Lessons Learned from Witnessing Constructive Interparental Conflict and the Beneficial Implications for Children

Abstract

This study examined children’s emotion situation knowledge and social problem-solving abilities as mediators in the prospective link between constructive interparental conflict (IPC) and children’s psychological adjustment in a sample of 238 preschool children (M_age = 4.38 years; 52% female children; 28% Black; 14% multi-racial or other race; 16% Latinx) and their mothers. The methodological approach consisted of multiple methods and informants in a longitudinal design with three annual measurement occasions (i.e., preschool, kindergarten, first grade). Results of the path analysis indicated that Wave 1 constructive IPC predicted residualized increases in children’s emotion situation knowledge and social problem-solving at Wave 2 with inclusion of hostile IPC, demographic factors, and prior child characteristics (i.e., child age, race, internalizing and externalizing symptoms, social competence) as simultaneous predictors. In turn, children’s emotion situation knowledge at Wave 2 predicted residualized decreases in their internalizing and externalizing symptoms at Wave 3. Wave 2 social problem-solving also predicted residualized increases in their social competence at Wave 3. Results are discussed in the context of how they advance developmental models of constructive family conflict.

Although witnessing hostile interparental conflict (IPC) manifested in intense, prolonged, and escalating bouts of anger is associated with children’s distress and mental health problems, not all ways of handling disagreements are harmful to children (Cummings & Davies, 2010). Research has indicated that constructive IPC characterized by problem-solving, cooperation, validation, and progress in resolving disputes may be benign or even beneficial for children (Harold & Sellers, 2018; Jouriles et al., 2016). For example, meta-analytic findings indicated that the negative association between constructive IPC and children’s internalizing and externalizing symptoms is small in magnitude and highly variable across studies (van Eldik et al., 2020). In supporting the conclusion that constructive IPC exposure is benign, studies have reported that greater exposure to constructive IPC is unrelated to children’s vulnerability to distress and psychopathology (e.g., Cheung, 2021; McCoy et al., 2013; Zemp et al., 2020). Still other empirical findings suggest that constructive IPC has salubrious developmental consequences in predicting children’s social competence and lower risk for internalizing and externalizing symptoms (Davies et al., 2016; Frosch & Manglesdorf, 2001; Miller-Graff et al., 2016).

Differences in findings on whether constructive IPC is beneficial or harmless for children’s psychological well-being may reflect the possibility that cooperative problem-solving between parents indirectly increases children’s psychological adjustment by altering how children process and respond to emotion-laden events. Studies designed to identify child mediators of constructive IPC have largely focused on the role of children’s positive and diminished negative emotional reactivity to conflict as possible explanatory mechanisms underpinning IPC (Cummings & Davies, 2010). Using a variety of methodological (e.g., daily diary, analogue simulations of adult conflict, prospective field designs) approaches, Cummings and colleagues have shown that constructive couple conflict is associated with lower levels of insecurity (e.g., negative emotional reactivity) and greater positive affect in response to conflicts (Cummings et al., 2002; 2003; 2004; Goeke-Morey et al., 2003). However, mediational tests directly examining whether children’s emotional responses (e.g., lower insecurity) to conflict mediate prospective pathways among their exposure to constructive IPC and their psychological adjustment have yielded inconsistent findings. For example, some studies have shown that children’s diminished negative (e.g., insecure) emotional responses to IPC account for relations between constructive IPC and children’s better psychological functioning when constructive IPC examined in models without the inclusion of destructive IPC as a predictor (Cheung, 2021; McCoy et al., 2009). In contrast, results from the five studies that have directly examined constructive and destructive IPC as simultaneous predictors have shown that constructive IPC is not a predictor of child emotional responses to IPC when the potent risk of destructive conflict is included in the analyses (Davies et al., 2012; 2016; Du Rocher Schudlich & Cummings, 2007).

Against this backdrop, other child mechanisms likely underpin the psychological adjustment of children’s exposure to constructive IPC. Supporting this premise, conceptual models of family conflict have posited that witnessing constructive IPC may promote children’s mental health over time by teaching them important socio-cognitive lessons about emotions and social relationships (Davies & Cummings, 1994; Grych & Fincham, 1990; Zemp et al., 2016). However, over three decades later, it remains that “little is known about what specific lessons children learn when exposed to resolved conflict and how these lessons, in turn, may mediate psychological adjustment” (Davies & Cummings, 1994, p. 405). Thus, the goal of this study is to provide the first longitudinal test of whether constructive IPC may be related to children’s social (i.e., social competence), emotional (i.e., lower internalizing symptoms), and behavioral (i.e., lower externalizing symptoms) adjustment by predicting their greater social understanding in domains of: (1) emotion situation knowledge, defined by their capacity to anticipate and identify emotions based on contextual cues in emotionally-evocative social situations (Fine et al., 2006); and (2) social problem-solving, consisting of their ability to generate constructive solutions to challenging social events (Youngstrom et al., 2000; Su et al., 2020).

Conceptual Bases for Mediational Role of Emotion Knowledge and Social Problem-Solving

Our hypothesis that constructive IPC is beneficial for children’s mental health because it, in part, promotes their social-cognitive understanding of emotions and social problems is rooted in the tripartite model of family emotions (Morris et al., 2007) and the interparental positivity spillover theory (IPST; Don et al., in press). According to both models, constructive interparental interactions may increase children’s social-cognitive understanding through vicarious learning and emotion contagion processes. As part of these learning processes, witnessing a wide array of constructive conflict behaviors (e.g., support, validation, active listening skills) and problem-solving tactics (e.g., non-defensive assertion of perspectives, identification of problems, efforts to understand the partner’s feelings, generation of effective solutions) may stimulate children’s acquisition of knowledge of emotions and problem-solving strategies in interpersonal contexts (Denham & Kochanoff, 2002). Given that children are more likely to model interactions when they produce positive outcomes (Don et al., in press), they may exhibit biases in emulating the effective nature of constructive IPC behaviors and, in the process, internalize scripts of emotions and solutions. Through emotion contagion processes consisting of reflexive tendencies to experience the emotions of others (e.g., Jouriles et al., 2016; Morris et al. 2007), exposure to well-regulated adult emotions (e.g., sympathy, positive regard, contentment) in constructive conflicts may also create optimal learning conditions by eliciting comparable levels of child emotional arousal, positivity, and engagement with the environment (Davies et al., 2016). Observing well-regulated parental emotions (see Cummings et al., 2002; 2003), in turn, may expand children’s attention, processing, and consolidation of social information in ways that further stimulate their acquisition of knowledge on emotions and problem-solving strategies in interpersonal contexts (e.g., Don et al., in press; Fredrickson et al., 2008; Stifter et al., 2020).

