The Moderating and Mediating Role of Responsive Parenting Behavior in Explaining the Link between Intimate Partner Violence and Posttraumatic Stress Symptoms in Young Children

Abstract

Responsive parenting serves an influential role in explaining the link between children’s exposure to intimate partner violence (IPV) and children’s mental health impairment, but how this occurs is not well elucidated. In some cases, researchers examine parenting as a mediator to explain how IPV leads to maladaptive outcomes (i.e., IPV negatively impacts one’s capacity for responsive parenting, which in turn impacts children), whereas others examine moderation in which either the absence of responsive parenting exacerbates adverse outcomes or increased responsive parenting buffers risk. Mediation addresses theoretical questions about how or why IPV leads to maladaptive outcomes, whereas moderation addresses who might be most impacted. However, responsive parenting has rarely, if ever, been tested as both a mediator and moderator of the link between IPV and posttraumatic stress symptoms (PTSS) within the same sample. The current study examined the mediating and moderating role of responsive parenting on physical IPV exposure and child PTSS in a longitudinal sample of 391 children ages 3 to 5 years (M = 4.74, SD = 0.89). Self-report measures of physical IPV exposure, parenting practices, and PTSS were completed by mothers. We found that responsive parenting significantly moderated and mediated the association between physical IPV exposure and child PTSS over time. Studies that include tests of both moderation and mediation are critical for advancing mechanistic insight into the role of parenting in the etiology of mental health impairment in children exposed to IPV.

Introduction

Intimate partner violence (IPV) and parenting have a significant impact on child well- (Artz et al., 2014; Graham-Bermann & Levendosky, 1998; Howell et al., 2016). While there are large literatures documenting that IPV and parenting are independent predictors of children’s mental health outcomes, especially posttraumatic stress symptoms (PTSS). PTSS in children includes of a variety of symptoms such as avoidance, hypervigilance, increased arousal, re-experiencing, and other cognitive and emotional changes (Connor et al., 2015; Ehrensaft et al., 2017). There is consistent evidence that IPV and non-optimal parenting (defined here as more harsh and/or less responsive parenting) commonly co-occur and contribute to cumulative risk (Briggs-Gowan et al., 2019; Knickerbocker et al., 2007; Levendosky & Graham-Bermann, 2000; Slep & O’Leary, 2005). Various analytic models can be used to examine how IPV and parenting impact child outcomes; however, these disparate analytic models are poorly integrated, which can result in underdeveloped theories. Given the significant public health implications of IPV and parenting, specifying the nature of the relationship between these experiences and how they independently and together confer risk is critical for identifying modifiable targets for early intervention on children’s behavior and PTSS (Cameranesi & Piotrowski, 2020; Labella & Masten, 2018).

In the current study, we focus on how exposure to physical IPV and low levels of responsive parenting contribute to the development of PTSS in young children by integrating analytic models often conducted separately – moderation and mediation. We begin by briefly reviewing the evidence examining the bivariate effects of IPV and parenting on PTSS. Next, we discuss key differences between mediation and moderation analyses and the implications of these two approaches. We then review existing studies that have applied a mediation or moderation framework to explore the relationships between IPV, responsive parenting, and child psychopathology.

IPV and Parenting Influences on Childhood Psychopathology and PTSS

There is a vast literature documenting the association between childhood IPV exposure and psychopathology in children (Lee et al., 2022; Vu et al., 2016). Meta-analytic results demonstrate that the effect size of IPV on PTSS is significantly larger compared to effect sizes of IPV relative to other psychopathology symptoms (Evans et al., 2008). IPV contributes to the development of PTSS in very young children (Graham-Bermann & Levendosky, 1998; Levendosky et al., 2013), with exposure to severe IPV associated with emergent PTSS as early as infancy (Bogat et al., 2006). IPV is widely viewed as a contributor to PTSS and maladaptive outcomes more generally, even when accounting for other risk factors (Holmes, 2013; Lee et al., 2022).

Similarly, responsive parenting also impacts PTSS. Responsive parenting is often characterized by consistent, sensitive, and contingent caregiving behaviors to address children’s needs and involves parental warmth, involvement, and appropriate monitoring of the child (e.g., Bornstein et al., 2008; Eshel et al., 2006; Landry et al., 2006). Although PTSS research has often focused on negative parenting, recent meta-analytic findings show small, but significant effects of responsive parenting on PTSS in children (Williamson et al., 2017). High levels of responsiveness have been found to lower risk for PTSS and mental health symptoms in general (Ehrensaft et al., 2017; O’Campo et al., 2010). Perhaps even more importantly is that studies of violence exposed families, maternal responsive parenting is inversely associated with children’s PTSS (Greene et al., 2020).

