Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jun 1.
Published in final edited form as: Fam Process. 2021 Aug 5;61(2):841–857. doi: 10.1111/famp.12703

Family and Individual Risk Factors for Triangulation: Evaluating Evidence for Emotion Coaching Buffering Effects

Devin M McCauley 1,2, Gregory M Fosco 1,2
PMCID: PMC8816974  NIHMSID: NIHMS1737310  PMID: 34355393

Abstract

Adolescents who are triangulated into interparental conflict are at increased risk for psychological maladjustment. However, little is known about factors that may predict family risk for triangulating adolescents, or protective factors that can off-set this risk. In this study, we conducted longitudinal tests of family, parent, and adolescent factors that might predict increases in triangulation over time. The sample included 174 adolescents and their mother figures from two-parent families (58% female; Mage = 14.75 years) who provided data on two occasions, six months apart. Hierarchical linear regression models evaluated family, parent, and adolescent risk factors for triangulation into interparental conflict, and subsequently parent’s emotion coaching and adolescent gender as potential moderators of risk for triangulation. Findings revealed that low family cohesion, parent depression, and adolescent difficulties with emotion regulation represented risks for triangulation. Parent emotion coaching moderated the association between low interparental love and triangulation differentially based on adolescent gender.

Triangulation generally refers to the involvement of a third party to defuse dyadic conflict; however, most attention has been given to the triangulation of children in their parents’ conflicts (Buchanan & Waizenhofer, 2001; Minuchin, 1974). Adolescents who are frequently triangulated into hostile interparental conflict (IPC) are at heightened risk for psychological and social maladjustment outcomes, including internalizing problems (Buchanan & Waizenhofer, 2001; Grych, Raynor, & Fosco, 2004), externalizing problems (Fosco, Lippold, & Feinberg, 2014; Gerard, Buehler, Franck, & Anderson, 2005) and greater conflict in relationships with parents (Fosco & Grych, 2010; Fosco et al., 2014) and peers (Buehler, Franck, & Cook, 2009).

Despite documented risks for child development, triangulation remains a fairly common response to IPC (Fosco & Grych, 2008; Jenkins, Smith, & Graham, 1989). This may be explained by the function triangulation serves for the broader family system. Structural and Bowen family systems theories suggest that couples who struggle to resolve chronic arguments may draw in their child as a means of redistributing tension away from the couple relationship, thereby creating temporary stability (Bowen, 1978; Minuchin, 1974). Although displacing conflict tension may provide temporary relief to parents, reliance on children to resolve IPC may undermine parents’ ability to directly address and resolve relational conflict, precipitating future conflicts (McCauley, Sloan, Xia, & Fosco, 2021; Minuchin, 1974). Rigid patterns of routinely triangulating children in IPC are prognostic of family stress and anxiety (Bowen, 1978).

Understanding the factors that precipitate a shift to more frequent, patterned triangulation represents an important yet underexplored dimension of family research. A hallmark indication of risk for triangulation is chronic, hostile, and poorly resolved IPC (e.g., Camisasca, Miragoli, & Di Blasio, 2019; Fosco & Grych, 2008, 2010; Grych et al., 2004). Beyond this risk, the interdependence inherent in family systems suggests that triadic conflict processes are likely shaped by vulnerabilities and characteristics throughout the family (Cox & Paley, 1997). Studies exploring additional sources of risk within the family are few, yet have identified parent-adolescent alliances (Grych et al., 2004) and parenting stress (Camisasca et al., 2019) as risk factors. Additional sources of risk remain undiscovered, representing a notable gap in family research. Thorough documentation of risk and protective factors for triangulation represents a critical step in progressing family science and identifying avenues of prevention for family programs. Accordingly, this study was designed to: a) identify risk factors for triangulation across multiple family domains, b) evaluate parent emotion coaching as a potential protective factor, and c) explore gender differences in these processes.

Which Families are Vulnerable to Triangulating Adolescents?

As a framework for conceptualizing predictors of triangulation, we propose that family, parent, and adolescent characteristics may increase families’ risk for triangulation.

Family risk for triangulation.

