Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Apr 1.
Published in final edited form as: J Fam Psychol. 2018 Apr;32(3):289–298. doi: 10.1037/fam0000346

Family Cohesion and Enmeshment Moderate Associations between Maternal Relationship Instability and Children’s Externalizing Problems

Jesse L Coe 1, Patrick T Davies 1, Melissa L Sturge-Apple 1
PMCID: PMC5926812  NIHMSID: NIHMS951750  PMID: 29698005

Abstract

This study examined the moderating roles of two different types of family-level closeness (i.e., family cohesion and enmeshment) in associations between maternal relationship instability and children’s externalizing problems in early childhood. Participants in this longitudinal (i.e., two waves of data collection spaced 2 years apart), multi-method (i.e., survey, observations), multi-informant (i.e., parent, teacher, observer) study included 243 preschool children (M age = 4.60 years) and their parents. Findings from the lagged, autoregressive tests of the predictive pathways indicated that family cohesion and enmeshment moderated associations between maternal relationship instability and increases in children’s externalizing problems. Maternal relationship instability was a significantly stronger predictor of children’s externalizing problems when cohesion was low or enmeshment was high. Follow up analyses revealed that cohesion predicted decreases in externalizing problems only at higher levels of instability. Conversely, higher levels of enmeshment predicted increases in children’s externalizing problems at high instability but decreases in externalizing symptoms under more stable family conditions.

Keywords: maternal relationship instability, family systems theory, family cohesion, family enmeshment, externalizing problems

Externalizing symptoms characterized by conduct problems, oppositional defiance, and overt hostility pose a critical problem for children in school settings and have substantial societal costs (Doshi et al., 2012; Foster, Jones, & the Conduct Problems Prevention Group, 2005). In the search for family risk factors, maternal relationship instability, characterized by intimate partner relationship transitions (i.e., starts of new relationships, break-ups, moving in with a partner), has been identified as a consistent precursor of externalizing problems in childhood (e.g., Ackerman, Brown, D’Eramo, & Izard, 2002; Bachman, Coley, & Carrano, 2011; Fomby & Cherlin, 2007; Osborne & McLanahan, 2007). During early childhood, children are forming expectancies about the consistency of their care and salient figures in their lives. Therefore, disruptions in the home environment and caregiving arrangements are posited to be especially potent in this time period (Bachman et al., 2011). For example, empirical documentation of heightened and protracted vulnerability of young children exposed to instability in family contexts suggests that the first five years of life (i.e., the time before children enter formal schooling) serve as a period of greater susceptibility for children (Belsky, Schlomer, & Ellis, 2012; Cavanagh & Huston, 2006; Donahue et al., 2010; Doom, Vanzomeren-Dohm, & Simpson, 2016; Simpson, Griskevicius, Kuo, Sung, & Collins, 2012). In fact, some studies have shown that exposure to instability in early childhood serves as a stronger and more consistent predictor of psychological problems later in childhood and adolescence than subsequent or concurrent assessments of instability (Cavanagh & Huston, 2008; Donahue et al., 2010; Simpson et al., 2012).

By the same token, significant variability is evident in the outcomes of children who experience high levels of maternal relationship instability (Amato, 2010; Bachman et al., 2011; Davies & Sturge-Apple, 2014). For example, experiencing multiple relationship transitions has not uniformly been linked with worse adjustment for children than exposure to a single transition (Carlson & Corcoran, 2001). However, in spite of apparent variability in the sequelae of children exposed to maternal relationship instability, little is known about the role of the broader family climate in altering the risk associated with instability. In fact, although significance is ascribed to family-level processes as sources of heterogeneity (see Hetherington & Stanley-Hagan, 1999), only one study has explored family climate as a moderator of instability. Specifically, Cavanagh and Huston (2006) found that general family support served as a protective factor, reducing the magnitude of the association between parental relationship transitions and teacher reports of first grade children’s problem behavior.

In building on this research direction, the goal of the current study was to provide the first longitudinal test of how different patterns of family relatedness (i.e., cohesion, enmeshment) identified in family systems theory (FST; Cox & Paley, 1997) may moderate associations between children’s early experiences with maternal relationship instability and their later externalizing problems during the transition from preschool to first grade. According to FST, the holistic climate of the family can be understood based on organizations of interactions within and across dyads or subsystems (e.g., mother-father, mother-child, father-child). Metaphorical boundaries allow for differentiation between these smaller subsystems within the larger family system. They serve as implicit rules for defining and understanding family relationships according to both the amount and quality of resources (e.g., warmth, autonomy support) and information transmitted across family subsystems (Minuchin, 1974). In other words, differences in types of boundaries guide how family members naturally interact with one another and the degree to which they involve one another in family activities (e.g., decision-making, conflict). In this boundary framework, FST differentiates between two types of family closeness which may alter associations between maternal relationship instability and children’s externalizing problems: family cohesion and enmeshment.

Family Cohesion

Families high in cohesion exhibit clear and flexible boundaries that permit each family member to function within subsystems without the interference of other members but also access resources from the larger family system when needed (Cox & Paley, 1997; Jacobvitz, Hazen, Curran, & Hitchens, 2004; Minuchin, 1974; Sturge-Apple, Davies, & Cummings, 2010). Thus, cohesiveness is marked by the provision of support, warmth, and intimacy across different family subsystems without compromising the autonomy or emotional well-being of any member (Cox & Paley, 1997; Minuchin, 1974; Sturge-Apple et al., 2010). When disruptive events in the family (e.g., maternal relationship transitions) do occur, the well-defined boundaries between family subsystems are reflected in the encapsulation of the difficulties within the specific subsystem (e.g., interparental subsystem). Thus, the problems experienced by the subsystem do not proliferate in a way that entangles other family members (e.g., child; Davies & Sturge-Apple, 2014; Kerig, 1995). Given that these families are able to successfully encapsulate problems, we hypothesize that family cohesion will serve as a protective factor, buffering children from the risk conferred by maternal relationship instability.

Drawing on the well-established taxonomy of protective factors in developmental psychopathology (Luthar, Cicchetti, & Becker, 2000), the protective role of family cohesion may operate in a couple of different ways, depending on how it confers protection in the presence of different levels of risk. First, as shown in Figure 1a, the protective-stabilizing model proposes that family cohesion is associated with uniformly low levels of externalizing problems across levels of maternal relationship instability. Whereas instability is a relatively weak predictor of changes in externalizing problems at high levels of cohesion (see solid line in Figure 1a), it may take an exponentially greater toll on children when cohesion is low (see dotted line in Figure 1a). Therefore, because families high in cohesion may be able to manage the emotional ramifications of unstable events in the interparental subsystem, it is possible that children from these families may be less adversely affected by high maternal relationship instability. Alternately, the protective-reactive form of moderation shown in Figure 1b posits that the protection afforded by cohesion is particularly robust when instability is low but loses much of its buffering power for children faced with high instability. In other words, although children from more cohesive families tend to be at lower risk for developing externalizing problems than children from less cohesive homes, the difference becomes less pronounced as children’s exposure to instability and the corresponding potency of its risk increases. Given the paucity of research exploring family cohesion as a moderator of instability, we test the relative value of protective-stabilizing versus protective-reactive models in examining cohesion as a buffer in the prospective association between maternal relationship instability and children’s externalizing problems.

