Select Page

Parent/Child Stress in Families Experiencing Domestic Violence

Shawn Worthy

Brian Bagwell

Tricia Hudson

Metropolitan State University of Denver

Author Note: 

Shawn Worthy, Ph.D. is an Associate Professor in the Human Services Department at Metropolitan State University of Denver with a focus on the concentration area of Mental Health and Counseling.

Brian Bagwell, Psy.D., is an Assistant Professor in the Human Services Department at Metropolitan State University of Denver and is co-director of the Department’s Center for Addiction Studies.

Tricia Hudson, Ed.D., is and Assistant Professor in the Human Services Department at Metropolitan State University of Denver with a focus on the concentration area of Mental Health and Counseling.

Correspondence concerning the content of this article can be addressed to Shawn Worthy, Ph.D. at


This analysis is based on a study conducted by an “alliance” of agencies who served the non-offending parent/guardian of children who witness domestic violence. Parents were assessed pre-treatment and post-treatment using the Parent Stress Index-Short Form (PSI-SF) and other measures. Findings indicated that parents in this study reported higher levels of parental distress, parent/child interaction, and reported their child as being more difficult than the normative sample of the PSI-SF. However, they did not report significantly higher Total Stress scores. The parent-child families who received treatment showed significant positive change as a group in comparison to the normative sample. This study supports the findings of current research, which suggests that children who witness domestic violence benefit from early intervention.

Parent/Child Stress in Families Experiencing Domestic Violence 

The National Coalition Against Domestic Violence (NCADV) indicated that in 2006 there were 14,123 domestic violence cases filed in Colorado county courts. Additionally, 19 children were killed during incidents of family violence from 2000 to 2006, and the NCADV reported that boys who witness domestic violence are twice as likely to abuse their own partners and children when they become adults. It is commonly known that domestic violence is pervasive in our culture, crossing geographic, ethnic, and social class lines. In many of the domestic violence situations, children are present to view the verbal and physical violence between the two adults involved. Some of the results of children watching family violence are known and others are still unclear. Of the known outcomes, all are negative including depression, aggression, anxiety, and a greater likelihood of being a participant in domestic violence as an adult (Jaffe, 1990; Kolbo, 1996; Margolin, 2000). In this study, we examine the parent/child relationships in families experiencing domestic violence and the effect of parental stress on parent/child interactions.

The Alliance 

The data presented in this paper is the result of funding provided by the Department of Health and Human Services, Administration for Children and Families. The funding was provided to a collaboration of agencies located south of Denver, Colorado. The participating agencies were AMEND, Arapahoe/Douglas Mental Health Network, The Colorado Coalition Against Domestic Violence, Douglas County Department of Human Services, the State of Colorado Department of Human Services Child Welfare, and the lead agency, The Women’s Crisis and Family Outreach Center. Each of these agencies provided different areas of expertise and treatment modalities.

The Alliance, as this group called itself, was given the charge to provide direct services to children who witness domestic violence and their non-offending parent/guardian.


Domestic violence occurs in all communities in our culture. In families in which domestic occurs, there is a high likelihood that children are present (Fantuzzo J. B., 1997). Some researchers theorize that the greatest frequency in domestic violence occurs early in the familial process, which is a time when children are more typically present (O’Leary, 1989). These findings seem to indicate that in families where domestic violence occurs, there are likely children around to observe its occurrence.

There have been a plethora of studies conducted on the consequences to children who witness domestic violence (Fantuzzo & Lindquist, 1989). A summary of this data suggests that the observation of violence generates emotional, behavioral, social and academic deficits in these children (Jaffe, 1990; Kolbo, 1996; Margolin, 2000). Kitzmann (2003) and colleagues conducted a meta-analysis of research on children who witness domestic violence. Findings from the meta-analysis indicated a significant correlation between children exposure to violence and the child’s problems. The children exposed to the violence had significantly worse outcomes than children who did not observe domestic violence (Kitzmann, 2003).

More specifically, DeLange (1986) reported the following observations about children who witness domestic violence; low self-esteem, lack of impulse control, short attention span, an inability to control anger, physical aggression, and a likelihood of pseudo-maturity. More recently, Grych et. al (2000) conducted a cluster analysis to identify distinct patterns of adjustment in child witnesses.