Family cohesion – the degree of support, connectedness, and emotional bonding within the family (Olson, Waldvogel, & Schlieff, 2019) – may shape how families engage in IPC. Low cohesion is associated with increased tension, hostility, and criticism (Fosco, Caruthers, & Dishion, 2012), factors which may undermine parents’ ability to contain and resolve conflict within the interparental dyad. Low family cohesion also exacerbates adolescents’ negative interpretations of IPC (Lindahl & Malik, 2011), which may in turn motivate their desire to intervene to help mitigate conflict tension (Fosco & Grych, 2010). Conversely, IPC may be less distressing to parents and adolescents in highly cohesive families, as norms of supportiveness and connectedness may mitigate conflict intensity and facilitate healthy conflict resolution within the interparental dyad (Driver & Gottman, 2004).

Couple relationships play a central role in guiding overall family functioning, including parenting quality and managing family conflict (Feinberg, 2003; Minuchin, 1974). Beyond IPC, positive qualities within the couple relationship - love, positivity, and commitment - may also shape risk for triangulation. Couples’ warmth and affection are key predictors of both marital quality and stability (Gottman, Coan, Carrere, & Swanson, 1998) and thus low levels may contribute to poor conflict management (e.g., verbal hostility, withdrawal). Poor relationship bonds, combined with poor conflict management, may in turn lead parents to displace hostility from their relationship toward adolescents (Cox, Paley, & Harter, 2001) or to recruit adolescents for support not adequately provided within the marital relationship (Kerig, 2005).

Parent risk for triangulation.

Parent depression may increase risk for triangulation by shaping a family’s interaction patterns around IPC. Parents struggling with depression tend to report lower marital satisfaction (Kronmüller et al., 2011), and are more likely to engage in withdrawal, avoidance, and verbal hostility during conflicts with their spouses (Marchand & Hock, 2000). This combination of factors may yield poorly resolved conflict scenarios in which distressed parents turn to their adolescent children for emotional support (Kerig, 2005). Furthermore, adolescents witnessing their parents’ hostile interactions may feel compelled to intervene on behalf an emotionally vulnerable parent in a protective role (Van Parys & Rober, 2013).

A parent’s general angry mood may also increase risk for triangulation by contributing to conflict dynamics. Parents’ general anger may heighten the intensity and hostility of IPC, while also making it more difficult for parents to reach effective resolution – factors which increase the likelihood that children will become drawn into conflict (e.g., Fosco & Grych, 2008). Parents’ anger may also shape the way adolescents respond to hostile IPC. Conflicts in which anger remains unresolved, in particular, are more likely to solicit adolescents’ intervention (Cummings, Ballard, El-Sheik, & Lake, 1991). High levels of anger in one parent may motivate adolescents to lend support to the opposing parent in conflict, paving the way for parent-adolescent alliances.

Adolescent risk for triangulation.

Interparental conflict research emphasizes that children may take an active role during conflict episodes by directly intervening or behaving disruptively to distract parents from their dispute (e.g., Schermerhorn et al., 2007). Adolescents’ individual characteristics are therefore likely to shape their behavioral responses to IPC. In particular, adolescent difficulties with emotion regulation may represent one risk factor for triangulation. Intense and unresolved IPC tends to elicit emotional distress in adolescents (Fosco & Grych, 2008). Adolescents who are able to effectively regulate their emotions may successfully abstain from engaging directly with feuding parents. However, those who typically experience greater difficulty may respond disruptively when distressed by IPC (e.g., arguing with parents) and risk becoming entangled in conflict.

Adolescents high in dispositional anxiety may also be at risk for intervening in IPC. Children’s intervention in parents’ conflicts is often motivated by concerns that conflict will threaten personal safety or disrupt the integrity of the family (Davies et al., 2015). Highly anxious adolescents are more likely to interpret threat in social scenarios and anticipate negative outcomes (Vassilopoulos & Banerjee 2008), suggesting they may experience heightened distress regarding the potential consequences of hostile IPC (e.g., divorce). These adolescents may anticipate their parents’ disputes and intervene in a mediational role in an attempt to mitigate conflict intensity and prevent negative consequences to the family (Davies et al., 2015).