Mediation and Moderation

Mediation and moderation analyses are two common approaches for studying developmental psychopathology, including associations between IPV, parenting and childhood outcomes (e.g., Levendosky et al., 2002;, 2006). However, mediation and moderation are not simply different statistical analyses, but have distinct theoretical and practical implications (Baron & Kenny, 1986; Holmbeck, 1997).

Parenting as a Mediator of the IPV-child Psychopathology Relationship

Mediation analyses are used to understand intermediate processes between variables, the how and why a relationship occurs. In developmental psychopathology, mediation holds an important role, for instance, to show that parenting may be a pathway through which a risk factor, such as IPV, impacts child outcomes. The discovery of such a mediator might suggest a prime target for intervention. Within a mediational framework, the role of parenting behavior in the current study might be conceptualized as consisting of two parts: (1) physical IPV may predict less responsive parenting, and this in turn may (2) predict greater PTSS in children. If the effect of IPV on children’s outcomes is exerted via responsive parenting (i.e., the mediator) then warmth, involvement, and monitoring might be important targets for intervention.

Mediation models are represented in studies that theorize a “spillover” effect, in which IPV impacts parenting, which in turn impacts children’s mental health (Grasso et al., 2016; Knickerbocker et al., 2007; Levendosky et al., 2006; Pu & Rodriguez, 2021). For example, IPV may compromise a victimized parent’s personal resources (e.g., adaptive emotion regulation strategies, tolerance), which are necessary for optimal parenting. Alternatively, the pattern of aggressive and hostile behavior that characterizes the partner relationship may repeat in the parent-child relationship (Slep & O’Leary, 2005). The spillover hypothesis asserts parenting as a mediator and several studies support this model (Grasso et al., 2016; Greene et al., 2018; Huang et al., 2010; Levendosky et al., 2006; Levendosky & Graham-Bermann, 2001). Although most studies have focused on harsh parenting, a growing body of literature examines responsive parenting including parental warmth or involvement (Miller-Grafet et al., 2016; Greeson et al., 2014). Miller-Graf and colleagues (2016), for example, found that the relationship between IPV exposure in childhood and adult psychopathology symptoms was mediated by parent-child warmth. However, this finding is not always replicated (Greeson et al., 2014). Taken together, there is evidence that parenting may be a mediator of the link between IPV exposure and children’s mental health, although more research is needed.

Parenting as a Moderator of the IPV- psychopathology Relationship

In moderation analyses, a third variable affects the strength of the relationship between two variables. Because of this, moderation can be used to examine when or for whom the strength of an association between two variables is altered. In this way, moderators can be useful for identifying those who might be at greatest risk and benefit the most from intervention. For instance, if IPV is associated with the worst outcomes in children who experience lower levels of responsive parenting, selecting these children might maximize prevention and intervention efforts.

Parenting is often thought of as a moderator of the relationship between IPV exposure and child mental health impairment, as higher levels of responsive parenting may serve as a protective factor or “buffer” against risk associated with IPV exposure. One school of thought suggests that the presence of a responsive parent may provide essential scaffolding of emotional processes that may buffer the impact of IPV on children’s emotion regulation and threat-response processes. Indeed, there is evidence of responsive parenting as a moderator of the link between IPV and children’s mental health in a number of studies (Ehrensaft et al., 2017; Levendosky et al., 2002; Rossman & Rea, 2005; Skopp et al., 2007). In particular, Ehrensaft and colleagues (2017), for example, found that parenting characterized by child-centeredness, time spent with the child, and closeness was protective against child PTSS.

The Present Study

Overall, there is evidence that parenting might mediate and moderate the relationship between IPV and children’s functioning; however, to the best of our knowledge, we are unaware of existing studies that have tested whether parenting is a mediator and/or moderator within the same sample or within the same analytic model. Indeed, testing whether a single variable is both a mediator and moderator in the same analysis is relatively uncommon in developmental psychopathology research even when there is clear evidence that a variable might possess both qualities, such as with parenting practices (Goldstein et al., 2021). Such a model has important theoretical implications. If responsive parenting is a mediator of the link between IPV and child psychopathology, it implies that responsive parenting itself should be targeted and changed to improve outcomes. If responsive parenting is a moderator variable it provides information about who should be prioritized for treatment. The possibility that responsive parenting plays both a mediating and moderating role may suggest that it is a particularly important etiological variable. Additionally, most studies have historically focused on harsh parenting, rather than looking at responsive parenting alone.

