Introduction

 

 

This overview is an updated version of the 1992 publication, "Children Witnessing Parental Violence: Current and Future Effects" (Cooper, 1992), produced by the BC Institute Against Family Violence. The original version covers a range of issues within this subject area including the effects of witnessing inter-parental conflict and violence, treatment and prevention measures, legal issues, and a discussion of the intergenerational transmission of violence. In the past decade, there has been a significant amount of research contributing to our knowledge in this area. This updated version will discuss highlights of research produced since 1992. However, in order to obtain a broad view of the development of the literature on children witnessing domestic violence, it is suggested that these two overviews be read consecutively in chronological order. Readers wishing to make a comprehensive review of the research in this area are encouraged to also consult recent publications by Gayla Margolin, including "The effects of family and community violence on children", and "Effects of domestic violence on children" (consult references section of this document for publication data).

 

Violence among intimates is a serious social problem. According to the 1993 National Violence Against Women Survey (NVAWS), 29% of Canadian women have experienced violence at the hands of a spouse or common-law partner (Statistics Canada, 1993), and 88% of victims of spousal violence incidents reported to Canadian police agencies in 1997 were women (Statistics Canada, 1999). Eight (8%) per cent of Canadian women and 7% of men sampled in the 1999 General Social Survey on Victimization reported experiencing some form of partner violence at some point during the previous five years. In the same study, we learned that in approximately 460,000 Canadian homes, children were exposed to partner violence. Only some of these children take refuge in transition homes and shelters: more than 39,000 children were accompanied by their mothers to 448 safe homes across Canada during a one-year period spanning 1999-2000 (Statistics Canada, 2001). However, these figures, stunningly large though they may be, are certainly an underestimate of how many children hear or see inter-parental assaults.

 

Despite the growing attention given to family violence issues since the 1970s, children exposed to domestic violence have only comparatively recently been recognized as victims. Once this neglected area was recognized, however, a body of research documenting the harmful effects of childhood exposure to domestic violence soon developed. This paper presents highlights of recent research findings describing the impact of partner violence on children, and explores, in brief, currently advocated therapeutic approaches for children coping with the effects of exposure to partner violence.

 

Effects on Children of Inter-parental Conflict and Divorce

 

"Marital conflict" is a broad term that can encompass a range of behaviours, from expressing anger and verbal disagreement to acts of physical violence. Before examining children's experiences of physically violent parental conflict, we will review research concerning the effects of non-violent marital conflict. In the following section, we will explore the ways which children appraise and respond to inter-parental conflict, the effects of this conflict on parent-child relations, and differences in children's adjustment patterns by age and gender.

 

There is general agreement that an association exists between marital conflict and adjustment and development problems in children (Wallerstein & Lewis, 1998; Cummings & Davies, 1994; Grynch and Fincham, 1990). However, our understanding of the mechanisms by which maladjustment occurs is far from clear. Negative effects may be a direct result of exposure to marital conflict, an indirect outcome of the conflict (possibilities include poverty upon separation, lack of social support, decline in parenting quality, frequent moves or disruption, time spent away from school, etc.), or the result of an interaction of both direct and indirect influences. This paper will examine both questions with the understanding that each are important contributors to the overall experiences of children (Fauber & Long, 1991: Osbourne & Fincham, 1996).

 

Two factors that have been identified as affecting children's experiences of marital conflict are the severity of the conflict and the way in which disputes are resolved. Rogers & Holmbeck (1997) found that parental conflict of greater frequency and intensity leads to greater maladjustment problems in children, and more specifically, to increased externalizing behaviour and depression (Grynch & Fincham, 1990; Amato & Keith, 1991). Many studies have concluded that the degree of severity of conflict is the most influential factor in determining children's well being. This factor has been shown to supersede many other factors expected to have an equivalent effect, such as marital dissolution, economic disadvantage or parental absence (Amato & Keith, 1991; Depner et al, 1991; Lutzke et al, 1996; Lee, 1997; Emery, 1998).

 

The second factor influencing the way in which a child experiences marital conflict is the manner in which the parental figures resolve the conflict. In a study directed by Davies, Myers & Cummings (1996), it was shown that both younger children and adolescents respond to the emotionality displayed by parents in the process of conflict resolution - that is, the kind and intensity of emotion characterizing parental interactions. From their findings, Davies et al made the following conclusions about the effects that positive resolution, or a "harmonious ending" can have on a child's experience of marital conflict:

 

1.      Children's negative emotionality was reduced.

2.      Children and adolescents were less motivated to become involved in the resolution of the conflict.

3.      Secure representations or appraisals of adult relations were promoted by de-escalation of hostility. These conclusions are consistent with the cognitive-contextual framework introduced by Grynch and Fincham (1990) which suggests that the impact of marital conflict on children is in part determined by children's appraisal of the situation.

 

Even in circumstances where inter-parental conflict appears to be minimal, children may be profoundly affected by divorce and familial dissolution. Judith Wallerstein and her colleagues have followed a sample of children of divorced parents into adulthood for the last 25 years. In the short-term, children reported feeling lonely, ashamed, and afraid of being abandoned. The break-up also led to real financial difficulties and a lowered standard of living for the children and their primary caregiver (typically mothers). These influences were observed to increase in their impact over time. As the children reached their teens, half reported serious drug/alcohol use and early sexual activity. By adulthood, approximately half of these children of divorce were earning less income than their parents and had obtained less education. Little money was available from parents to pay for post-secondary education in cases where children did complete high school. Most had difficulty establishing long-term relationships and expressed feeling apprehensive about marriage. Relationships with their fathers continued to be problematic, even in their adult years (Wallerstein & Lewis, 1998).

