Intimate partner violence (IPV) is an undesirable, yet common social phenomenon. It has been defined as “a pattern of assaultive and coercive behaviors” that includes threats, psychological abuse, physical aggression, and other hostile behaviors (Kiely et al. 2010, p. 1). Verbal (e.g., insults, yelling, humiliation) and physical (e.g., pushing, shoving, choking) behaviors are included as components of IPV (e.g., Black et al. 2011). It is estimated that 33 % of women and 28 % of men are targets of physical violence and 48 % of women and men are targets of verbal aggression by an intimate partner during their lifetimes (Black et al. 2011). Prevalence rates of IPV are particularly elevated in economically disadvantaged segments of society (Jones et al. 1999), among Black families (Black et al.), amongst partners with children (Fantuzzo et al. 1997), and amongst cohabitors (Stets and Straus 1989). For example, in a national survey, Black et al. reported that 41 % of Black women had experienced physical violence from a partner with 24 % reporting severe physical violence. Families with the four characteristics just noted (cohabitating, Black, low-income, and experienced IPV) are the focus of this paper.

Cohabiting Families and IPV

Family conflict and violence is not limited to married couples. It also occurs in cohabiting relationships—structures in which two romantically involved but unmarried individuals live together (McDonald et al. 2006). Such unions have become considerably more prevalent over the past decades (e.g., Rinelli and Brown 2010; Skinner et al. 2002). In fact, it is estimated that nearly 50 % of U.S. children will live in a cohabiting family constellation before reaching adulthood (Bumpass and Lu 2000; Kennedy and Bumpass 2008). This trend is particularly pronounced in Black families, which have both the lowest rate of marriage and the highest rate of children born outside of marriage of any racial and ethnic group in the U.S. (see Dunlap et al. 2010; Martin et al. 2013). Despite these demographic trends, surprisingly little is currently known about cohabiting family structures (Sassler 2004; Sassler and Miller 2011) or IPV in these families.

Of relevance to the present discussion on IPV and child outcomes is the finding that cohabiting family structures appear to experience higher rates of conflict than any other family constellation. Stets and Straus (1989) reported that rates of physical aggression were the highest for cohabitors (35 %), followed by daters (20 %), and married coupled (15 %). Similarly, Brown and Bulanda (2008) found that, among women, cohabitors had the highest rates of relationship violence, both as perpetrators and as victims. By extension, such results suggest that a larger percentage of children residing in cohabiting unions may be exposed to IPV than those being raised by married or single parents. The findings of Rothman et al. (2007) reinforce this conclusion: cohabiting males were less concerned than biological fathers about the effects of the violence on children in the family.

Considering both the increasing prevalence of cohabiting family unions as well as the elevated rates of violence in these constellations, it is surprising how little research has explicitly focused on cohabitating structures. Most research in this area has classified married and cohabiting partners in the same rubric as “intimate partners” or “couples” (e.g., Heyman and Slep 2002; McDonald et al. 2006). As such, little is known about the link between partner violence and developmental outcomes specifically of children living in cohabiting families. Other lines of research suggest that children in cohabiting unions are at higher risk for a variety of unfavorable developmental outcomes (e.g., higher rates of delinquency, lower self-esteem, and inferior academic performance) than their peers from married families (e.g., Brown 2004; Manning and Brown 2006; Manning and Lamb 2003). IPV may play a role in these outcomes.

Forms and Prevalence of IPV

There is general agreement that exposure to violence by children can take various forms, ranging from overhearing the conflict to becoming an active participant in it (e.g., when trying to intervene) (Edleson 1999). O’Leary (1993) has argued that there is a continuum of aggression between partners, such that verbal aggression is followed by physical aggression and severe physical aggression. According to O’Leary’s conceptualization of violence, we may expect that verbal aggression exerts a less noticeable impact on family members, including affected children, than physical aggression. This conceptualization has been supported in a meta-analytic review (Kitzmann et al. 2003).