Figure 1.

Figure 1

Figure 1

Figure 1

Figure 1

Open in a new tab

a. Conceptual illustration of protective-stabilizing model.

b. Conceptual illustration of protective-reactive model.

c. Conceptual illustration of vulnerable-reactive model.

d. Conceptual illustration of vulnerable-adaptive model.

Family Enmeshment

Highly enmeshed families tend to have boundaries that are overly diffuse and permeable, causing family members to become emotionally entangled with one another (Minuchin, 1985). Enmeshment can take the form of (a) conditional access to resources (e.g., conditional support) that occurs at the cost of hindering individual autonomy or (b) distress and hostility that seamlessly spills over across subsystems and immerses individuals in the ongoing issues in the family (Minuchin, 1974, 1985). A main premise is that the high level of interdependency increases children’s preoccupation with and sensitivity to family stress (Davies & Sturge-Apple, 2014; Forman & Davies, 2005; Kerig, 2005). Therefore, we hypothesize that family enmeshment will serve as a potentiating factor, exacerbating the risk posed by maternal relationship instability for the development of children’s externalizing problems.

The potentiating effect may assume two primary forms within moderator models of developmental psychopathology (Davies, Coe, Martin, Sturge-Apple, & Cummings, 2015; Luthar et al., 2000). On the one hand, the vulnerable-reactive form of risk (see Figure 1c) proposes that high enmeshment amplifies the risk of maternal relationship instability. Thus, although instability is theorized to predict greater behavior problems for children across different levels of enmeshment, it is assumed to pose a particularly pronounced risk when enmeshment is high (see solid line in Figure 1c). On the other hand, the vulnerable-adaptive model (see Figure 1d) posits that tendencies for enmeshed children to exhibit high concern for family members’ welfare may confer significant psychological disadvantages when maternal relationship instability is high but also potential developmental benefits in homes with more stable maternal relationships. Therefore, as is shown by the steep slope of the solid line in Figure 1d, the vulnerable-adaptive and vulnerable-reactive models share the assertion that associations between maternal relationship instability and externalizing problems will be particularly pronounced for children in highly enmeshed families. However, the key difference is that the vulnerable-adaptive model proposes that higher levels of family enmeshment may be associated with better child adjustment in stable homes. We test the relative viability of the vulnerable-reactive and vulnerable-adaptive models for family enmeshment.

Present Study

The present study is the first longitudinal test of how FST characterizations of the emotional climate of the family (i.e., family cohesion and enmeshment) may moderate associations between children’s early experiences with instability in maternal intimate relationships and their later externalizing problems during the transition from preschool to first grade. As an important developmental milestone, the transition into formal educational settings provides a unique opportunity to understand how children’s experiences with early family relationships may shape how they adjust to extra-familial contexts. Thus, by virtue of their novelty and challenges, school settings are salient contexts for children to draw on prior family experiences as a way to understand the unfamiliarity and complexity of new situations (e.g., Davies, Winter & Cicchetti, 2006). Supporting this conceptualization, prior research has identified characteristics of children’s early family relationships to be significant predictors of their behavioral and emotional adjustment to school (e.g., Clark & Ladd, 2000; National Institute of Child Health and Human Development Early Child Care Research Network, 2004). Additionally, quantitative psychologists have emphasized the importance of utilizing longitudinal data in which the predictor and outcome are temporally separated with autoregressive controls specified for the outcome in obtaining accurate assessments of predictor-outcome relationships (e.g., Cole & Maxwell, 2003; Gollob & Reichardt, 1991). Therefore, in our two-wave longitudinal design, we examined instability and its interplay with family enmeshment and cohesion at the first measurement occasion as predictors of subsequent changes in children’s externalizing symptoms 2 years later.

To reduce the operation of common method and informant variance in the analyses, we utilized a multi-method (surveys, observations), multi-informant (observer, parent, and teacher) approach to assessing the key constructs. Analyses control for child sex and total household income per capita based on empirical documentation of associations with family processes (Conger & Conger, 2002; Maccoby & Jacklin, 1974; Wauchope & Straus, 1990) and child adjustment (Ackerman, Brown, & Izard, 2004; Campbell, Shaw, & Gilliom, 2000; Conger, Conger, & Martin, 2010).

Methods

Participants

Participants included 243 families (mother, intimate partner, and preschool child) residing in a moderate-sized metropolitan area in the Northeastern United States. The sample was recruited through fliers and informational presentations at local preschools, Head Start agencies, Women, Infants, and Children (WIC) programs, and public and private daycares serving children and families from a variety of demographic backgrounds. To be eligible for participation, primary caregivers had to currently have an intimate partner who evidenced regular contact (average of 2 to 3 days per week) with the caregiver and child over the majority of the year (i.e., at least 10 months; not necessarily continuously) prior to Wave 1. Children must have been within one year of enrolling in kindergarten, and only one child per family was eligible to participate. The two measurement occasions for this study were spaced 2 years apart, and the retention rate from Wave 1 to Wave 2 was 91%. Comparisons of families lost to attrition and those who participated in both waves along the 11 study variables used in the primary analyses yielded only one significant difference: families participating in both waves evidenced significantly lower total household income per capita (in thousands: M = 9.23, SD = 5.81) than those who dropped from the study (in thousands: M = 13.35, SD = 8.83), t(237) = 2.94, p < .01.

The average age of child participants was 4.60 years (SD = .44) at Wave 1 and 6.81 years (SD = .48) at Wave 2. At Wave 1, the average age of participating mothers was 30.76 years (SD = 6.31), and the average age of participating partners was 33.63 years (SD = 7.91). Of child participants, 47% were the mother’s first-born child, and 21% were the mother’s only biological child. Almost half (48%) of the families were Black or African American, 43% White, and 9% multi-racial or another race. Approximately 16% of family members were Latino. Median household income of families was $36,000 per year (range = $2,000 - $121,000), with most families (69%) receiving public assistance. Median education for parents consisted of a GED or high school diploma. At Wave 1, parents had lived together an average of 3.36 years and had, on average, daily contact with each other and the child (range = 2 or 3 days a week to daily). Of adults, 99% of mothers and 74% of partners were biological parents of the target child. Forty-seven percent of the adults were married.