In highlighting the viability of the second links in the proposed mediational chain, social information processing models also provide conceptual bases for expecting that children’s emotion situation knowledge and ability to generate constructive solutions to social challenges will predict their better social, emotional, and behavioral functioning. According to these models (e.g., Crick & Dodge, 1994; Lemerise & Arsenio, 2000), children use emotion knowledge as a guide for accurately encoding social and affective cues and formulating attributions for the causes and intentions of people in interpersonal settings. From this perspective, accurately identifying the emotions of others based on contextual cues may promote social competence and protect against the development of behavior problems by organizing more nuanced, empathetic, and socially oriented responses to social interactions (Denham et al., 2012; Thompson et al., 2020). In addition, the other-oriented perspective-taking aspects of emotion situation knowledge may decrease children’s vulnerability to internalizing problems by preventing self-focused rumination (e.g., Ştefan & Avram, 2021). As another significant aspect of social information processing, children’s capacity to formulate constructive solutions to social problems may also enhance their social, emotional, and behavioral functioning by increasing their flexibility in enacting coping behaviors that match the interpersonal demands (Lopes et al., 2012).

Research on Emotion Situation Knowledge and Social Problem-Solving as Mechanisms

Although research has yet to examine children’s emotion situation knowledge and social problem-solving abilities as explanatory processes in the prospective association between constructive IPC and children’s psychological adjustment, there is preliminary, piecewise support for viability of each link in the proposed mediational chain. Supporting the first part of the developmental cascade, findings from a cross-sectional study indicated that father reports of conflict resolution was related to higher levels of children’s interpersonal problem solving (Goodman et al., 1999). In the same investigation, maternal report of reasoning during conflicts was associated with better child problem-solving ability but only when IPC frequency was low. Although no studies have examined constructive IPC as a predictor of children’s emotion situation knowledge, Harrist and Ainslie (1998) found that an interparental discord composite consisting of low consensus and greater conflict intensity as concurrently associated with children’s difficulties with emotion understanding. Despite indirect support for links between constructive IPC and the two dimensions of children’s social-cognitive understanding, no studies to our knowledge have examined whether constructive IPC is a distinctive predictor of: (1) children’s emotion situation knowledge and social problem-solving abilities over time within a longitudinal design or (2) over and above their variation in exposure to hostile IPC.

Research on the psychological sequelae of children’s emotion situation knowledge and social problem-solving provide consistent support for their roles as predictors of better social competence and lower psychopathology. For example, studies have shown that children’s emotion situation knowledge is a precursor of lower risk for internalizing and externalizing symptoms and better social competence (e.g., Shields et al., 2001; Thompson et al., 2020; Yang & Wang, 2022). Likewise, research findings have indicated that children’s constructive solutions to social situations prospectively predicted lower levels of externalizing (e.g., Youngstrom et al., 2000) and internalizing (e.g., Quamma & Greenberg, 1994) symptoms and greater social competence (Youngstrom et al., 2000). Moreover, research supports that emotion knowledge and social-problem-solving are distinct in their composition and roles as predictors of child adjustment (e.g., Dodge et al. 2002; Nix et al., 2013). However, the research findings have yet consistently identified that the two dimensions of social-cognitive knowledge are broadly related to children’s psychological adjustment across emotional, social, and behavioral domains, particularly within the limited studies that employed longitudinal designs or analytic models examining emotion knowledge and social problem-solving as simultaneous predictors (Cunningham et al., 2009; Nix et al., 2013; Sette et al., 2017; Trentacosta & Fine, 2010). Therefore, there are insufficient bases for formulating hypotheses on whether social problem-solving and emotion situation knowledge selectively or broadly mediate children’s psychological adjustment across emotional, social, and behavioral domains.

The Transition to the Early School Years as a Developmentally Salient Period

We examined the mediational pathways involving constructive IPC, children’s emotion situation knowledge and social problem-solving, and their psychological adjustment during the transition from preschool to the early elementary school years. Developmental models have designated this time frame as a sensitive period for children’s adaptation to IPC due to their increasing sensitivity to family relationships outside the parent-child dyads and their reliance on primary caregivers as models for interpreting and responding to novel socialization contexts (e.g., Cummings & Davies, 2010; Davies et al., 2016; Jouriles et al., 2016). The early school years are also key period for examining the family origins and mental health sequelae of the two proposed mediators. For example, the transition to formal schooling is marked by considerable plasticity in children’s emotion understanding and social problem-solving (Heinze et al., 2015; Sette et al., 2017). In this same period, studies have identified these two dimensions of social understanding as antecedents of children’s psychological adjustment (e.g., Thompson et al., 2020; Youngstrom et al., 2000). Thus, there are strong developmental bases for examining our process model of constructive IPC as children progress into the early elementary school years.

The Present Study

Although theoretical models provide firm grounds for proposing that children’s emotion situation knowledge and social problem-solving underpin the beneficial sequelae of constructive IPC (Don et al., in press; Morris et al., 2007), studies have not tested these mediational hypotheses. The small corpus of studies most relevant to elucidating emotion situation knowledge and social problem-solving as mediators of constructive IPC have utilized broad composites of interparental discord within cross-sectional analyses (Harrist & Ainslie, 1998; Marcus et al., 2001). More specifically, Harrist and Ainslie (1998) documented interparental discord assessed by low consensus on relationship issues and high perceived impact of arguments was concurrently associated with lower child emotion situation knowledge and, in turn, greater child social withdrawal. Findings from another cross-sectional study indicated that a latent construct of aggressive social cognitions (e.g., aggressive solutions to social problems) mediated the association between a broad composite of interparental hostility (largely consisting of aggressive conflict measures) and children’s aggressive behavior (Marcus et al., 2001). Because these studies do not distinguish between constructive and destructive IPC as predictors, it is unclear whether the findings are attributable to variation in children’s exposure to constructive IPC, hostile IPC, or both. Moreover, given that cross-sectional analyses produce highly biased mediation findings (Maxwell & Cole, 2007), the existing research cannot generate firm conclusions about the roles of children’s emotion situation knowledge and social problem solving as mechanisms accounting for the sequelae of constructive IPC.