Research has struggled to accurately estimate how many children are exposed to IPV (McDonald et al. 2006). Such inconsistencies can be largely attributed to varying definitions of IPV (e.g., inclusion of verbal aggression versus only physical aggression), who the informant is (e.g., mother versus child report), and research methodology (e.g., past year versus lifetime) (Carlson 2000). Nevertheless, conservative calculations suggest that 15.5 million American children live in homes where IPV has occurred in the last year; 7 million of these children experienced severe partner violence (McDonald et al. 2006). Further, it is estimated that at least one in five children will witness IPV during their childhood or adolescence (Carlson 2000; Finkelhor et al. 2009). Such numbers highlight that exposure to IPV affects a considerable number of children.

Theoretical Perspectives on IPV and Child Mental Health

Both the family and the child’s appraisal of IPV appear important in the child’s psychosocial adjustment. Specifically, research indicates that the way in which a child interprets the adult conflict is predictive of his/her ensuing psychosocial adjustment (Fosco et al. 2007; Grych et al. 2000). As such, we consider both a family systems perspective (Cox and Paley 1997) and a cognitive-contextual framework, which posits that a child’s appraisal of the conflict is linked to the child’s (mal)adjustment (Fosco et al. 2007; Grych and Fincham 1990). Family systems theory suggests that children who witness IPV are likely to experience stress. Experimental research has long documented that overt parental conflict is a direct stressor for the child (e.g., Cummings et al. 1981). Cross sectional and prospective studies have also established that children exposed to parental conflict are more likely to develop both internalizing and externalizing problems than their non-exposed peers (for reviews see Buehler et al. 1997; Evans et al. 2008; Jouriles et al. 1991). In particular, stress may occur through “triangulation”—the process by which a child is drawn into parental conflict (Grych 2005). Triangulation can manifest itself in several different ways, all of which have the potential to adversely impact the child. For example, a “cross-generational coalition” describes the process whereby a child forms a close bond with one parent while distancing him or herself from the other caretaker (Grych 2005). The detrimental effects of marital disagreement on child adjustment are more pronounced when the child is triangulated into the parental conflict (Buehler et al. 1997).

From a family systems perspective, IPV appears not only to shape the child directly but also indirectly through family-level processes. Such effects have been described in the spill-over hypothesis—the claim that conflict in one family system (e.g., the marital relationship) can interfere with other parts of the family structure (e.g., Erel and Burman 1995; Repetti 1987). For example, a parent’s capacity to provide adequate care for their child(ren) appears to be compromised in situations of partner discord (e.g., Belsky 1984; Buchbinder 2004; Grych and Fincham 1990). Further, research suggests that the quality of the parent-child attachment is compromised in families that experience partner violence (Cleaver et al. 1999; Levendosky et al. 2003). Grych’s review (Grych 2002) supports the spill-over hypothesis by concluding that parents have less positive relationships with their children when also dealing with partner conflict. Taken together, IPV has considerable undesirable effects on the youth through the family system.

Turning to the child, theories from the stress and coping literature (e.g., Lazarus and Folkman 1984) have emphasized that an individual’s subjective appraisal of a given event is related to ensuing developmental outcomes (e.g., Compas et al. 1987; Compas et al. 1988). Focusing specifically on interparental conflict, Grych and Fincham (1990) and later Fosco et al. (2007) articulated the cognitive-contextual framework, which posits that a child’s appraisal of the conflict is linked to a child’s (mal)adjustment. More specifically, when children experience IPV, they try to “understand how it may affect them (perceived threat), why it is happening (attribution), and what they can do in response (coping efficiacy)” (Fosco et al. 2007, p. 7). Such appraisals of IPV are associated with children’s problem behaviors (e.g. Grych et al. 2000). Of particular relevance, youth self-blame for the IPV has received substantial attention (Kilpatrick and Williams 1998). Although it has been assumed that, as children age, they are less likely to blame themselves for the domestic conflict because of more accurate attributions for the violence (Carlson 2000), even older children and adolescents who experience self-blame have more problem behaviors (Cummings et al. 1994; DeBoard-Lucas et al. 2010; Jouriles et al. 2000). Furthermore, but not surprising, when considering the role of cognitive processes, there appears to be a stronger association between a child’s report of IPV and child (mal)adjustment than between IPV reports of other family members and child psychosocial development (e.g., Litrownik et al. 2003; Sternberg et al. 2006). The cited empirical work on a child’s cognitive appraisal of adult conflict highlights the importance of considering the child’s perspective in research on IPV.