 

 

Exposure to Marital Conflict Versus Marital Violence

 

How might the effects of marital distress be different from marital conflict that is not physically violent? According to Margolin (1998), the research methodology in this area has progressed from the use of uncontrolled case studies to case studies that control for specific factors such as marital discord versus abuse. These studies have illustrated the fact that exposure to violent conflict tends to predict more severe short and long-term problems for children (McNeal & Amato, 1998). Many studies, such as the one conducted by Jouriles, Murphy and O'Leary (1989), have shown that marital violence contributes to the prediction of conduct disorder for boys and "inadequacy-immaturity", or developmentally inappropriate behavior, in both boys and girls. This is consistent with the findings of Adamson & Thompson (1998), who reported that children who had witnessed spousal abuse, as opposed to non-violent conflict, displayed more emotional intensity and greater sensitivity to scenarios depicting marital conflict. Margolin warns, however, that other studies have failed to obtain these results (Hershorn & Rosenbaum, 1985; Rosenbaum & O'Leary, 1981). Thus, this remains an area for further study.

 

 

Effects on Children of Exposure to Partner Violence

 

Before discussing the behavioral outcomes of children exposed to partner violence, it is important to explain why this terminology was selected and what is meant by this phrase. Much of the literature on this subject has concerned "children who witness domestic violence", a term meant to indicate that the child is within visual range of the altercation (Edleson, 1998). We believe this assumption is far too limited as it excludes a significant range of possible ways in which children may be exposed to partner violence. Children may, for example, be physically present during the violence, overhear the violence, witness the outcome of an assault, be used to intimidate or control the mother, experience diminished caregiving as a result of the violence, or be themselves direct victims of abuse (McGee, 1997). Thus, we have sought to use a term that has a broader connotation in this paper. Also, in the interest of clarity, we have chosen to refer to "domestic" or "family" violence as partner violence in this review so as to not confuse the many forms of family violence (which include child abuse) with childhood exposure to inter-parental violence, and to be respectful of the varieties of adult partnerships that may be involved.

 

Domestic violence threatens the lives and well-being of a large proportion of Canadians, and the healthy development of many of our children. As presented earlier, the Violence Against Women Survey estimated that 4 in 10 women who reported having experienced violence in their relationships also reported that their children had witnessed the violence (Statistics Canada, 1993). But many experts (e.g., Jaffe, Wolfe & Wilson, 1990) suggest that the rate of child exposure to domestic violence in these families is much higher - perhaps as high as 80%. Furthermore, it can be argued that children are affected by domestic violence as long as their parents are, whether children hear, see, or otherwise directly experience violent interactions between parents. For several reasons, it is difficult to estimate a reliable and accurate figure representing the experiences of children. First of all, there are problems with collecting sensitive information directly from children. Secondly, information collected from parents can be unreliable, as parental agreement concerning whether and what kind of violence has occurred is generally low (Margolin, 1998). Despite the lack of precision in measuring rate of exposure, available evidence suggests that this is a critical area for both investigation and intervention.

 

Heterogeneous Outcomes

 

Not all children exposed to domestic violence are affected in similar or measurable ways. Reviews of research in this area make it clear that that there is no single pattern of behaviour exhibited by children exposed to partner violence. What accounts for this variation in outcomes? Firstly, findings are unclear in terms of which features of the abuse, as well as which characteristics of the child or other factors, determine whether a child will be significantly affected by childhood exposure to violence. And a key question to ask in these circumstances is whether negative effects on children are attributable to exposure to violence alone, or to exposure to violence and other possible problems that can be encountered in a violent domestic setting (Margolin, 1998).

 

Children's symptoms of distress may be systematically assessed by looking at the child's functioning in several domains (Kolbo, Blakely, & Engleman, 1996; Wolak & Finkelhor, 1998). Children exposed to partner violence have shown in numerous studies to be more likely to experience a range of problems, including:

 

·         Behavioural problems, which can include aggression, tantrums, acting out, immaturity, truancy, delinquency and hyperactivity.

·         Emotional problems, including anxiety and prolonged stress responses, depression, isolation/alienation, low self-esteem, anger, among other effects. According to Sudermann & Jaffe (1999), children exposed to partner violence are 10-17 times more likely to suffer emotional and behavioral problems than children from non-violent domestic settings.

·         Interpersonal difficulties, such as poor social skills, peer rejection, mistrust of others and a lack of empathy.

·         Cognitive problems can include poor academic performance, difficulty with language development and poor concentration.

·         Possible physical outcomes are failure to thrive, sleeplessness, eating disorders, poor motor skills and psychosomatic symptoms.

·         In addition, "subtle" symptoms of an attitudinal nature (Sudermann & Jaffe, 1999) may be present, including condoning the general use of violence, dismissing the effect of violence on victims (including victim blaming - "she deserves it"), and believing violence is a way of life in intimate relationships. These effects of exposure to partner violence can be especially dangerous if they are maintained when the child grows older and forms intimate relationships with others.

 

This is a comprehensive list that broadly outlines the possible difficulties in adjustment many children may express. Recently, a number of reviews have been generated on this topic that discuss the influence of variables such as age, gender, frequency and severity of violence, aspects of the parent-child relationship, and the co-occurrence of direct child abuse on outcome (Margolin, 1998; Edleson, 1999). The next sections will explore how several of these factors have been observed to contribute to the development and adjustment of children exposed to parental violence.

 

 

Moderating and Mediating Variables Influencing Children's Response

to Partner Violence

 

As mentioned, not all children exposed to partner violence experience adverse effects, nor do those who are negatively impacted express their distress in the same way. As consideration of the wide variety of possible short-term and long-term outcomes for these children has become more sophisticated, researchers have realized that a broader perspective needs to be adopted to address the many factors that contribute to a child's response to partner violence. It is hoped that this approach will help us to understand why some children appear unaffected, why others appear to be greatly impacted, and why still others show little sign of distress in the short-term, but experience difficulties later on.