Therefore, the current study examined whether three aspects of responsive parenting (parental involvement, parental monitoring, and warm parenting) serve as potential mediators and moderators of the link between physical IPV and PTSS in a sample of 391 mother and child dyads. Given existing evidence, we hypothesized that responsive parenting would serve both roles. If responsive parenting is indeed a moderator and mediator it suggests that parenting is not only an important variable for identifying and selecting children who could benefit most from intervention, but also that responsive parenting itself is a key variable to target within those intervention efforts.

Methods

Participants included in the current analysis are part of the (withheld for blind review) study. Mothers were initially recruited for a survey study from pediatric primary care settings if they had a child between 3 and 5 years of age, yielding a large diverse sample (N = 1,857). Informed consent was obtained and all procedures were approved by Institutional Review boards at (withheld for review) Universities. Subsequently, a subset of mother-child dyads oversampled for experiences of IPV or child behavior problems were invited for an intensive assessment longitudinal cohort, and 497 participated. This portion of the study took place during the years of 2011–2014. The current study utilized data from two time points of the intensive cohort separated by 6–9 months. At baseline, mothers reported their experiences of physical IPV in the past year and recent parenting behaviors. At follow-up, mothers reported on their children’s PTSS.

Of the 497 mother-child dyads, 391 had complete data (the remainder did not complete a measure of physical IPV, of which, 32 were partnered, 12 identified a partner, but reported being separated, and 62 did not report having a partner). Children in the final sample was on average 4.74 years of age at the first assessment (SD = 0.89; range 3.10–7.05) and 5.49 years at the follow-up assessment (SD = 0.92; range 3.75–7.82). Children in the final sample were evenly divided by sex (male; N = 201), most children were identified as a racial or ethnic minority (N = 295), and most were living in or near poverty (N = 234). Those missing data were more likely to have female children (participants with complete data: N = 190, 48.59% female; incomplete: N = 63, 59.43% female, χ = 3.92, p = .05), identify as a racial or ethnic minority (completers: N = 295, 75.5% minority; non-completers: N = 95, 89.6% minority, χ = 9.92, p < .01), or be impoverished (completers: N = 234, 59.9% impoverished; non-completers: N = 76, 71.7% impoverished, χ = 4.99, p < .05). Including these variables in the primary analytic model did not alter results and were thus, not included in the final model.

Measures

Intimate Partner Violence

The Revised Conflict Tactics Scale (CTS2; Straus et al., 1996) was used to measure the presence and frequency of physical IPV within the past year. We used 48 items in which participants were asked to rate the frequencies at which behaviors had occurred between themselves and their partner (e.g., never, once, twice, 3–5 times, 6–10 times, 11–20 times). Examples items include, “I twisted my partner’s arm or hair.” For this paper we collapsed the physical abuse partner to self and self to partner scales which both had good psychometric properties (alpha = 0.88 and 0.80). The CTS has good psychometric properties and is a widely used measure of IPV.

Child Posttraumatic Stress Symptoms

The Trauma Symptom Checklist for Children (TSCYC; Briere et al., 2001) was used to measure PTSS via maternal report over the past month. The TSCYC is widely used and appropriate for children aged 3 to 12. For the current study we used 27 items. The TSCYC does not require a PTSD qualifying event in order to be completed. Mother’s used a 4-point frequency scale ranging from “not at all” (1) to “very often” (4) to rate their child’s arousal, avoidance and intrusive symptoms. Internal consistency was excellent (alpha = 0.92).

Responsive Parenting

The Parenting Practices Questionnaire (Gorman-Smith et al., 2010) was used as a self-report measure of several parenting domains including positive parenting, monitoring, and involvement. These domains are assessed by three subscales within the Parenting Practices Questionnaire including positive parenting (e.g., “Praise or give approval to your child?”), monitoring (e.g., “Supervise an activity with your child and his/her friends?”), and involvement (e.g., “Read stories to your child?”). For this study, the original monitoring scale was supplemented with items appropriate for younger children, e.g., “Check on your child regularly while playing, even if he or she is quiet”. Mothers used a 5-point frequency scale ranging from “never” (0) to “always” (5) to describe their parenting behavior over the past month. The Parenting Practices Questionnaire has good psychometric properties; internal consistency was good for all subscales (positive parenting alpha = 0.85; monitoring = 0.75; involvement = 0.80).