Agreement among Informants

Research has shown that in married families different informants have low to moderate agreement of reported violence (see Hungerford et al. 2010). Partners appear to disagree as to whether violence occurred, what form of violence occurred (e.g., “threw something” versus “pushed”), and whether the child was exposed to the violence (e.g., Caetano et al. 2002; O’Brien et al. 1994; Schafer et al. 1998). To further complicate assessments of IPV, it appears that children’s reports diverge from those provided by adults; children generally indicate higher rates of IPV than do adults (Berger et al. 1988; Litrownik et al. 2003). It has been hypothesized that children are less reluctant than adults to disclose domestic victimization and violence (Kolko et al. 1996). As concluded by O’Brien et al. “even when both parents’ reports are obtained, parents’ combined reports of children’s exposure to marital physical aggression are not ideal proxies for children’s reports” (p. 58). As a consequence, scholars have recommended using child and adolescent reports when assessing IPV (e.g., Grych et al. 1992).

The Current Study

The current study addressed the foregoing gaps in the literature by examining a sample of youth from low-income, urban Black cohabiting families. In all families, the adolescent was living with her or his biological mother as well as a biologically unrelated male cohabiting partner (MCP). We examined two specific research questions. First, do adolescents report more adult-to-adult verbal and physical violence than mothers and MCPs? Second, is adolescent report of verbal and physical violence related to their report of psychosocial functioning in three domains: (1) Attributions for mother-MCP violence (i.e., self-blame); (2) internalizing and externalizing problem behaviors; and (3) their relationships within the family (i.e., mother and MCP)? We offer the following hypotheses: (1) A larger percentage of adolescents will report violence than either mother or MCP; and (2) Adolescents will report more difficulties in all three domains when reporting IPV, especially physical violence. Of note, as mothers, but not MCPs, in our sample are biologically related to the adolescent, a “cross-generational coalition” may form between mother and adolescents in the face of IPV (Grych 2005). Thus, the youth’s relationship with the MCP, but not the mother, may be associated with adult partner violence.

Although some (see Evans et al. 2008; Kitzmann et al. 2003), but not all (see Wolfe et al. 2003), reviews suggest that child age does not qualify the association of partner violence and developmental outcome, scholars have noted that “domestic violence may manifest differently in children of different developmental stages” (Evans et al., p. 133). As a consequence we limited our sample to one age group—adolescence—for two reasons. First, adolescents are more likely than younger children to have been exposed to IPV (Finkelhor et al. 2009). Second, youth at this developmental stage are ideally suited for studies focusing on family conflict, as they are more likely than younger children to have the cognitive abilities to self-report both IPV and their own psychosocial adjustment. As such, we viewed our adolescent self-report measures of IPV and psychosocial functioning as valid indexes of these variables. However, it should be noted that, because of the advanced age of the participating youth (i.e., adolescents), self-blame in the face of IPV might be less pronounced than it is with younger children.

Method

Participants

Participants in the original sample were 121 Black single-mother families with a male cohabiting partner and a child in the 10 to 17 year age range living in New York City (Forehand et al. 2014). Ninety-two MCPs agreed to participate. As this study focused on IPV reports from all three members of the triad, the sample consisted of these 92 families.

The mean ages of participating youth, mothers, and MCPs were 13.25 years (SD = 2.04; 58.7 % girls), 39.12 years (S.D. = 7.90), and 40.66 years (S.D. = 10.74), respectively. Of the mothers, 34.8 %, 35.9 %, and 29.3 % did not complete high school, completed high school/GED, and had some college/vocational school after high school, respectively. Of the MCPs, 28.6 %, 57.1 %, and 14.3 % did not complete high school, completed high school/GED, and had some college/vocational school after high school, respectively. Household incomes averaged $26,519 per year (SD = $19,318; range = $720–$96,180; median = $21,960). All families had at least one adult who self-identified as Black. The mother-MCP relationship was roughly split between “established” (greater than 12 months) (57.3 %) and “new” (12 months or less) (42.8 %).