 

Moderating Variables. Influences that are present in the child's life at the time of exposure to violence can significantly moderate the child's response to violence exposure - these influences are termed "moderating variables" in developmental psychology literature. One commonly referred to and powerful moderating variable is social support, which can act as a buffer between the child and the stresses he or she might be undergoing. Children exposed to partner violence may be fortunate to have a positive relationship with a family member, teacher, or friend with whom they can confide, and this relationship may serve to bolster their coping ability. They may also perform well in activities at school, or have a special talent that provides for them a sense of belonging and promotes self-esteem.

 

Mediating Variables. Similarly, circumstances characterizing the child's life some time after exposure to partner violence may also play a role in determining the child's adjustment. These circumstances may be present soon afterward, or much later. For example, experiences in adolescence may help the child cope with childhood adversity, and ease the transition into adulthood. These distal influences are termed "mediating variables", because they act as mediators between prior and present experience. In this way, social support may be a mediating as well as a moderating variable, if the impact of relationships developed later in the individual's life influences his or her ongoing adjustment.

The effects of a number of mediating and moderating variables on outcome following childhood exposure to partner violence have been investigated. Many of these are outlined in the illustration found on the following page, and are discussed in the following passages.

 

 

General Familial Functioning and Conditions

 

It is difficult, if not impossible, to determine with confidence the relative weight contributed to adverse impact on children by violence and abuse compared to other factors that can negatively characterize the family environment. Non-violent partner conflict and separation and divorce have been discussed, but other important variables need also to be considered. Although partner violence occurs across all social strata, violence that is reported to police and government agencies is more likely to occur in chaotic families characterized by poverty and unemployment, lower levels of educational achievement, single parent or blended families, and families undergoing changes in structure/membership. In addition, there is some evidence that parenting quality suffers in violent homes (e.g., Levendosky & Graham-Bermann, 2001; Statistics Canada, 2001). This finding was disputed, however, in one recent study that observed no significant variance in maternal care and form of child discipline used between comparison groups of homes in which violence was present and in which it was absent (Sullivan & Bybee, 2001). A number of researchers are currently attempting to identify the many familial and environmental factors that may influence how a child responds to partner violence, and disentangle these factors from the effect violence exposure alone may have on children.

 

 

Co-occurrence of Direct Forms of Child Abuse

 

It's been estimated that children experience direct forms of abuse in 30-70% of homes in which partner violence also occurs (Edleson, 2001). As can be expected, research has shown that children who witness the abuse of a parent and also experience direct abuse exhibit greater difficulties (Sternberg et al, 1994; O'Keefe, 1994; Edleson, 1999) when compared to children experiencing single abuse forms. Studies have also found that adults, when asked to describe their experience of both witnessing domestic violence and direct child abuse, specifically acknowledged exposure to partner violence and attributed many of their problems to this experience  (e.g., Silvern et al, 1995). Silvern and her colleagues propose that these two experiences have somewhat independent effects on children. Other research has suggested that the experience of child abuse interacts with the experience of witnessing violence, and increases the severity of trauma resulting from that exposure (O'Keefe, 1996; Edleson, 1999).

 

Children residing with mothers who are being abused by their partners are also 12-14 times more likely to experience sexual abuse by their mother's partner, and 7 times more likely to be sexually abused by someone outside of the family home (Margolin, 1998). Exposure to parental violence, then, might reasonably be considered a risk factor for child abuse in the present and poor outcome in the future.

 

If the young person is also being maltreated by siblings or peers (i.e., bullying), greater challenges to the child's adjustment may be expected.

 

Age of the Child

 

Infants who witness violence have been found to experience poor health, poor sleeping habits and increased agitation (Jaffe et al, 1990; James, 1994). Younger children, such as preschoolers, are more likely to exhibit emotional distress, immature behaviour and somatic complaints. School-aged children were more likely to evaluate the violence negatively and to attempt purposive behavioral interventions, resulting in greater risk for direct harm. Adolescents exhibited increased anxiety and aggression (Margolin, 1998). These findings notwithstanding, Margolin explains that despite evidence pointing to the connection between age and children's adjustment, there are still no clear patterns of age-related effects of witnessing violence. Further, there is some question whether the effects of age and developmental stage are confounded with the effects of repeated exposure and the level of development the child was at when he or she was initially exposed. For an in-depth discussion of effects on children of partner violence by age and developmental level, please see Sudermann & Jaffe, 1999.

 

 

Gender of the Child

 

Children's adjustment patterns frequently vary as a function of their gender, as well as age and developmental stage (Davies, Myers & Cummings, 1996). Many child development researchers have categorized children's behavioral responses to distress as either predominantly externalizing or internalizing in nature (Achenbach, 1991; Crockenberg & Forgays, 1996; Adamson & Thompson, 1998). "Internalizing" behaviours or symptoms include depression, withdrawal, and lowered levels of motor activity. "Externalizing" behaviours include generally more manifest signs of distress, such as anger, physical aggression and hyperactivity (Crockenberg & Forgays, 1996; Davies, Meyers & Cummings, 1996).

 

In general, research has concluded that witnessing violence predicts an increase in internalizing and externalizing behaviours for both boys and girls (Jouriles & Norwood, 1995; O'Keefe, 1994). Exposed boys tend to demonstrate higher levels of aggression, externalizing behaviours, and total behaviour problems (Doumas et al, 1994; Jaffe et al; 1986; Kerig, 1998). In a number of studies, girls exhibited greater internalizing behaviours -anxiety, lowered self-esteem and propensity to taking on responsibility for their mother's well-being (Holden & Ritchie, 1991; Spaccarelli et al, 1994; Kerig, 1998). However, several researchers have failed to find gender-related differences when assessing internalizing and externalizing behaviours, social competence and depression in children exposed to partner violence (Fantuzzo et al, 1991; Jaffe et al, 1986; Margolin, 1998; Spaccarelli et al, 1994). In sum, research findings on gender differences in children's response to adversity are not entirely consistent, but there appears to be some support for the internalizing/externalizing distinction.