Procedures and Measures

Parents and children visited our research center laboratory for two waves of data collection. At each wave, teachers completed and mailed questionnaires about children’s behavior. Families and teachers were compensated monetarily for participation. Parents gave written consent for themselves and for their child to participate. Ethical permission for the study was granted by the university’s Institutional Review Board.

Maternal relationship instability

At Wave 1, mothers completed the Family Instability Questionnaire (FIQ; Ackerman, Kogos, Youngstrom, Schoff, & Izard, 1999; Forman & Davies, 2003). Three items assessed the total number of intimate relationship transitions experienced in the last year (i.e., start of new relationships, break-ups, moving in with a partner). The scale consisted of the sum of items (sample range = 0 to 8 transitions). Speaking to the variability in instability reported by mothers in our sample, each item was endorsed by a sizable portion of mothers (i.e., 46% of mothers reported at least one unstable event over the previous year). Specifically, 35% of mothers reported that they experienced at least one instance of becoming involved in a serious romantic relationship in the last year (sample M = .56, SD = 1.24), 30% reported at least one instance of moving in with a serious romantic partner (sample M = .34, SD = .59), and 20% reported at least one instance of breaking up with or separating from a romantic partner (sample M = .42, SD = 1.58). Prior research supports the validity of the maternal relationship instability assessment (Bachman et al., 2011; Cavanagh & Huston, 2006; Donahue et al., 2010).

Family cohesion and enmeshment

At Wave 1, families (mother, partner, and child) participated in a family interaction task. In this 10-minute task, families were instructed to work together to build a house out of Legos that closely resembled a picture version. The family was then left alone to complete the activity, and the task was video recorded for subsequent coding. Given the dearth of observational assessments of family boundaries, we used prior family interaction coding systems (Jacobvitz et al., 2004; Lindahl & Malik, 2000) as a guide in developing two 9-point scales ranging from 1 (not at all characteristic) to 9 (highly characteristic) to assess family boundary patterns.

The cohesion scale assesses the extent to which the family’s interaction reflects a balanced concern for all family members’ emotional and psychological well-being so that there is some flexibility, permeability, and access to resources (support, warmth, intimacy) across relationships, but it occurs within appropriate, well-defined boundaries that do not compromise any one member’s autonomy. In families rated highly in cohesion, flexibility in boundaries is reflected in the ability of all three members to be responsive to each other’s bids for attention and comments in warm or supportive ways, work as a team to build the house, and respect and support each other’s individuality, personal decisions, and contributions. For example, all three members interact in ways that encourage and facilitate their contributions to the activity (e.g., mother may make comments to build the father up to the child, “[Child], see how daddy found this blue piece? Can you find one like it?”, complimenting the father while giving the child suggestions without being over-controlling). Additionally, cohesive families often display positive affect (e.g., smiling, nods, jokes, signs of enjoyment) and displays of affection (e.g., pet names, compliments, statements of liking or appreciation) throughout the interaction. The “well-defined” part of the descriptor reflects that teamwork and synchrony is expressed in a way that also supports the individual contributions and personal decisions of each family member. For example, parents may recognize the need to support exploration of their child within the context of the task by structuring activities that offer choices for the child in their role in the task (e.g., “Do you think we should put this yellow piece here?”) and praising the child for accomplishments in the task. In a cohesive family, members may direct one another’s behavior by giving choices for what to do next, therefore providing direction without compromising individual contributions.

In contrast, the enmeshment scale assesses the extent to which boundaries are overly diffuse or permeable. High scores on enmeshment are reserved for families where all members are overly involved in one another’s activities (e.g., repeatedly interrupt or talk over one another, overly rely on others for help, grab Legos out of each other’s hands). While members may express some degree of conditional warmth and positivity when the activity is going well (e.g., parent may say “We’re doing such a good job! We’re building the best house!” when the family has made progress in making a house that looks like the picture), they may also quickly become critical or hostile until other members meet their standards or acknowledge their perspective (e.g., parent may say “I told you we shouldn’t put that piece there. Now it looks bad. Put the red piece on instead” when another family member put a piece on the house that the parent did not like). Family members may also impede on one another’s individual contributions to the task. For instance, whereas in a cohesive family, family members may give direction by giving the others a choice of what to do next (e.g., “Do you think we should put the yellow piece here?”), in an enmeshed family, family members may hover over one another and give commands or direct requests, denying the others of opportunities to make their own individual contributions to the task (e.g., “You really need to put the yellow piece here”).

Two trained coders independently rated 20% of the videos throughout the entire coding period to assess interrater reliability and to monitor and address rater drift. Intraclass correlation coefficients were .92 for cohesion and .88 for enmeshment. Although these codes are being used for the first time in the present study, findings from supplementary analyses support their validity. For example, consistent with family systems theory and previous empirical work (e.g., Johnson, Cowan, & Cowan 1999; Kerig, 1995; Richmond & Stocker, 2006; Sturge-Apple et al., 2010), higher ratings on our Cohesion code at Wave 1 were significantly correlated with teachers’ reports of lower levels of child classroom difficulties in preschool and first grade (i.e., lower inattention, fewer externalizing problems, higher prosocial behavior, and greater school engagement). Also consistent with theory and prior research (e.g., Barber & Buehler, 1996; Jacobvitz et al., 2004; Minuchin, 1974; Sturge-Apple et al., 2010), higher ratings of Enmeshment at Wave 1 were significantly correlated with teacher-reported classroom difficulties (i.e., less engagement in classroom activities and higher rates of teacher-child conflict) and indices of internalizing symptoms (i.e., asocial behavior and social withdrawal). Additionally, our codes were found to be related to other measures of family processes. For instance, higher Cohesion ratings were significantly and positively correlated with both maternal and paternal reports of family cohesion on the Family Adaptation and Cohesion Scales (FACES III; Olson, 1986).

Children’s externalizing problems

Teacher reports on three scales from the MacArthur Health and Behavior Questionnaire (HBQ; Ablow et al., 1999) were used as indicators of a latent construct of children’s externalizing problems at Wave 1 (i.e., when children were in preschool) and Wave 2 (i.e., when children were in first grade). These included Overt Hostility (four items; e.g., “kicks, bites, or hits other children”), Conduct Problems (10 items; e.g., “physically attacks people”), and Oppositional Defiant (nine items; e.g., “defiant, talks back to adults”) scales. Each scale consisted of the sum of items, and internal consistencies for the scales across the two waves ranged from .80 to .92. Prior research supports the reliability and validity of the HBQ scales for assessing young children’s psychological adjustment (see Ablow et al., 1999).