Procedure

The National Development and Research Institute (NDRI) Institutional Review Board reviewed and approved the study. All participants initially signed consent (mother and MCP) and assent (adolescent) forms. Field staff members who had experience working with low-income Black residents of New York City recruited study participants. Field staff members used existing networks of research study participants, field informants, street recruiting, and social services agency contacts to recruit potential participants. Families completed the assessment either at a community site or in their home, according to the preferences of each family. Family members completed the assessments separately and privately with interviewers, who entered the responses into a computer database. Mothers, MCPs, and adolescents completed measures assessing a range of variables related to personal and family psychosocial functioning, including the variables of focus in the current study. Each interview took approximately 60 min to complete; adults were compensated $40, and adolescents were compensated $20, for their participation.

Control Variables

Key demographic variables that may influence family violence (e.g., mother and MCP education, adolescent age and gender) were assessed.

Demographic Information

Mothers and MCPs responded to demographic questions about themselves (e.g., race, education), their families (e.g., family income), and the length of their relationship (established = greater than 12 months or new =12 months or less). Youth reported on their gender and age.

Study Variables

Intimate Partner Violence

Mother report and MCP report of IPV was determined by each of their responses on two items from the National Survey of Families and Households (Brown 2000, 2003) regarding how they dealt with serious disagreements: (1) Argue heatedly or shout at each other; and (2) end up hitting each other or throwing things at each other. Each item was scored on a 5-point scale: 0 = never; 1 = seldom; 2 = sometimes; 3 = very often; and 4 = always.

Adolescent report of IPV was determined by their responses to two items from the Children’s Perceptions of Interparental Conflict scale (Grych et al. 1992): (1) When my mother and her partner have an argument, they yell a lot; and (2) my mother and her partner have pushed or shoved each other during an argument. Each item was scored on a 3-point scale: 0 = true; 1 = sort of true; and 2 = false.

Self-Blame

The self-blame subscale (the degree to which adolescents blame themselves for their parents’ conflict) of the Children’s Perception of Interparental Conflict Scale (CPIC; Grych et al. 1992) was used to assess youth’s self-attributions for perceived adult conflict. Adequate reliability and validity have been reported in adolescents (Bickham and Fiese 1997). Internal consistency for this sample was .84.

Adolescent Internalizing and Externalizing Problems

The Youth Self-Report (YSR; adolescent-report) assessed child internalizing and externalizing problems (Achenbach and Rescorla 2001). Reliability and validity of the YSR are well-established for the internalizing and externalizing dimensions (Achenbach and Rescorla 2001). The alpha coefficient for the YSR internalizing and externalizing problems was .82 and .87, respectively, for the current sample.

Adolescent-Mother and Adolescent-MCP Relationships

The Acceptance vs. Rejection subscale of the 30-item version of the Children’s Report of Parental Behavior Inventory (CRPBI-30; Schuldermann and Schuldermann 1988) was used to assess the youth’s perspective of the mother-adolescent and MCP-adolescent relationships. Adolescents indicated whether the mother is “like,” “somewhat like,” or “not like” such statements as “Enjoys talking things over with me.” Higher scores indicate a more positive mother-adolescent relationship (i.e., more acceptance and closeness). The youth also completed the same measure on the MCP. The Acceptance/Rejection subscale has been shown to be internally consistent, and convergent and discriminant validity has been well demonstrated in prior research (see McKee et al. 2013, for a review). Internal consistency for adolescent report on the mother (α = .72) and MCP (α = .74) was adequate for the current sample.

Data Analytic Plan

Preliminary Analysis of Control Variables

The relationship between categorical demographic variables (e.g., parent education) and continuous demographic variables (e.g., youth age) with primary outcome measures (i.e., CPIC, YSR, and CRPBI) was examined by analysis of variance or correlations, respectively. If significant associations emerged, these demographic variables were included in the analyses as covariates.