 

Nature of the Violence: Frequency, Severity and Time Since Violent Incidents

 

Relatively few studies have been able to obtain detailed information about the nature of the domestic violence that children witness. For reasons previously discussed, this information can be difficult to acquire, particularly within a research setting. However, in the few studies that have aimed to assess the degree to which these variables affect outcome, findings suggest that the severity, chronicity and recency of partner violence exposure exert considerable impact on children's overall adjustment. One of the ways in which the effects of the nature of the violence may be gauged is by comparing known high-violence samples with known lower-violence samples. In comparison with abused women living in the community, women in shelters report a much higher number and greater frequency of violent incidents of partner abuse (Straus, 1990). Typically, children temporarily living with an abused parent in places of refuge - transition homes, shelters and other safe homes -exhibit greater social and personal problems than do those exposed to partner violence, but who remain resident in their own communities (Edleson, 1999; Jaffe et al, 1986). Based on this knowledge, and other anecdotal evidence, we may assume that maladjustment, as a rule, is more likely given exposure to violence of greater frequency and severity and that the longer the period of time that has passed, the less likely a child is to exhibit problems (Wolfe et al, 1986; Edleson, 1999). This position is consistent with the literature on non-violent martial conflict.

 

 

Homicide of a Parent

 

Children who witness parental homicide are exposed to the most extreme act of domestic violence. These children are emotionally traumatized by the event, and may also be at risk for neglect, as attention becomes focused on the crime, the perpetrator and the victim (Burman & Allen-Meares, 1994). A study conducted by Eth & Pynoos (1994) looked at the immediate psychological outcomes of witnessing parental homicide by interviewing young victims within hours or weeks of the event. They found that these children experience a particular variety of post-traumatic stress disorder.  The longer-term outcome of this psychic trauma was influenced by both personal and environmental mediating and moderating variables, such as self-esteem, self-efficacy and coping skills and social support.

 

 

Influence of the Parent-Child Relationship

 

In general, the quality of children's relationships with their mothers has been identified as a key factor in determining a child's well-being (Edleson, 1999). But how do aspects of the mother-child bond affect children's outcomes in cases where partner violence is present? It is clear that violence can place enormous stress on women, therefore adding additional challenges to their parenting. Data show that women who are victims of abuse are more likely to experience maternal stress, mental health problems and alcohol misuse (Levendosky & Graham-Bermann, 2001; Margolin, 1998; McCloskey et al, 1995; Spaccarelli et al, 1994; Wolfe et al, 1995).  All of these factors can potentially challenge the mother-child relationship. Several studies have confirmed a connection between mothers' stress levels and emotional and behavioral problems in children (Graham-Bermann, 1996). This connection has not been consistently observed across all studies, however. For example, McCloskey and colleagues (1995) found that children's mental health scores are more highly correlated to incidents of violence in the home, than to maternal distress.

 

Children's relationships with their fathers have been the subject of much less scrutiny by social scientists, with a few recent exceptions. Edleson (1999) conducted a review of the current literature on parent-child relationships as a mediating factor of child adjustment and obtained provocative findings. Edelson summarized the findings by Peled (1998), who argues that children's relationships with their violent fathers are generally confusing and problematic. Peled found that children expressed both positive emotion, such as affection, as well as negative emotion, such as pain, resentment and disappointment. Sternberg and her colleagues (1994) conducted a study of the effects of violence on children's perceptions of perpetrating and non-perpetrating parents. Their results confirmed Peled's conclusions that children are able to assign both positive and negative characteristics to the violent parent, but that this duality, particularly among younger children, can cause confusion and distress.

 

 

The "Violence Breeds Violence" Hypothesis

 

We know that parents are very influential behavioral models for their children, thus it makes sense that witnessing parental violence could shape the behavior of the children as they grow up. The extent to which children exposed to partner violence replicate that experience in their later intimate relationships has been referred to as the rate of "intergenerational transmission of violence" (whose explanation is provided by the "violence breeds violence" hypothesis). In general, this hypothesis has received some research support - in approximately 30% of cases this continuity, or "transmission", of tendencies to receive or perpetrate violence in relationships - is observed. However, as evident in considering the inverse of this figure, the majority of children exposed to partner violence do not go on to perpetrate or become victims of violence in their adult relationships.

 

There is some predictive utility to this hypothesis, however. Hotaling & Sugarman (1986) found childhood exposure to partner violence to be the most reliably identified risk factor for adult perpetration and receipt of relationship violence. These results were confirmed in a more recent retrospective study conducted by Doumas, Margolin & John (1994). Several recent studies have looked specifically at the use of violence in the intimate relationships of adolescents and young adults. This research confirms that exposure to inter-parental violence is predictive of violence in adolescent and young adult intimate relationships. It's worth noting that Doumas et al's findings were discovered to vary by gender. Several studies have found a relationship between violent behaviour and violent parental models in boys, but not in girls (Breslin et al, 1990; O'Keefe, 1998). This is consistent with other research which has found that fathers are a more "powerful models/legitimizers of violent behaviour" than are mothers (Cantrell et al, 1995, p. 39). Both social learning and feminist theory provide a theoretical basis for understanding these research results.

 

Is Exposure to Partner Violence a Form of Child Abuse?