Covariates

Two demographic covariates, derived from a maternal interview at Wave 1, included (a) children’s sex (1 = girls; 2 = boys) and (b) total household income per capita, calculated by dividing the total annual household income by the number of residents in the home.

Results

Descriptive Analyses

Table 1 provides the means, standard deviations, and intercorrelations for the variables used in the primary analyses.

Table 1.

Means, Standard Deviations, and Intercorrelations of the Main Variables

Mean SD 1 2 3 4 5 6 7 8 9 10
Wave 1
1. Child Sex – – –
2. Household Income Per Capita (thousands) 9.59 6.22 .06 –
3. Maternal Relationship Instability 1.23 1.86 .04 –.24* –
4. Enmeshment 3.33 2.57 −.15* .07 −.01 –
5. Cohesion 3.75 2.56 .04 .35* −.34* .03 –
6. Overt Hostility .99 1.74 .09 −.12 .13† .05 −.14† –
7. Conduct Problems 1.52 3.27 .13 −.11 .09 .10 −.12 .80* –
8. Oppositional Defiant 2.77 3.57 .06 −.11 .22* .10 −.18* .80* .80* –
Wave 2
9. Overt Hostility 1.09 1.74 −.00 −.26* .38* −.07 −.27* .34* .43* .41* –
10. Conduct Problems 1.59 3.01 −.04 −.28* .43* .01 −.27* .32* .41* .43* .89* –
11. Oppositional Defiant 2.71 3.97 −.04 −.26* .39* .02 −.27* .38* .49* .46* .84* .82*
Open in a new tab

Note.

*

p < .05.

†

p < .10

Primary Analyses

To test associations involving the interplay of maternal relationship instability and family boundaries (i.e., enmeshment, cohesion), and children’s externalizing problems, we utilized autoregressive structural equation modeling (SEM) through the Amos 22.0 statistical software program. Full-information maximum likelihood (FIML) was used to estimate missing data (data were missing for 17% of all values and for 29% of teacher-reported data) and retain the full sample for primary analyses (Enders, 2001). To maximize measurement equivalence in latent constructs of externalizing problems from Wave 1 to Wave 2, we specified strong factorial invariance constraints on the analyses (Widaman, Ferrer, & Conger, 2010). Therefore, factor loadings and intercepts of the same indicators of externalizing problems were fixed to be invariant across time. Adding these constraints did not result in significant change in fit from the unconstrained model: Δχ2 = 3.98, df = 4, p = .41.

To test our primary research questions, maternal relationship instability, family cohesion, family enmeshment, the multiplicative interactions between instability and each boundary pattern, and the two covariates (total household income per capita and child sex) were simultaneously specified as predictors of children’s externalizing problems at Wave 2 (see Figure 2). Additionally, the autoregressive path at Wave 1 was estimated for externalizing problems to control for stability in the proposed outcome. To reduce multicollinearity, we centered the predictors (instability, enmeshment, cohesion) and created interaction terms from the centered variables (Aiken & West, 1991). All correlations among exogenous predictors were also estimated. However, for clarity, only significant correlations are shown in Figure 2.

Figure 2.

Figure 2

Open in a new tab

Structural equation model examining interactive effect of enmeshment, cohesion, and relationship instability on children’s Wave 1 and Wave 2 externalizing problems. Parameter estimates for structural paths are standardized path coefficients. Dashed lines indicate non-significant pathways. *p < .05. ***p < .001. †p < .10

The resulting model provided a good fit with the data: χ2 (40, N = 243) = 52.89, p = .08; RMSEA = .04; CFI = .99; χ2/df ratio = 1.32. In support of the measurement model, the standardized loadings of the manifest indicators onto their latent constructs were all significant and strong in magnitude (≥ .87). The autoregressive path was moderate in magnitude and significant for children’s externalizing problems, β = .33, p < .001. Consistent with prior empirical work, children’s externalizing problems at Wave 2 were predicted by higher levels of maternal relationship instability at Wave 1, β = .18, p < .05, lower household income per capita, β = −.13, p < .10, and lower family cohesion at Wave 1, β = −.20, p < .05. Also consistent with theory and previous research (e.g., Barber & Buehler, 1996), enmeshment and cohesion ratings were unrelated to one another. Of relevance to our aims, Wave 2 externalizing symptoms were significantly predicted by the interactions between: (a) maternal relationship instability and family cohesion, β = −.24, p < .05 and (b) maternal relationship instability and family enmeshment, β = .27, p < .001.

To dissect the interactions, we conducted simple slope plots and analyses of maternal relationship instability at lower (−1 SD) and higher (+1 SD) levels of each family boundary pattern (Aiken & West, 1991). First, the findings indicated that maternal relationship instability was significantly associated with increases in externalizing problems for children who experienced lower (−1 SD), b = 0.81, p < .001, but not higher (+1 SD), b = −.15, p = .63, levels of family cohesion. The graphical plot depicted in Figure 3a appeared to most closely resemble the protective-stabilizing (see Figure 1a), rather than protective-reactive (see Figure 1b), model. As a more authoritative quantitative test, we followed statistical guidelines for calculating regions of significance on X (RoS on X) tests (Dearing & Hamilton, 2006). RoS on X tests invert the predictor and moderator to yield analyses of the significance of the association between the moderator (i.e., family cohesion) and outcome (i.e., externalizing symptoms) within the bounded regions (i.e., −/+ 1 SD) of the proposed predictor (i.e., maternal relationship instability). Support for the protective-stabilizing model would be evidenced by findings indicating that cohesion was significantly associated with decreases in externalizing problems only at higher levels of instability. In contrast, the protective-reactive model would be supported if cohesion predicted decreases in externalizing problems only when instability was low. Consistent with the protective-stabilizing form of moderation, the results indicated that cohesion predicted decreases in externalizing problems at higher (+1 SD), b = −0.61, p < .001, but not lower (−1 SD), b = 0.09, p = .66, levels of maternal relationship instability.

Figure 3.

Figure 3

Figure 3

Open in a new tab

a. Graphical plot of interaction between maternal relationship instability and family cohesion in predicting changes in children’s externalizing problems over 2-year period.

b. Graphical plot of interaction between maternal relationship instability and family enmeshment in predicting changes in children’s externalizing problems over 2-year period.