They found five patterns: multiproblem-externalizing, multiproblem-internalizing, externalizing, mild distress, and no problems reported (Grych, 2000). With the five patterns identified, the researchers (Grych, 2000) found that the clusters were distinguished by how often the child was exposed to the violence, if the parent was aggressive towards the child, and how the child understood the violence. Given the work of both of these researchers, it appears that children who witness domestic violence are impacted negatively in many areas of their life. In addition, it is likely that the depth and type of impact is moderated by variables such as age, interpretation of the violence, and frequency and intensity of the violence.

Whipple (2002) found that parental stress plays an important role in abusive families and those parents in physically abusive families were more likely to experience stress from a number of stressors including low income, younger mothers with less education, and substance abuse. Abusive mothers reported more marital discord and social isolation than mothers who were not abusive (Whipple, 2002). Crouch and Behl (2001) found similar results in a study in which it was noted that situational factors, including high levels of stress, may increase the risk of maltreatment of children including the potential for physical child abuse and that the level of stress related to parenting was significantly associated with child abuse potential. These results provide further support that elevated levels of parental stress have a negative impact on parent/child interactions and increase the potential for maltreatment of the child.

While the data support that elevated levels of parental stress have a negative impact on parent/child interactions, we ask the question: “What is the impact of domestic violence in the home on the non-abusive parent/child relationship?”

In their book, The Batterer as Parent, Addressing the Impact of Domestic Violence on Family Dynamics, Bancroft and Silverman (2002) note that the destructive patterns that batterers and battering can create between mothers and children alternate “with patterns of caring, rescuing, nurturing, playfulness, and cooperation” (Hurley & Jafffe, 1990, p. 473). Other researchers point out that mothers sometimes succeed in finding ways to increase their parenting effectiveness in response to their awareness of the effects of domestic violence (Levendosky, Lynch, & Graham-Bermann, 2000). Bancroft and Silverman identified specific factors that can contribute to the resilience of familial relationships and in the maintenance of healthier family dynamics. These factors associated with increasing family resilience are: (a) a mother who is an unusually competent and caring parent and is able to combine kindness with strong discipline, and does particularly well at shielding her children from the effects of the batterer’s abuse; (b) mothers and children who receive particularly good support from friends, relatives, their religious organization, or other community resources; (c) a batterer who is not a skilled manipulator or whose violence is obvious and extreme and is less successful at causing mutual blame, self-blame, and other divisive and unhealthy dynamics; (d) a batterer who is highly neglectful and uninvolved in any aspect of parenting or who abuses his children directly, physically or sexually (although this behavior can also lead to the most divided families, depending on the batterer’s ability to manipulate some family members while terrorizing, violating, or ignoring others); and (e) a family that receives an especially constructive response from law enforcement, courts, or child protective services, in a way that holds the batterer fully accountable for his actions and that offers support to mothers and children to help them remain close (Bancroft & Silverman, 2002, p. 81-82).

Current Study 

The alliance of agencies set forth to use the best practices of treating children who witness domestic violence and determine the effectiveness of their intervention, as well as the effect of increased levels of parental stress on the parent/child relationship. To that end, parents and children were treated as the need arose, with the tools that clinicians believed would be most effective. Therefore, there was no control group and no single treatment protocol. In this research, there was no attempt to ascribe significant client improvement to any specific treatment or set of treatments. It was also understood that without a control group, change could also be due to unobserved variables. However, the importance of this research lies in the relevance to the effectiveness of cooperation between vested social agencies (public and private) and professionally provided intervention in a community setting.

Course of Treatment and Services 

As indicated above, the non-offending parent/guardian and child/children were provided services based on their needs. The Women’s Crisis and Family Outreach Center (WCFOC) provided the majority of direct services. The staff at the WCFOC is comprised of master’s level therapists who work in residential and out-client care. They referred parents and children to other agencies when the need arose (e.g., psychiatric consultation).

Services that were available to the non-offending parent/guardian included parenting support services, individual support/therapy, legal support, and psychiatric support. Services that were available to children were filial therapy, individual play therapy, group support/therapy, psychiatric support, and equine therapy. As indicated above, the therapist following the case determined which interventions were required. There was no specific intervention protocol. In all cases the parent received parenting support and children received group support/therapy.

The Clients 

In this study, data was collected on 119 children. There were 51 males and 68 females. The ages ranged from 1 to 17 years old. The ethnic makeup of the children was as follows: 69% Caucasian, 10% Hispanic, 9% multiracial, 3% Asian or Pacific Islander, 2% American Indian, 1% African American, 5% other, and 2% unspecified. All of the children had been exposed to substantial verbal abuse in their homes, 50% had been exposed to physical abuse, and 8% had been exposed to life threatening abuse by the non-offending parent in their home. Only 20% of the children had physical violence directed toward them.