Primary Analyses

First, we examined the report of mother, MCP, and youth for the report of verbal and physical violence by grouping each reporter into one of three groups: Non-occurrence of IPV, verbal IPV, and physical IPV. We report percentages for each informant and conduct selected chi-square tests. Second, we examined the agreement within family of the report of IPV for mother-adolescent and MCP-adolescent dyads. Third, using adolescent report for the three groups (non-occurrence of violence, verbal violence, and physical violence), we conducted one-way analysis of variance (or covariance if covariates were included) examining the adolescent’s report in three domains of psychosocial functioning: (1) Attributions (i.e., self-blame) for IPV; (2) problem behaviors (internalizing and externalizing); and (3) family relations (adolescent-mother and adolescent-MCP relationship).

Post-Hoc Analyses

We re-classified mother and MCP report of verbal violence to determine if this would change our conclusions.

Results

Preliminary Analyses

Group Formation

Three groups were formed based on the report of IPV by each informant (mother, MCP, & adolescent). For both the mother and MCP, the groups were as follows: Minimum Family Violence (score of 0 or 1 on “argue heatedly or shout at each other” and score of 0 on “end up hitting each other or throwing things at each other”); Verbal Violence only (2, 3, or 4 on “argue heatedly or shout at each other” and score of 0 on “end up hitting each other or throwing things at each other”); and Verbal + Physical Violence (score of 1, 2, 3, or 4 on “end up hitting each other or throwing things at each other”). The cut point of 1 (seldom occurrence of IPV) for the Physical Violence group identifies those families in which any physical IPV occurred. There were no families that had Physical Violence only.

For the adolescent, the groups were as follows: Minimum Family Violence (scores of 2 on “when my mother and her partner have an argument, they yell a lot” and on “my mother and her partner have pushed or shoved each other during an argument”); Verbal Violence only (score of 0 or 1 on “when my mother and her partner have an argument, they yell a lot”, and score of 2 on “my mother and her partner have pushed or shoved each other during an argument”); and Verbal + Physical Violence (score of 0 or 1 on “my mother and her partner have pushed or shoved each other during an argument.”). The cut point of 1 (sort of true) for the Physical Violence group identifies those families in which any physical IPV occurred.

These criteria for group inclusion resulted in 51, 55, and 33 families included in the Minimum Violence group; 27, 30, and 39 included in the Verbal Violence group; and 14, 7, and 20 included in the Verbal + Physical Violence group based on mother, MCP, and adolescent report, respectively. Of interest, in the Verbal + Physical Violence group, 50 % and 71 % of mothers and MCPs, respectively, reported that IPV occurred “seldomly” and none reported that it occurred “always.” Further, more than half (55 %) of adolescents in the Verbal + Physical Violence group reported that the occurrence of IPV was “sort of true” and 45 % reported it was “true.”

Preliminary Analysis of Control Variables

None of the key outcome variables for the second research question (i.e., adolescent self-blame, internalizing and externalizing problems, adolescent-mother relationship, and adolescent-MCP relationship) differed by youth gender, youth age, mother education, or MCP education. Adolescent-MCP relationship quality, but no other outcome variable, differed by the length of the mother-MCP relationship (F (1,80) = 8.67, p < .01) such that adolescents reported higher relationship quality with MCPs who had been in a relationship with their mother for over a year. Thus, when adolescent-MCP relationship quality was the outcome of interest, mother-MCP relationship length was entered as a covariate.

Primary Analyses

The first research question examined the percent of families in the Minimum Violence Group based on mother, MCP, and adolescent report. These data are reported in Table 1. 56 % of mothers and 60 % of MCPs were in this group based on their self-report whereas 36 % of adolescents were in this group based on their self-report (x2 (1) > 7.10, p < .01, for mother vs. adolescent report and for MCP vs. adolescent report). The percentage of reporters in the Verbal Violence only group did not differ among mothers, MCP’s, and adolescents. A larger percentage of adolescents than MCPs reported being in the Verbal + Physical Violence group (22 % vs. 8 %; x2 (1) = 7.31, p < .01). Adolescent report did not differ from mother report.