Next, we conducted simple slope plots and analyses of maternal relationship instability at −1 SD and +1 SD from the mean of family enmeshment. Maternal relationship instability was significantly associated with increases in externalizing problems for children who experienced higher (+1 SD), b = 0.80, p < .001, but not lower (−1 SD), b = −.14, p = .56, levels of family enmeshment. The graphical plot shown in Figure 3b appeared to more closely resemble the vulnerable-adaptive (see Figure 1d) than vulnerable-reactive (see Figure 1c) model. To more authoritatively test the nature of the interaction, we proceeded to conduct the RoS on X test. Support for the vulnerable-reactive model would be obtained if enmeshment was significantly associated with increases in externalizing problems only when instability was high. In contrast, the vulnerable-adaptive model would be supported if enmeshment predicted significant increases in externalizing problems at high instability and significant decreases in externalizing problems at low instability. Consistent with the vulnerable-adaptive form of risk, the results indicated that enmeshment predicted decreases in externalizing problems at lower (−1 SD) levels of maternal relationship instability, b = −0.35, p < .01, and increases in externalizing problems at higher (+1 SD), b = 0.34, p < .01, levels of maternal relationship instability.

We also conducted follow-up analyses to test whether our pattern of results would be the same for children’s concurrent experiences with maternal relationship instability. Therefore, we tested an additional model in which we examined instability experienced between the two waves. The model fit the data well: χ2 (40, N = 243) = 50.09, p = .13; RMSEA = .03; CFI = .99; χ2/df ratio = 1.25. Consistent with our a priori decision to focus on the influence of children’s early experiences with maternal relationship instability on their later adjustment outcomes, instability experienced between Wave 1 and Wave 2 and its interactions with cohesion and enmeshment failed to predict changes in children’s externalizing problems from Wave 1 to Wave 2, all ps > .61.

Discussion

Although maternal relationship instability has repeatedly been shown to increase children’s risk for developing behavior problems (e.g., Bachman et al., 2011; Cavanagh & Huston, 2008; Donahue et al., 2010), little is known about the family characteristics that may moderate this association. Guided by family systems theory (FST; Cox & Paley, 1997), our longitudinal study examined how the interplay between children’s early experiences with maternal relationship instability and two different types of family closeness (i.e., family cohesion and enmeshment) informs an understanding of subsequent changes in their externalizing problems over a 2-year period representing the transition from preschool to first grade. Findings indicated that both family cohesion and enmeshment moderated the association between maternal relationship instability and children’s externalizing problems in unique ways.

Consistent with previous research and theory (Ackerman et al., 2002; Bachman et al., 2011; Belsky et al., 2012), maternal relationship instability experienced in the preschool, but not early school, years uniquely predicted increases in children’s externalizing problems over a 2-year period, even with the inclusion of family boundary patterns, demographic characteristics, and prior externalizing problems as predictors. In accord with previous empirical documentation of the healthy adjustment of children from cohesive families (e.g., Johnson et al., 1999; Kerig, 1995; Richmond & Stocker, 2006), family cohesion was also a unique predictor of decreases in externalizing problems over time. Of relevance to the main aims of the study, our findings further revealed that family cohesion moderated the association between maternal relationship instability and children’s externalizing problems in a way that corresponded with the protective-stabilizing (Figure 1a) model (Luthar et al., 2000). For children from highly cohesive homes, exposure to maternal relationship instability was not a significant predictor of their externalizing problems. In other words, children from cohesive families exhibited low levels of externalizing problems regardless of their exposure to maternal relationship instability. Conversely, higher levels of maternal relationship instability predicted increases in externalizing problems for children who experienced lower family cohesiveness.

Our results are consistent with the notion that parents in cohesive families are better at compartmentalization. Parents who are able to compartmentalize can effectively separate their roles as spouse and parent (Grych, 2002; Sturge-Apple, Davies, Cicchetti, & Fittoria, 2014). Thus, even when they are experiencing higher levels of discord and instability in their intimate relationships, parents in cohesive families may be able to successfully contain the distress to the interparental subsystem and, in the process, still provide resources (e.g., protection, warmth, support) to children. In accord with the concept of compartmentalization, cohesive families are characterized by well-defined boundaries that serve to insure that difficulties in any given relationship (e.g., interparental) do not proliferate to adversely impact the broader family unit (Cox & Paley, 1997; Davies & Sturge-Apple, 2014).

Our results also indicated that family enmeshment moderated the relationship between maternal relationship instability and children’s externalizing symptoms in a vulnerable-adaptive form (see Davies et al., 2015; Luthar et al., 2000). Maternal relationship instability was only a significant predictor of greater externalizing symptoms when children experienced high family enmeshment. In further dissecting the cross-over interaction, the results support the thesis that growing up in a highly enmeshed family alters the impact of maternal relationship instability in a “for better or for worse” fashion. For children from highly enmeshed families, relatively high levels of maternal relationship instability predicted greater increases in their externalizing problems compared to their counterparts in less enmeshed families. However, relatively low levels of maternal relationship instability (i.e., exposure to greater stability in maternal relationships) was associated with greater decreases in externalizing problems over time for children from highly enmeshed families.

Previous studies have specifically identified child attributes (e.g., child temperament, involvement in interparental conflict, histories of insecurity) as conferring vulnerable-adaptive forms of moderation in associations between family characteristics and children’s adjustment (Belsky & Pluess, 2009; Davies et al., 2015; Davies, Sturge-Apple, Bascoe, & Cummings, 2014). Thus, our findings differ in the identification of a family-level (i.e., enmeshment) variable as a susceptibility factor. Guided by FST, it is possible that the heightened entanglement in enmeshed homes may amplify children’s emotional investment and sensitivity to variations in family stability in a “for better or worse” fashion. On the “for worse” side of the equation, high immersion in the family may sensitize children to highly unstable events and, as a result, amplify their vulnerability to behavior problems (Kerig, 2005; Minuchin, 1985). On the “for better” side of the equation, children in enmeshed families may benefit disproportionately from greater exposure to the resources of stable maternal relationships (e.g., Chase-Lansdale, Wakschlag, & Brooks-Gunn, 1995). For example, given the weaker boundaries between interparental and parent-child relationships, these children may learn important social lessons through greater exposure to parental displays of cooperation and conflict resolution strategies in their intimate relationship (Cummings &Davies, 2010).

With regard to the potential operative processes underlying the moderating roles of the two family boundaries, it is also important to highlight that family cohesion and enmeshment vary systematically in their levels of warmth and autonomy support (Cox & Paley, 1997; Minuchin, 1974). Framed in this way, it is possible that the findings from our use of pattern-based assessments were due to differences in warmth, boundary permeability, or their combination. Thus, sources of the protective effects of cohesion under conditions of high maternal relationship instability may be rooted in the greater capacity for the family to exhibit warmth and sensitivity in times of distress, encourage autonomy and independence, or both. In further considering the full unfolding cascade of processes, these types of resources may help to offset the risk posed by instability by reducing children’s appraisals of the family as sources of threat and insecurity (Ackerman et al., 1999; Forman & Davies, 2003).

