Introduction
It is well established that exposure to domestic violence has a significant negative impact on children’s development, affecting their emotional, social, and cognitive functioning and interfering with their ability to learn. Children are affected by violence as early as infancy. For more than 30 years, researchers have published accounts about children exposed to domestic violence, with the rate of publication increasing dramatically in the past 10 years. Numerous books and countless papers and articles have described the cognitive, emotional, and behavioral impact of witnessing violence on children as well as the mediators and moderators of these effects. The sheer volume of this material has given rise to conceptual and methodological disagreements about the optimal course of research in this area. It has been challenging at times to draw definitive conclusions from the numerous studies because there is a lack of agreement about what descriptive terms are best and the meanings of these terms. , Methodological differences have also delayed consensus because researchers have drawn on different samples and have not consistently examined conditions such as poverty, parental mental illness, parental substance abuse, and child maltreatment that often co-exist with domestic violence and have independent impact or act as moderators of children’s functioning. Even with these limitations, it is well established that exposure to domestic violence has independent effects relative to other negative factors in the child’s environment.
In this chapter we briefly review the incidence and prevalence of exposure to domestic violence and describe some of the challenges that inconsistent terminology brings to the research. We also examine the ecological contexts that mediate and moderate the impact of exposure to domestic violence and review, by developmental stage, the known impacts of exposure. We describe models of intervention shown to be effective with exposed children and discuss the features that these intervention models have in common. Finally, we will evaluate the strength of the evidence about children’s exposure and suggest directions for future research.
Terminology and Taxonomy
In spite of two excellent critiques published 14 years apart, , the field still struggles with inconsistency in how to think about what it means to witness or be exposed to domestic violence and what we mean by domestic violence. These inconsistencies present interpretive challenges. Holden notes that the move to the use of the word “exposed” rather than “witnessed” or “observed” is an improvement in describing children’s experience of domestic violence because the term is inclusive of different types of experience and does not assume that the child must actually see the violence to be affected. Many studies, however, do not describe the nature of the participating children’s exposure. One review of 22 studies published between 1987 and 1997 found that only 43% included a description of the nature of the exposure.
To improve reporting of the nature of exposure, Holden proposes a 10-category descriptive taxonomy of children’s exposure to domestic violence. The categories are: (1) prenatal exposure, (2) intervening verbally or physically to stop the assault, (3) being physically or verbally assaulted during a domestic violence incident, (4) joining the assault, either through force or voluntarily, (5) being an eyewitness to the assault, (6) hearing, but not seeing, the assault, (7) seeing the immediate consequences of the assault, (8) experiencing life changes as a result of the assault, (9) being told or overhearing conversation about the assault, and, (10) being ostensibly unaware of the assault.
The second definitional challenge is the lack of unanimity in the literature about what constitutes domestic violence. Based on a review of the research, Holden suggests nine dimensions around which domestic violence can be organized, and suggests that researchers use these descriptive dimensions when reporting study findings. The first dimension describes the type of violence that occurred, including whether the violence is physical or psychological, minor or severe, or mutual vs. male-initiated violence. The second dimension is the nature of the particular act of violence that occurred and whether the perpetrator of the violence intended injury. The third dimension considers the nature and severity of injuries. The fourth dimension consists of timing variables such as the frequency of violence, the duration of single incidents of violence, the child’s age at the time the exposure began (important because duration of exposure has a significant effect on the accumulation of posttraumatic stress disorder [PTSD] symptoms), and the amount of time that elapses between assaults. The fifth dimension captures the extent to which incidents escalate in frequency or intensity over time. The sixth dimension examines the type of perpetrator, using the typology proposed by Holtzworth-Munroe and Stuart of family assaults only, antisocial, and dysphoric/borderline. The seventh, eighth, and ninth dimensions describe, respectively, the perpetrator’s relationship to the child, the victim’s role in the assault, and the way in which the assault is resolved.
These two taxonomies, describing the nature of the child’s exposure and the nature of the violence itself, are based on research studies but require empirical validation. If researchers consistently gathered and reported study data along these descriptive dimensions, more definitive conclusions could be drawn from the literature describing the cognitive, behavioral, and emotional impact of children’s exposure to domestic violence. The taxonomies can also be used by clinicians to collect information about an individual child’s experience that will enable them to choose the interventions most likely to be effective.
In this chapter, we include in our literature review only work that describes the psychological impact of children’s exposure to violence between the child’s mother and a current or former adult partner. Other forms of violence to which children may be exposed in their homes, such as violence between siblings or physical violence directed at the child, are outside the scope of this chapter.
Prevalence of Children Living in Violent Households
An accurate estimate of the prevalence of children’s exposure to domestic violence is challenging because researchers have used different definitions of domestic violence, different measures for collecting data, and different methodologies. Field and Caetano reviewed nationally representative estimates of assaultive behavior between partners in the United States and found generally low prevalence rates (e.g., < 1%) when respondents must self-identify as crime victims before questions about domestic violence are asked. In contrast, survey instruments that measure behaviors in the context of family rather than crime generally result in higher prevalence rates (e.g., 12-20%). This is not necessarily a consistent pattern, however; for example, the National Violence Against Women Survey used behavioral queries and found a prevalence rate of 1.4%, only slightly greater than the crime surveys.
One early attempt to estimate the number of children living in violent households was Carlson’s 1984 study, still frequently cited by researchers and policy makers. Extrapolating from the 1975 National Family Violence Survey and U.S. Census data, Carlson estimated that approximately 3.3 million children in the United States are exposed to domestic violence each year. Because this estimate was created in the context of an early literature review and with limited data available, it is beset with a number of methodological challenges that are acknowledged by the author. First, Carlson’s estimate was based on reports from one family member. Disagreement between partners on the occurrence of acts of domestic violence is common, casting doubt on the reliability of data from a single reporter. Second, the estimate was limited to households with a child between 3 and 17 years of age, which could be problematic because children 5 years old or younger appear to be disproportionately represented in homes with domestic violence. , Finally, Carlson considered only acts of severe violence (e.g., punching, kicking, threat or actual use of weapons), acts that are less common than milder forms of violence (e.g., pushing, slapping, grabbing). These milder forms of violence, however, are associated with behavioral problems in children exposed to them. All the methodological flaws of this early review tend to underestimate prevalence of domestic violence.
More recent population-based surveys attempt to find a more accurate estimate of the extent of children’s exposure. McDonald et al estimated that 15.5 million children (29.4%) are exposed to domestic violence annually, including 7 million children (13.3%) exposed to severe violence in their homes. The researchers conducted in-home interviews with 1615 married or cohabitating couples. Both members of the couple were interviewed about specific violent behaviors, both mild and severe, that they had committed or sustained in the year preceding the interview. This study revealed that there is more likely to be domestic violence in households with children; it does not, however, report the ages of children living in the home, so it cannot be used to confirm earlier findings that children under 6 are disproportionately exposed to domestic violence. , Nor does it report on number and frequency of violent events or other contextual risk factors such as community violence, poverty, or child maltreatment.