Accordingly, this study tested children’s social problem-solving and emotion situation knowledge as mediators in the link between constructive IPC and their internalizing symptoms, externalizing difficulties, and social competence. To increase methodological rigor, we used a multiple methods (i.e., semi-structured interviews, structured interviews, surveys) and informants (i.e., mother, coder, teacher, child) design within a longitudinal study with three annual measurement occasions. Following quantitative recommendations for testing mediation pathways (Maxwell & Cole, 2007), our analyses specified children’s exposure to constructive IPC as a predictor of residualized increases in their emotion situation knowledge and social problem-solving abilities over a one-year period. In the same analysis, we examined the two dimensions of children’s socio-emotional understanding as predictors of teacher reports of their internalizing and externalizing symptoms and social competence one year later after controlling for initial levels of their psychological problems. Analyses also included hostile IPC, household income, parent education, and child race and age as covariates to determine if constructive IPC was a distinctive predictor of the proposed mediational cascade in the context of possible confounding factors. Guided by theory (Don et al., in press), we hypothesized that emotion situation knowledge and social problem-solving would mediate links between constructive IPC and children’s psychological adjustment. Given the lack of research on the two dimensions of social-cognitive understanding as distinctive precursors of children’s psychological functioning, we did not formulate any hypotheses on their relative roles as predictors of children’s internalizing symptoms, externalizing problems, of social competence. Although we developed our hypotheses a priori, the predictions and design of this study were not preregistered.

Method

Participants

Participants included 238 mothers and preschool children from a moderate-sized metropolitan area and surrounding county in the Northeastern United States. We recruited through multiple agencies including pre-K programs, Head Start agencies, and public and private childcare providers. The average age of children at Wave 1 was 4.38 years (SD = .32), with 52% of the sample consisting of female children. The median household income of the families was $72,000 per year (range = $500 – $260,000), with 32% of families receiving public assistance. Parent educational attainment was as follows: less than a high school diploma or equivalent (5%), high school diploma or equivalent (18%); some college (23%), bachelor’s degree (28%), and graduate training (26%). Children were primarily White (68%; 61% non-Latinx White), followed by children who were Black or African American (18%), multi-racial (10%), or another race (4%). Approximately 16% of children were Latinx. At Wave 1, 100% of the mothers were biological parents. Mothers, partners, and children lived together in the same household for at least 10 months and all had regular contact as a triad over the past year (i.e., at least 2 to 3 times per week across the prior year). Most mothers were married (73%), with the remaining couples in long-term committed relationships (20%), engaged to be married (5%), or in dating relationships (2%). Retention across the three annual measurement occasions was 88%. Data collection took place between 2018 and 2022.

Procedures and Measures

Parents and children participated in visits to a research center laboratory at three waves of data collection. Institutional Review Board at the University of Rochester approved the research procedures under the title “Family Relationships and Children’s Functioning” prior to conducting the study (Approval # 00070387). Teachers completed surveys at Waves 1 and 3. Families were compensated for their participation ($225 for 2 visits at Wave 1; $250 for 2 visits at Wave 2; $200 for 1 visit at Wave 3) and children received small toys at each visit. Teachers who participated received $30 at each wave.

Interparental Conflict

At Wave 1, mothers completed the Interparental Disagreement Interview, a semi-structured, narrative interview with the mother designed to assess the frequency, nature, course, and aftermath of interpartner conflicts witnessed by child participants (e.g., Davies et al., 2009; 2012). After selecting interparental disagreement topics that commonly take place in front of the child, a trained experimenter presented mothers with a series of open-ended questions (e.g., “How would you describe your disagreements over [topic]?”; “How do you typically feel during these disagreements?”; “How does the disagreement end?”). Experimenters used additional probes when necessary to maximize the richness of maternal narratives. Video-records of the interview were collected for subsequent coding.

To assess constructive IPC, a pair of trained coders rated three codes along seven-point dimensions ranging from 0 (None) to 6 (High). As the first two codes, maternal and partner supportive problem-solving were characterized by constructive conflict tactics that included expressing perspectives in a productive way, engaging in behaviors that help identify the problem or partner’s perspective, demonstrating empathy, and generating and implementing constructive solutions to the problem. The third code, conflict resolution, assessed the degree to which the couple as a dyad made progress toward resolving the problem over the course of the disagreements. High conflict resolution (i.e., 6) was evidenced by parents’ joint efforts to resolve disagreements in a mutually satisfactory way by maintaining composure, discussing perspectives in constructive ways, and exhibiting reciprocal cooperation and appreciation. To assess hostile IPC, a pair of trained coders rated three codes along the seven-point dimensions (0 = None; 6 = High). Coders specifically rated maternal and partner anger separately based on maternal narratives about the characteristics and course of the conflicts. High ratings reflect intense, dysregulated, and frequent forms of parental anger characterized by yelling in conjunction with the use of obscenities, physical violence, insults, and/or intense forms of angry withdrawal (e.g., storming off and leaving the house). As the third code, coders rated negative escalation displayed by the dyad during the interparental conflict. Negative escalation consisted of couple’s tendencies to exhibit mutual negativity toward each other in a cyclical way that prolongs or intensifies interparental discord. ICCs, reflecting interrater reliabilities for the coders on 21% of the interviews, ranged from .78 to .90 for the six codes assessing constructive and hostile IPC.

To determine if the coder ratings yielded two factors indexing constructive and hostile IPC, we submitted the six codes to an exploratory factor analysis using principal axis factor and a varimax rotation. Analyses of the eigenvalues (i.e., > 1) and scree plot supported a two-factor solution that explained 74.6% of the variance. Consistent with our measurement aims, the three constructive IPC codes and the three hostile IPC codes each loaded onto separate factors (all loadings > .51). Thus, we averaged each of the three codes to obtain composites of constructive (α = .66) and hostile (α = .66) IPC. Prior research supports the validity of coder ratings of interparental conflict derived from the IDI (e.g., Davies et al., 2009).