Conversely, the role of family enmeshment as conferring potential benefits only in the context of stable interparental relationships may be the result of the lower and more conditional expressions of family warmth that also co-occur with restrictions in autonomy and possible spillover of negativity across relationships. For example, on the one hand, conditional warmth expressed in enmeshed families may assume a more positive meaning for children when there is high stability and predictability in the interparental relationship. On the other hand, when relationships are less stable, the conditional warmth may also serve to immerse children in the highly volatile and discordant interparental subsystem in ways that exacerbate children’s negative appraisals of the family unit. As another potential operative process, the control, immersion, and intrusiveness exhibited in enmeshed families might also be effective, at least temporarily, in impeding the development of behavior problems (e.g., externalizing symptoms) within more structured (i.e., stable) family conditions. However, as a possible tradeoff, it is plausible that this configuration of enmeshment and stability might actually undermine children’s adjustment in other domains (e.g., internalizing symptoms). Consistent with this notion, previous research has found that children from highly enmeshed families are at heightened risk for developing internalizing problems (Barber & Buehler, 1996; Jacobvitz et al., 2004; Kerig, 1995; Sturge-Apple et al., 2010).

Discussion of the study limitations is also warranted. First, as we previously noted, our findings may not generalize to all types of adjustment difficulties (e.g., internalizing symptoms). Second, although participants in our study were from diverse racial, ethnic, and socioeconomic backgrounds, the results may not necessarily generalize to high risk families, highly affluent families, or children in different developmental periods (e.g., adolescence). Third, our eligibility criterion that mothers had to currently have an intimate partner who evidenced regular contact with the caregiver and child for at least 10 months prior to Wave 1 may have somewhat constrained the amount of instability in our sample. Although a substantial portion (i.e., 46%) of mothers did report that they experienced at least one relationship transition over the course of the previous year, and the average number of transitions experienced by families was comparable to or somewhat higher than those reported in previous studies (e.g., Ackerman et al., 2002; Bachman et al., 2011; Belsky et al., 2012), it will be important for future research to test whether our findings generalize to families experiencing even higher levels of maternal relationship instability.

In addition to limitations regarding the potential generalizability of our findings, it is also important to acknowledge other study limitations. First, although our new observational assessment of family boundaries may be regarded as a strength of the current study, it will be critical for future work to further explore its psychometric properties (e.g., examine correlation between the enmeshment code and more traditional assessments of enmeshment). Similarly, replication is an important next step given the novel nature of this measure. Second, future research would benefit from expanding the search for family-level moderators of maternal relationship instability. Our focus on two forms of family closeness was guided by theory, but there are a number of other family factors (e.g., family emotional expressiveness, disengagement) that may also inform our understanding of the pathways of risk experienced by children from highly unstable homes. Moreover, exploring parent (e.g., sensitivity) and child (e.g. reactivity to family events) characteristics as explanatory mechanisms for the moderating effects of family cohesion and enmeshment is an important future direction for research. Finally, the modest effect sizes for the interaction findings underscore that other family and child characteristics may amplify or reduce the risk posed by maternal relationship instability.

In summary, findings from our multi-method, multi-informant, longitudinal study indicated that both family cohesion and enmeshment moderated the link between maternal relationship instability and children’s externalizing problems over time in distinct ways. On the one hand, maternal relationship instability predicted increases in externalizing problems only when cohesion was low. On the other hand, maternal relationship instability predicted increases in externalizing problems only when enmeshment was high. Furthermore, enmeshment predicted decreases in externalizing problems at low instability but increases in externalizing problems at high instability. Although replication and extension of our findings is necessary before we can offer definitive clinical recommendations, our results provide some empirical support for the value of incorporating a family systems approach to interventions aimed at couples and parenting (e.g., Cowan & Cowan, 2006; Cowan, Cowan & Heming, 2005). Targeting the family as a whole in a way that enhances cohesiveness has the potential to not only reduce young children’s risk for behavior problems, but also alter their vulnerability to other family stressors (i.e., instability of maternal intimate relationships).

Acknowledgments

This research was supported by the Eunice Shriver Kennedy National Institute of Child Health and Human Development (R01 HD065425) awarded to Patrick T. Davies and Melissa L. Sturge-Apple. The authors are grateful to the children, parents, teachers, and community agencies who participated in this project. We would also like to thank Mike Ripple, the Mt. Hope Family Center Staff, and the graduate and undergraduate students at the University of Rochester who assisted on this project.

Footnotes

Author’s Note: This study was presented as a poster at the 2017 Society for Research in Child Development biennial conference and as a part of two research talks given at the Mt. Hope Family Center and SUNY Potsdam.

