hilton (Ed.), Legal responses to wife assault. Newbury Park, CA: Sage.

Geffner, R., & Mantooth, C. (2000). Ending spouse/partner abuse: A psychoeducational approach for individuals and couples. New York: Springer.

Geffner, R., Mantooth, C., Franks, D., & Rao, L. (1989). A psychoeducational conjoint therapy approach to reducing family violence. In P. Caesar & L. Hamberger (Eds.), Treating men who batter: Theory, practice and programs (pp. 103-133). New York: Springer.

Gelles, R. J., & Straus, M. A. (1988). Intimate violence. New York: Simon & Schuster.

Gingerich, W. J., & Eisengart, S. (2000). Solution-focused brief therapy: A review of the outcome research. Family Process, 39, 477-498.

**Goldner, V., Penn, P., Sheinberg, M., & Walker, G. (1990). Love and violence: Gender paradoxes in volatile attachments. Family Process, 29(4), 343-364.

Gondolf, E. W. (1988). Who are those guys? Toward a behavioral typology of batterers. Violence and Victims, 3(3), 187-203.

Gondolf, E. W. (1998). The victims of court-ordered batterers: Their victimization, helpseeking, and perceptions. Violence Against Women, 4(6), 659-676.

Gondolf, E. W. (2002). Batterer intervention systems: Issues, outcomes, and recommendations. Thousand Oaks, CA: Sage.

Gottman, J. M. (1999). The marriage clinic: A scientifically-based marital therapy. New York: Norton.

**Greenspun, W. (2000). Embracing the controversy: A metasystemic approach to the treatment of domestic violence. In P. Papp (Ed.), Couples on the fault line: New directions for therapists (pp. 152-177). New York: Guilford Press.

Hamberger, L. K., & Hastings, J. E. (1988). Skills training for treatment of spouse abusers: An outcome study. Journal of Family Violence, 3, 121-130.

Harrell, A. (1991). Evaluation of court-ordered treatment for domestic violence offenders. Final report to the State Justice Institute. Washington, DC: The Urban Institute.

Harris, J. (1986). Counseling violent couples using Walker’s model. Psychotherapy, 23(4), 613-621.

Harris, R., Savage, S., Jones, T, & Brooke, W. (1988). A comparison of treatments for abusive men and their partners within a family-service agency. Canadian Journal of Community Mental Health, 7(2), 147-155.

Healey, K. M., & Smith, C. (1998). Batterer programs: What criminal justice agencies need to know. Washington, DC: National Institute of Justice.

**Heyman, R. E., & Neidig, P. H. (1997). Physical aggression couples treatment. In W. K. Halford & H. J. Markman (Eds.), Clinical handbook of marriage and couples interventions (pp. 589-617). Chichester, England: John Wiley & Sons.

Holtzworth-Munroe, A., & Stuart, G. L. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psychological Bulletin, J 16(3), 476-497.

Jaffe, J. H., & Sudderman, M. (1995). Child witnesses of woman abuse: Research and community responses. In S. Stith & M. Straus (Eds.), Understanding partner violence: Prevalence, causes, consequences and solutions. Minneapolis, MN: National Council on Family Relations.

Jaffe, P. G., Wolfe, D., & Wilson, S. (1990). Children of battered women. Newbury Park, CA: Sage.

Johannson, M. A., & Tutty, L. M. (1998). An evaluation of after-treatment couples’ groups for wife abuse. Family Relations, 47(1), 27-35.

Johnson, S. M. (1996). The practice of emotionally focused marital therapy: Creating connection. Brighton, NY: Brunner-Routledge.

Liddle, H. A., & Dakof, G. A. (1995). Efficacy of family therapy for drug abuse: Promising but not definitive. Journal of Marital & Family Therapy, 21, 511-543.

Lindquist, C. U., Telch, C. F, & Taylor, J. (1983). Evaluation of a conjugal violence treatment program: A pilot study. Behavioral Counschild management pmblemseling & Community Interventions, 3(1), 76-90.

**Lipchik, E., & Kubicki, A. (1996). Solution-focused domestic violence views: Bridges toward a new reality in couples therapy. In S. D. Miller, M. A. Hubble, & B. L. Duncan (Eds.), Handbook of solution-focused brief therapy (pp. 65-97). San Francisco: Josey-Bass.

Neidig, P. H. (1992). Participants’ workbook for PACT. State University of New York at Stony Brook.

**Neidig, P., & Friedman, D. (1984). Spouse abuse: A treatment program for couples. Champlain, IL: Research Press.

Neidig, P. H., Friedman, D. H., & Collins, B. S. (1985). Domestic conflict containment: A spouse abuse treatment program. Social Casework, 66(4), 195-204.

Newell, R. G. (1994). The effectiveness of court-mandated counseling for domestic violence: An outcome study. Dissertation Abstracts International, A. The Humanities and Social Sciences, 55(5-A), 1193.

O’Farrell, T. J., Van Hutton, Y, & Murphy, C. (1999). Domestic violence before and after alcoholism treatment: A two-year longitudinal study. Journal of Studies on Alcohol, 60, 317-321.

O’Leary, K. D. (1988). Physical aggression between spouses: A social learning theory perspective. In V. B. Van Hasselt & R. L. Morrison (Eds.), Handbook of family violence (pp. 31-55). New York: Plenum Press.

O’Leary, K. D., Heyman, R. E., & Neidig, P. H. (1999). Treatment of wife abuse: A comparison of gender-specific and conjoint approaches. Behavior Therapy, 30, 475-505.

O’Leary, K. D., Vivian, D., & Malone, J. (1992). Assessment of physical aggression against women in marriage: The need for multimodal assessment. Behavioral Assessment, 14(1), 5-14.

Palmer, S. E., Brown, R. A., & Berrera, M. E. (1992). Group treatment program for abusive husbands: Long-term evaluation. American Journal of Orthopsychiatry, 62(2), 276-283.

Pan, H. S., Neidig, P. H., & O’Leary, K. D. (1994). Predicting mild and severe husband-to-wife physical aggression. Journal of Consulting and Clinical, Psychology, 62(5), 975-981.

Ferez, P. J., & Rasmussen, K. (1997). An ounce of prevention: A model for working with couples at-risk for battering. Contemporary Family Therapy, 19(2), 229-251.

Riza, W. R., Stacey, W. A., & Shupe, A. (1985). An evaluation of the effect of the Family Preservation Program in Tyler, Texas (Vol. 35). Arlington, TX: University of Texas at Arlington, Department of Sociology, Anthropology, and Social Work, Center for Social Research.

Rusinoff, J. S. (1990). Men’s treatment handbook (2nd ed.). Minneapolis, MN: Domestic Abuse Project.

Saunders, D. G. (1992). A typology of men who batter: Three types derived from cluster analysis. American Journal of Orthopsychiatry, 62, 264-275.

Saunders, D. (1996). Feminist-cognitive behavioral and process-psychodynamic treatments for men who batter: Interactions of abuser traits and treatment models. Violence and Victims, 11(4), 393-414.

Shupe, A., Stacey, W. A., & Hazelwood, L. R. (1987). Violent men, violent couples. Lexington, MA: Lexington Books.

Stith, S. M., McCollum, E. E., & Rosen, K. H. (Unpublished). The effectiveness of an integrated solution-focused couples treatment program for domestic violence.

**Stith, S. M., McCollum, E. E., Rosen, K. H., & Locke, L. D. (2002). Multicouple group treatment for domestic violence. In F. Kaslow (Ed.), Comprehensive textbook of psychotherapy (Vol. 4). New York: John Wiley & Sons.

Stith, S. M., & Straus, M. A. (1995). Introduction. In S. M. Stith & M. A. Straus (Eds.), Understanding partner violence: Prevalence, causes, consequences, and solutions (pp. 1-11). Minneapolis, MN: National Council on Family Relations.

Stosny, S. (1995). Treating attachment abuse: A compassionate approach. New York: Springer.

Straus, M. A. (1979). Measuring intrafamily conflict and violence: The conflict tactics (CT) scales. Journal of Marriage & the Family, 41(1), 75-88.

Straus, M. A. (1993). Physical assaults by wives: A major social problem. In R. J. Gelles (Ed.), Current controversies on family violence (pp. 67-87). Newbury Park, CA: Sage.

Straus, M. A., & Gelles, R. J. (1990). Physical violence in American families: Risk factors and adaptations to violence in 8,145 families. New Brunswick, NJ: Transaction.

**Stuart, G. L., & Holtzworth-Munroe, A. (1995). Identifying subtypes of maritally violent men: Descriptive dimensions, correlates and causes of violence, and treatment implications. In S. M. Stith & M. A. Straus (Eds.), Understanding partner violence: Prevalence, causes, consequences, and solutions (pp. 162-172). Minneapolis, MN: National Council on Family Relations.

Taylor, J. (1984). Structured conjoint therapy for spouse abuse cases. Social Casework: The Journal of Contemporary Social Work, 65, 11-18.

Tolman, R. (1990). The impact of group process on outcome of groups for men who batter. Paper presented at the European Congress on the Advancement of Behavior Therapy, Paris, France.

Tolman, R. M., & Edleson, J. L. (1995). Intervention for men who batter: A review of research. In S. M. Stith & M. A. Straus (Eds.), Understanding partner violence: Prevalence, causes, consequences, and solutions (pp. 262-272). Minneapolis, MN: National Council on Family Relations.

Walker, L. E. (1979). The battered woman. New York: Harper and Row.

Wexler, D. (1999). Domestic Violence 2000: An integrated skills program for men. New York: Norton.

Sandra M. Stith, Karen H. Rosen, and Eric E. McCollum Virginia Polytechnic Institute and State University

Sandra M. Stith, PhD, Karen H. Rosen, EdD, and Eric E. McCollum, PhD, Virginia Polytechnic Institute and State University, Falls Church, Virginia.

This article was also published as chapter 8, “Domestic Violence,” in the book edited by D. H. Sprenkle (2002), Effectiveness Research in Marriage and Family Therapy (pp. 223-254). Alexandria, VA: American Association for Marriage and Family Therapy, web site: www.aamft.org, phone: (703) 838-9808.

Correspondence concerning this article should be addressed to Sandra M. Stith, Virginia Polytechnic Institute and State University, Human Development Department, 7054 Haycock Road, Falls Church, Virginia, 22043. E-mail:sstith@vt.edu

Copyright American Association for Marriage and Family Therapy Jul 2003

Provided by ProQuest Information and Learning Company. All rights Reserved