Data Analysis

Descriptive statistics were performed in SPSS version 21 (IBM). Given that physical IPV was not endorsed by the vast majority of participants (N = 289; 73.9%), we dichotomized the variable. PTSS were skewed (skewness = 3.95), a commonly observed distribution for symptom data, and therefore, were log transformed. The three parenting scales were used to create a latent responsive parenting variable that fit the data well as the model was just identified (Comparative Fit Index and Tucker Lewis Index = 1.00); all variables loaded highly and significantly on the latent variable (standardized loadings ranged from 0.71 to 0.85).

Figure 1 depicts the primary analysis, which examines moderation and mediation in a single model (for descriptions of this model see Goldstein et al., 2021; Hayes, 2015). In this model, physical IPV is the independent variable, which predicts the latent responsive parenting variable (the mediator) and this predictive pathway is labeled a. Physical IPV directly predicts PTSS, the dependent variable and this pathway is labeled c. Parenting predicting PTSS is labeled b. Additionally, an interaction term, which is the product of physical IPV and the latent responsive parenting variable also predicts PTSS and is labeled b₂. Importantly, moderation and mediation are not evaluated as independent effects, but rather are determined by the Index of Moderated Mediation (Hayes, 2015), a statistic that is calculated by multiplying the path coefficient of physical IPV on responsive parenting (a) and the path coefficient of the interaction term predicting the outcome (b₂). As is standard practice in mediation models, an asymptotic 95% confidence interval is used to determine if the index of moderated mediation is significant. The model was examined in Mplus version 8 (Muthén & Muthén, 2017) using Maximum Likelihood estimation and Bootstrapped 10,000 times. Pathways are reported as unstandardized estimates, as standardized estimates are not provided when estimating latent variable interactions in Mplus.

Fig. 1.

Physical IPV, Responsive Parenting, and IPV as a longitudinal moderation and mediation model. Note: All pathways are unstandardized estimates. CI = Confidence Interval; * = p < .05; **= p < .01; *** = p < .001

Results

Table 1 reports sample demographics, responsive parenting, and PTSS for those children with and without exposure to physical IPV. Family poverty significantly differed between the groups, with those exposed to physical IPV being more likely to be impoverished (χ = 12.35, p < .001). As a result of this observation, we further examined the relationship between poverty and parenting, PTSS, and considered it for inclusion in our primary model. Poverty was not significantly associated with any parenting subscale (t = 0.34, p = .71, for parental involvement; t = -1.53 p = .13, for parental monitoring; and t = 0.28, p = .78, for positive parenting), but was significantly associated with PTSS (t = -2.01, p = .045). However, poverty status did not influence the results of the final analytic model and was therefore not included. All aspects of responsive parenting and child PTSS significantly differed between children with vs. without exposure to physical IPV (parental involvement: t = 3.52, p < .001; parental monitoring: t = 2.19, p < .05; positive parenting: t = 2.17, p < .05; PTSS: t = 2.68, p < .01). Table 2 shows the correlations among the three parenting scales and log transformed PTSS. Each parenting scale was significantly and highly correlated with the other (rs range = 0.51 − 0.60). However, the three responsive parenting scales were not significantly correlated with child PTSS.

Table 1.

Descriptive data for children with vs. without exposure to physical IPV

	Non-exposed N = 289	IPV exposed N = 102
Child Sex (%, Female)ns	48.79%	48.04%
Child Age at baseline (Mean and SD)ns	4.74 (0.88)	4.75 (0.90)
Child Age at follow-up (Mean and SD)ns	5.48 (0.90)	5.54 (1.00)
Child Race/Ethnicity (%)ns
Black	41.18%	55.88%
Latinx	30.80%	23.53%
White	26.30%	19.61%
Other/Prefer not to say	1.73%	0.98%
Poverty (%, impoverished)***	54.67%	74.51%
Maternal Education (%)ns
Less than High School	2.81%	2.91%
HS Graduate/GED	19.65%	22.55%
Associate degree	17.19%	18.63%
Some college	30.18%	35.29%
Bachelors Degree	18.95%	8.82%
Graduate Degree	11.23%	8.82%
	Mean (SD)	Mean (SD)
Parental Involvement***	4.26 (0.62)	3.98 (0.83)
Parental Monitoring*	4.04 (0.69)	3.86 (0.72)
Positive Parenting*	4.32 (0.68)	4.14 (0.81)
Child PTSS (values range 27–90)**	30.16 (5.56)	32.11 (7.24)