References

  1. Ablow J, Measelle JR, Kraemer HC, Harrington R, Luby J, Smider N, Kupfer DJ. The MacArthur Three-City Outcome Study: Evaluating multi-informant measures of young children’s symptomatology. Journal of the American Academy of Child and Adolescent Psychiatry. 1999;38:1580–1590. doi: 10.1097/00004583-199912000-00020. [DOI] [PubMed] [Google Scholar]
  2. Ackerman BP, Brown ED, D’Eramo KS, Izard CE. Maternal relationship instability and the school behavior of children from disadvantaged families. Developmental Psychology. 2002;38:694–704. doi: 10.1037/0012-1649.38.5.694. [DOI] [PubMed] [Google Scholar]
  3. Ackerman BP, Brown ED, Izard CE. The relations between persistent poverty and contextual risk and children’s behavior in elementary school. Developmental Psychology. 2004;40:367–377. doi: 10.1037/0012-1649.40.3.367. [DOI] [PubMed] [Google Scholar]
  4. Ackerman BP, Kogos J, Youngstrom E, Schoff K, Izard C. Family instability and the problem behaviors of children from economically disadvantaged families. Developmental Psychology. 1999;35:258–268. doi: 10.1037/0012-1649.35.1.258. [DOI] [PubMed] [Google Scholar]
  5. Aiken LS, West SG. Multiple regression: Testing and interpreting interactions. Thousand Oaks: Sage; 1991. [Google Scholar]
  6. Amato PR. Research on divorce: Continuing trends and new developments. Journal of Marriage and Family. 2010;72:650–666. doi: 10.1111/j.1741-3737.2010.00723.x. [DOI] [Google Scholar]
  7. Bachman H, Coley RL, Carrano J. Maternal relationship instability influences on children’s emotional and behavioral functioning in low-income families. Journal of Abnormal Child Psychology. 2011;39:1149–1161. doi: 10.1007/s10802-011-9535-1. [DOI] [PubMed] [Google Scholar]
  8. Barber BK, Buehler C. Family cohesion and enmeshment: Different constructs, different effects. Journal of Marriage and Family. 1996;58:433–441. doi: 10.2307/353507. [DOI] [Google Scholar]
  9. Belsky J, Pluess M. Beyond diathesis stress: Differential susceptibility to environmental influences. Psychological Bulletin. 2009;135:885–908. doi: 10.1037/a0017376. [DOI] [PubMed] [Google Scholar]
  10. Belsky J, Schlomer GL, Ellis BJ. Beyond cumulative risk: Distinguishing harshness and unpredictability as determinants of parenting and early life history strategy. Developmental Psychology. 2012;48:662–673. doi: 10.1037/a0024454. [DOI] [PubMed] [Google Scholar]
  11. Campbell SB, Shaw DS, Gilliom M. Early externalizing behavior problems: Toddlers and preschoolers at risk for later maladjustment. Development and Psychopathology. 2000;12:467–488. doi: 10.1017/S0954579400003114. [DOI] [PubMed] [Google Scholar]
  12. Carlson MJ, Corcoran ME. Family structure and children’s behavioral and cognitive outcomes. Journal of Marriage and Family. 2001;63:779–792. doi: 10.1111/j.1741-3737.2001.00779.x. [DOI] [Google Scholar]
  13. Cavanagh SE, Huston AC. Family instability and children’s early problem behavior. Social Forces. 2006;85:551–581. doi: 10.1353/sof.2006.0120. [DOI] [Google Scholar]
  14. Cavanagh SE, Huston AC. The timing of family instability and children’s social development. Journal of Marriage and Family. 2008;70:1258–1270. doi: 10.1111/j.1741-3737.2008.00564.x. [DOI] [Google Scholar]
  15. Chase-Lansdale PL, Wakschlag LS, Brooks-Gunn J. A psychological perspective on the development of caring in children and youth: The role of the family. Journal of Adoelscence. 1995;18:515–556. doi: 10.1006/jado.1995.1037. [DOI] [Google Scholar]
  16. Clark KE, Ladd GW. Connectedness and autonomy in parent-child relationships: Links to children’s socioemotional orientation and peer relationships. Developmental Psychology. 2000;36:485–498. doi: 10.1037/0012-1649.36.4.485. [DOI] [PubMed] [Google Scholar]
  17. Cole DA, Maxwell SE. Testing mediational models with longitudinal data: Questions and tips in the use of structural equation modeling. Journal of Abnormal Psychology. 2003;112:558–577. doi: 10.1037/0021-843X.112.4.558. [DOI] [PubMed] [Google Scholar]
  18. Conger RD, Conger KJ. Resilience in Midwestern families: Selected findings from the first decade of a prospective, longitudinal study. Journal of Marriage and Family. 2002;64:361–373. doi: 10.1111/j.1741-3737.2002.00361.x. [DOI] [Google Scholar]
  19. Conger RD, Conger KJ, Martin MJ. Socioeconomic status, family processes, and individual development. Journal of Marriage and Family. 2010;72:685–704. doi: 10.1111/j.1741-3737.2010.00725.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Cowan PA, Cowan CP. Developmental psychopathology from family systems and family risk factors perspectives: Implications for family research, practice, and policy. In: Cicchetti D, Cohen DJ, editors. Developmental Psychopathology. 2. Vol. 1. New York: Wiley; 2006. pp. 530–587. [Google Scholar]
  21. Cowan CP, Cowan PA, Heming G. Two variations of a preventive intervention for couples: Effects on parents and children during the transition to school. In: Cowan PA, Cowan CP, Ablow JC, Johnson VK, Measelle JR, editors. The family context of parenting in children’s adaptation to elementary school. Mahwah, NJ: Lawrence Erlbaum Associates; 2005. pp. 277–312. [Google Scholar]
  22. Cox MJ, Paley B. Families as systems. Annual Review of Psychology. 1997;48:243–267. doi: 10.1146/annurev.psych.48.1.243. [DOI] [PubMed] [Google Scholar]
  23. Cummings EM, Davies PT. Marital conflict and children: An emotional security perspective. New York, NY: Guilford Press; 2010. [Google Scholar]
  24. Davies PT, Coe JL, Martin MJ, Sturge-Apple ML, Cummings EM. The developmental costs and benefits of children’s involvement in interparental conflict. Developmental Psychology. 2015;51:1026–1047. doi: 10.1037/dev0000024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Davies PT, Sturge-Apple ML. Family context in the development of psychopathology. In: Lewis M, Rudolph KD, editors. Handbook of Developmental Psychopathology. Springer; US: 2014. pp. 143–161. [Google Scholar]
  26. Davies PT, Sturge-Apple ML, Bascoe SM, Cummings EM. The legacy of early insecurity histories in shaping adolescent adaptation to interparental conflict. Child Development. 2014;85:338–354. doi: 10.1111/cdev.12119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Davies PT, Winter MA, Cicchetti D. The implications of emotional security theory for understanding and treating childhood psychopathology. Development and Psychopathology. 2006;18:707–735. doi: 10.1017/S0954579406060354. [DOI] [PubMed] [Google Scholar]
  28. Dearing E, Hamilton LC. Contemporary advances and classic advice for analyzing mediating and moderating variables. Monographs of the Society for Research in Child Development. 2006;71:88–104. doi: 10.1111/j.1540-5834.2006.00406.x. [DOI] [Google Scholar]