Children’s Emotion Situation Knowledge

At Waves 1 and 2, children completed two emotion situation knowledge tasks to assess their capacity to identify the emotions of others based on situational cues (Strand et al., 2016). For the first assessment, children watched 18 videos from the Emotions Subtest of the Schultz Test of Emotion Processing (STEP; Schultz et al., 2010). Each video of the Emotions Subtest depicts situations that prototypically elicit happiness, sadness, and fear (6 videos per emotion). After each video, the experimenter asks the child, “Does [the protagonist] feel happy, sad, angry, or scared?” For example, one of the videos designed to evoke happiness involves a girl complimenting another girl (i.e., the protagonist) on her drawing. An overall emotion perspective taking score was indexed by the proportion of correctly attributed emotions across the 18 videos. As the second measure, experimenters administered the Social Situations Scale from the Assessment of Children’s Emotion Skills (ACES; Izard et al., 2001) consisting of 12 verbal vignettes describing social situations that prototypically produce happy, mad, sad, and scared responses (3 for each emotion). After hearing each brief vignette (i.e., 2 or 3 sentences; “Lonnie is in line for lunch. Darren steps in front of him without asking”) from the experimenter, children are asked “Do you think [protagonist’s name] feels mad, scared, happy, or sad?” Children’s emotion situation knowledge score reflected the percentage of correct responses on the STEP and ACES, respectively. Given the conceptual and empirical correspondence between the ACES and STEP (r = .56 and .46 at Waves 1 and 2, respectively), we averaged the two measures together to obtain a parsimonious composite of emotion situation knowledge at each wave.

Children’s Social Problem-Solving

To assess children’s social problem-solving, experimenters administered two tasks with children at Waves 1 and 2. First, experimenters presented children with five stories portraying peer challenges that were adapted from the Social Problem Solving (SPS) vignettes (Davies et al., 2020; Dodge et al., 1995). The stories consisted of transgressions and peer group entry challenges (e.g., you let a kid play with your favorite toy while you are out of the room and when you get back you realize your toy is broken; you wanted to play a game with other kids, but they haven’t invited you to play). Experimenters encouraged children to pretend they were the protagonist in each story. After each vignette, experimenters asked children: “After [description of the specific peer event in the story], what would you do next?” Consistent with previous research (Davies et al., 2020), trained coders rated the recorded responses to each story along a four-point scale (1 = “none;” 4 = “high”) indexing constructive problem-solving. Whereas the “none” code was reserved for solutions that would likely worsen the stressful situation or the quality of peer relationships for children (e.g., aggressiveness, crying, loss of self-control), the “high” code captured responses that reflected a deep understanding of the complexity of the situation in a way that is likely to be effective in substantially reducing the stressful event and improving peer relations (e.g., “go to another group and try to be friends”). Interrater reliability coefficients based on two trained ratings of at least 21% of the children’s responses were good, with ICCs ranging from .81 to .95 across the five stories at each wave. We averaged the ratings across the five stories to form a composite at Waves 1 (α = .79) and 2 (α = .80).

Second, children completed an adapted version of the Provocation Subtest of the STEP (Schultz et al., 2010). Children watched 6 videos from the STEP depicting relational (3 videos; e.g., A child in a white shirt responds “No, not right now” to a bid from another child to join him in play) or physical (3 videos; e.g., Kids are playing catch when another child in a blue shirt walking by is hit with their ball) transgressions between children. After each video, children responded to the open-ended question: “What do you think the [boy/girl victim] in [color of clothing] will say or do next?” Following the SPS coding approach, trained coders rated the transcribed responses on a similar four-point scale (1 = “none;” 4 = “high”) indexing constructive problem-solving. Whereas the “none” code was reserved for solutions that would likely worsen the stressful situation (e.g., aggressiveness, crying), the “high” code was reserved for responses that reflected a sophisticated understanding of the situation that is likely to substantially reduce the stressful event and improve peer relations. ICCs, indexing interrater reliability coefficients based on two trained ratings of at least 34% of the children’s narrative responses, ranged from .75 to .92 across the six vignettes at each wave. Ratings across the six videos were averaged together to form a composite at Waves 1 (α = .71) and 2 (α = .72). Because SPS and STEP measures of constructive problem-solving were moderately correlated at Waves 1 (r = .30, p < .001) and 2 (r = .40, p < .001), we averaged the measures together at each measurement occasion to create a composite of constructive problem-solving at each wave.

Children’s Social Competence

For the battery measures assessing children’s social competence, teachers completed five scales at Waves 1 and 3, including the Communication, Empathy, and Assertiveness Scales from the Social Skills Improvement System - Rating Scales (SSIS-RS; Gresham et al., 2010) and the Prosocial Behavior and Peer Acceptance Scales from the MacArthur Health and Behavior Questionnaire (HBQ; Ablow et al., 1999). Whereas the SSIS-RS Communication scale reflects children’s tendencies to successfully exchange information and perspectives through verbal and nonverbal means (7 items, e.g., “The child takes turns in conversations”), the SSIS-RS Empathy scale assesses children’s concern and support for others (6 items, e.g., “The child tries to comfort others”). In addition, the SSIS-RS Assertiveness Scale assesses children’s abilities to communicate with others directly and authoritatively (7 items, e.g., “This child stands up for others who are treated unfairly”). As the fourth measure, the HBQ Prosocial Behavior Scale indexes children’s support, cooperation and affiliation in peer and classroom activities (20 items, e.g., “Offers to share materials or tools being used in a task”). Finally, the HBQ Peer Acceptance scale assesses the children’s successful formation of friendships and strong connections with peers (8 items, e.g., “Has lots of friends at school”). Research supports the reliability and validity of the SSIS-RS and HBQ scales (e.g., Ablow et al., 1999; Gresham et al., 2010). Internal consistencies (α) at Waves 1 and 3 respectively were as follows: .86 and .84 for the SSIS-RS Communication Scale, .92 and .92 for the SSIS-RS Empathy Scale, .82 and .84 for the SSIS-RS Assertive Scale, .94 and .94 for the HBQ Prosocial Behavior Scale, and .90 and .87 for the HBQ Peer Acceptance Scale. To adjust for the different response alternatives used in the SSIS-RS and HBQ, we standardized and averaged the five scales to create composites of children’s social competence at Waves 1 (scale level α = .86) and 3 (scale level α = .89).

Children’s Externalizing Symptoms

To assess children’s externalizing symptoms at Waves 1 and 3, teachers completed the HBQ Oppositional Defiant (9 items, e.g., “Has temper tantrums or hot temper”), Conduct Problems (11 items, e.g., “Lies or cheats”), and ADHD Symptoms (15 items, e.g., “Interrupts or butts in on others”) Scales (Ablow et al., 1999). Response options for the items were: 0 = Never or not true, 1 = Sometimes or somewhat true, 2 = Often or very true. Internal consistencies (α) in this sample at Waves 1 and 3 respectively were .90 and .89 for Oppositional Defiant, .90 and .85 for Conduct Problems, and .94 and .94 for ADHD. We averaged three measures together within each measurement occasion to create a composite of externalizing symptoms at Waves 1 (scale level α = .87) and 3 (scale level α = .76).