 

Provincial child protection statutes define the circumstances in which members of the public have a duty to report and child protection workers have a duty to intervene to protect a child from abuse. Individual assessment of each child is crucial in considering whether any state intervention is required and, if so, what form of intervention should occur (e.g., Edleson, 1999). Some service providers and child protection workers take the view that once there is evidence that a child has been exposed to partner violence, the odds of harm to the child are sufficiently high to warrant intervention. From this perspective, exposure to inter-parental violence is always considered child maltreatment (e.g., Tomkins et al, 1994; Haddix, 1996).

 

Although favouring the interests of the child, others advocate that supporting the battered woman is of prime importance in ensuring the safety and well-being of the child. Proponents of this position believe the best way to help the children is to provide services to the battered woman and caution that there is a danger in labelling children exposed to partner violence a "deviant" population when in fact many cope quite well without outside intervention (e.g., Peled, 1993).

 

In our view, in many instances, the safety and well-being of the child are promoted when the safety of the non-abusive is ensured. The abused partner may require supportive services for a time.  In some cases, children may need to be temporarily cared for by others or by child protection agencies. Individual assessment of each child is crucial in considering which arrangement is the best option (e.g., Edleson, 1999). Such assessments should determine the impact of exposure on the child, threat of future exposure and possible risks to safety that may result from continued co-habitation or access visits with the abusive parent. Safety plans for both the abused partner and child should be discussed if the child is to remain in contact with the abusive parent.

 

 

Children Exposed to Partner Violence and the Law

 

As noted above, a child's exposure to partner violence may be the subject of a complaint to a child protection worker, resulting in a child protection investigation, and possible apprehension of the child. Provincial legislation requires that anyone who has reason to believe that a child has been or will be harmed must report his or her concerns to a child protection worker. The worker will then investigate and determine what type of intervention, if any, is indicated in the circumstances.

 

The second circumstance in which a child exposed to partner violence may become involved in the justice system is if a parent is charged with assaulting either the child or the other parent. Young children may not sufficiently understand the concept of truth and the consequences of lying to be allowed to testify. Older children may be allowed to testify in circumstances where measures are taken to make the experience less traumatic, such as placing a screen between the child and the abuser, or allowing videotaped testimony.

 

A third way in which the child's exposure to violence may become the subject of a legal proceeding is if the parents separate and one parent applies for a court order for custody of  or access to the child. In weighing the evidence to determine what would be in the best interests of the child, the judge considers factors including the impact of the partner violence on the child, the risk of future impact on the child, the child's need for stability and security, the child's need for time with both parents, and the parenting ability of each parent. To prevent further violence, the judge may order that the abusive parent can only visit the child under the supervision of a third party. In cases of severe violence, the judge may deny the abuser any access where the risk of harm to the child is too great. (For a detailed exploration of this issue, please consult "Child Custody and Access in the Context of Family Violence: A Review of the Literature and Annotated Bibliography" [Braun, 2001], available through the BC Institute Against Family Violence).

 

 

Assessment, Treatment and Intervention

 

The information that we have explored has formed a basis for the development of prevention and intervention measures. In the following section, we will explore many of these measures and look at the way in which they propose to be a part of a coordinated effort to maintain children's safety and well-being.

 

One common intervention for children who have witnessed violence is group treatment. A "structured group treatment" model, as developed by Jaffe et al. (1990), includes the following components: "labeling feelings, dealing with anger, safety skills, social support, social competence and self-concept, responsibility for parent/violence, understanding about family violence and wishes about family" (p.63). Evaluation has revealed a significant improvement in children's emotional state and coping skills following exposure to this treatment appproach. However, the results also show that this type of intervention may be best suited to children who have mild to moderate behavioural problems. A second, more formal, evaluation conducted by Wagar and Rodway (1995) found that in their 6-month follow-up interview, children and parents were able to give concrete examples of how they have maintained their improved attitudes, anger management and support skills.

 

Other models have been proposed, such as family systems intervention (Gentry & Eaddy, 1980; Lehmann et al, 1994), or the sibling group approach (Frey-Angel, 1989). One example of a family-oriented approach, developed by Lehmann and his colleagues, proposed a "multi-dimensional family systems approach". This program is designed for mothers and their children, in father-absent situations. Lehmann incorporates three dimensions to the therapy; emotional and behavioural impact of violence on children, the significance of PTSD, and feminist family therapy practices.

 

On a community level, there has been a great deal of emphasis on providing coordinated programming and services. This suggests a collaborative effort among social service agencies, the educational sector, the criminal justice system and government. One example of a community-based approach is the model discussed by Riva (1998). This children who have witnessed violence program functions under the mandate of the Domestic Violence Free Zone initiative. This integrated community response was piloted in Cambridge, Massachusetts and includes the following components: school programming, group treatment, enhanced programs at the transition house, and improve collaboration and training within government services and agencies.

 

School programs have also been identified as an important primary prevention tool. At the high school level, youth are beginning to get involved in intimate relationships and are establishing behavioural patterns within those dating situations. Sudermann and Jaffe (1995), among others, have proposed that this is an excellent time to provide education on matters relating to healthy relationship building. There have been a number of different programs developed in this area, and although there has been a lack of longitudinal studies to evaluate these programs, the results of less formal evaluations have shown us that the outcomes are mixed (Sudermann & Jaffe, 1995). This points to the fact that there is room for further development in this area.

 

As we have learned, exposure as a child to inter-parental violence can have both short-term and long-term deleterious effects on many children and youth. Similarly, we know that different children react in different ways to exposure to that violence.

 

Most research described earlier in this publication examines the effects of exposure on groups of children. The trend in this decade has been to refine the overall finding of deleterious effects to determine if sub-groups of these children in differing circumstances are more or less affected than other sub-groups. Sometimes findings are described in terms of some children being more negatively affected than others, while other times they are described in terms of some sub-groups of children being more or less vulnerable to this exposure. Since many of these studies are small in sample size, and typically are based on samples convenient to the researcher, it should not be surprising that sub-group results are often contradictory. Indeed, it is doubtful that one could ever design an ecologically valid study that would reliably disentangle confounding effects in children's lives of important factors such as exposure to inter-parental violence, child abuse, poverty, cultural interpretations of violence, gender of child and parent, age of child, coping style, etc.

 

This last observation raises the question of whether it really matters, for example, that we know whether boys of a certain age group are likely to be affected more or less than girls, or any such similar comparison. Does knowing a group's characteristics help us decide how to support any particular child?  Surely the answer to this question must be "no", albeit with some qualifications. The proportion of children adversely affected by domestic violence is sufficiently high that we must assess each child individually to determine how he or she perceives the violence, in what ways he or she may be affected, and in what ways he or she requires intervention. What does matter is that we know the right questions to ask the child, and the right observations to make and that we know what action is likely to help or harm that particular child.

Assessment

 

1.      Determine whether children have been exposed to inter-parental violence

When inter-parental violence has occurred, the odds are high that children living with at least one of the parents have some awareness of that violence (Edleson, 1999; McGee, 1997). As we stated earlier, the best assumption, therefore, may be to assume that any children involved have been exposed.

 

There are many groups of professionals who may first become aware of inter-parental violence: shelter workers, police and other emergency workers, and medical service providers, to name just a few. Current practice, at least in the United States as described in a survey reported by Tomkins et al (1992), is for any inquiries about children's exposure to be made of the mother. Because there are many good reasons for under-reporting by the mother (such as fear of child protection intervention or lack of awareness of deleterious effects due to the personal trauma she is experiencing), service providers must broaden their inquiries to include children themselves, as well as other family members or close associates, where possible  (Wolak & Finkelhor, 1998). These authors (1998) do warn, however, that it is important not to probe prematurely for disclosures before a respectful, understanding relationship has been established and a level of trust built with the child.

 

2.      Ascertain whether the children are safe

 

If inter-parental violence has occurred, the first order of business must be to determine where the children should be living in the immediate future and with whom (Wolak & Finkelhor, 1998). At this time, an assessment should be made of risk for future vicarious or direct child maltreatment and/or neglect. In other words, what is the potential for harm resulting from either directly or vicariously experienced violence?  The significant overlap between partner abuse and child abuse is a factor that cannot be ignored. If the service provider has reason to believe that the child has been harmed or that there is significant risk of future harm, he or she has a duty to report the concerns to a child protection worker.

 

Regardless of living arrangement, service providers should discuss with the non-abusive parent his or her and the children's well-being. The service provider should help the parent create a safety plan outlining steps family members can take should danger threaten in the future. Older children should rehearse each step of the plan.

 

3.  Ascertain whether the children have been traumatized

 

After determining that the children are safe, the service provider should investigate any signs that the child has been traumatized by the experience (Wolak and Finkelhor, 1998). If there is any sign from the child or other family members that the violence has been very frightening or disturbing to the child, Wolak and Finkelhor recommend that a professional trained in crisis counselling and assessment be involved. Rudo & Powell (1996) also advise that choice between assessment techniques should depend on how severe the violence has been. Specifically, they suggest that the methods developed by Pynoos & Eth (1986) are appropriate when trauma is suspected, and the system developed by Jaffe, Wilson & Wolfe (1989) when the violence or its impact has been less extreme.

 

Assessment after extreme or traumatic violence. Eth and Pynoos (1994) describe the use of their assessment system with 55 children and adolescents who witnessed the killing of a parent. This is an excellent article to help assessment professionals and service providers understand how exposure to what for most of us is unimaginable trauma may vary by age. It is recommended reading for all interested in this topic.

 

The protocol they have developed uses semi-structured interview techniques developed specifically for this type of situation. Interviews may be completed from hours to weeks after the violent event in a two-hour session. The interview is divided into three stages:

 

(1)  An opening stage which allows the interviewer to gauge the child's current coping style through the use of art materials. The child is asked to draw whatever he or she wishes and then tell a story about the drawing. The interviewer asks the child to elaborate on any features that seem conspicuous or a source of anxiety.

(2)  The central stage starts when the interviewer identifies a traumatic reference and interprets it to the child. The role of the interviewer is to help the child recount all aspects of the violence, its antecedents and its aftermath.

(3)  The closure stage reviews the themes of the session and present symptoms and future concerns. The interviewer emphasizes how understandable, realistic and universal the child's responses are. The goal is to help the child feel less isolated and more willing to accept support from others. An attempt is made to assess family relationships, changes in the home or school, attention from media coverage and ongoing involvement in legal proceedings. Feedback is also sought from the child about the effects of the interview itself.

 

If significant psychopathology is noted, the child should referred for formal psychiatric evaluation and treatment. For other children, Eth and Pynoos suggest that networking with social services agencies may be most helpful, sometimes in addition to brief or long-term psychotherapy and family therapy.

 

With violence as extreme as the homicide of a parent, the following reactions among the 55 children whose assessments are reviewed are noted by Eth and Pynoos (1994):

 

·         Memories of the event are highly accurate, consistent with the flashbulb theory of traumatic memories (Brown & Kulik, 1977) but not with the unreliable testimony position advocated by authors such as Loftus (1979).

·         Identification with the deceased is more common and dramatic in children than in adults and can result in the child imagining a shared destiny.

·         Identification with the aggressor or some other person prominent in the event such as a police officer, doctor or attorney can also occur.

·         Children sometimes complain about the stigma of being the child of a parent who died by murder and this is dreaded as a life-long burden.

·         The worst moment recalled is usually an emotional exchange between the child and either the victim or the assailant.

·         The funeral can be of special importance, especially if there has been physical mutilation. A photograph of the deceased can be very comforting.

·         A very stressful aspect of the experience is the struggle with allocation of responsibility for the homicide - when the murderer is the other parent, there can be an intense conflict of loyalty and assignment of blame may change as time passes. Some children blame themselves for not having intervened effectively or for having provoked the killing by their behaviour.

 

Eth and Pynoos (1994) go on to describe the way the child's age affects their assessment responses after homicide of a parent.

 

Pre-schoolers may need the most help; they usually feel defenseless and often simply cannot imagine how they will manage without the parent. They typically act as passive observers, maybe trying to escape or turning away but then lying awake, trying to sleep. They can appear withdrawn, subdued or even mute. Some will not speak of the event at all. They do typically re-enact the event and play around the traumatic themes, with a focus on the actual killing. They can be prone to regressed behaviours (e.g., intensified separation and stranger anxiety, tantrums, lapses in toileting, sleep talking etc.) and tend to use denial to ease their pain.

 

School-age children show a wider range of responses, and usually have some degree of cognitive impairment and dulling of intelligence. There may be decreased ability to concentrate in class because of the intrusion of traumatic memories. They are not limited to passive spectator roles but can become participants if only in their imagination. Revenge fantasies may relieve guilt feelings. There can be a wide range of behavioural changes and peer relationships can suffer. This age group is especially susceptible to development of psychosomatic complaints (e.g., stomach pains and other bodily symptoms).

 

Adolescents' responses to trauma are more similar to adults than to those of younger children. The trauma often precipitates an early entrance into adulthood and gives them a false sense of readiness to leave home and community. Teenagers who have survived a parental homicide often act out, have poor impulse control and may potentially re-enact life-threatening behaviour. They often appear uncooperative, suspicious and guarded during the interview. Relative to children of other ages, youth show less interest in the homicide per se and more closely examine the behaviour of the murderer, victim, themselves and others. Eth and Pynoos (1994) say teenagers seem compelled to judge the courage of their own behaviour and may be reluctant to criticize the behaviour of the victim. They can accurately identify how their actions figured in the chain of events but may inflate their own guilt feelings. They are often sensitive to the stigmatization that may accompany a homicide.

Assessment in non-crisis scenarios. The instrument most frequently recommended for situations where the inter-parental violence has been less extreme or traumatic impact is less evident is the Child Witness to Violence Interview, developed by Jaffe, Wilson and Wolfe (1989). Jaffe, Sudermann and Reitzel (1992) comment on assessment of this population in general. They strongly advocate comprehensive assessment that not only examines behavioural and emotional problems but also less obvious, more specific, subtle symptoms that may accompany exposure (e.g., conflict resolution beliefs, assigning blame for violence and knowledge and skill in dealing with violence). It is also important to consider the effects of any social and environmental disruption on children and their caregivers, including stress-buffering child protective factors such as intelligence, school achievement, and positive relationships with peers and adults. Observed effects in the child may well vary depending on whether the violence has ended and on the child's perception of whether it will start again.

 

There is often a code of secrecy that may limit social support and make children suspicious of persons outside the family. There is a need to obtain evidence from other members of the social group.

The interview schedule is for children over age 6. The three main areas of questioning are:

 

a)     Attitudes and responses to anger. Children are asked about their responses to anger-provoking situations, how they would go about resolving interpersonal conflicts and react to violent behaviour between parents and among others. Specific appropriate strategies are probed as well as inappropriate responses. Jaffe et al. note that preference for violent television shows and a history of exposure to inter-parental violence may be an especially strong predictor of maladaptive interpersonal relationship strategies. Opinions about normality, incidence and acceptability of violence are sought.

b)     Responsibility for violence. Interviewees are asked what they think parents fight about and are asked about specific problems. Many believe they are responsible for the violence.

c)     Safety skills. Children are asked about knowledge of sources of help and emergency phone numbers as well as predictability of inter-parental violence.

 

Results are used to plan interventions, assess outcomes and assign clients to group as opposed to individual treatment. In addition, comprehensive assessments such as those resulting from the Child Witness to Violence Interview may be used in the following situations:

 

1)     for juvenile court judges when delinquent behaviour may be linked to exposure to inter-parental violence

2)     for parents, child protection workers, and shelter workers to determine whether placement is appropriate in light of the effects on children

3)     for family court judges in custody disputes, and

4)     to assess the appropriateness of using children as court witnesses

 

Other Assessment Instruments

Wolak and Finkelhor (1998) note that if there is any evidence of direct violence against the child, a medical examination is warranted. Additional instruments that may be useful include the Survey of Children's Exposure to Community Violence (Martinez & Richters, 1993), the Conflict Tactics Schedule (Straus, 1979; Straus & Gelles, 1990), and the Child Behaviour Checklist (Achenbach & Edelbrock, 1984).

 

 

Treatment and Intervention

Is treatment necessary?

In many clinical models, there is no sharp distinction between assessment and treatment. This is especially the case when the assessment is comprehensive. Whatever the form of assessment, one of the most important conclusions emanating from it should be a recommendation as to whether any further formal or informal intervention is indicated.

 

Although it is important that all children exposed to inter-parental violence be assessed, some children and youth, perhaps even the majority depending on the assessment location, will neither need nor benefit from direct treatment (Wolak and Finkelhor, 1998). These authors list the following groups as potentially not benefiting from treatment:   

 

·         children who show no symptoms of behavioural or emotional distress

·         children who have good coping abilities and support systems

·         children who have not been exposed to lengthy or highly disturbing violent episodes, or

·         children who do not show any interest in therapy

 

In such cases, the children or youth should be given some brief practical information about how to seek help should it be required.

 

Alternatively, a decision about the type and length of intervention may be based on an assessment of the child's problems, developmental level and the family context.

Length of treatment or intervention

Wolak and Finkelhor suggest that short-term treatment may be enough for a child with symptoms of traumatic stress and adjustment problems but with no evidence of "more deeply rooted behavioural problems". Short-term treatment usually focuses on trauma processing - getting the child to describe all of the details of the violence and the emotions felt, the goal being to help the child manage the strong feelings and images evoked by the events. Attention is also paid to reducing any feelings of responsibility or self-blame.

 

Longer-term treatment is indicated for those who show serious depression, risk of suicide or self-harm and for those who have conduct disorders or show aggressive behaviours. Typically, issues other than exposure to inter-parental violence are present when these more serious symptoms are evident and coping abilities in general may need to be bolstered.

 

Developmental considerations in planning treatment or intervention

Wolak and Finkelhor (1998) provide the following guidelines for treatment according to developmental age:

 

·         infants' primary need is for a safe and secure environment where their needs for food, sleep and physical contact can be met reliably and according to their needs - hence placement is the most important issue

·         toddlers and preschoolers will usually be most responsive to play therapy, either individually or in a family/caregiver setting

·         school-aged children and adolescents usually will be best suited to group therapy. We would add that individual sessions may also be indicated if there is high risk of self-harm or conduct disorder.

Individual versus group treatment

A decision concerning individual or group treatment will be very much guided by available resources, preference of the child or youth, and preference of his or her family. We have not located any research that discusses the benefits of these modalities, other than the suggestion by some authors, such as Wolak and Finkelhor (1998), that a child who has been exposed to more severe or prolonged violence might require individual treatment as well as or instead of group treatment.

 

Parenting support as intervention

When the children involved are young, there may be situations when increased parenting support or education would be the most appropriate intervention. Much will depend on the current living arrangement, as well as on the preferences of individual family members. Parenting support might be especially appropriate when an assessment of the child/parent relationship indicates either that the custodial parental skills are lacking or when the disruption related to the violence is such that parental awareness of child needs and reactions to the exposure is dimmed. Groves (1996), for example, feels that additional parental support might serve to remind parents that they are the most important emotional protectors for their children. Rudo and Powell (1996) add that the mother's health and well-being are critical to the well-being of the children.

 

It should be noted, however, that it cannot be assumed that provision of services to parents will automatically have beneficial effects for the children. Individual assessment of the child's progress must accompany any intervention aimed at support of the parent(s).

 

Conversely, it is also important that any intervention with the child be accompanied by some form of support for the parents (Wagar and Rodway, 1995). These authors point out that, without coordinated support for parent(s), any learning by the children may be undermined by the parents, be this intentional or not. Parental cooperation may not, however, always be forthcoming (Wolak and Finkelhor, 1998). Absence of parental cooperation calls for additional coordination and intervention on the part of the agencies and professionals involved with the family.

Should the abusive partner be involved in the child's treatment?

There is no easy answer to this question. There are those who advocate that the violent partner should never be involved, having shown himself to be an unfit parent by virtue of exposing the child to the violence (e.g., Haddix, 1996). Others are more pragmatic, recognizing the variation that may exist among families as to the current living and visitation arrangements with the violent partner. Wagar and Rodway (1995) suggest that, if cooperation is forthcoming and the violent partner still has contact with the child, and if it is safe for the non-violent partner, child and professionals involved, every opportunity should be sought to educate the parent as to the potentially deleterious effects of his or her actions on the child.

Group approaches to treatment

Various models have been proposed to guide group therapy or psycho-educational groups, but most seem to cover similar types of issues. Below are three examples where goals are relatively clearly defined. We have not located articles that discuss structural features such as group size, number of sessions, etc.

Lehmann et al's (1994) multi-dimensional family model

Their approach assumes the father is not involved with the mother or child. They take into account the negative effects of being exposed to inter-parental violence, post-traumatic stress disorder and feminist family therapy. They focus on the assault, deal with accountability, engage in open family discussion and adopt a personal and political position, involving the following components:

·         working with the larger system -  e.g., on housing, finances, day care and legal problems

·         addressing traumatic after-effects  - e.g., memory management strategies

·         rebuilding relational imbalances among family members - e.g., teaching basic safety skills, responsibility, building social support systems, differentiating conflict and abuse, opening up family secrets, and discussing loyalty conflicts and permanent family splits

 

Kirkpatrick and Williams'(1997) clinical model

This approach assumes that the violence has been sufficiently severe and long-lasting that the child is vulnerable to post-traumatic stress disorder. They advocate an age-determined group setting that involves gestalt-based play and art therapy, cognitive behaviour therapy and structured physical challenges. The goal is to help the children

·         help identify and change destructive belief patterns

·         identify and express emotional residue from the violence

·         exercise personal control

·         plan for the future, and

·         break down their sense of isolation

 

 

A psychoeducational model - the Minneapolis Domestic Abuse Project

(described by Rudo and Powell, 1999)

In line with their much admired community coordination approach, the Minneapolis Domestic Abuse Project's programming goals for children exposed to inter-parental violence involve a significant psycho-educational component. Group goals include:

·         to define violence and responsibility for it

·         to express feelings including anger

·         to improve communication, problem-solving and cognitive coping skills

·         to increase self-esteem

·         to develop social support networks, and

·         to develop safety plans and trust during positive group experiences

 

 

The BC/Yukon Society of Transition Houses' "Children Who Witness Abuse Programs"

 

Children Who Witness Abuse programs in BC are delivered by a variety of community agencies and organizations such as transition houses, family serving agencies and women's resource societies. They exist in cities, small towns and rural areas. They are based on the programs developed by the London Family Court Clinic in Ontario, and on research that shows there are emotional, cognitive, and behavioral impacts on children as a consequence of being exposed to violence.

 </