Note. Superscripts indicate significant differences between those with vs. without exposure to physical IPV. ns = non-significant; * = p < .05; ** = p < .01; *** = p < .001

Table 2.

Correlations among individual parenting scales and child Posttraumatic stress symptoms

	Involvement	Monitoring	Positive	Child PTSS
Parental Involvement	-	0.60***	0.51***	− 0.02
Parental Monitoring		-	0.60***	− 0.03
Positive Parenting			-	− 0.07
Child PTSS				-

Note: * = p < .05; ** = p < .01; *** = p < .001

The primary analytic model examining responsive parenting as a moderator and mediator of the link between physical IPV and child PTSS is shown in Fig. 1. As can be seen in the figure, past year physical IPV exhibited a significant negative association on recent responsive parenting (unstandardized b = -0.40, standard error = 0.15, Z = -2.73, p < .01). Since past year physical IPV is a dichotomous variable and responsive parenting is a standardized latent variable, the b = -0.40 can be considered a mean difference such that those with past year IPV experiences are about half of a standard deviation lower on responsive parenting compared to mother’s without past year IPV experiences. The direct path from past year physical IPV predicting child PTSS was also significant (b = 0.02, standard error = 0.01, Z = 2.28, p < .05). While the direct path from responsive parenting predicting child PTSS was not significant (b = 0.003, standard error = 0.004, Z = 0.60, p > .05), the interaction between IPV and responsive parenting was significant (b = -0.02, standard error = 0.01, Z = -2.19, p < .05). The moderation effect forms a pattern suggesting that the presence of IPV and lower levels of responsive parenting were associated with higher levels of PTSS. The mediation effect, the indirect effect of physical IPV on PTSS via responsive parenting, as determined by the product of a and b paths is not significant (b = -0.001, 95% CI = -0.006–0.002). However and most importantly, the 95% confidence interval of the index of moderated mediation did not include zero (b = 0.008, 95% CI = 0.002–0.020). In other words, responsive parenting functions as a mediator of the link between physical IPV and PTSS, only when physical IPV is present in families. This finding supports a moderation and mediation model explaining the link between physical IPV and child PTSS via responsive parenting.

Discussion

The current study supported a combined moderation and mediation framework in which the link between past year physical IPV and child PTSS 6–9 months later was moderated and mediated by responsive parenting. Specifically, our results revealed that past year physical IPV predicted less responsive parenting (less parental monitoring, involvement, and positive parenting) and that responsive parenting interacted with physical IPV such that children who experienced both less responsive parenting and physical IPV had greater PTSS. We found that there was a significant test of the index of moderated mediation suggesting that responsive parenting had a significant indirect effect on PTSS, but only when physical IPV was present.

Mediation alone was not significant if the moderating effect was not also included (as is done within the index of moderated mediation). What this means is that responsive parenting was not a mediator for our sample on average. However, when physical IPV was present and responsive parenting was low, responsive parenting also acted as a mediator (i.e. the interpretation of the index of moderated mediation). This is a noteworthy finding. If we tested mediation alone we would have reported a non-significant result and concluded that responsive parenting was not functioning as a mechanism explaining the link between physical IPV and PTSS. This is precisely the shortcoming of not considering both moderation and mediation simultaneously within the same analytic model. Indeed, by instead reporting the index of moderated mediation, responsive parenting takes on more nuanced implications as moderators may be useful for determining who would benefit from intervention and mediators suggest which variables should be targeted within an intervention program itself.

Our results are largely consistent with existing research showing that responsive parenting is a mediator (Grasso et al., 2016; Levendosky et al., 2006), as well as a moderator (Ehrensaft et al., 2017; Levendosky et al., 2002; Rossman & Rea, 2005) of the relationship between IPV and child PTSS. What makes the current paper unique is that we examined the possibility of moderation and mediation in the same sample within the same analytic model. The fact that we observed both moderation and mediation in this model suggests that viewing parenting solely as a mediator or a moderator yields an incomplete picture.