Children’s Internalizing Symptoms

At Waves 1 and 3, teachers also completed the HBQ Overanxious (8 items; “Nervous, high-strung, or tense”), Depression (6 items; e.g., “Unhappy, sad, or depressed”), and Social Withdrawal (9 items; e.g., “Withdraws from peer activities”) Scales (Ablow et al., 1999) using the same response alternatives from the externalizing items. Internal consistencies (α) at Waves 1 and 3 respectively were .76 and .82 for Anxiety, .72 and .79 for Depression, and .79 and .84 for Social Withdrawal Scales. We averaged the scales together to create measures of internalizing symptoms at Wave 1 (scale level α = .74) and 3 (scale level α = .77).

Demographic Characteristics (Covariates)

Four demographic covariates, based on a maternal interview consisted of: (1) children’s minority racial and ethnic identity (0 = White, not Latinx; 1 = racial minority or Latinx); (2) Wave 1 annual household income in units of $1000; and (3) the mean of maternal and partner education levels on a 7-point scale (1 = less than 7^th; 7 = graduate degree) at Wave 1. Finally, because children’s emotion understanding and social problem-solving vary by age during this period, we included children’s age at Wave 2 (in years) as a concurrent covariate of the two dimensions of their social-cognitive understanding and sequelae.

Transparency and Openness

Sample size was determined prior to conducting analyses based on power analyses in the original grant proposal and feasibility considerations. Access to the research materials, data, and statistical code for the primary and sensitivity analyses in this paper upon request by e-mailing the corresponding author. The data are not publicly available because the identities of the participants are potentially discoverable, particularly when integrated with other data from the project. The study design and analysis plan were not preregistered.

Results

Table 1 shows the means, standard deviations, and correlations for the variables included in the primary analyses. Across the one-year period from preschool to kindergarten, children exhibited increases in their social problem-solving skills (Cohen’s d = .71) and emotion situation knowledge (Cohen’s d = .96). Although constructive and hostile IPC were significant correlates of children’s social problem-solving and emotion situation knowledge at Wave 1, only constructive IPC was prospectively associated with the two putative mediators. In addition, children’s social problem-solving and emotion situation knowledge at Wave 2 were consistently associated with teacher reports of their psychological adjustment at Wave 3.

Table 1.

Means, Standard Deviations, and Correlations Among the Primary Variables in the Study.

	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
Covariates
1. Child Race	--	--	--
2. Wave 2 Child Age	5.62	0.41	.00	--
3. Parent Education Level	5.49	1.18	−.59*	−.11	--
4. Annual Income	80.35	49.12	−.49*	−.09	−.63*	--
Wave 1 Interparental Conflict
5. Constructive	2.19	1.56	−.23*	−.02	.28*	.30*	--
6. Hostile	2.10	1.20	.10	.07	−.25*	−.21*	−.45*	--
Wave 1 Child Mediators
7. Social Problem-Solving	2.42	0.60	−.22*	.09	.29*	.24*	.17*	−.19*	--
8. Emotion Situation Knowledge	0.69	0.16	−.35*	.07	.45*	.38*	.20*	−.23*	.26*	--
Wave 2 Child Mediators
9. Social Problem-Solving	2.76	0.64	−.17*	.06	.22*	.20*	.25*	−.14	.25*	.29*	--
10. Emotion Situation Knowledge	0.81	0.09	−.34*	.02	.40*	.29*	.23*	−.13	.30*	.62*	.29*	--
Wave 1 Child Psychological Adjustment
11. Externalizing Symptoms	0.25	0.32	.06	.00	−.17*	−.06	.01	.04	−.21*	−.04	−.25*	−.14	--
12. Internalizing Symptoms	0.27	0.23	.04	.07	−.08	−.01	−.06	.12	.02	.08	−.10	.08	.33*	--
13. Social Competence	−0.01	0.81	−.17*	.12	.22*	.14	.12	−.16*	.24*	.15*	.23*	.17*	−.62*	−.36*	--
Wave 3 Child Psychological Adjustment
14. Externalizing Symptoms	0.21	0.27	.05	−.02	−.12	−.11	−.08	.16*	−.16*	−.13	−.28*	−.23*	.33*	.06	−.25*	--
15. Internalizing Symptoms	0.26	0.27	−.02	.05	−.05	−.01	−.04	.09	−.01	−.03	−.08	−.21*	.07	.17*	−.16*	.50*	--
16. Social Competence	0.00	0.83	−.04	−.01	.14*	.09	.11	−.09	.11	.10	.30*	.21*	−.28*	−.18*	.34*	−.66*	−.49*

Note. Annual Income is indexed in units of $1000.

^*p < .05. Child race: 0 = White, non-Latinx; 1 = 1 = racial or ethnic (i.e., Latinx) minority.

Missing data for the variables in the primary analyses were minimal (14.1 %) and Little’s test, χ² = 265.50, df = 246, p = .19, supported the conclusion that the data were missing completely at random (Little, 1988). Given that full-information maximum likelihood (FIML) methods for estimating data successfully minimizes bias in the results for all types of missing data when the amount of missing data is less than 20% (Schlomer et al., 2010), we used FIML to estimate all parameters while retaining the full sample for all subsequent analyses. For all primary analyses, we conducted path analyses using a structural approach and observed variables with Amos 29.0 software (Arbuckle, 2021).

Primary Analyses: Emotion Situation Knowledge and Social Problem-Solving as Mediators

To test the mediational model, our path analysis specified Wave 1 constructive and hostile IPC and the four demographic covariates (i.e., child race and age, household income, parent education) as predictors of the two Wave 2 mediators (i.e., children’s emotion situation knowledge and social problem-solving) and the three proposed outcomes at Wave 3 (i.e., internalizing symptoms, externalizing problems, and social competence). The Wave 2 mediators, in turn, were estimated as predictors of the three forms of children’s psychological adjustment at Wave 3. In addition, we specified autoregressive paths using Wave 1 assessments of each of the mediators and the three dimensions of children’s psychological adjustment. To account for the possibility that mediational pathways are artifacts of concurrent associations among the IPC dimensions and children’s psychological outcomes, we also specified structural paths between each of the three forms of psychological adjustment at Wave 1 and the children’s emotion knowledge and social problem-solving at Wave 2. Finally, although not shown in the figures for clarity, we estimated correlations between all pairs of: (1) exogenous variables; and (2) residuals for the two mediators at Wave 2 and the three Wave 3 outcomes.

As depicted in Figure 1, the model provided a good representation of the data, χ²(14) = 21.56, p = .09, CFI = .99, RMSEA = .05. Autoregressive paths were significant for four of the five endogenous variables including: children’s emotion situation knowledge, β = .54, p < .001; internalizing symptoms, β = .17, p = .008; externalizing symptoms, β = .26, p < .001; and their social competence, β = .27, p < .001. Structural paths for all the demographic covariates were not significant. Although hostile IPC was significantly correlated with lower social competence at Wave 1 and greater externalizing symptoms at Wave 3 (see Table 1), hostile and constructive IPC were not significant predictors of any of the three forms of psychological adjustment in the model with the mediators and covariate paths. In addition, mediational pathways were not significant for hostile IPC due to its negligible role as a predictor of children’s emotion situation knowledge, β = .08, p = .23, and social problem-solving abilities, β = −.01, p = .86.

Figure 1.

Results of the Path Model Testing Children’s Social Problem-Solving and Emotion Situation Knowledge as Mediators in the Prospective Associations among Dimensions of Hostile and Constructive Interparental Conflict and Their Psychological Adjustment with Inclusion of Demographic and Child Adjustment Covariates. * p < .05.

Of relevance to our primary research questions, the results of the mediational analyses for constructive IPC were significant for three of the six pathways. Wave 1 constructive IPC was a predictor of children’s emotion situation knowledge at Wave 2, β = .14, p = .049. Children’s emotion situation knowledge, in turn, significantly predicted lower levels of their internalizing, β = −.23, p = .007, and externalizing, β = −.20, p = .02, symptoms at Wave 3. As follow up analyses, we conducted asymmetrical confidence interval analyses to determine if the two indirect paths were significant (MacKinnon et al., 2007). In supporting mediation, the results indicated that the indirect paths for Wave 1 constructive IPC and Wave 2 emotion situation knowledge were significant for both internalizing, b = −.005, 95% CI [−.0133, −.0001], and externalizing, b = −.005, 95% CI [−.0112, −.0001], symptoms at Wave 3. In addition, Wave 1 constructive IPC predicted greater social problem-solving abilities at Wave 2, β = .20, p = .02. Children’s social problem-solving abilities, in turn, were prospectively related to residualized increases children’s social competence in school at Wave 3, β = .20, p = .01. Asymmetrical confidence interval analyses indicated that the indirect path for constructive IPC, child social problem-solving, and their social competence was significant, b = .020, 95% CI [.0475, .0019].

Sensitivity Analyses

To examine the consistency of the mediational findings across different multivariate conditions, we conducted two sensitivity analyses. First, it is possible that the mediational findings may be idiosyncratic products of including child psychological adjustment variables as predictors of the proposed mediators and the demographic characteristics as predictors of the proposed mediators and outcomes. To address this possibility, the first analysis re-estimated all the structural paths in the primary model except that we excluded the paths involving the: (a) demographic covariates and the Wave 1 assessments of child psychological adjustment as predictors of Wave 2 social problem-solving and emotion situation knowledge; and (b) the demographic covariates as predictors of three dimensions of children’s psychological adjustment at Wave 3. The resulting model, depicted in Figure 2, fit the data well, χ²(40) = 47.79, p = .19, CFI = .99, RMSEA = .03. All five autoregressive paths for the endogenous variables were significant: children’s social problem-solving, β = .19, p = .01; emotion situation knowledge, β = .63, p < .001; internalizing symptoms, β = .17, p = .009; externalizing symptoms, β = .26, p < .001; and their social competence, β = .27, p < .001. In accord with the primary results, hostile IPC did not predict children’s social problem-solving or emotion situation knowledge at Wave 2. Moreover, the results of the mediational tests were consistent with the primary analytic findings. Wave 1 constructive IPC significantly predicted children’s emotion situation knowledge at Wave 2, β = .16, p = .02. Emotion situation knowledge, in turn, predicted residualized decreases in children’s internalizing, β = −.16, p = .04, and externalizing, β = −.18, p = .02, symptoms at Wave 3. Asymmetrical confidence interval analyses also indicated that the indirect paths involving constructive IPC and children’s emotion situation knowledge were significant for internalizing, b = −.004, 95% CI [−.0109, −.0001], and externalizing, b = −.005, 95% CI [−.0115, −.0003], problems. Likewise, Wave 1 constructive IPC predicted Wave 2 social problem-solving, β = .22, p = .009, which in turn, selectively predicted residualized increases in their social competence at Wave 3, β = .21, p = .006. Asymmetrical confidence interval analyses also indicated that the indirect path involving constructive IPC, children’s social problem-solving, and their social competence was significant, b = .024, 95% CI [.0535, .0040].

Figure 2.

Results of the Sensitivity Analyses Testing Children’s Social Problem-Solving and Emotion Situation Knowledge as Mediators in the Prospective Associations among Constructive and Their Forms of Psychological Adjustment without Inclusion of the Predictive Paths of Various Covariates. Values to the left and right of the backlashes reflect standardized path coefficients for sensitivity analysis I and II, respectively. n.e. = not estimated. * p < .05.

Another possibility is that the mediational paths for constructive IPC are an artifact of the inclusion of hostile IPC as a predictor of the endogenous variables in the model. Thus, in addition to excluding the same paths as the first sensitivity analysis, the second sensitivity analysis omitted the predictive paths running from Wave 1 hostile IPC to the Wave 2 mediators and Wave 3 outcomes. Thus, in addition to estimating the autoregressive paths for all exogenous variables, the paths specified included: (a) constructive IPC as a predictor of both Wave 2 mediators and the three forms of child adjustment at Wave 3; and (b) the two mediators at Wave 2 as a predictor of the three forms of child adjustment at Wave 3. The model shown in Figure 2 provided a satisfactory representation of the data, χ²(45) = 51.62, p = .23, CFI = .99, RMSEA = .03. The mediational findings for constructive IPC were mostly consistent with the results from the prior analyses. First, constructive IPC at Wave 1 continued to predict greater social problem-solving for children at Wave 2, β = .22, p = .003. Of the three outcomes, Wave 2 social problem-solving remained a significant predictor of child social competence at Wave 3, β = .22, p = .005. Asymmetrical confidence interval analyses continued to support the significance of the indirect path involving constructive IPC, child social problem-solving, and their social competence, b = .025, 95% CI [.0531, .0053]. Second, constructive IPC was also prospectively related to children’s better emotion situation knowledge at Wave 2, β = .13, p = .04. Emotion situation knowledge, in turn, significantly predicted residualized decreases in children’s internalizing, β = −.16, p = .04, and externalizing, β = −.18, p = .02, symptoms at Wave 3. Consistent with the previous analyses, asymmetrical confidence interval results indicated that the indirect path involving constructive IPC, child emotion situation knowledge, and externalizing symptoms was significant, b = −.004, 95% CI [−.0097, −.0001]. However, as an exception to the consistency in findings across analyses, asymmetrical confidence interval analyses were not significant for the indirect path of constructive IPC, child emotion situation knowledge, and internalizing symptoms, b = −.004, 95% CI [−.0092, .0001].

Discussion

Although research supports the premise that children may learn important social lessons from witnessing constructive IPC (Cummings & Davies, 2010; Jouriles et al., 2016), little is known about the social knowledge they acquire from these interactions and its implications for their mental health. Guided by theory (e.g., Don et al., in press), the goal of this study was to examine children’s emotion situation knowledge and social problem-solving as mediating mechanisms in the prospective links between constructive IPC and children’s social competence, internalizing symptoms, and externalizing difficulties. Consistent with hypotheses (Don et al., in press), results of the cross-lagged autoregressive path analyses indicated that exposure to constructive IPC predicted children’s greater emotion situation knowledge and social problem-solving over a one-year period from preschool to kindergarten after inclusion of hostile IPC, prior psychological adjustment, and demographic variables as predictors. In turn, our results supported conceptually guided predictions that emotion knowledge and problem-solving were precursors of better psychological adjustment (e.g., Nix et al., 2013; Ştefan & Avram, 2021; Thompson et al., 2020). Children’s emotion knowledge specifically predicted lower levels of their internalizing and externalizing symptoms while social problem-solving predicted higher child social competence over a one-year period. Indirect path analyses further revealed that children’s emotion situation knowledge was a significant mediator of their internalizing and externalizing problems and their social problem-solving was a significant mediator of their social competence. Finally, sensitivity analyses designed to examine the reproducibility of the findings across varying multivariate conditions supported the stability of the indirect pathways with one exception: one of the two sensitivity analyses failed to replicate the significant indirect path for constructive IPC, children’s emotion situation knowledge, and their internalizing symptoms.

Emotion Situation Knowledge and Social Problem-Solving as Sequelae of Constructive IPC

Our findings beg the question of why witnessing constructive IPC is a distinct predictor of children’s subsequent emotion situation knowledge and social problem-solving in the context of their exposure to hostile IPC, sociodemographic characteristics, and their pre-existing psychological adjustment. IPST and the tripartite model of family emotions share the thesis that the vicarious learning component of social learning theory may account for part of the link between constructive interparental interactions and children’s better social-cognitive capacities (Don et al., in press; Morris et al., 2007). Family figures who children perceive to be competent, warm, and interpersonally effective are regarded to be the most salient models for emulation (Don et al., in press). Therefore, children may be particularly likely to enact the constructive demeanor and conflict tactics of parents who are progressing toward resolving disagreements in satisfactory ways. Although these theories do not characterize the process through which children’s emulation of behaviors becomes embedded in their advanced social-cognitive understanding (Don et al, in press; Morris et al., 2007), other social learning formulations propose that acquiring generalized scripts and abstract rules of behavioral conduct are a key part of the vicarious learning process (Margolin et al., 2001). Thus, by reenacting parents’ successful conflict resolution tactics, children may internalize procedural scripts for formulating solutions to effectively cope with peer transgressions (Don et al., in press; Margolin et al., 2001). Relatedly, children’s emulation and internalization of affective behaviors in social contexts may also facilitate the development of emotion knowledge abilities centered on understanding meaningful connections between contextual cues and the experience of emotions (Vaughn et al., 2019).

Emotion contagion is another mechanism in IPST and the tripartite model of family emotions that may explain why children experience gains in emotion situation knowledge and social problem-solving following exposure to constructive IPC. Through their reflexive experience of the emotions displayed by their parents during constructive IPC, children may exhibit greater contentment, acceptance, compassion, engagement, and perseverance (Don et al., in press). Supporting this explanation, findings from studies using different methods (e.g., parent daily diaries; analogue simulations of conflict) have shown that witnessing constructive tactics between couples predicts higher levels of children’s positive emotionality (Cummings et al., 2002; 2003). In turn, these affective states are theorized to promote encoding, processing, and integration of a broader array of environmental (e.g., resources, stressors, challenges) cues and facilitate inclusive and flexible response decisions (Cummings & Rennels, 2014; Frederickson, 2013; Stifter et al., 2020). Accordingly, the interplay between children’s emotional experiences and greater information processing may accelerate their understanding of the linkages between contextual stimuli and emotions and their formulation of solutions to social problems.

Beyond the possible operation of emotion contagion and social learning processes, other mechanisms might account for the social-cognitive sequelae of constructive IPC. Spillover models propose that IPC indirectly affects children’s psychological functioning by altering parental caregiving practices with children (Cummings & Davies, 2010; Harold & Sellers, 2018). Thus, it is possible that constructive IPC may enhance children’s emotion understanding and social problem solving by promoting sensitive dimensions of parenting (McCoy et al., 2013; Swerbenski et al., in press). Likewise, although behavioral genetics studies support the notion that environmental processes partially account for associations between interparental and child functioning, there is also evidence that genetic factors shared by the biological parents and children give rise to better interparental and child psychological adjustment (Harold & Sellers, 2018). Thus, a genotype-environment correlation may exist whereby the intergenerational transmission of genetic characteristics is a source of children’s better social-cognitive understanding and parental use of constructive conflict tactics.

Mental Health Implications of Emotion Situation Knowledge and Social Problem-Solving

Supporting multiple mediational pathways, our path analytic findings further showed that children’s emotion situation knowledge and social problem-solving were predictors of better psychological functioning one year later even after controlling for their prior levels of psychological adjustment, exposure to IPC dimensions, and several demographic characteristics. Emotion situation knowledge was a precursor of children’s lower levels of internalizing and externalizing symptoms with the inclusion of the negligible role of social problem-solving as a predictor. Social information processing models offer several explanations for why emotion knowledge may reduce children’s risk for psychopathology. According to these models (e.g., Crick & Dodge, 2004; Lemerise & Arsenio, 2000), emotion situation knowledge provides an essential information base for successfully responding to multiple aspects of stressful situations, including: (1) adeptly encoding interpersonal cues, (2) accurately attributing causes and interpersonal motives for the events, (3) responding with well-regulated emotions, and (4) formulating and selecting effective solutions to managing the challenges (Denham et al., 2020). Thus, proficiency in interpersonal encoding and appraisals afforded by emotion situation knowledge may inhibit children’s social disorientation and confusion and, in turn, reduce their internalizing and externalizing symptoms by inhibiting social disorientation and confusion (Denham et al., 2012; Heinze et al., 2015). Likewise, skillfully identifying the contextual cues underpinning emotions may buffer children from psychopathology by increasing understanding and regulation of their own emotional reactions and reducing the enactment of hostile or dysregulated solutions in response to stressful events (Boden & Thompson, 2017; Di Maggio et al., 2016; Thompson et al., 2020). Finally, the greater emotional and behavioral regulation may prevent children’s involvement in the transactional cycles of interpersonal hostility and distress that further increase their vulnerability to psychopathology (Lopes et al., 2012).

Although research has yet to rigorously evaluate the viability of these explanations, there is emerging, preliminary support for some of the hypothesized processes underpinning the sequelae of emotion situation knowledge. For example, in a cross-sectional study, Di Maggio and colleagues (2016) reported that children’s better emotion regulation was both a correlate of their emotion knowledge and a predictor of their lower risk for internalizing and externalizing symptoms. Findings from the present study also showed that children’s emotion situation knowledge was concurrently associated with their constructive social problem-solving abilities which, in turn, predicted residualized increases in their social competence one year later. Finally, documented links between children’s emotion knowledge and harmonious peer relationships increase the plausibility that emotion knowledge might reduce psychopathology by forestalling their involvement in reciprocal, escalating episodes of interpersonal discord (Heinze et al., 2015).

In complementing the mediational pathways for emotion situation knowledge, our findings also revealed that social problem-solving was a selective mediator in the prospective link between constructive IPC and children’s social competence. Consistent with social information processing theory (Crick & Dodge, 1994; Lemerise & Arsenio, 2000), the ability to generate constructive solutions to peer challenges may provide children with a proximal knowledge base for enacting effective communicative, empathetic, assertive, and prosocial behaviors that strengthen their peer bonds (Hubbard et al., 2023; Youngstrom et al., 2000). Although replication is necessary before systematically examining why social problem-solving may mediate children’s social competence but not their internalizing or externalizing problems, one explanation is that social competence is a more proximal outcome of children’s capacity to generate effective solutions to peer problems than their behavioral and emotional problems (Barnes et al., 2017; Mooij et al., 2020), Thus, social problem-solving may be weakly and indirectly related to diminished risk for externalizing and internalizing symptoms through the more proximal mediating role of better social skills and competence (Hubbard et al., 2023). In accord with this interpretation, meta-analytic findings indicated that the effect sizes for social problem-solving interventions have a larger effect on increasing children’s social competence than they do in decreasing child behavior problems (Barnes et al., 2017; de Mooij et al., 2020). Relatedly, it is possible that emotion situation knowledge may be a more proximal mediator of relations between social problem-solving ability and children’s psychopathology. Supporting this explanation, although W2 social problem-solving was significantly correlated with Wave 3 child externalizing symptoms at a bivariate level in Table 1, it was no longer related to children’s externalizing problems when the significant role of emotion situation knowledge as a predictor was included in the path analyses.

Study Limitations and Conclusions

Study limitations also merit discussion. First, future research would benefit from expanding the examination of social-cognitive mediators beyond emotion situation knowledge and constructive problem-solving abilities. For example, emotion situation knowledge is part of a broader emotion understanding construct that encompasses emotion recognition and expressive knowledge dimensions (Heinze et al., 2015; Thompson et al., 2020). Likewise, although social problem-solving is a central construct in theory and intervention work (Lemerise & Arsenio, 2000; de Mooij et al., 2020; Su et al., 2020), inclusion of other social problem-solving dimensions (e.g., hostile and passive solutions, flexibility, evaluation of responses) may expand an understanding children’s psychological adjustment, particularly as precursors of their behavioral or emotional problems (Crick & Dodge, 1994; Hubbard et al., 2023). Second, although children in our sample were more racially diverse and had less economic resources than the US population, examining the comparability of our mediational findings in other samples of children (e.g., other cultures and nations, greater risk exposure) is an important task for future research. Third, although our sensitivity analyses primarily supported the stability of the mediational results, the indirect path involving constructive IPC, children’s emotion situation knowledge, and internalizing symptoms was no longer significant in one of the sensitivity models. Therefore, we recommend caution in interpreting these results until the field conducts further research. Finally, we aimed to provide a rigorous test of our hypotheses by implementing a fully lagged mediational path model with multiple methods and informants. However, our non-experimental design does not address all the confounding variables that might operate in the mediational paths. Therefore, implementing complementary methodological designs (e.g., interventions to increase constructive IPC) in future work may help to address this limitation.

In conclusion, despite advances in understanding the implications of constructive IPC for children’s emotional responding to conflict (Cummings et al., 2002; 2003; 2004; Cummings & Davies, 2010; Goeke-Morey et al., 2003), research has yet to identify the lessons children learn from witnessing constructive IPC and their implications for their psychological adjustment. To address this gap, our study was the first to examine children’s emotion situation knowledge and constructive social problem-solving ability as mediators of the prospective associations among their exposure to constructive IPC and their psychological adjustment. The findings supported the notion that children’s emotion situation knowledge and social problem-solving mediated children’s better outcomes following exposure to constructive IPC. If replicated, our results may have implications for future theory and practice. At a theoretical level, prior research has shown that the diminished vulnerability of children exposed to constructive IPC operates through mechanisms that do not involve negative valence systems designed to defend against threat (e.g., Davies et al., 2016). Thus, our findings highlight the value of examining how constructive IPC exposure may distinctively affect approach-oriented systems like affiliation and its theorized products in the form of social and emotional understanding. At the level of practice, the findings may have translational utility for developing and adapting prevention and intervention programs designed to increase supportive problem-solving and conflict resolution (e.g., Cummings et al., 2008). In contexts where programs for improving constructive conflict tactics between distressed caregivers is not feasible, our findings also suggest that school or community-based social and emotional learning interventions may be useful in interrupting risk cascades for children who witness low levels of constructive conflict with their caregivers (Cipriano et al., 2023).

Public Health Significance Statement:

Although developmental theory proposes that children benefit psychologically from witnessing parents constructively resolve their conflicts, little is known about the specific lessons they learn from these experiences. Findings from this study support the notion that constructive interparental conflict may reduce children’s risk for psychopathology and increase their social adjustment by increasing their likelihood of acquiring social problem-solving skills and knowledge about emotions. Identifying these skills may advance prevention targets and tools designed to promote children’s coping and adjustment.