Several outcome measures were used, however, in this paper we will be focusing on the results from the Parenting Stress Index-Short Form (PSI-SF) scores. The PSI-SF consists of 36 items derived from the Parenting Stress Index which comprise three scales: Parental Distress, Difficult Child Characteristics, and Dysfunctional Parent-Child Interaction. Normed on more than 2,500 parents, the PSI and PSI-SF are robust measures that maintain their validity with diverse non-English speaking cultures. Both have proven useful in programs aimed at early identification and prevention of family problems, and are helpful in planning intervention and treatment. (Abdinl, 1999). This assessment was given to participants at the beginning and at the end of treatment. As noted above, there was no standard treatment length. Treatment length was based on the needs of the child as determined by their therapist.


The magnitude of the pre-to-post changes were determined by using paired sample T-tests. The alpha level was set at the .05 level to determine statistical significance. The PSI-SF yielded the following results. There was a significant difference in Total Stress t(26) = 2.87, p < .00, which demonstrated a reduction in total parenting stress.











Total Score

Parental Distress

Parent Child Interaction

Difficult Child

Figure 1 

PSI-SF Pre- Post Mean Scores 

Pre-test Mean Score

Post-test Mean Score

There was also a significant decrease in all of the subscale variables: parent distress t(26) = 4.24, p < .00, parent child interaction t(26) = 2.68, p < .01, attention and difficult child t(26) = 3.30, p < .00 (see Table 1, & Figure 1).

Table 1 PSI-SF Pre-Post Mean Scores and Standard Deviation Subscale Pre-test Mean Standard Deviation Post-test Mean Standard Deviation
Total Score 75.36 20.91 66.81 19.03
Parental Distress 24.55 8.10 23.04 8.83
Parent Child Interaction 20.92 6.62 18.37 6.04
Difficult Child 29.98 9.16 25.41 7.27

Using a two sample t-test, comparisons were conducted between the normative sample (Abdinl, 1990) of the PSI-SF and the sample obtained in this research. This analysis was done to compare the parental stress between the families who experienced domestic violence in this study and those who had not.

For those families evaluated with the PSI-SF at pretest there were significantly elevated scores for the research sample in parental distress: t(98) = -2.02, p < .05, parent/child interaction t(92) = 2.99, p < .00, and difficult child t(93) = 2.92, p < .00. However, there was no significant difference for total score t(10) = 2.12, p < .05. In order to detemine if treatment was effective in reducing parent stress to normal levels, the post-test means were compared to the normative sample means. There was no significant difference found for total stress, parent/child interaction, or difficult child (see Figure 2). However, the parental distress subscale was significnaly lower t(27) = -1.99, p < .05 than the normative sample (see Figure 2).


As indicated above, domestic violence can affect the parent-child relationship. However, how domestic violence in a family affects the relationship between the non-perpetrating parent and child remains unclear (Lapierre, 2010). The authors of this study examined parenting stress between the non-perpetrating parent and their children to determine the intensity of parenting stress and the effect of treatment on parenting stress.

The results of this study indicated that the non-perpetrating parent does not experience the parent-child interaction as substantially more stressful than the normative sample. This finding may be due to the fairly large variability of scores, which may suggest that violence in the family effects different parents in different ways. In fact, some researchers (Letourneau, 2007) suggested mothers who experience domestic violence my respond by becoming more sensitive and responsive to their children than those in the general population. Other data suggested that mothers who have experienced abuse may become more abusive toward their own children (Huth-Bocks, 2004). This variability of parenting style may account for a large variability in parenting stress.

Although in total there was no significant difference in the total score for parent stress, significant differences were found for subscales. The somewhat contradictory findings of this study suggested that parents who have experienced partner violence expressed greater parental distress, negative parent/child interaction, and reported the child as being more difficult than the normative sample. It is understandable that these parents, fresh out of a situation in which they experienced violence, may feel distressed about a variety of things that their children have experienced. During this period of transition, they may experience more difficult parent-child interactions because of factors such as having to move out of their home, not seeing the other parent, and dealing with numerous logistical issues. The children in this group were also identified as being more difficult than the normative sample. In a previously reported portion of this study, evidence suggested (Worthy, 2011) that the parents in this study may identify their children as demonstrating more negative behaviors, particularly boys.

An encouraging outcome of this study suggests that mental health treatment has a substantial effect on all aspects of parent-child relationships.