While the current study is just one set of findings and clearly replications are necessary, the presence of moderation and mediation simultaneously may be important for establishing the most accurate theoretical framework. Studies that only consider parenting as an etiological stepping stone (i.e. mediator) or instead as only a buffering protective factor or exacerbating risk factor (i.e. moderator) reveal only part of the picture. The fact that prior studies have not simultaneously tested parenting as a mediator and moderator is an impediment to developing the most accurate theoretical models of the link between IPV and children’s PTSS and, ultimately, slows progress toward mental health interventions for children exposed to IPV. Rather than this piecemeal approach, future research should consider testing both moderation and mediation possibilities simultaneously. Indeed, studies like this one support responsive parenting as playing a complex role that further elevates responsive parenting as a key target for intervention in homes where physical IPV occurs. It is important to emphasize that in our paper we examined a latent variable comprised of components of responsive parenting (parental monitoring, involvement, and warm parenting), and that less of this responsive parenting was maladaptive for children. However, an absence of responsive parenting (e.g., low warmth) may not be the same as excessive negative parenting (excessively harsh discipline and corporal punishment) such that results may differ depending on the way in which parenting is measured.

The current study has several strengths including its large sample size, longitudinal design, and integrative analytic framework, but also has several limitations. First, the study relied entirely on maternal report. Future work could be improved by using observational measures of parenting or interview-based assessments of symptoms. Second, we only examined physical IPV as opposed to other forms of IPV (e.g., verbal/psychological), which means that our findings may be specific to physical IPV exposure. Third, we were unable to examine physical IPV severity, chronicity, or early life IPV exposure (only past year IPV). However, our inability to examine these other factors likely had a conservative effect on our results. Indeed, it is possible that more severe and chronic IPV could have larger impacts on parenting and PTSS than we were able to observe. Fourth, the follow-up period was relatively short and effects might be reduced at longer intervals. In particular, the recency effect of IPV on parenting might wane with time (e.g., mothers’ responsive parenting may improve as more time passes since an IPV incident). Future studies would benefit from more clearly delineating the temporal order of variables by using three time points over a longer period of time. Lastly, we were not able to control for the effect of PTSS at baseline, which may impact parental behavior.

Overall, this study contributes to the broader IPV literature by demonstrating that responsive parenting is both a mediator and a moderator of the link between IPV and PTSS. While previous studies have found that responsive parenting may be a moderator and mediator, this paper extends those existing findings by testing moderation and mediation in a single analytic framework. While much of the literature on IPV focuses on harsh parenting, our results also emphasize the importance of responsive parenting in understanding outcomes following IPV exposure. Moreover, if these findings replicate, responsive parenting may impact approaches to intervention in two ways. In general, moderators suggest who should be selected for interventions and mediators suggest what processes should be targeted within the intervention itself. In this case, since responsive parenting was found to be a moderator, it suggests that children from homes with less responsive parenting, should likely be targeted for interventions (e.g., those with lower levels of warmth, monitoring, and involvement), as these children had the worst PTSS outcomes. Additionally, since responsive parenting was also found to be functioning as a mediator, then responsive parenting should be a direct target in intervention programs, as improving responsive parenting may disrupt the link between IPV and PTSS. Fortunately, there are already several intervention programs with relevance to the problem of IPV and improving parenting practices. Interventions such as Triple P (Sanders, 2008; Sanders et al., 2014) and the Nurse Family Partnership (Olds, 2006) are well validated programs that have been found to improve parenting practices and reduce indicators of child abuse and neglect. Based on our work, there may be a public health benefit for screening families for the presence of IPV and responsive parenting at pediatric clinics and then providing referrals to these or other relevant intervention programs.

In sum, future etiological research would benefit from taking an integrative modeling approach to mechanisms and risk factors of children’s outcomes following IPV exposure. In the event that replications support responsive parenting as functioning simultaneously as a moderator and a mediator of the link between physical IPV and PTSS, responsive parenting may be an important variable to target within an intervention program as well as for determining which children may benefit most from intervention.

Funding

Financial support for the study was provided by grants from the National Institute of Mental Health (U01MH090301; R01MH082830) and by the National Institute of Mental Health Intramural Research Program.

Declarations

Conflict of Interest

The authors do not have any conflicts of interests to report.