Finkelhor and colleagues examined exposure to domestic violence as one of a variety of types of victimization in 2030 children age 2 to 17 years living in the United States. This survey found that 71 children per 1000 witness domestic violence each year, or 2,190,000 children. Girls were more likely to witness domestic violence than boys, and children under 13 were more likely to be exposed.
Several methodological differences between McDonald and Finklehor may explain the lower prevalence rate in the latter study. First, Finklehor eliminated children aged birth to 2 years from the sample. Second, domestic violence was defined far more narrowly in Finklehor’s study, with a single question asking whether the child had seen a parent get hit (e.g., slapped, hit, punched, or beat up) by another parent or by the parent’s boyfriend or girlfriend in the past year. All types of exposure except direct observation were eliminated, as were all types of violence except hitting. Other questions covered the child’s direct observation of assaults with weapons and the murder of someone close to the child. These types of victimization might have involved domestic assaults or murders, but they were not included in the domestic violence totals.
Fantuzzo and his colleagues took a public health approach to estimating the prevalence of children exposed to domestic violence crimes investigated by police officers. Prevalence findings showed that children were present for 44% of the investigated domestic violence incidents, with slightly over 1000 children exposed across the period of the study. Children under 6 were overrepresented. Of the children present during the domestic violence event, 81% were assessed by the investigating police officer to have experienced direct sensory exposure to the event; 4% of the exposed children were physically injured in the domestic violence event.
These three recent studies use divergent methodologies and, not surprisingly, find different prevalence rates. They leave no doubt, however, that large numbers of children are exposed to domestic violence each year.
Domestic Violence Exposure in an Ecological Context
Children’s development is shaped by a vast array of factors, and domestic violence is not generally an isolated stressor. These two statements point to the undeniable truth that when children present with cognitive, emotional, or behavioral problems, the quest to understand the etiology of the problems does not end with asking about whether the child has been exposed to violence at home. This is an important factor, but only one of many that must be assessed. Bronfenbrenner , and others proposed an ecological-transactional approach to understanding children’s development that conceptualizes the child’s developmental environment as an interaction among the immediate settings in which the child develops ( microsystems ), the relationships between these microsystems ( mesosystems ), and settings in which children are not necessarily found (e.g., government agencies, parent’s workplaces) but that affect their development in important ways ( exosystems). The decisions at the level of exosystems can have profound impacts on microsystems, as when a legislative body passes funding to subsidize child care for young children, which in turn alleviates economic stress in children’s immediate families.
In an ecological-transactional analysis of development, children are at the heart of the interlocking relationships between systems at these multiple levels. Metaphorically, it can also be said that children are themselves microsystems. Children’s temperaments, personalities, patterns of attachment, capacities for self-regulation, and cognitive potential are all the result of an interplay between children’s genetic makeup and the environment in which that genetic makeup is expressed. Development, even at the cellular level, is experience dependent, and all of the environmental contexts that, directly or indirectly, affect children have a role in determining who the child will become.
Factors Intrinsic to the Child
Pynoos and colleagues suggest a number of factors intrinsic to the child that may affect children’s responses to potentially traumatic stressors such as exposure to domestic violence. These include the child’s genetic makeup, temperament, quality of attachment, repertoire of coping strategies, acquired developmental competencies, and a range of attributes having to do with the organization of the child’s stress response system. (See Chapter 54, “Effects of Abuse and Neglect on Brain Development.” ) Even these so-called intrinsic factors, however, develop in relationships with caregivers, and parents hold a central place within the ecological world of the developing child. Especially for young children, parents place their imprint on every aspect of the child’s development, both directly through their behavior toward the child and indirectly as they bring the influences that the larger world has upon them into their relationships with their children.
The Power of Parents
Parents have the most powerful direct effect on children’s development and are themselves acted upon by interlocking systems that determine their individual and parental functioning. Belsky proposed three factors that interact with one another, predicting compromised parenting in families where there is violence. First, parenting quality is tied to the parent’s personal psychological resources, and both victims , and perpetrators of violence often have diminished psychological well-being. Second, children’s behavior has a powerful role in shaping parenting, and the behavioral difficulties of children exposed to violence may lead parents to adopt authoritarian parenting strategies. Third, a satisfying relationship between the parents, lacking in violent families, provides a major contextual support for good parenting. Other models of the determinants of parenting behavior include additional factors (e.g., the parents’ social network, parents’ view of the child), but all of the models give parents’ functioning and behavior a central place in children’s development. These theoretical models are supported by ample empirical evidence of the importance of the quality of the parent–child relationship in mediating and moderating children’s outcomes after exposure to violence.
Factors Outside the Family: Society and Culture
James Garbarino labels as “socially toxic” the combination of social-cultural conditions that combine to deprive children of opportunities to learn and thrive. These conditions include economic inequality, racism, community violence, and the legitimization of aggression and violence in the mass media. These societal risk factors act together, with children from racial minorities more likely to be poor, more likely to be exposed to violence and aggression in the home and the community, and less likely to have access to services. The impact of social toxicity on children is nowhere clearer than in the stark differences in the incidence of domestic violence among socioeconomic groups. The likelihood of domestic violence increases as family income decreases. The incidence of domestic violence is 3% for families with a yearly income of more than $75,000 a year, and rises to 20% in families with a yearly income of less than $7500.
These social forces act on children’s development both through direct exposure and as the communities in which parents live and work affect parents’ moods and sense of support and security, feelings that are brought to their parenting. To buffer children from stress, parents need to rely on society as a whole to help them protect their children from the toxic effects of violence and inequality. When society fails families, the reverberations are felt at every level.
Children’s roles in their families, the meaning of their experiences, and the meanings that their caregivers attach to the supports and stressors in the larger community are determined by cultural factors. To give the force of culture its due in an ecological model, we must go beyond an assessment of the family’s current environment and ask additional questions: From where did the family come; why, how, and when did it get here; and how does its culture view and cope with stressful and potentially traumatic experiences?