There was a significant positive change within the treatment group for the total score and for all subscales of the Parent Stress Index-Short Form. Not only was there a significant change within the treatment group, but participants who received treatment ended with mean scores below the normative level for all subscales, although the differences were not statistically significant (refer to figure 2).

Limitations of the Study 

One major limitation to this study is the lack of control group. Because of the absence of a control group, we cannot be certain that the changes observed where due to mental health treatment, or were merely changes that naturally occurred when a non-perpetrating parent and their children leave a physically abusive home environment.

Another limitation for this study was the lack of control over which support service and/or mental health services were provided. This study took place in the natural environment of active agencies providing service to individuals in crisis; as such, the primary goal was to provide integrative services to assist clients. This study was an attempt to understand the outcome effect of the intervention provided on the families being served. The authors placed greater importance on assisting families through their crisis, as opposed to experimental control; therefore, no conclusive statements can be made about the cause of the improvement in parental stress. However, the results of this study are encouraging because of the positive relationship found between mental health and supportive interventions, and the possibility of improving parent-child relationship for individuals coming out of situations involving family violence.

Recommendations for Future Research 

The results of this study suggest that mental health treatment may be an important part of improving the life of parents and children as they exit a physically abusive situation. If additional research supports the findings in this paper, we will need to conduct research to determine the most salient components of treatment. Also, we may begin to answer the question as to whether treatment can be used to reduce parent-child stress and conflict prior leaving the abusive home environment. These finding can assist us in reducing the impact on children and parents in homes with domestic violence.


Abdinl, R. (1990). Parenting Stress Index Test Manual, 3rd ed. Charlottesville, VA: Pediatic Psychology Press.

Achenbach. (1991). Manual for the child behavior checklist and 1991 profile. Burlington, VT: University Associates in Psychiatry.

Briere, J. J. (2001). The Trauma Symptom Checklist for Young Children (TSCYC): Reliability and association with abuse exposure in a multi-site study. Child Abuse & Neglect , 25, 1001-1014.

Briere, J. (1996). Trauma Symptom Checklist for Children (TSCC) Professional Manual. Odessa, FL: Psychological Assessment Resources.

DeLange, C. (1986). The family place children’s therapeutic program. Children Today , 15, 12-15.

Fantuzzo, J. B. (1997). Domestic violence and children: Prevelence and risk in five major U.S. cites. Journal of the American Academy of Children and Adolescent Psychiatry , 36, 116-122.

Fantuzzo, j. D. (1991). Effects of interpersonal violence on the psychological adjustment and competence of young children. Journal of Consulting and Clinical Psychology , 59, 258-265.

Grych, L. J. (2000). Patterns of adjustment amoung children of battered women. Journal of Consulting and Clinical Pyschology , 68, 84-98.

Huth-Bocks, A. L. (2004). The impact of domestic violence on mothers’ prenatal representation of their infants. Infant Mental Health Journal, Vol 25 , 79-98.

Jaffe, P. W. (1990). Children of Battered Women. Newbury Park, CA.: Sage.

Kitzmann, K. G. (2003). Children witnesses of domestic violence: A metapanalytic review. Journal of Consulting and Clinical Psychology , 71, 339-352.

Kolbo, J. B. (1996). Children who witness domestic violence: A review of empirical literature. Journal of Interpersonal Violence , 11, 281-293.

Kot, S. L. (1998). Intensive child-centered play therapy with child witnesses of domestic violence. International Journal of Play Therapy , 7, 17-36.

Lapierre, S. (2010). Striving to be “Good” Mothers: Abused women’s experience of mothering. Child Abuse Review Vol. 19 , 342-357.

Letourneau, N. F. (2007). Mothering and domestic violence: A longitudianal analysis. Journal of Family Violence, Vol 22 , 649-659.

Margolin, G. &. (2000). The effects of family and community violence on children. Annual Review of Psychology , 51, 445-479.

O’Leary, K. B. (1989). Prevalence and stabilty of physical aggression between spouses: A longitudinal analysis. Journal of Consulting and Clinical Psychology , 57, 263-278.

Smith, N. &. (2003). Intensive filial therapy with child witnesses of domestic violence: A comparison with individual and sibling group therapy. International Jounral of Play Therapy , 12, 67-88.

Worthy, S. (2011). Children who witnessed domestic violence. 2011 NOHS National Conference Proceedings (pp. 32-37). San Antonio: National Organization of Human Services