Parental Emotion Coaching May Buffer Risk for Triangulation

In addition to identifying risk factors for triangulation, we evaluated parent emotion coaching as a practice that might mitigate family and individual vulnerabilities in a protective fashion. Emotion coaching refers to parents’ sensitive and constructive orientation toward their children’s experience and expression of emotions, and is a core component of a broader meta-emotion philosophy (Katz; Maliken, & Stettler, 2012). Parents who practice emotion coaching tend to be highly attuned to their children’s emotions, validate and support them around emotional experiences, help them label and understand their emotions, and teach strategies for coping with emotional experiences (Katz et al., 2012).

There are at least two ways in which parental emotion coaching may mitigate risk for triangulation. First, parent emotion coaching is positively associated with adolescents’ emotion regulation (Shortt et al., 2010), which may reduce adolescents’ tendencies to intervene or behave disruptively when emotionally distressed by IPC. Parents who regularly engage in emotion coaching may facilitate their adolescents’ coping strategies for distress caused by IPC, which may be particularly beneficial for adolescents who typically struggle with dysregulated emotions or anxiousness. Second, emotion coaching is a process by which parents maintain strong attunement to adolescents’ emotions as well as their own (Katz et al., 2012). Parents who consistently use emotion coaching parenting may be more aware of distress experienced by their adolescent during IPC episodes and be more likely to compartmentalize IPC within the couple relationship for the sake of their child. Even in the context of a poor couple relationship or low family cohesion, these parents may be particularly motivated to contain and resolve IPC within the parental dyad, rather than risk further distress to their adolescents by enlisting their support or otherwise permitting their involvement.

Adolescent Gender, Emotion Coaching, and Triangulation

The role of adolescent gender in understanding the impact of IPC and triangulation remains inconclusive, prompting calls for further investigation of how specific IPC processes and outcomes vary by gender (van Eldik et al., 2020). Prior studies note similar rates of triangulation for sons and daughters (e.g., Fosco & Grych, 2010; Gerard et al., 2005), yet others suggest that specific patterns of triangulation differ by gender (Bell, Bell, & Nakata, 2011). Girls are also thought to experience greater emotional distress in response to IPC (van Eldik et al., 2020), which may be attributable to cultural and familial expectations on adolescent girls to prioritize harmonious social relationships above individualistic pursuits (Davies & Lindsay, 2004). Furthermore, parents often socialize emotion differently based on child gender, for example, discouraging boys’ expressions of fear and sadness but encouraging the same emotions when displayed by girls (Garside & Klimes-Dougan, 2002). Due to potential gender differences in experiences of IPC and triangulation, and the central role of emotion coaching in the present study, we evaluated whether differences in how boys and girls experience risk and protective processes for triangulation were evident in our sample.

The Current Study

The goal of this study was to identify factors beyond IPC that place families at risk for triangulation and to evaluate parent emotion coaching as a potential buffer of this risk. We also evaluated gender differences in risk and protective factors. Accordingly, we tested hypotheses in three domains: a) Family Factors: we hypothesized that lower levels of family cohesion and interparental love would predict increases in triangulation over time, as low levels of these qualities may hinder a family’s ability to contain and resolve IPC within the interparental dyad. However, we hypothesized that high levels of parent emotion coaching would mitigate these sources of risk. b) Parent Factors: we hypothesized that higher levels of parent angry mood and depression would predict increases in triangulation over time by compromising parents’ ability to directly resolve conflicts with their partners. We also hypothesized that parent emotion coaching might benefit parents in a way that would mitigate these individual sources of risk. c) Adolescent Factors: We expected that higher levels of adolescent dysregulated emotion and anxious mood would predict increases in triangulation over time by elevating the likelihood that adolescents would intervene or behave disruptively in response to IPC. We hypothesized that parent emotion coaching would provide adolescents with the skills to overcome individual risk factors, representing a protective factor. Finally, due to socialized gender differences, particularly in domains of family relationships and emotion socialization, we evaluated adolescent gender as a potential moderator of risk factors and triangulation in an exploratory fashion within each model.