Co-Occurrence of Stressors
Domestic violence generally occurs as one of several stressors with which families and children must cope. As stated earlier, children exposed to domestic violence are more likely to suffer poverty and racial inequities. , In addition, children exposed to domestic violence are more likely than nonexposed children to suffer child maltreatment. In a metaanalysis of the literature related to children’s exposure to domestic violence, Kitzmann and colleagues found a 50% overlap between child maltreatment and child exposure to domestic violence. Researchers have offered two theoretical models for these high rates of co-occurrence. The first is that particular person-based variables are responsible for both types of aggression. These variables may include some that are intrinsic to the parent (personality characteristics such as impulsivity or hostility, biological characteristics such as genetic loading for aggression or physiological reactivity to stress, psychological functioning, and historical risk factors), and some that are environmental or contextual risk factors (stressful events, financial stress, and lack of social support). The second model is the spillover hypothesis, positing that one type of aggression contributes to the other (e.g., that the impact of victimization increases the victimized parent’s potential for maltreating the child).
Domestic violence and child maltreatment are not the only co-occurring stressors that children suffer. Finklehor and colleagues measured a variety of assaults and violence to which children are exposed, including exposure to domestic violence, and found that any child who reported being victimized had an average of three types of victimization within the year. They also found, consistent with other research, that poor and minority children were more likely to report victimization in the forms of sustaining or witnessing violence. This finding is of critical importance because a substantial body of research indicates that victimized children’s risk for poor outcomes both during childhood , and adulthood.
Given the complexity of the variables that contribute to creating challenges and vulnerabilities in children’s developmental outcomes, it stands to reason that isolating exposure to domestic violence as the responsible variable would be a difficult task. Nevertheless, three metaanalytic studies using different methodologies and selecting studies from different periods established that exposure to domestic violence makes a contribution to the variance in children’s outcomes over and above the variance contributed by other factors.
Mechanisms of Action: The Dual Lenses of Attachment and Trauma
Whether exposure to domestic violence is traumatic for a particular child depends upon an array of constitutional and experiential factors. Exposure often raises to the level of traumatic stress, as evidenced by the high number of exposed preschool children as well as school-age children and adolescents who meet criteria for PTSD. For children younger than 4 years old, witnessing assaults against their mothers is associated with more symptoms than are other types of traumas, including assaults sustained by the children themselves. These high rates of PTSD suggest that for children of all ages, exposure to acts of violence between their parents can give rise to the kind of overwhelming fear and terror that, by definition, accompanies a traumatic event. Such events, with their overwhelming sights, sounds, and even smells, shatter the child’s developmentally based expectation that their caregivers will protect them from pain and injury. This is especially the case for young children, who are unprepared to appraise risk and take protective action on their own. ,
Young children’s experience of domestic violence must be viewed through the dual lenses of attachment and trauma because young children organize their responses to fear and danger around their relationships with their attachment figures, seeking protection from their caregivers in times of threat. Attachment figures can be the child’s strongest buffer against stress, but can also present children with developmentally salient challenges. An attachment figure who is available to the child under conditions of risk and stress can buffer the child’s response, both physiologically and emotionally, , contributing to the child’s recovery after the traumatic event. In contrast, an attachment figure who is not attuned and reassuring or who is the source of the child’s fears can exacerbate the child’s response. , Although intertwined, attachment and trauma also exert separate influences on children’s development, but research and clinical thought in each of these two areas have, until recently, proceeded without regard to the impact of the other. ,
The Attachment Lens
Attachment theory posits that a primary mother figure is central to normal early development, asserting that systematic links exist between quality of caregiving, resulting patterns of attachment, and the developing child’s emotional health. The original research establishing the connection between quality of attachment and infant mental health was based on careful observation of infant behavior in a variety of ecologically valid settings. , Main and colleagues , greatly increased the clinical relevance of attachment theory when they moved away from purely behavioral observations and toward the conceptualization of states of mind related to attachment in both adults and children. They asserted that frightening or frightened behavior in the parent—the hallmark of caregiving behaviors linked to disorganized behavior and increased risk for mental health disturbances in the child—stemmed from the parent’s unresolved state of mind about her own traumatic childhood experiences. Thus, an attachment-based understanding of young children’s symptoms holds that frightening or frightened parental behavior is the mechanism responsible for transmitting from parent to child incoherent and contradictory states of mind regarding attachment. These contradictory states of mind, in turn, are manifested in the child by disorganized behavior as the child attempts to resolve the paradox of fearing the person from whom protection is sought.
Lyons-Ruth and colleagues elaborated these ideas in a way that is particularly relevant to the child’s direct exposure to traumatic events such as domestic violence. They propose a relationship diathesis model that focuses on the modulation of fear and places it in a relational context. In the relationship diathesis model, vulnerability to stress-related dysfunction is determined by at least three factors: the characteristics of the stressor, the individual’s genetic vulnerability to stress, and the capacity of the attachment system to modulate the high levels of arousal that accompany stress. Children’s emotional and behavioral symptoms emerge when the stressor is too overwhelming or when the attachment relationship is unable to modulate the child’s overwhelming affective response to the stressor. The authors proposed that parents with unresolved fear dating back to childhood traumatic experiences have difficulty helping their children modulate strong emotions such as fear because the parents curtail their conscious attention to the child’s fear signals in order to not reevoke their own early traumatic responses. Fear signals left unattended are not modulated in the relationship, leaving children alone with their own unresolved traumatic experiences.
Attachment theory thus predicts two explanations for young children’s symptoms. The first is rooted in children’s responses to their parents’ frightened/frightening behavior, stemming in turn from the parents’ unresolved childhood traumatic experiences of trauma. The second, explicated in the relationship diathesis model, predicts that the parent’s own experiences of childhood trauma interfere with the parent’s capacity to soothe the child in the face of present stress, leading to emotional and behavioral dysfunction in the child. The relationship diathesis model provides a bridge to trauma theory, which offers its own explanation for children’s symptoms after a stressful event.
The Trauma Lens
Trauma theory addresses the individual’s response to direct exposure to an overwhelming stressor. The specific symptoms and developmental challenges that a child may face after a trauma depend, among other factors, on the child’s developmental stage at the time of the experience. Infants, toddlers, and preschoolers may be particularly negatively affected because of the impact that trauma has on every aspect of the infant and young child’s development. There are three fundamental developmental tasks in infancy and early childhood: forming a hierarchy of attachment relationships and other close interpersonal relationships; developing the capacity to experience, regulate, and express a variety of emotions; and exploring the environment and learning. Trauma, which involves multiple and intense negative emotions, can damage the developing mechanisms of emotional regulation. It threatens the developing child’s ability to maintain a sense of security in attachment relationships as the child experiences failure of protection at the moment of trauma and grows to expect that the pain and fear of the original trauma will be repeated in other relationships. Finally, children’s ability to explore the environment and learn can be diminished by the fearfulness, constricted and repetitive play, and hypervigilance that can follow trauma. As the child grows, reminders of the original trauma reawaken the negative emotions that were part of the original event, further distorting the child’s development.
The Dual Lens
Trauma theory posits that symptoms flow from the dysregulation, relational disturbances, and inhibitions of exploration that are part of the trauma response not only in early childhood but at all developmental stages. Attachment theory sees the source of children’s symptoms as misattuned caregiving patterns in which frightened or frightening parents do not meet the children’s needs for comfort but are, instead, the source of their fear. We argue that in order to understand the source of any individual child’s difficulties and intervene effectively, clinicians should perform their assessments using both attachment and trauma lenses because children’s attachment relationships and their responses to traumatic events are inextricably intertwined. Whether the traumatic experience is exposure to domestic violence or some other event, it is essential to examine the quality of the child’s caregiving relationships before the trauma in order to understand the degree of interpersonal security that the child brings to the experience and how the child’s relationships might assist or derail recovery. It is also essential to understand the nature of the traumatic stressor, the child’s experience of the event itself, the developmental and constitutional coping resources that the child possesses, and the social and cultural context in which the trauma took place.
Children’s Responses to Domestic Violence
Understanding the effect of domestic violence on children is a complex enterprise that demands understanding of the nature and extent of the violence, the nature of children’s sensory experience of the violence, the quality of caregiving relationships, the extent to which the violence co-occurs with other stressors, and the environmental and cultural context within which the violence occurs. As we examine the effects of domestic violence on children, we will, therefore, focus on three questions: (1) Is exposure to domestic violence associated with negative outcomes for children even when controlling for the impact of other stressors? (2) Are there variables that mediate or moderate the impact of exposure to violence for children generally or for particular groups of children? and (3) Are there differences in the effect of exposure to violence across developmental stages?
It is clear from the literature that children exposed to violence generally have worse outcomes than nonexposed children. An excellent way to understand the effect of violence generally is to examine metaanalyses—namely, studies that examine a number of individual studies in an area of interest in order to determine effect sizes in the population. Three recent metaanalyses examine children’s exposure to domestic violence as the variable of interest. Kitzmann et al included in their metaanalysis shelter samples, community samples, and clinical samples. They reported no significant differences in effect sizes among the samples and found that across the samples, about 63% of children exposed to domestic violence were faring more poorly than nonexposed children. This finding is consistent with earlier studies of shelter samples indicating that about one third of children in shelter were functioning as well as nonexposed children. , This metaanalysis and two others , examined age and gender as moderators of children’s internalizing problems (anxious, depressed, and withdrawn behaviors) and externalizing problems (aggressive and destructive behaviors) and found no significant differences by either age or gender. There was an indication, however, that preschool girls exposed to domestic violence had lower scores on measures of social competence and showed greater distress in response to adult conflict than did nonexposed preschool girls. This finding did not exist for preschool boys or for girls in middle childhood.
One striking result from the two metaanalyses that obtained effect sizes for symptoms of PTSD in addition to internalizing and externalizing behaviors was that the effect sizes for PTSD were larger. , This leads to the question of whether internalizing and externalizing behaviors, the most commonly used measures of child functioning in the literature on children exposed to violence, are the most valid measures of outcome.
Kitzmann et al also examined the impact of other adversities in addition to child exposure to physical violence. They found that children exposed to physical violence had more negative outcomes than children who were exposed only to verbal discord in their homes, but the outcomes of children exposed to physical violence between their parents did not differ significantly from the outcomes of physically abused children or of children who were both exposed to interparental violence and physically abused. The same metaanalysis found, however, that when individual studies controlled for the presence of multiple stressors, they found smaller effect sizes for exposure to domestic violence than did studies that did not control for these variables. This finding leads to the conclusion that multiple stressors have a cumulative effect on children’s development.
These three metaanalyses offer strong support for the proposition that children exposed to domestic violence fare generally worse than do nonexposed children; that across samples, children of all ages and boys and girls are equally affected; and that PTSD symptoms may be a stronger predictor of outcome than internalizing and externalizing behaviors. One of them begins to answer the first question posed above with the finding that significant effects of exposure to domestic violence remain even when controlling for co-occurring stressors, although the effects are attenuated. The other questions, however, remain unanswered. The metaanalyses do not examine moderators other than age and gender, nor do they take into account that violence might affect children differently at different developmental stages.
Developmental psychopathology and trauma theory both suggest that childhood exposure to violence should be viewed in the context of normal development. At each stage of development, children are faced with different challenges, and exposure to violence may disrupt a child’s capacity to meet those stage-specific challenges. , In the sections that follow, we examine the literature on children’s exposure to domestic violence across three different developmental stages: (1) infancy and early childhood, (2) middle childhood, and (3) adolescence.
Exposure to Violence in Infancy and Early Childhood
Children are affected by exposure to violence as early as infancy. Because of their physical vulnerability, infants and very young children are at increased risk for physical injury during episodes of violence; in addition, they are faced during these episodes with the unsolvable problem of needing to seek protection from the very caregivers who are the source of their fear. One of the elementary developmental tasks of infancy and early childhood is forming trusting, secure relationships with attachment figures. Infants and young children rely on their attachment figures to protect them from danger and to make the world predictable for them. Domestic violence can shatter the developing trust that young children have in their caregivers, leading to insecurities not only in their attachments, but in other relationships in their lives. In addition to forming attachment relationships, infants and young children must navigate a series of anxieties that emerge sequentially as they develop. These are anxiety about loss of the caregiver, anxiety about loss of the caregiver’s love, anxiety about physical damage to the self, and anxiety about not meeting social standards that takes the form of guilt and shame. Witnessing domestic violence during infancy and early childhood can interact with these developmentally normative anxieties and amplify them because witnessing actual assaults on caregivers makes the child’s internal fears all too real. The exacerbation of developmentally expectable fears may be at the root of the symptoms that have been observed clinically, including irritability and difficulty being soothed, sleep disturbances, emotional distress, somatic complaints, fears of being alone, and regression in language and toileting. Finally, one of the most salient developmental tasks of toddlerhood is managing the conflict between emerging urges toward independence and autonomy and the wish to stay close to and protected by the attachment figure. Assaults on an attachment figure at this stage of development can make a child so anxious about the attachment figure’s safety that the child’s developmental push for autonomy is derailed.
Empirical evidence exists that the risk to infants from domestic violence begins before birth. Pregnant women who are victims of domestic violence have more difficult time bonding with their babies, increased potential for perpetrating acts of child abuse, and higher rates of child abuse and punitive parenting, with young minority mothers having increased risks. ,
This risk continues after the birth of the baby. Consistent with attachment theory, most studies of the impact of domestic violence on the youngest children indicate that negative outcomes are tied to parenting behaviors and/or the quality of the parent–child relationship. In one community sample of 203 12-month-old infants and their mothers, the infants’ direct and indirect risk and protective factors for externalizing behaviors included exposure to domestic violence, maternal parenting behavior, maternal mental health, and maternal social support. The occurrence of domestic violence was measured twice. During the third trimester of pregnancy, domestic violence prior to and during pregnancy was assessed (past domestic violence); when the baby was 12 months old, domestic violence during the first postpartum year was assessed (current domestic violence). Although both past and current domestic violence were associated with mental health problems in the mothers, only current domestic violence was related to observed parenting. Women who suffered domestic violence during the postpartum period were less able to respond warmly and sensitively to their babies, showed more hostility toward the child, and demonstrated more emotional disengagement in the mother–child relationship. On the other hand, infant externalizing behavior was predicted independently by both current and past domestic violence, raising the possibility that past domestic violence occurring during pregnancy affected the fetus through stress-induced cortisol changes that are still manifested in externalizing behavior at age 1.
Researchers have also considered the impact of violence exposure on infants’ internalizing behaviors. Crockenberg and colleagues suggest a possible mechanism for development in infancy of internalizing behavior in response to marital aggression. She found that infants exposed to paternal aggression displayed distress at novel stimuli and withdrew, and this response was stronger if the infant’s father was actively involved in caregiving.
The importance of relationship continues into the preschool years. Levendosky et al suggest that in the preschool years, children’s behavioral difficulties are more pronounced in their interactions with their mothers than in their general functioning. Exposure negatively affected children’s observed interactions with their mothers, but not maternal reports of problem behaviors. The authors interpret this finding as suggesting that early effects of domestic violence might originate in the realm of relationships rather than the realm of children’s mental health. This study also examined observed parenting variables and found that the quality of parenting in mothers who were victimized by domestic violence was tied to their mental health. Women who had elevated symptoms of depression and PTSD also had parenting problems. Women who did not have these mental health challenges seemed to compensate for the effects of violence by being more effective and responsive in their relationships with their children.
Two studies by Lieberman and colleagues , examined different sets of relationship variables in a group of preschool age children exposed to domestic violence and their mothers. One study found that the quality of the mother–child relationship as rated by clinicians, mothers’ attunement to their children’s sad and angry emotions, and the severity of domestic violence reported by the mother each made an independent contribution to children’s externalizing behavior problems; together, these three variables explained 55% of the variance in externalizing behavior. Similar relationships did not exist, however, in relation to children’s internalizing problems because there was no significant link between mothers’ attunement to children’s sad and angry feelings and children’s internalizing behavior. In the second study, children’s total behavior problems were significantly related to maternal life stress, a relationship that was completely mediated by mothers’ symptoms of PTSD and the quality of the mother–child relationship.
Exposure to domestic violence also has negative effects on children’s cognitive development. In a study of exposed preschoolers and their mothers compared with nonexposed child–mother pairs matched for child’s age, gender, and ethnicity; mother’s age and education; and annual family income, the exposed children scored significantly lower than the nonexposed ones on a measure of verbal intelligence. The nonexposed children had been exposed to equivalent community violence, however, indicating that domestic violence exposure has a unique role in explaining negative cognitive outcomes over and above other stressors. A large twin study that controlled for genetic factors found an 8-point IQ loss associated with childhood exposure to domestic violence. There is some indication in the literature that even these cognitive deficits may have their roots in the quality of children’s caregiving relationships. One study found that domestic violence was negatively correlated with preschool children’s performance on explicit memory tasks, with children who were exposed to higher levels of violence performing more poorly. This relationship was moderated, however, by the children’s mothers’ positive parenting practices.
Finally, there is some evidence that preschool children exposed to violence have more difficulty than their nonexposed peers in their relationships outside their families. In one study that included observations of group play and assessment of children’s relationships with their preschool teachers, children exposed to domestic violence were more likely to exhibit negative affect, respond less appropriately to situations, be more aggressive with peers, and have more ambivalent relationships with their caregivers. The most significant predictors of children’s adjustment were psychological violence against the mother and the mother’s self-esteem.
The complexity of these findings indicates that there are still many questions to be answered about the links between young children’s exposure to domestic violence and emotional and behavior problems. It does appear certain, however, that children’s relationships with their caregivers, and the emotional well-being of the caregivers are essential pieces of the puzzle and should be considered when assessing young children exposed to domestic violence.
Exposure to Domestic Violence in Middle Childhood
School-age children face the developmental challenge of adapting to environments and relationships outside the immediate family. In meeting this challenge, school-age children must be able to regulate emotions, to show empathy, and to accomplish increasingly complex cognitive material. As children move to these broader environments, children’s development may be derailed if they are preoccupied with issues of security or if their heightened sensitivity to risk and danger causes them to process social information with a bias toward seeing ambiguous actions as hostile. The most overwhelming anxiety faced by children in this age group is the fear of reemergence of longings and urges that belong to a younger age and now are be felt to be too babyish or dependent, undermining the child’s emerging sense of competence and autonomy. The research on school-age children’s approach to domestic violence is in part responsive to these developmental issues, as researchers examine some child characteristics that may mediate or moderate the impact of children’s exposure. In the main, however, studies have compared outcomes of exposed vs. nonexposed children, often considering the impact of family variables such as quality of parenting and parental mental health.
Although child behavior problems continue to be the outcome of interest in many studies of school-age children, researchers also investigate the level of PTSD symptoms in this population. When compared with nonexposed children, school-age children exposed to domestic violence have both higher levels of internalizing and externalizing behavior problems and higher levels of PTSD symptoms.
McCloskey et al examined the relationship between maternal mental health, family environment, and children’s outcomes. The researchers found that although women who sustained domestic violence had more mental health problems than did nonvictimized women, these problems did not mediate children’s response to family violence. Further, they found that although violent families had lower levels of both sibling and parental warmth than nonviolent ones, even when these social supports were present, they did not buffer exposed children from behavior problems. These results may be something of an anomaly, however, as other studies have found family-related variables to be more impactful.
Margolin and colleagues have done extensive research on how parenting variables mediate and moderate children’s outcomes in families with domestic violence. Margolin’s research has notable strengths: she relies on community samples that include both violent and nonviolent families and she obtains her data on parents’ and children’s behavior through coded observations rather than self-report. The observational data are obtained from dyadic conversations between the child and one parent and from triadic conversations among the child and both parents.
In a study that involved both dyadic and triadic interactions Margolin and colleagues found that fathers’ physical aggression against marital partners was associated with the fathers exhibiting more authoritarian behaviors, fewer authoritative behaviors, and more negative affect in dyadic interaction with their sons. These fathers also exhibited more controlling behaviors in triadic interactions that included sons. This pattern of findings was not evidenced in interactions with daughters. Two other findings emerged from this study: father-to-mother aggression was not associated with any negative parenting behaviors by mothers, and mothers’ physical aggression in the marital relationship was not associated with any hostile, controlling, or angry behavior toward the children. Two other studies also found that violence between parents was associated with specific parenting behaviors and, in one case, with child functioning. In the first study, parent-to-child hostility in triadic interactions was related to boys’ increased anxiety and distracting behavior. In the second study, aggression between parents was associated with lower levels of father-to-child empathy and higher levels of mother-to-child negative affect in dyadic interactions. Taken together, these findings illustrate that in homes with interparental violence, parents seem to be less emotionally available to their children, and children’s repeated exposure to parents’ negative affect might influence the children’s own emotional reactions and behaviors. These findings appear to be stronger for boys than girls.
A study by Margolin and John is unique in the literature on children’s exposure to domestic violence because it relies entirely on children’s own reports of the violence to which they were exposed, their parents’ behaviors toward them, and their own emotional well-being. The study involved 108 children between 8 and 11 years of age recruited in the community. There were three major findings. First, aggression between parents directly influenced parenting, although the influences are different for boys and girls. Both boys and girls reported that interparental aggression was associated with parenting behaviors that were negativistic, controlling, and punitive, although this association was stronger for boys than it was for girls. Girls also reported a strong association between aggression between parents and positive parenting. Second, both boys and girls in aggressive families reported high levels of hostility (girls’, but not boys’, hostility was linked to negative parenting as well as to aggression between their parents), and high levels of anxiety and depression. For both boys and girls, difficulties in functioning were nearly entirely mediated by negative parenting, meaning that once negative parenting practices were taken into account, the effects of domestic violence on children’s behavior were much less pronounced. Third, violence between parents and negative parenting explained considerably more variance in boys’ behavior than it did in girls’.
Maternal warmth has been found to protect children from the effects of domestic violence. One study of children between ages 7 and 9 found that both boys and girls from domestically violent homes reported high levels of externalizing problems only in homes that were low in maternal warmth. The children from high-warmth homes did not have externalizing behavior problems. On the other hand, in this sample fathers’ warmth interacted with children’s behavior problems in the opposite way. Children whose violent fathers were warm had more behavior problems. The researchers used social learning theory to explain this latter finding, asserting that when warmer bonds exist between a child and a violent parent, the child might view aggression in a more positive light, leading them to behave aggressively themselves.
Researchers have also investigated parenting variables using shelter samples. One such study investigated whether women’s self-reported parenting stress was associated with children’s behavior problems. In this study, sheltered women and their 7- to 12-year-old children were compared with nonsheltered women and children from the same community. Physical and psychological violence against the mothers and parenting stress were measured in both samples. Interestingly, one third of the families in the nonsheltered comparison group reported domestic violence. In this study, domestic violence had a significant effect on both parenting stress and on children’s behavior. Parenting stress did not differ between the sheltered and nonsheltered women, but when violence was used as a continuous variable it was significant in predicting higher levels of parenting stress. Two categories of domestic violence, physical and psychological, were examined to determine the relative effects of both on children’s functioning. Although both types of violence significantly and negatively affected parenting stress, psychological violence was the stronger predictor of children’s functioning. Parenting stress predicted children’s functioning, over and above the effects of domestic violence. Children whose mothers suffered higher levels of parenting stress exhibited more internalizing, externalizing, and total behavior problems.
Similar findings emerged in another sample of sheltered women compared with community women in which parenting stress and violence were the variables of interest. The children in this study were between 2 and 8, spanning the preschool to middle childhood stage; older children in the study, however, had more behavior problems than younger ones and girls had more problems than boys. Among both the battered women and the comparison women, parenting stress predicted children’s behavior problems. Among battered women, the amount of violence in the home in the last year also predicted children’s behavior problems, but even in this group, mother’s parenting stress was the more robust predictor of children’s functioning.
Although most of the literature on school-age children looks at the effect of adult behavior and well-being on children’s functioning, there is evidence from one study of sheltered women and their children that supports the assertion of the ecological-transactional model of development that parents and children affect one another. The researchers examined the relationships among the following variables: extent of interparental physical violence; children’s PTSD symptoms, behavior problems, and intervention in violence between parents; and mothers’ depression, anxiety, and anger. Children’s PTSD symptoms were associated with the amount of physical violence that occurred; their behavioral problems, however, were related to maternal anxiety and anger. Mother’s depression was associated with child intervention in episodes of violence and the quality of the mother–child relationship; anxiety was related to witnessing child abuse, child age, and child internalizing behaviors; and anger was associated with violence-related injuries, violence frequency, and children’s internalizing behaviors. Although the sample for this study was small, and the resulting findings should be interpreted with caution, this study provides valuable preliminary information about the interconnected emotional functioning of battered women and their children that should be validated through further research.
A few researchers have examined child characteristics to determine their relationship with functioning after exposure to violence. The child’s ability to regulate emotional response has been found to mediate the relationship between children’s exposure to domestic violence at age 5 and positive play at age 9, as well as negative peer group interactions, social problems, and internalizing and externalizing behavior problems at age 11. Children’s appraisals of violence also influence their outcomes. Children who feel more threatened by the violence, or who blame themselves for it, experience more adjustment problems. In a separate study, perceived threat and self-blame were found to mediate the relationship between interparental conflict (including but not limited to physical assault) for internalizing problems.
The Effect of Exposure on Adolescents
Adolescence is marked by profound changes in biological, psychological, and social functioning. Much of the internal tension experienced by adolescents is evidenced in increasingly conflicted relationships with parents as adolescents strive to achieve independence and cement important new relationships. Researchers interested in the effect of domestic violence on children have studied adolescents less frequently than younger children; because adolescents are forging new relationships and becoming ever more involved in a world wider than the family. The research concerning adolescents has often examined different outcomes, including dating as well as other forms of violence committed by adolescents.
In one study of dating violence, exposure to domestic violence, destructive communication, gender stereotyping, and attitudes accepting of domestic violence mediated the observed relationship between ethnic minority status, low levels of parent education, and expressed dating violence in 13- to 19-year-olds. When domestic violence and community violence are combined for analysis, violence exposure and PTSD symptoms account for as much as 50% of the variance in adolescents’ violent behaviors. In one large sample of African-American males living in or near public housing projects, violence exposure was strongly associated with expressed violence, but the effects were moderated when youth were less depressed and had a stronger sense of purpose in life.
Researchers who separated the effects of domestic and community violence found that only domestic violence affected the functioning of a group of high-risk adolescents and that the impact was moderated by the adolescents’ self-reported social support. These results have limited generalizability, however. Participants were 65 inpatients between the ages of 13 and 17. Two thirds were hospitalized for suicidal ideation or behavior. The remaining third were hospitalized for homicidal ideation, combined suicidal and homicidal ideation, and assaultive behavior. Prior to hospitalization, 12 of the 65 participants lived in a residential treatment facility. Adolescents whose emotional problems are less extreme might respond differently both to community violence as a stressor and social support as a buffer. In another high-risk population of adolescents (13-18 years of age) drawn from residential treatment agencies and shelters, youth exposed to violence between their parents were more likely to be depressed, run away, and be violent with parents than nonexposed youth. Exposed youth were also more likely to approve of and use violence toward dating partners, but all findings were modest and moderated by gender. The impact of witnessing violence was significant only for males and had no impact on the well-being or behavior of females.
A number of studies have looked exclusively at adolescent’s internalizing, externalizing, and PTSD symptoms, with a range of findings. In a child welfare population, psychological maltreatment, as opposed to experiencing or witnessing physical or sexual assault, had the most profound effect on youth’s internalizing and externalizing behaviors; witnessing family violence had a modest effect only for boys. A study of a community sample of young adolescents (age 11-15) found that fathers’ self-reported physical and verbal violence against mothers’ significantly affected internalizing, externalizing, prosocial behavior, and cognitive functioning as reported by teachers for both boys and girls. Mothers’ self-reported physical and verbal violence did not affect any of the dimensions for either boys or girls in spite of the fact that mothers and fathers reported similar rates of violence. In young adolescents, the use of social supports buffered the negative effects of witnessing domestic violence on participants’ self-reported sense of self-worth, externalizing, and depression. Witnessing violence was associated with symptoms of PTSD in both boys and girls in one study, and to higher internalizing and externalizing problems for only girls in another.
Interventions for Children Exposed to Violence
Several intervention models have demonstrated efficacy for children exposed to domestic violence across developmental levels. In this section, we discuss those models as well as some basic principles of intervention with exposed children and their nonoffending parents.
Child–Parent Psychotherapy
Child–Parent Psychotherapy (CPP) for preschoolers (age 3-5) exposed to domestic violence is a relationship-based intervention grounded in psychoanalytic, attachment, and trauma theory that also includes interventions based in social learning and cognitive behavioral theories. , Because young children organize their responses to stress and danger around their caregiving relationships, CPP promotes repair of the parent–child relationship after the trauma of domestic violence. Weekly treatment sessions involve play to promote parent understanding of the child’s internal world, unstructured reflective developmental guidance, emotional support, concrete assistance with problems of living, and interpretation of behavior and feelings. CPP helps young children and their parents regulate affect, co-construct a narrative of their experiences of violence that have meaning for both of them, and restore a relationship in which both parent and child have confidence in the parent’s capacity to protect the child from harm. In a randomized controlled trial comparing CPP to case management plus community intervention as usual, CPP was more effective than individual treatment in the control group in reducing children’s behavior problems and PTSD symptoms as well as mother’s symptoms of avoidance, with improvements in behavior problems sustained at a follow-up conducted 6 months after the end of treatment. PTSD symptoms were not measured on follow-up for either mother or child, but general functioning continued to improve in mothers who received CPP after the conclusion of treatment, whereas comparison group mothers’ functioning did not. It is noteworthy that while all of the children in this study had been exposed to at least one incident of physical violence perpetrated against their mothers by an intimate partner, they and their mothers had all suffered multiple other traumatic events as well, such as physical or sexual abuse.
Project SUPPORT
Project SUPPORT is designed for children ages 4 to 9 and their mothers as they leave domestic violence shelters. Children with conduct disorders and oppositional defiant disorders and their mothers were included in a randomized controlled trial comparing the intervention to existing services. Children meet with a supportive mentor; mothers meet weekly with a therapist for parent coaching. On average, the intervention lasted 8 months with a mean number of 23 home visits per family. Trained paraprofessionals also provided advocacy and role modeling for the mothers. Children in both the intervention and comparison groups improved in their externalizing behaviors, but the children in the intervention group improved at a faster rate. Participating mothers’ child management skills were enhanced, and treatment gains persisted 24 months after treatment when compared with children in the control group. In addition, at follow-up intervention children showed fewer internalizing problems. There had been no significant improvements in internalizing in either group at the conclusion of treatment.
Kids’ Club and Kids’ Club Preschool
These are group interventions designed for children age 5 to 13. The intervention is offered in both shelter and community settings. These psycho-educational groups for children, with separate groups for mothers, meet weekly for 10 weeks. Children’s groups help children recover from the traumatic effects of exposure to domestic violence, and aim to prevent future problems by discussing feelings and concerns related to violence, increasing coping skills, and addressing distorted thoughts and assumptions about violence. Mothers’ groups provide empowerment, parenting support, and information about the impact of violence on children. An efficacy study sequentially assigned families to child only, mother and child, and wait list conditions. Children in both treatment groups had fewer PTSD diagnoses after treatment than did control children. Only the mother and child group showed significant improvement in externalizing behaviors and violence attitudes compared to wait list controls.
The Learning Club
This 16-week program was developed to provide advocacy services for mothers as they leave shelter and a mentoring experience plus a 10-week educational group for children ages 7 to 11. Contact with families is intense, with an average of 9 hours per family per week. Children’s mentors transport the children to the educational groups and attend the groups with them. When compared with control group children, children in the intervention group improved their feelings of self-confidence. Mothers experienced increased social support. In addition, child abuse was less likely in the intervention group. These changes were maintained at 6-month follow-up.
Youth Relationships Project
This group intervention, originally designed for adolescents with a history of violence exposure and risk factors for abuse, has been used in general school populations as well. Its goal is to prevent intimate partner violence and victimization and to promote healthy relationships. The group activities include psycho-education, skills training, and community involvement; youth are taught that aggression is a choice and are helped to examine the attitudes and power dynamics that foster and sustain relationship violence. In a randomized controlled trial, 14- to 16-year-olds with maltreatment histories including exposure to domestic violence showed a greater decline in PTSD symptoms than did controls.
Other Interventions
Although the previously mentioned interventions are the only ones demonstrated in randomized controlled trials to be efficacious with children exposed to domestic violence, other trauma-focused interventions might also be effective for these children, given the traumatic nature of domestic violence for many children and the high rates of PTSD found in some samples. The authors of two recent reviews of interventions for exposed children , suggest several features that should be included in interventions with this population: (1) a thorough assessment that takes the circumstances of exposure, family members’ individual needs, co-occurring risk factors, and protective factors into consideration, (2) a focus on reexposure or similar interventions shown to be effective to treat the symptoms of PTSD, (3) education about violence and normative responses to violence, (4) a focus on processing emotional cues and affect regulation, (5) a focus on social problem-solving and interaction skills, (6) a focus on safety planning and coping with violence, and (7) inclusion of parents in the interventions. Each of the empirically validated intervention models described above includes some or all of these intervention characteristics.
Implications for Health Care Providers
Although it is clear that witnessing violence between parents has a negative effect on some children, it is impossible to predict from the current literature for whom the effect will be greatest, what conditions may contribute to or even mediate the child’s distress after witnessing violence, and what conditions buffer the effect of violence on the child. What does emerge from the literature is the complex nature of exposure on developing children. How an individual child will be affected depends on the nature of the violent incident and the child’s sensory exposure to it; the child’s gender, individual characteristics, past experiences, relationships both inside and outside the family, and other current stressors; the stressors and protective factors that exist for the family as a whole and for the community in which it is embedded; and on what the child, family, and larger culture think about who is responsible for the violence and whether the violence is justified. Understanding all of these factors is a daunting task for an individual clinician facing an individual child and family.
HCPs do not need to complete such a nuanced picture in order to make effective interventions for the children in their care. The evidence is sufficiently strong to justify screening for domestic violence during well-child check-ups at all stages of a child’s development. Internalizing and externalizing behavior problems and symptoms of PTSD are indicators of distress for children exposed to violence. There are a number of concrete things clinicians can do to help.
Ask about changes in children’s moods and behavior at every well-child visit. When you hear that children are distressed, ask about stressors in the child’s life at school, at home, and in the community. Tell parents how important they are to their children’s recovery. Specific parenting behaviors that help are warmth, ability to listen to children’s concerns and answer their questions, and ability to help children find ways to calm and soothe themselves. Become familiar with interventions for exposed children that are available in your community and take the time to make referrals.
When a child has been exposed to violence or trauma, explain to parents that when their children have been frightened, parents are their most important source of security. Children, especially young ones, feel safer when they are close to their parents. Explain that children may be more anxious than usual about separations. Tell parents that when children have been frightened, they can help by speaking with children about what happened and explaining what will happen next. Predictability and consistency help children feel secure. Parents’ willingness to speak with their children about what happens in their lives gives the child a sense of being cared for and increases their feeling that the world is predictable.
Parents can enhance their children’s sense of control by giving them choices that are consistent with their age and developmental stage. Remember that parents or foster parents might not understand the effect of violence or other forms of trauma on children’s behavior. Their annoyance when their children misbehave or act immature and clingy does not make them bad parents. Parents need information and reassurance from their HCPs so that they can help their children.
When dealing with traumatized toddlers and school-age children, explain to parents and foster parents that their own well-being is important to their children. Help parents find intervention if they need it. Encourage parents to tell their children that the violence is not the child’s fault and that the parent is trying to find ways to be safe. Encourage parents to engage in safety planning with their children so that children know where to go and what to do if they are frightened. Encourage parents to monitor their children’s moods and behavior and obtain help for them if they need it. Explain to parents and foster parents that young children need predictable routines as well as physical closeness and comforting. Encourage them to spend time with their toddlers and preschoolers engaging in joint activities such as reading aloud, singing, and playing that will help restore feelings of togetherness and safety.
For adolescents, social supports are especially important. Group interventions can help tie violence-exposed adolescents to others who can understand their dilemma. Tell parents of violence-exposed adolescents that there are increased risks for self-endangering behaviors and that it is critical that they monitor changes in their adolescents’ mood and behavior.
Strength of Evidence and Future Directions for Research
Although there is substantial evidence that exposure to domestic violence has a detrimental effect on most children, the evidence is not entirely consistent on the pattern of outcomes that children suffer. One major concern about the evidence of impact is that it may be overstated because researchers do not measure co-occurring risks and adversities that may explain more of the variance in children’s outcomes than does exposure to domestic violence. The three metaanalyses of this literature all showed that effect sizes for children’s exposure ranged from small to moderate. One study that did examine co-occurring stressors found that the effect of domestic violence on children’s functioning was reduced to zero for boys when the effects of other stressors were taken into account. On the other hand, this finding is inconsistent with another study that found unique affects when comparing exposed children with a matched group of children who were not exposed but who had equivalent risks for poverty, ethnic minority status, living in a single mother household, and exposure to community violence. Researchers would do well to recall Rutter’s finding that although children can tolerate the effect of one severe stressor without declines in functioning, the experience of two stressors increases their risk fourfold. This is consistent with the findings of Felitti and colleagues, that children who suffered four or more adverse childhood experiences were at greatly increased risk for negative physical and mental health outcomes that extended into adulthood. In studies controlled for co-occurring adversities, researchers and clinicians would be in a better position to assert that experiencing domestic violence affects children’s development.
We have pointed out some additional problems in the research earlier in the chapter. Researchers are not consistent about the way in which they define domestic violence or children’s exposure to violence. Future studies should state clearly the type or types of violence to which child victims have been exposed and describe the range of participant children’s sensory exposure to the violence. The two taxonomies suggested by Holden are exemplary in this regard. Understanding the nature of children’s exposure would help make meaning of the literature. Further, researchers in this field have greatly relied on maternal report. Observational data, and data from a variety of reporters, would enhance the validity of the findings and strengthen the evidence that witnessing violence has a genuine impact on children. Finally, the evidence would be clearer if a wider variety of outcomes were explored. In spite of the fact that trauma response symptoms are widespread among exposed children, relatively few researchers measure either symptom level or presence/absence of PTSD diagnosis. In addition, a variety of outcomes beyond behavior problems and symptoms would shed light on the effects of exposure. It is becoming easier to collect the kind of psychophysiological data that could shed light on the way exposure to domestic violence effects children’s stress response systems (see Chapter 54, “Effects of Abuse and Neglect on Brain Development” ). Relatively few studies have examined the effect of violence exposure on children’s peer relationships. In short, we will not truly understand the phenomenological impacts of children’s exposure until large enough samples of children are recruited to allow researchers to examine the effects of co-occurring stressors, to assess risk and resilience factors, and to evaluate multiple outcomes.
In addition to addressing the weaknesses in the evidence set forth above, future research might focus on four areas that have heretofore been ignored or given short shrift. First, we need longitudinal studies of children’s functioning after exposure to domestic violence. With the exception of treatment outcome studies that conduct follow-ups, the literature on children’s exposure to domestic violence is cross-sectional and provides only a snapshot of the child’s functioning at the time. This leaves unanswered the question of whether the passage of time or developmental changes will ameliorate or exacerbate the effect of violence exposure on children.
Second, given the complexity of children’s experience, intervention research is needed that can establish the extent to which interventions that are effective with populations of children exposed to domestic violence are equally effective for children with the single stressor of violence exposure and children with multiple traumatic stressors in their lives.
Third, mixed design research that captures qualitative as well as quantitative data would give a richer picture of children’s responses to domestic violence. And finally, given the evidence that psychological violence between parents can have a greater effect on some children than physical violence, research is needed to examine the impact of psychological assault on large, heterogeneous groups of children, and to examine the mediators and moderators of this impact.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