Table 1 Intimate partner violence based on mother, MCP, and adolescent report
Full size table

We further examined the agreement for IPV between adolescent and mother and between adolescent and MCP by calculating the concordance across these two dyads within each family. Specifically, when a mother reported minimum partner violence, verbal violence only, or verbal plus physical violence, did the adolescent in the family agree with her? Similarly, when an MCP reported minimum partner violence only, verbal violence only, or verbal plus physical violence, did the adolescent in the family agree with him? The percent agreement for each dyad was as follows: mother-adolescent, 48 %; and MCP-adolescent, 48 %.

The second research question examined adolescent self-report of self-blame attributions for adult-to-adult violence, internalizing and externalizing problems, and relationships with adults in the home (i.e., mother and MCP) based on her or his report of mother-MCP violence. As shown in Table 2, adolescents in the Verbal + Physical Violence group reported more self-blame, more internalizing and externalizing problems, and a poorer relationship with the MCP, but not the mother, than those in the Minimum Violence group and in the Verbal Violence only group.

Table 2 Self-reported outcomes M and (SD) based on adolescent’s perception of mother-MCP violence group
Full size table

Post-Hoc Analyses

In the primary analyses, we classified mothers and MCPs who had a response of 1 (“seldom”) on the Verbal Violence item into the Minimum Violence Group. We re-classified these participants into the Verbal Violence group. As expected, this resulted in a smaller percentage of mothers (23 %) and MCPs (26 %) in the Minimum Violence group but did not change the mother-adolescent (48 %) and MCP-adolescent (48 %) agreement on violence occurrence or the findings regarding the Verbal + Physical Violence group.

Discussion

In this article, we examined data from 92 Black single-mother families with a male cohabiting partner and adolescent to (1) determine if adolescents report more IPV than mothers and MCPs and (2) investigate association(s) between youth exposure to IPV and youth psychosocial outcomes. An overarching goal of this research was to expand what is known about cohabiting family structures by examining individual and interpersonal experiences of youth living in such constellations. Prior work had emphasized that cohabiting family structures are becoming increasingly prevalent, but that little is currently known about IPV and its impact on youth living in cohabiting families.

Results suggested that the occurrence of intimate partner verbal and physical violence in our cohabiting sample (i.e., 15 % verbal + physical IPV, according to mother report) was below that reported by Black women in a national sample (41 %; Black et al. 2011), by cohabiting adults (35 %; Stets and Straus 1989), and by children (33 %; Litrownik et al. 2003). Furthermore, both mothers and MCPs reported that physical violence “seldomly” occurred and the majority of adolescents conveyed that their report of intimate partner physical violence was “sort of true.” Data also indicated that a higher percentage of youth in our sample reported violence than either adult and, in particular, a higher percentage of youth than MCP reported physical violence. When concordance of IPV reports was examined within families, there was slightly less than 50 % agreement between adolescents and both adult figures—the biological mother, as well as the MCP.

Using adolescents’ report of IPV, our analysis revealed that youth who were exposed to verbal and physical violence showed greater signs of maladjustment compared to non-exposed individuals. Significant group differences were apparent on measures of cognitive attitudes (i.e., greater levels of self-blame) and psychological well-being (i.e., greater internalizing and externalizing symptomatology). In addition to these individual factors, we also found that exposure to verbal and physical IPV was associated with less favorable outcomes at the family-level. When compared to the group of minimally-exposed and verbally violence-exposed youth, those who witnessed verbal and physical IPV reported less favorable relations with the MCP. Interestingly, those exposed to verbal violence did not differ from adolescents in the minimally-exposed group. The ensuing paragraphs review these findings within the context of previous empirical work and extant theoretical models, highlighting both points of convergence and divergence.

The lower occurrence of physical IPV in our cohabiting families than previous reports from national samples of married or cohabiting couples may have emerged for several reasons. First, the majority of our sample had been cohabiting for longer than a year. Couples in which verbal and physical violence occurred may have transitioned out of cohabitation. Second, single mothers who re-partner after having children “partner with men who are good providers” (Bzostek et al. 2012, p. 829). Our prior research with this sample indicates that MCPs are involved in daily childrearing of the adolescents (Forehand et al. 2014). Thus, the single mothers in our study may well have selected men who were not physically violent but rather actively involved in the family as contributing partners. Third, a mother with an adolescent-aged child is relatively old (mean age of current sample = 39 years), suggesting a longer history of selecting partners. This experience may lead to avoiding physically-violent men. Finally, methodological differences in our study and prior research (e.g., time frame sampled, recruitment procedure) could account for the noted differences. Regardless of the reason, the low occurrence of physical IPV in our sample is particularly impressive considering that prior research suggests a high incidence of physical IPV in Black, urban, economically-challenged, cohabiting families (e.g., Dunlap et al. 2010; Stets and Straus 1989).

The low concordance between youth and adult reports of physical and verbal IPV directly aligns with previous research. Studies have generally found that children’s reports of IPV diverge from those provided by adults; children tend to indicate higher conflict rates (Berger et al. 1988; Litrownik et al. 2003). There are several possible explanations for such discrepancies. Jaffe et al. (1990) found that parents believed that their children were not aware of the parental conflict, either because children were engaged in another activity, or because they were not in the home at the time of the conflict. Scholars have also argued that children may be less reluctant than adults to disclose IPV (Kolko et al. 1996). Finally, adult and adolescent assessments of IPV were obtained through items with different response scales—a factor that could help account for the low inter-rater agreement. Regardless of what causes discrepant youth-adult reports of IPV, our findings (along with previous research) emphasize the critical importance of consulting affected children and youth when assessing incidence of IPV.

The significant relationship between exposure to verbal and physical IPV and youth’s internalizing and externalizing symptomatology directly aligns with the extensive body of literature documenting pernicious effects of exposure to IPV for affected children (e.g., Evans et al. 2008; Kitzmann et al. 2003). Our findings, therefore, support the assertion that exposure to adult intimate partner verbal and physical violence places youth at risk for psychological maladjustment. Whereas the cited literature has predominantly focused on traditional family systems—that is, a child who lives with her or his biological parents—our research found comparable effects in cohabiting unions. As such, our results, in combination with the existing literature, emphasize that exposure to IPV is associated with a variety of unfavorable outcomes for affected youth, regardless of whether one of the adult figures is or is not biologically related to the child.

Moving from the individual-level to the family-level, our results also suggest that IPV disrupts the broader family system. Youth who witnessed verbal and physical IPV reported less favorable relations with the MCP than adolescents exposed to either verbal violence or only minimum violence. These findings align with family-systems theories, such as the spill-over hypothesis, which describes how conflict in one family system (e.g., marital relationship) can compromise other domains of family functioning. (e.g., Erel and Burman 1995; Repetti 1987). The strained youth-MCP relationship could also reflect a “cross-generational coalition”—the process whereby a child forms a close bond with one parent while distancing him or herself from the other caretaker (Grych 2005). The fact that violence-exposed youth in our sample reported less favorable relations with the MCP but not their biological mother is not surprising and suggests that a “cross-generational coalition” may exist in violence-prone cohabiting family structures.

Although data from the present study were collected cross-sectionally, which does not allow for mediation analyses (Maxwell and Cole 2007), our results provide preliminary insight into potential underlying mechanisms. As was reviewed above, scholars have contended that a youth’s cognitive appraisal of the adult conflict mediates the link(s) between IPV exposure and maladjustment (Grych and Fincham 1990). For example, Fosco et al. (2007) reviewed both theoretical and empirical evidence on how a child’s appraisal of the adult conflict could mediate the link between exposure to IPV and ensuing child well-being. In our sample, youth exposed to physical (but not verbal) violence reported higher levels of self-blame. It is therefore conceivable that these cognitions mediated the significant association between IPV exposure and psychological symptomatology. These results could be of relevance to applied work; clinical interventions might target a youth’s cognitions, particularly self-blame, in an effort to support psychosocial adjustment. It is important to recognize that additional processes—such as those related to emotion-regulation and social learning—can account for the link between IPV exposure and psychological maladjustment. However, the data collected in the present study cannot provide information on such potential alternative mechanisms.

It is noteworthy that significant group differences were only evident for adolescents exposed to verbal physical violence. Stated differently, youth who reported “only” verbal IPV did not differ significantly from the minimal-violence exposed group on any of the examined measures. These results indicate a differential effect of verbal and verbal + physical IPV, such that the latter appears more harmful than the former. Our findings are congruent with O’Leary’s (1993) conceptualization of violence; this scholar has posited that there is a continuum of aggression between partners, such that verbal aggression is followed by physical aggression and severe physical aggression, and progression on the continuum is associated with more disruption in psychosocial adjustment.

The present study included several limitations, which should be noted. First, data were collected at one time point, thereby limiting the degree to which causal inferences can be made. Although other strands of empirical work (e.g., Edleson 1999; Katz et al. 2007), as well as theoretical models (e.g., Davies and Cummings 1994; Gottman and Katz 1989; Grych and Fincham 1990), suggest that exposure to IPV precedes and predicts ensuing child maladjustment, our data cannot directly address causal claims. Similarly, our data collection methods did not allow for conclusive mediation analyses. Large-scale, prospective research designs with cohabiting families could address these limitations by examining causal relations and elucidating underlying mechanisms between IPV exposure and youth psychosocial functioning. Gaining a more comprehensive understanding of underlying processes could be particularly valuable from an intervention standpoint. Such knowledge could directly inform applied clinical work, by allowing mental health providers to target mediating variables (e.g., a youth’s self-blame).

A second limitation pertains to our measurement of verbal and physical violence; these two study variables were assessed with one item each. Although each of these items measured the construct of interest, a more thorough assessment by multiple items may have yielded different results. For example, our assessment of both types of violence did not measure: a) intent to harm your partner, or b) attempts to exert control over your partner, although these features are common components of IPV (Centers for Disease Control and Prevention, CDC 2010). Our assessment also did not assess the frequency of occurrence for each type of IPV to capture the range of physical and verbal violence. Our reliance on a narrow assessment of violence might provide a credible explanation for the lower occurrence of physical violence in our sample than in prior reports (e.g., Black et al. 2011). Future work can address the noted measurement shortcomings by using a more comprehensive evaluation of verbal and physical violence (i.e., use of several items; assessment of frequency of violence). Finally, the response scale for adult (mother and MCP) and adolescent assessment of violence differed (5-point versus 3-point scales). This difference makes comparable classifications of violence by adults and the adolescent difficult. Future research should use the same response scale across reporters.

Third, our analyses were also limited by our sample size. Although some previous research has identified potential moderator variables, such as a child’s gender (see Edleson 1999, for review), age of first exposure (e.g., Graham-Bermann and Perkins 2010), and developmental stage (e.g., Wolfe et al. 2003), we were not able to scrutinize such factors. Future work with larger samples may benefit from examining whether gender, age, and duration of exposure moderate the links between IPV and (mal)adjustment. Future work could also distinguish between sources of violence—that is, whether the IPV was perpetrated by the mother, the MCP, or bidirectionally (i.e., by both adults). Research has indicated that maternal- and paternal-perpetrated conflict have similar but distinct relations with child outcomes, suggesting that not all IPV is equal (Harding et al. 2013). Our sample size did not allow for such a nuanced analysis of IPV. A final limitation relates to our sample characteristics. Our analyses were based on a sample of low-income, urban, Black individuals, thereby limiting generalizability of findings. Future investigations could recruit participants from a more diverse and nationally representative group to address these concerns of external validity.

Despite these limitations, our study made an important contribution to the field by examining cohabiting unions—an increasingly common family constellation that appears to be particularly prone to IPV. Our findings revealed considerable discrepancies in adult and youth report of IPV. These results have important implications for applied work; they suggest that assessment of IPV may benefit from considering both adults and youth reports. In fact, only considering adult report may result in an underestimate of IPV. Our findings also suggested that intimate partner verbal and physical violence does not occur in the great majority of these families; however, youth who experience verbal and physical partner violence between their biological mother and her MCP report impairment in various domains of functioning. These findings highlight that, although adult-to-adult physical violence is not common, it is a risk factor for adolescents exposed to it. As such, prevention and intervention efforts should attempt to reduce or eliminate domestic IPV, particularly the incidence of physical aggression. One promising approach could include educating the public about the adverse effects of IPV in cohabiting families. Such efforts are critical in ensuring the psychosocial adjustment of children and youth.