  29. Donahue KL, D’Onofrio BM, Bates JE, Lansford JE, Dodge KA, Pettit GS. Early exposure to parents’ relationship instability: Implications for sexual behavior and depression in adolescence. Journal of Adolescent Health. 2010;47:547–554. doi: 10.1016/j.jadohealth.2010.04.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Doom JR, Vanzomeren-Dohm AA, Simpson JA. Early unpredictability predicts increased adolescent externalizing behaviors and substance use: A life history perspective. Development and Psychopathology. 2016;28:1505–1516. doi: 10.1017/S0954579415001169. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Doshi JA, Hodgkins P, Kahle J, Sikrica V, Cangelosi J, Neumann PJ. Economic impact of childhood and adult attention deficit/hyperactivity disorder in the United States. Journal of the American Academy of Child and Adolescent Psychiatry. 2012;51:990–1002. doi: 10.1016/j.jaac.2012.07.008. [DOI] [PubMed] [Google Scholar]
  32. Enders CK. A primer on maximum likelihood algorithms available for use with missing data. Structural Equation Modeling. 2001;8:128–141. doi: 10.1207/S15328007SEM0801_7. [DOI] [Google Scholar]
  33. Fomby P, Cherlin AJ. Family instability and child well-being. American Sociological Review. 2007;72:181–204. doi: 10.1177/000312240707200203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Forman EM, Davies PT. Family instability and young adolescent maladjustment: The mediating effects of parenting quality and adolescent appraisals of family security. Journal of Clinical Child and Adolescent Psychology. 2003;32:94–105. doi: 10.1207/S15374424JCCP3201_09. [DOI] [PubMed] [Google Scholar]
  35. Forman EM, Davies PT. Assessing children’s appraisals of security in the family system: The development of the Security in the Family System (SIFS) scales. Journal of Child Psychology and Psychiatry. 2005;46:900–916. doi: 10.1111/j.1469-7610.2004.00385.x. [DOI] [PubMed] [Google Scholar]
  36. Foster EM, Jones DE, The Conduct Problems Prevention Research Group The high costs of aggression: Public expenditures resulting from conduct disorder. American Journal of Public Health. 2005;95:1767–1772. doi: 10.2105/AJPH.2004.061424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Gollob HF, Reichardt CS. Interpreting and estimating indirect effects assuming time lags really matter. In: Collins LM, Horn JL, editors. Best methods for analysis of change: Recent advances, unanswered questions, future directions. Washington DC: American Psychological Association; 1991. pp. 243–259. [Google Scholar]
  38. Grych JH. Marital relationships and parenting. In: Bornstein MH, editor. Handbook of parenting: Vol 4: Social conditions and applied parenting. 2nd. Mahwah, NJ: Lawrence Erlbaum Associates Publishers; 2002. pp. 203–225. [Google Scholar]
  39. Hetherington EM, Stanley-Hagan M. The adjustment of children with divorced parents: A risk and resiliency perspective. Journal of Child Psychology and Psychiatry. 1999;40:129–140. doi: 10.1111/1469-7610.00427. [DOI] [PubMed] [Google Scholar]
  40. Jacobvitz D, Hazen N, Curran M, Hitchens K. Observations of early triadic family interactions: Boundary disturbances in the family predict symptoms of depression, anxiety, and attention-deficit/hyperactivity disorder in middle childhood. Development and Psychopathology. 2004;16:577–592. doi: 10.1017/S0954579404004675. [DOI] [PubMed] [Google Scholar]
  41. Johnson VK, Cowan PA, Cowan CP. Children’s classroom behavior: The unique contribution of family organization. Journal of Family Psychology. 1999;13:355–371. doi: 10.1037/0893-3200.13.3.355. [DOI] [Google Scholar]
  42. Kerig PK. Triangles in the family circle: Effects of family structure on marriage, parenting, and child adjustment. Journal of Family Psychology. 1995;9:28–43. doi: 10.1037/0893-3200.9.1.28. [DOI] [Google Scholar]
  43. Kerig PK. Revisiting the construct of boundary dissolution: A multidimensional perspective. Journal of Emotional Abuse. 2005;5:5–42. doi: 10.1300/J135v05n02_02. [DOI] [Google Scholar]
  44. Lindahl KM, Malik NM. The system for coding interactions and family functioning. In: Kerig P, Lindahl KM, editors. Family observational coding systems: Resources for systemic research. Mahwah, NJ: Lawrence Earlbaum Associates Publishers; 2000. pp. 77–91. [Google Scholar]
  45. Luthar SS, Cicchetti D, Becker B. The construct of resilience: A critical evaluation and guidelines for future work. Child Development. 2000;71:543–562. doi: 10.1111/1467-8624.00164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Maccoby EE, Jacklin CN. The psychology of sex differences. Stanford, CA: Stanford University Press; 1974. [Google Scholar]
  47. Minuchin P. Families and individual development: Provocations from the field of family therapy. Child Development. 1985;56:289–302. doi: 10.2307/1129720. [DOI] [PubMed] [Google Scholar]
  48. Minuchin S. Families and family therapy. Cambridge, MA: Harvard University Press; 1974. [Google Scholar]
  49. National Institute of Child Health and Human Development Early Child Care Network. Fathers’ and mothers’ parenting behavior and beliefs as predictors of children’s social adjustment in the transition to school. Journal of Family Psychology. 2004;18:628–638. doi: 10.1037/0893-3200.18.4.628. [DOI] [PubMed] [Google Scholar]
  50. Olson DH. Circumplex model VII: Validation studies and FACES III. Family Process. 1986;25:337–351. doi: 10.1111/j.1545-5300.1986.00337.x. [DOI] [PubMed] [Google Scholar]
  51. Osborne C, McLanahan S. Partnership instability and child well-being. Journal of Marriage and Family. 2007;69:1065–1083. doi: 10.1111/j.1741-3737.2007.00431.x. [DOI] [Google Scholar]
  52. Richmond MK, Stocker CM. Associations between family cohesion and adolescent siblings’ externalizing behavior. Journal of Family Psychology. 2006;20:663–669. doi: 10.1037/0893-3200.20.4.663. [DOI] [PubMed] [Google Scholar]
  53. Simpson JA, Griskevicius V, Kuo SI, Sung S, Collins WA. Evolution, stress, and sensitive periods: The influence of unpredictability in early versus late childhood on sex and risky behavior. Developmental Psychology. 2012;48:674–686. doi: 10.1037/a0027293. [DOI] [PubMed] [Google Scholar]
  54. Sturge-Apple ML, Davies PT, Cummings EM. Typologies of family functioning and children’s adjustment during the early school years. Child Development. 2010;81:1320–1335. doi: 10.1111/j.1467-8624.2010.01471.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Sturge-Apple ML, Davies PT, Cicchetti D, Fittoria MG. A typology of interpartner conflict and maternal parenting practices in high-risk families: Examining spillover and compensatory models and implications for child adjustment. Development and Psychopathology. 2014;26:983–998. doi: 10.1017/S0954579414000509. [DOI] [PubMed] [Google Scholar]
  56. Wauchope B, Straus MA. Physical punishment and physical abuse of American children: Incidence rates by age, gender, and occupational class. In: Straus MA, Gelles RJ, editors. Physical violence in American families: Risk factors and adaptations to violence in 8,145 families. New Brunswick, NJ: Transaction Publishers; 1990. pp. 133–148. [Google Scholar]
  57. Widaman KF, Ferrer E, Conger RD. Factorial invariance within longitudinal structural equation models: Measuring the same construct across time. Child Development Perspectives. 2010;4:10–18. doi: 10.1111/j.1750-8606.2009.00110.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES