The majority of women entering battered women’s shelters bring children with them (Jaffe et al. 1990). Children who are temporarily residing in shelters may have needs related to the impact of witnessing or being targeted by violence, as well as adjustment with leaving their homes and living in a shelter (Bennett et al. 1999). Several programs targeted toward children who have been exposed to domestic violence have been developed and implemented (Ezell et al. 2000); however, few have been empirically evaluated (Ezell et al. 2000; Jouriles et al. 1998, 2001; Peled 1997a; Rosenberg 1987). These programs vary considerably as to their purpose, the specific types of problems they target, their format (i.e., individual, family, or group), length of program, and participants included (Ezell et al. 2000). The most widely suggested program for children is group counseling, which teaches and reinforces new behaviors and skills (Jaffe et al. 1990). However, support groups are the most prevalent, which provide emotional support (Groves and Gewirtz 2006; Hester et al. 2000; YM–YWCA of Winnipeg [YM–YWCA] 1990).

In this area of research, terminology proves challenging as each term has philosophical implications (Ashcraft 2000; Eliasson and Lundy 1999). Although previous research has referred to wife abuse, woman abuse, partner abuse, or intimate violence, we opted to use the term domestic violence. By using domestic violence, we refer to all forms of abuse and violence in the family, household or dependent relationships. Domestic violence may be perpetrated and experienced by both men and women; however, women continue to be the predominant victims, reporting higher rates of severe violence than men (Krug et al. 2002). Domestic violence also proves fatal more often for women than men (Federal-Provincial-Territorial Ministers Responsible for the Status of Women 2002; Krug et al. 2002).

When children are involved in or witness domestic violence, they often experience adverse effects. Children may experience a variety of problems including severe mood swings, sleep disturbances, somatic complaints, and school phobia, although consequences clearly vary from child to child (Dolon and Hendricks 1991). Therefore, there is a need for children’s services, as well as a need for research on existing children’s services. The purpose of this paper is to critically examine the literature on services for children who have been exposed to domestic violence. There were three objectives of this review: (a) to summarize intervention approaches tailored directly towards children and indirectly through their parents; (b) to review specific programs and their effectiveness; and (c) to identify barriers to service provision and participation.

Articles included in this review were selected from a search of electronic databases PsychInfo, Child Development and Adolescent Studies, and Psychology and Behavioral Sciences Collection using the following search terms: children’s services, transition homes, shelters, domestic violence, and mothers. Sources were also taken from relevant reference lists. Twenty-eight articles met the following criteria: they focused on programs for children who witnessed domestic violence, and they discussed transition homes, women’s shelters, or community family violence programs.

Services for Children

Presently, virtually all shelters offer child care services, that is, temporary care of children’s safety and well-being while mothers are otherwise occupied (Shostack 2001). Sometimes child care is provided by paid staff and other times by other mothers or by volunteers. While such activities may provide an opportunity for informal guidance and modeling to occur when staff members or adults and children interact (Stephens et al. 2000), children’s services in shelters should not be limited to such (Gundy 1981a). Structured daily activities are recommended as a minimum level of service for children (Gundy 1981a). Such activities provide children with positive adult attention and the opportunity to interact with positive adult role models (Stephens et al. 2000).

There are shelters that provide significantly more interventions than minimal child care; for example, child counseling, school integration, support for healing family relationships, recognizing feelings and behavior patterns, addressing hurt, adapting to change, child advocacy, screening and assessment of socio-emotional functioning, and legal advocacy for children’s needs (Groves and Gewirtz 2006; Shostack 2001). Child advocates who address children’s rights in the shelters as well as in legal or other interactions are considered especially important by some authors given that mothers’ and children’s interests may conflict (Peled 1997a, b).

The primary goals of different programs are often the same. Increasing safety is the most basic goal of the majority of children’s programs. This involves helping children identify people in their social networks who they can trust and confide in as well as develop a safety plan when they feel they may experience or witness violence. In addition to safety, it is generally agreed that a well-planned program allows children the chance to release physical tension, which requires adequate space and equipment. This also gives children the opportunity to freely act with few restrictions, try new activities and consider new behaviors, and to make mistakes and learn from them (Gundy 1981a). Thus, personal safety, unconditional acceptance, and physical and mental well-being are the main goals of most programs.

In addition to a group approach, specific strategies can be tailored to the needs of individual children. Accordingly, the first step in supporting a child is to assess the impact of trauma (Jaffe et al. 1990; Stephens et al. 2000). Once the child’s strengths and areas in need of support are identified, various services can be provided accordingly. Through assessment, multiple goals can be identified, including the types of therapeutic interventions that can be provided for specific behavioral and emotional problems (Groves and Gewirtz 2006). In addition to assessing the child’s needs, the various family factors and outside factors that impact the family might be considered for their impact on children. Overall, flexibility is required given the great variability in children’s reactions to exposure to domestic violence. Whereas vulnerable children may need extra support, clinical intervention, and ongoing therapy, resilient children also need support, although it may not be in the form of specialized or continuing services. Thus an understanding of child development, the developmental impact of trauma on children, and the patterns of domestic violence are necessary to plan effective interventions (Groves and Gewirtz 2006).

Developmental Levels

The ability to understand and cope with violence in the family varies as a function of a child’s developmental stage (Dolon and Hendricks 1991). Adolescents tend to be angry toward their mother for removing them from their friends and their routine when they leave an abuser and the family home. School-age children tend to be fearful, but they mainly need assistance in dealing with their new living arrangement. Preschoolers are in particular need of developing a sense of security and are at greatest risk for psychological maladjustment (Hughes 1981, 1986).

With younger preschoolers (3–4 years old), clinical programming tends to consist of interpretive play therapy, storytelling, movement exercises, projective drawing, and the identification of feelings (Ragg and Webb 1992). Safety programming consists of having children identify hiding places and helpers within the community or their social network to be used in conflict situations, and teaching them about personal boundaries. There are also teachable moments that arise spontaneously during group play. Prevention programming involves building non-violent coping skills, encouraging self-responsibility, and providing alternate perspectives on sex roles in relationships. With such young children it is important to implement an activity in a permissive and unstructured manner, and to structure the materials instead of the activities (Ragg and Webb 1992).

Older preschoolers (beginning at about age 4 years) are a distinct group of children, who have learned the basics of cooperation and rules (Ragg and Webb 1992). Their programming involves activities such as interpretive play therapy, structured group activities, puppetry, movement exercises, storytelling, and verbal expression. Safety programming is the same as with younger children, except for the introduction of reaction exercises. This training helps children to develop safety scenarios and then to react to them repeatedly. Prevention training involves group decision making, interactive coaching, and re-enactment, as seen with younger children; however, it also introduces group problem solving, self-responsibility, and process commentary (Ragg and Webb 1992). The most commonly reported activities are discussed next.

Play and Art

Play helps a child develop mastery in many areas: pleasure, exploration and orientation, control of feelings, social relationships, decision-making, body movement, problem-solving, and communication (Gundy 1981a). When incorporated into group activities, play is designed to provide children with an opportunity to learn and practice coping and social skills, which although not a form of play therapy, is considered crucial to gain mastery (YM–YWCA 1990). Group activities typically involve structured play. Such play can be used to shape behaviors or to reinforce adaptive behaviors. Generally, play transcends age (YM–YWCA 1990), but must be developmentally appropriate and attractive to be of value.

Play involves several different mediums, which may be more or less appropriate given certain activities or the age and maturity level of the children (YM–YWCA 1990). Art allows children to unconsciously express and resolve overwhelming problems through symbolic representation. Pieces of artwork are objects in themselves, but they also contain meaning in reference to outside events and objects (Arguile 1992). It is believed that children will transfer their feelings about their experiences to their artwork (Waller and Dalley 1992). Children may project their feelings and displace their conflicts onto puppets and dolls. Lifelike, malleable and anatomically correct figures lend themselves to reenactments of children’s experiences, whereas stuffed animals can be used in representational play. These play media are particularly appealing to preschool children and school-aged girls, whereas school-aged boys often reject such materials as ‘girls’ toys’ (Webb 1999). Play media can be used for learning or practicing new skills or social roles, conveying information, providing feedback, and facilitating discussion (YM–YWCA 1990). Storytelling is a useful method for distancing, identification, and projection. Stories can be therapeutic whether they are read to children or told by the children (Webb 1999). They can be used to deliver a message or moral, to provide empathy, or to demonstrate alternate ways of coping or solving problems (YM–YWCA 1990). Music can be used to convey a range of feelings or to illicit different behaviors in children. It is also particularly useful for children who tend to somatize their problems, as it can help them determine the location of a feeling in their bodies (YM–YWCA 1990). Dance/movement is closely tied to music and is also particularly effective with younger children. It can be used to express feelings and to identify the physical manifestations of feelings (YM–YWCA 1990). Acting allows children to express themselves and to share their experiences through fictional characters or while concealing themselves behind masks and costumes (Jennings 1990). Acting can be used in acquiring and practicing new behaviors or social roles, developing empathy for others, and changing patterns of relating to people. Although it can be used with all ages, it is especially powerful with older children and adolescents (YM–YWCA 1990). Imagery—visualizing a pleasant image or blocking out an unpleasant image or thought—can be used to effectively reduce anxiety (YM–YWCA 1990). Finally, as all games have rules, they have built-in control mechanisms which can be altered to meet the group objectives (YM–YWCA 1990). Organized games become more effective as children age, as children become more interested in reality-oriented organization than in imaginative play (Webb 1999).

Individual Services

Individual support is most essential for children who seem to need considerable attention and who have few sources of support, are blamed for their disruptive behavior, or who specifically express a desire to talk about their feelings (Hughes 1981). Children may also be in need of individual intervention if they have been physically abused themselves (Hughes 1986). When working with children individually—whether formally or informally—it is important to give them time and space to reflect on their experiences and to express their feelings (Hester et al. 2000). Discussing family violence with an adult provides an opportunity to acknowledge the secret of family violence, understand that they are not alone nor at fault, and validate their experiences (Hester et al. 2000).

Group Services

Although the literature is not always clear in what is meant by group work, it is widely accepted that it is possible to develop extensive group programs when children stay at a shelter for several weeks. Such programs should reflect the diversity of the children’s backgrounds and experiences and special care should be taken to ensure that no child is isolated or marginalized within the group (Hester et al. 2000). Groups are usually targeted at school-age children (Jaffe et al. 1990), and according to the YM–YWCA of Winnipeg (1990), 6–11 year olds make a natural grouping, as they are action-oriented and less inhibited than older children. However, Jaffe et al. (1990) stated that children’s groups are usually developed for 8–13 year olds, as these children are most accepting of this approach and comprise a large proportion of shelter residents. Thus, children in middle childhood years are generally considered to be an appropriate age for group services.

Children’s groups often provide the first opportunity for children to discuss the violence in their family. They are able to learn that they are not alone in their experiences (Jaffe et al. 1990). Although these groups are underdeveloped in comparison to services for abused women and their male perpetrators, children tend to welcome them when they are offered (Hester et al. 2000).

New behaviors and skills can be taught and reinforced through counseling group activities. These activities are structured to be short-term and to produce immediate, measurable results. As the focus of intervention is on skills and the resulting behaviors, group activities assume that emotional or psychological adaptation will succeed the acquisition of skills and behavior. Implementing a group activity is ideal in shelters as there is a captive audience on site and there are few practical considerations, such as transportation and attendance (YM–YWCA 1990). One of the challenges for group programs, however, is that many shelters do not provide accommodation for the “several” weeks required.

Support groups require greater commitment and motivation, as they are designed to be more long term than group activities and they demand more active participation. Such groups provide emotional support, with the understanding that such support will help children to act in healthy ways and to try new skills and behaviors over time. The results of support groups are evident over a longer period of time and tend to be variable across participants. In support groups, it is the group itself that serves as the medium of change. The focus of intervention is on emotional adaptation, and there is the assumption that positive emotional adjustments will precede the acquisition of new skills and behaviors (YM–YWCA 1990).

Despite their potential advantage, groups are not always feasible in shelters. As the traditional mandate of shelters is to provide emergency shelter to women and their children, connect them with the required community resources, and help resolve their immediate crisis, groups often exceed a shelter’s mandate. Unless adequate resources, such as space, transportation, materials, and staff, are available, expanding the mandate to provide long-term services that address broader issues may actually weaken the core residential program (YM–YWCA 1990).

Services for Parents

Ideal interventions involve both parents; however, given certain ethical and practical considerations, such as the potential risk to women and children, it is not always advisable to include the perpetrator (Jouriles et al. 2001). That being said, the active involvement of the non-offending parent, typically the mother, is an essential component of intervention with children (Groves and Gewirtz 2006). Working with mothers indirectly helps children, as children need to trust that at least one parent is a reliable protector (Groves and Gewirtz 2006; Gundy 1981b). Offering services to mothers also directly benefits children, and is commonly a goal for many shelters (Stephens et al. 2000). For example, The Child Trauma Research Project concentrates on restoring the positive developmental trajectories of children by improving their relationships with their primary caregiver (Groves and Gewirtz 2006). In addition, the Child Witness to Violence Project takes an ecological approach by helping children through interventions with both the mother and the child (Groves and Gewirtz 2006).

Supportive counseling and some parenting relief empower the mother, allowing her to function more effectively as a parent (Jaffe et al. 1990). Mothers are also taught parenting skills in order to strengthen the parent–child relationship (Shostack 2001). It is possible to reduce children’s conduct problems by teaching parents how to respond effectively to their children’s misbehavior, and how to foster and facilitate appropriate behavior (Jouriles et al. 2001). Such skill training is considered to be an effective intervention with preschoolers and young school-age children (Jouriles et al. 1998). Parenting sessions generally consist of discussions on a number of different topics, including limit setting, ways to increase positive interactions with their children, improved (non-violent) disciplinary techniques, normal developmental behaviors, and behaviors that may be a cause for concern. Parenting education programs that build on the mothers’ strengths tend to be appreciated by mothers (Bennett et al. 1999).

Intra- and Intersectoral Collaboration and Coordination

A number of recommendations for children’s services were identified in the literature. Children vary greatly in their response to domestic violence; therefore, there is no one intervention to meet the needs of all children (Jouriles et al. 1998; Stephens et al. 2000). As such it is important to collaborate with other service providers (Groves and Gewirtz 2006). For example, the Institute for Safe Families is the result of a collaborative effort among the child welfare system, domestic violence services, battered women services, child treatment providers, legal aid, and substance abuse providers. These services are intended to provide integrated and comprehensive support to families affected by domestic violence (Groves and Gewirtz 2006). Coordination with schools and child protection services is thought to be the most essential as they directly impact children (Jaffe et al. 1990). Also, it is important to consider the child’s adjustment within the broader context of the child’s and family’s life. Many families affected by domestic violence are also affected by poverty, homelessness, and other forms of violence (Groves and Gewirtz 2006). With this in mind, it is important to provide a comprehensive approach that addresses the various physical and psychological needs of children and their families.

Barriers to Services

Despite increased awareness of the potential negative impact on children of living with the abuse of their mothers, only a small portion of children of women who experience interpersonal violence actually receive help specific to their needs (Peled 1997a). There are several challenges to engaging families affected by domestic violence. These include stigma about help-seeking, concerns about state child protection involvement and apprehension, safety concerns, and transportation or other economic barriers (Groves and Gewirtz 2006). In addition, most families who turn to domestic violence shelters are primarily seeking refuge from violence, not counseling or other interventions (Stephens et al. 2000).

There are also several challenges to the implementation of children’s services: limited number of staff (Stephens et al. 2000); space (Hughes 1982; Stephens et al. 2000); funding (Groves and Gewirtz 2006; Stephens et al. 2000); and length of time spent at the shelter (Gibson and Gutierrez 1991; Groves and Gewirtz 2006; Hughes 1982; Jaffe et al. 1990; Stephens et al. 2000). The brevity of contact is a major barrier to service provision. On average, women and children stay at a shelter for 1 week (Jaffe et al. 1990), and some leave after only a day (Jaffe et al. 1990; Stephens et al. 2000). Some families leave the shelter with little or no notice. Given the uncertainty of a family’s stay at a shelter, it may be unethical to initiate a service without being able to adequately address complex problems. Children who have been exposed to domestic violence often experience problems that tend to take longer than a few days or weeks to address (Stephens et al. 2000). Hughes (1982) found that families who stayed at a shelter for 2–4 weeks received more services, and that those services were more effective than for families who stayed for a shorter period.

In a study by Peled and Edleson (1999) the most commonly identified barriers were technical difficulties, such as time conflicts, transportation, child care for siblings, and group waitlists. Violence-related stress was another factor. Many families experienced elevated levels of stress, and their energy and resources were depleted. Parents’ perceptions of their children’s needs were also important in determining whether or not children participated in services. Some parents believed that their children were unaware of or did not witness the violence, or that their children were too young to understand or be influenced by what they had witnessed. Others did not believe that the severity of the violence in their family warranted their children’s participation in a domestic violence program. Several parents believed that the program was only intended for children who experienced difficulties. Some parents believed that their children would be hurt by being exposed to others’ experiences, whereas others projected their own negative feelings about their group experience in the adult program. There was also resistance from fathers who denied being violent toward their partner. Some men felt rejected by the program, which seemed to impact the family’s experience of the program. Finally, children themselves sometimes resisted participation or were ineligible for the program due to behavioral and emotional difficulties.

In summary, several family-related factors were identified, including situational difficulties and parents’ perceptions of children’s needs and their own experiences in the adult program. There were also agency-related factors, such as timing and distance, resources, and the agency’s difficulty in accommodating abusive men. It is necessary to ensure that parents have a clear understanding of a program’s goals, the target population, and the ways in which a program may benefit their children (Peled and Edleson 1999).

Evaluation of Children’s Programs

Only three evaluations of children’s programs were identified in the published literature so each is reviewed briefly.

Project Support

The goals of Project Support were to reduce children’s conduct problems, to enhance parenting effectiveness (Groves and Gewirtz 2006), and to ultimately end the cycle of violence (Ezell et al. 2000). Designed specifically for mothers and children who have been affected by domestic violence, the program is intended to meet the needs of families who are leaving shelters, have few financial resources, and lack transportation. To be eligible for the program, mothers must be trying to leave the abusive partner and have at least one child (4–9 years old) who exhibits clinical levels of conduct problems (Ezell et al. 2000; Jouriles et al. 1998, 2001). Mothers are provided with instrumental and social support, given basic home supplies and referrals to community services, and are taught stress-reducing decision-making and problem-solving skills. Mothers are also trained in child management skills. The training focuses on enhancing parent-child communication, promoting prosocial behavior, and decreasing deviant, antisocial behavior (Ezell et al. 2000). Parenting coaching sessions with a therapist, advocacy, and mentoring for children are provided weekly over an average of 8 months. A thorough assessment of the child and family serves to guide individualized intervention plans (Groves and Gewirtz 2006).

The program was evaluated by conducting six comprehensive assessments, which included questionnaires concerning family functioning, child behavior, mother’s social support and psychological functioning, and mother’s use of child management skills. Families were randomly assigned to either the experimental or comparison condition. Videos of mothers and children interacting were analyzed and teachers provided data with regard to the children’s school behavior and academic performance. According to Ezell et al. (2000), the conduct problems of children involved in Project SUPPORT improved to a greater extent and at a faster rate than did the problems of children in a comparison group over the course of the follow-up assessments. Also, the psychological distress of mothers in the intervention group decreased more quickly than it did for mothers in the comparison group over the same time period; although, distress decreased equally across groups by the end of the six follow up assessments. Finally, mothers in the intervention group experienced greater and faster improvements in parenting skills than did mothers in the comparison group (Ezell et al. 2000). In a follow up study including both intervention and control mothers, McDonald et al. (2006) found that these results persisted at 24 months following the termination of services. Thus, this program has documented evidence of effectiveness in reducing children’s conduct problems and enhancing parenting effectiveness, but reduction of the cycle of violence was not studied.

Group Program for Children Exposed to Wife Abuse

Wilson et al. (1989) authored the Manual for a Group Program for Children Exposed to Wife Abuse for children ages 8–13 years old. The program addresses important areas of emotional, behavioral, and cognitive problems associated with witnessing domestic violence. It aims to address children’s adjustment difficulties, while also attending to the more subtle symptoms related to attitudes about violence and responsibility for adult behavior. The program consists of ten weekly sessions, which follow a structured format; however, individual needs and ongoing crises are addressed with considerable flexibility (Jaffe et al. 1990).

There are ten sessions to this program that begin with introducing the purpose of the group, explaining confidentiality, and discussing common experiences. Following sessions include labeling and expressing feelings, dealing with anger, and comparing anger and violence. Safety skills, social support, self-concept, and self-blame, common for these children, then become the focus. The final sessions focus on discussing myths that children tend to hold about domestic violence, acknowledging the reality of family life, and reviewing progress before the end of the group sessions (Jaffe et al. 1990; Wilson et al. 1986, 1989).

Wilson et al.’s (1989) group program focuses on adjustment problems as well as subtle attitudes about violence and responsibility for adult behavior. Wagar and Rodway (1995) conducted an evaluation of the program, finding a significant difference between pre- and post-group attendance. The difference was particularly apparent in relation to attitudes about and responses to anger, and in understanding that children should not be held responsible for adult perpetrated violence against them or another parent. The program was effective in attitudinal, rather than behavioral, changes. It should be noted that the study involved a small sample and lacked follow-up assessment and a comparison group (Ezell et al. 2000); therefore, there may be other explanations for the positive changes. This sort of group counseling program may be most appropriate for mild-to-moderate behavior problems in children, as children who have been exposed to, and have possibly themselves experienced, repeated acts of severe violence over several years typically require more extensive individual support (Jaffe et al. 1990).

A Model Preventive Program

This program intervenes with children, mothers, and school and shelter staff. Interventions with children are conducted on both an individual and group basis. The individual services for children range from a recreational approach to longer, more intensive interventions which resemble traditional child psychotherapy. If a child is particularly withdrawn or aggressive, or remains distressed after an extended period, it is recommended that the child attend professional counseling (Hughes 1982). There are also weekly “Children’s Group” meetings that address topics such as coping with change, developing friendships, exploring feelings about abusive fathers, and discussing attitudes about physical punishment (Hughes 1982, 1981).

Interventions with mothers take an educational approach which aims to improve parenting skills, and enhance mothers’ knowledge of their children’s emotional needs and how these needs vary according to age and sex. A weekly parenting group helps mothers learn how to improve their relationship with their children and how to set firm, consistent limits for their children. They are also introduced to basic principles of child development (Hughes 1982).

The program also provides school liaison services to help children transition to a new school. Additionally, one staff member is assigned the role of child advocate and is the primary staff member who attends to the child’s interests. It is also stressed that all staff members should be knowledgeable of appropriate ways of interacting with children. Furthermore, they need to be trained to act as models for mothers, while providing on-going support for the mothers’ attempts at improving their relationships with their children (Hughes 1982).

An evaluation of this program was found to result in lowered levels of child anxiety at the end of the program; although, no differences were found in children’s overt behavior or parent’s parenting skills (Hughes and Barad 1982). Again, no follow-up assessments were conducted and the study did not include a control group (Ezell et al. 2000). Thus, conclusions about effectiveness are limited. Given the complexity of the program and multiple participants, a large sample and extended time would be needed to detect significant changes.

Conclusion

This review has identified some of the primary needs of children who have been exposed to domestic violence, recommendations for service and some of the barriers to access and provision, as well as some of the general and specific services that have been developed to meet those needs. There is generally a lack of research on this topic and clearly a paucity of evaluations of services for children in shelters (Groves and Gewirtz 2006). What this means is that we not only do not know if services are beneficial and a good use of scarce resources, but we do not know if they do harm. Even if they do no harm, but do not actually improve outcomes more than doing nothing, these programs would result in opportunity cost. That is, resources could be better used where we know they have added benefit.

Shelters tend to offer services and training programs that meet residents’ basic needs; however, services to help break the cycle of violence are crucial as the ultimate form of prevention (Dolon and Hendricks 1991). Given the limited resources and short-term stay of most shelters, it may be that other community services should take responsibility for offering support to children. Without rigorous research, however, we cannot be sure what the program logic models should look like, that is, who should be served, over what period, and with what kind of staff. Clarity is needed in what specifically will change in children after an intervention. Since not all participants can be expected to accomplish the same degree of change, we need to consider what proportion of change would represent actual “success” for a program. There are many gaps in the literature and few clear directions for program planners who care about the needs of children who visit shelters. In addition, we need to understand whether the needs of children who witness domestic violence and never go to shelters are different.

Ethical issues also need attention in this field of research, including caution about pathologizing all children who witness domestic violence. As Jouriles et al. (1998) and Stephens et al. (2000) stated, children vary widely in their responses to domestic violence. This can be attributed in part to the resources provided to them by their mothers and other members of their social networks. It should not be assumed that all children who witness domestic violence have an emotional or psychological deficit; however, this is a trend.

In addition, researchers and program planners need to be clear on their stance on gender equality and parenting. It can be difficult for mothers and service providers to offer programs for children in a social climate where mothers in general tend to be blamed for their children’s problems and are subject to state interventions that punish them (Swift 1995). For instance, will a mother who refuses permission for assessment of her child in a shelter or who does not want a child admitted to a program be subject to referral to a state child protection agency because the program is considered “in the best interests of the child”? How will the most marginalized women (e.g., street workers, poor Indigenous women) be impacted by these programs?

Given the extent of domestic violence and the lack of success in lowering the prevalence internationally, programs that have a goal of breaking the cycle of violence are important. This review has identified that intervening with children is widely considered to have potential for prevention. The evidence that many children do suffer from social, emotional, and physical outcomes of witnessing domestic violence may need to be strengthened and clarified; however, the stories that are available point to a compelling need for successful programs with the goal of improving children’s well-being. There is a pressing need for research that builds an understanding of children’s needs, points to the elements of successful interventions, and evaluates the effectiveness of new and innovative approaches.