Method

Procedure

Data in this study came from the Penn State Family Life Optimizing Well-being (FLOW) study, in which adolescents and their primary caregiver completed assessments about family functioning and emotions. School principals emailed parents a website link which included study details and a brief eligibility screener. Families were eligible to participate if adolescents lived in one, two-caregiver household, were in 9th or 10th grade, and fluent in English. Following parent/adolescent consent/assent, each were sent separate links to T1 surveys and instructed to complete them independently and privately. Six months after T1, adolescents and parents were emailed a link to complete T2 follow-up questionnaires. Prior work has demonstrated meaningful change in triangulation over six months in families of adolescents (Fosco & Grych, 2010). Parents and adolescents were compensated with $25 at T1 and $35 at T2 in gift cards.

Participants

The initial T1 sample included one adolescent and one parent from 201 families in which the majority of participating parents were female (95%). Adolescent gender was an important moderator in this study; therefore, only families in which the participating parent was female were included in study analyses (n=190) to avoid potential confounds introduced by parents’ gender differences (e.g., depressive symptoms and conflict resolution strategies, Marchand & Hock, 2000). Also, families in which the two primary caregivers were not in a romantic relationship (e.g., a household with a mother and an aunt as caregivers; n = 6) were trimmed from the sample. Attrition for adolescents at T2 was 5.3% (n= 10), yielding a final analytic sample of 174 families (N = 174). A Little’s MCAR test indicated that data were missing completely at random (χ2(35) = 30.32, p = .69). Further, we created a dichotomous variable at T2 to indicate attrition (Miller & Wright, 1995). No demographic or predictor variables at T1 predicted attrition. Thus, no bias is expected due to differential attrition. The analytic sample included 58% female (n = 101) and 42% male (n = 73) adolescents ages 12 to 17 years old (Mage = 14.8, SDage = 0.8). Parents reported that their adolescents were White/European American (84%), Black/African American (2%), Native American (<1%), Asian American (3%), Hispanic/Latino/a (1%), multiple races (7%), or missing (<1%). Parents self-identified as White/European American (90%), Black/African American (3%), Native American (<1%), Asian American (2%), Hispanic or Latino/a (1%), multiple races (3%), or missing (1%). Parent ages ranged from 30 to 67 years (Mage = 43.2, SDage = 6.3). Most parents reported they were the mother of the adolescent in the study (98%), with other parents reporting they were either a stepmother (1%), aunt (<1%), or foster mother (<1%). As noted, caregivers within the analytic sample were in romantic relationships with their partners (Myears = 18.0, SDyears = 6.6), and most reported being married (93%). Household annual income ranged from less than $10,000 to over $125,000 (Median: $70,000-$79,000). Parents’ median education level was junior college or an associate’s degree.

Measures

Triangulation.

At T1 and T2, adolescents rated five items assessing their triangulation into interparental conflict from the triangulation subscale of the Children’s Perceptions of Interparental Conflict Scale (CPIC; Grych, Seid, & Fincham, 1992; Grych et al., 2004). This scale is widely used and correlates with both observational and parent reports of triangulation (Fosco & Grych, 2008; Lindahl, 1998). Questions are framed in terms of typical experiences during parental arguments to capture common patterns of triangulation into IPC including being caught in the middle of conflict and experiencing pressure to choose sides. A sample item was “I feel caught in the middle when my parents argue”. Items were rated from 1 (Strongly Disagree) to 5 (Strongly Agree). Cronbach’s alphas were 0.86 (T1) and 0.90 (T2).

Interparental conflict.

At T1, interparental conflict was measured by averaging adolescents’ responses to nine items from the CPIC conflict properties subscale (Grych et al., 1992), assessing the frequency, intensity, and (poor) resolution of conflict in the home. Items were rated from 1 (Strongly Disagree) to 5 (Strongly Agree) (α = 0.89). A sample item was “My parents get really mad when they argue”. This widely-used scale is correlated with observational and parent reports of IPC (Fosco & Grych, 2008; Grych et al., 1992). Four items were reverse coded so that higher values indicated higher levels of interparental conflict.

Interparental love.

At T1, parents’ love felt toward their partners during the last two months was measured by averaging parents’ responses to five items from the Love subscale of the Love and Relationships scale (Braiker & Kelley, 1979). Items were selected based on having strong factor loadings on the original scale and for capturing a breadth of positive qualities including love, belonging, closeness, attachment, and commitment. Items were rated from 1 (Not at All) to 5 (Very Much) (α = .96). Example items were “to what extent do you love your partner at this stage?” and “how committed do you feel toward your partner?”.

Family cohesion.

At T1, adolescents rated the five-item validated short form of the Family Environment Scale (Bloom, 1985) to indicate family cohesion over the last month. Items were rated from 1 (Almost never) to 5 (Almost Always) (α = 0.84). A sample item was “family members got along really well”. Higher values indicated higher family cohesion.

Parent general angry mood.

At T1, parent angry mood was measured as the mean of two items from the Profile of Mood States - Adolescents, a measure of mood validated for use with adults and adolescents (POMS-A; Terry, Lane, & Fogarty, 2003). Parents rated the extent to which they felt angry during the last month from 1 (None of the Time) to 5 (All of the Time) (α = 0.67). A sample item was “How much of the time in the last month did you feel angry?” The two items were moderately correlated, (172) = 0.50, p < .01.

Parent depression.

At T1, parents indicated their depression during the past week by responding to the validated 20-item Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977). Items were rated from 1 (Rarely) to 4 (Almost All the Time) (α = 0.90) and were averaged to create a score reflecting overall parent depression. A sample item was “I could not get going.” Though depression is often classified categorically, evidence supports its use as a continuous construct (Hankin, Fraley, Lahey, & Waldman, 2005); however, in this sample, 14% of parents exceeded clinical threshold for depression (Radloff, 1977).

Adolescent dispositional anxiety.

At T1, adolescent anxious mood was measured as the mean of two items from the Profile of Mood States-Adolescents (POMS-A; Terry et al., 2003). Adolescents rated the extent to which they felt anxious during the last month from 1 (None of the Time) to 5 (All of the Time) (α = 0.67). A sample item was “How much of the time do you feel worried?”. The items were correlated, (172) = 0.49, p < .01.

Adolescent difficulties with emotion regulation.

At T1, adolescents completed the 18-item short form of the Difficulties of Emotion Regulation Scale, validated for use with adolescents and adults to assess poor emotional awareness, acceptance, and control (Kaufman, Xia, Fosco, Yaptangco, Skidmore, & Crowell, 2015). Items were rated from 1 (Almost Never) to 5 (Almost Always) (α = 0.90). A sample item was “when I’m upset, I lose control over my behaviors”. Higher values reflected more frequent difficulty in regulating emotions.

Emotion coaching.

At T1, the participating parents responded to six items that assessed the previously established criteria for emotion coaching – awareness, acceptance, and coaching (Katz et al., 2012). Items were rated from 1 (Not at All) to 5 (Extremely) (α = 0.86), and were averaged to create an overall score for emotion coaching. A sample item was “I look for opportunities to help my child learn about his/her emotions”.

Data Analysis

Analyses were run using R version 1.1.463. As a first step, we computed correlations and descriptive statistics. Variables were then standardized to have a mean of zero. Then, to evaluate hypothesized relations, we computed three sets of hierarchical regression models – for family, parent, and adolescent domains separately in order to minimize Type II errors for two and three-way interaction terms (Beauchaine, Neuhaus, Brenner, & Gatzke-Kopp, 2008). In step 1, hypothesized main effects in each domain (family, parent, or adolescent factors) were tested. In the second step, two-way interactions were included for parent emotion coaching and adolescent gender. We tested parent emotion coaching moderation in step 2a, and adolescent gender as a moderator in step 2b (to minimize Type II error risk). In step 3, three-way interactions were tested between hypothesized main effect variables, emotion coaching, and gender to evaluate whether emotion coaching buffering effects varied by adolescent gender. Step 3 interactions were computed one at a time, and identified as separate steps (e.g., 3a, 3b).

Analyses used autoregressive methods in which T2 triangulation was regressed on T1 triangulation in all models. Thus, any other predictors in the model (e.g., family cohesion) that are significantly associated with T2 triangulation help explain residual variance in triangulation, after accounting for T1 levels. This process – referred to as residualized change – is commonly interpreted as indicating change in the T2 variable attributable to T1 variables (Castro-Schilo & Grimm, 2018). Although this method is not a direct statistical assessment of change, we refer to our analyses as predicting change in triangulation for clarity and readability. All models included interparental conflict, adolescent gender (coded 0 = female, 1 = male), family income, and emotion coaching as covariates.

Results

Correlations and descriptive statistics for all variables are reported in Table 1. In general, correlations were in the expected direction; family cohesion, interparental love, and emotion coaching were all negatively correlated with T2 triangulation. Parent depression, parent anger, adolescent difficulty in emotion regulation, and IPC were all positively correlated with T2 triangulation. Triangulation was moderately correlated at T1 and T2. Descriptive statistics indicate moderate levels of interparental conflict (M = 2.36, SD = 0.86) and relatively low levels of triangulation (M = 1.85, SD = 0.84), yet are consistent with samples used in other studies of triangulation (e.g., Camisasca et al., 2019; Fosco & Bray, 2016).

Table 1.

Correlations, Means, and Standard Deviations.

Variables 1 2 3 4 5 6 7 8 9 10 11 12
1. Gender -
2. Income .05 -
3. IPC, T1 −.09 −.04 -
4. Cohesion, T1 −.03 .20** −.57** -
5. IP_Love, T1 −.03 −.03 −.27** .19* -
6. P_Depression, T1 .06 −.19* .32** −.28** −.34** -
7. P_Anger, T1 .08 .05 .31** −.20** −.20** .38** -
8. Emo_Coaching, T1 −.07 .00 −.11 .15* .16* −.15* −.19** -
9. A_Anxious, T1 −.16* .14 .22** −.22* −.08 .10 .11 −.01 -
10. A_DERS, T1 −.07 −.10 .38** −.40** −.17* .36** .21** −.03 .47** -
11. Triangulation, T1 .01 −.03 .62** −.37** −.25** .20** .28** .00 .28** .37** -
12. Triangulation, T2 .06 .00 .38** −.34** −.27** .29** .33** −.15* .27** .37** .57** -
Mean - 8.00 2.36 4.19 4.42 1.46 2.23 4.23 2.10 2.07 1.85 1.59
SD - 3.10 0.86 0.68 0.89 0.42 0.60 0.67 0.91 0.72 0.84 0.78
Open in a new tab

Note. N = 174.

*

= p < 0.05

**

= p < 0.01; T1 = time 1, T2 = time 2; IPC = interparental conflict; Cohesion = low family cohesion, IP_Love = low interparental love; P_Depression = parent depression; P_Anger = parent angry mood; Emo_Coaching = parent emotion coaching; A_Anxious = adolescent dispositional anxiety; A_DERS = adolescent difficulties with emotion regulation; SD = standard deviation. Means and standard deviations are presented prior to z-score transformations for interpretability.

Family Models

Results from model 1 evaluating family-level risk and protective factors for adolescent triangulation at T2 are shown in Table 2. In step 1, family cohesion was marginally associated with decreases in triangulation, β5 = −0.15, p = .05, suggesting that adolescents in highly cohesive families experienced reduced risk for becoming involved in IPC over time. Couple love was not associated with changes in triangulation.

Table 2.

Family Predictors of Changes in Triangulation at T2

Block Predictors b SE b β
1 Main effects
T1 Triangulation (β1) .54** .08 .54
Gender (β2) .03 .10 .04
Income (β3) .01 .01 .04
IPC (β4) −.08 .08 −.08
Cohesion (β5) −.18† .09 −.15
IP_Love (β6) −.10 .06 −.11
E_Coach (β7) −.12 .07 −.11
R2 .37**
2 2-way interactions
a IPC* E_Coach (β8) .16 .11 .12
Cohesion*E_Coach (β9) .16 .12 .09
IP_Love*E_Coach (β10) .29** .08 .22
R2 .43**
b IPC*Gender (β8) −.06 .15 −.07
Cohesion*Gender (β9) −.07 .18 −.06
IP_Love*Gender (β10) .18 .11 .21
E_Coach*Gender (β11) −.29 .15 −.25
R2 .40**
3 3-way interactions
a IP_Love*E_Coach*Gender (β10) −.60** .16 .43
R2 .48**
b Cohesion*E_Coach*Gender (β10) −.13 .22 −.06
R2 .38**
Open in a new tab

Note. N = 174.

†

= p = .05

*

= p < 0.05

**

= p < .01; T1 = Time 1; IPC = interparental conflict; IP_Love = interparental love; E_Coach = parent emotion coaching.

In step 2a, emotion coaching significantly moderated the association between interparental love and T2 triangulation, β10 = 0.22, p < .01. This interaction was not probed because it was qualified by a three-way interaction in step 3a. None of the other two-way interactions were statistically significant. In step 3a, the three-way interaction between interparental love, emotion coaching, and adolescent gender was statistically significant, β12 = 0.43, p < .01, indicating gender differences in the extent to which emotion coaching moderated the association between love and change in T2 triangulation. As depicted in Figure 2, tests of simple slopes for interparental love and coaching at high (+1 SD), mean, and low (−1 SD) values of parent emotion coaching were conducted separately for adolescent girls and boys. For adolescent girls, a graded relationship between interparental love and change in T2 triangulation was evident as a function of emotion coaching. Interparental love was most negatively associated with change in T2 triangulation at high levels of emotion coaching (+1 SD; β = −.27, p < .05). At average levels of emotion coaching, interparental love was still significantly associated with change in T2 triangulation, although to a lesser degree (β = −.19, p < .05). At low levels (−1 SD) of emotion coaching the relation was no longer statistically significant (β = −.11, p = .33).

Figure 2.

Figure 2.

Open in a new tab

N = 174 * = p < 0.05; ** = p < .01. Three-way interaction between interparental love, parent emotion coaching, and adolescent gender. Parent emotion coaching moderates the association between interparental love and changes in triangulation at T2 differently based on adolescent gender.

The pattern of results differed for boys. Interparental love was most negatively associated with change in T2 triangulation at low levels of emotion coaching (β = −.24, p < .05) and most positively associated with change in T2 triangulation at high levels of emotion coaching (β = .48, p < .01). At mean levels of emotion coaching, the relation between love and triangulation was not statistically significant (β = .12, p = .19). In step 3b, the three-way interaction between family cohesion, emotion coaching, and adolescent gender was found to be non-significant.

Parent Models

Models evaluating parent risk factors followed the same procedure as above. Across all steps, only one statistically significant predictor emerged. Parent depression was marginally associated with increases in triangulation over time, β5 = 0.14, p = 0.05, suggesting that families with parents experiencing heightened depressed mood are at greater risk for increasing levels of triangulation. No other predictors or interaction terms were statistically significant.

Adolescent Models

The third set of models, evaluating adolescent risk factors for change in triangulation, were conducted the same as above. In step 1, adolescent difficulties with emotion regulation predicted statistically significant increases in triangulation, β6 = 0.17, p < 0.05, but adolescent dispositional anxiety did not. Parent emotion coaching was associated with decreases in triangulation, β7 = −0.14, p < 0.05. Parent emotion coaching did not moderate any effects in step 2a; however, the two-way interaction for adolescent gender and emotion coaching was statistically significant, β = −0.26, p < 0.05. Examination of simple slopes revealed that emotion coaching was associated with decreases in triangulation for boys (β = −.28, p < 0.01) but not for girls (β = −.02, p = 0.80). No other interactions were statistically significant.

Discussion

Despite established harmful effects of triangulation on adolescent development (e.g., Buchanan & Waizenhofer, 2001; Fosco et al., 2014), family research has yet to adequately identify the breadth of factors that precipitate a family’s shift toward triadic responses to IPC. The current study addressed this gap in the literature by evaluating potential family, parent, and adolescent risk factors for increasing triangulation, as well as considering parent emotion coaching as a potential buffer for these risks. Adolescent gender differences in risk and protective factors were also evaluated.

Key Predictors of Triangulation

In the family domain, low family cohesion increased risk for triangulation over time, offering partial support for the family hypothesis. This finding suggests that a family’s strategies for managing IPC are contextualized by broader family climate. In low-cohesion families, adolescents increased likelihood of being drawn in to IPC may reflect the family’s inability to contain and resolve dyadic conflict. Adolescents are also more likely to perceive IPC as threatening within distressed family environments (Lindahl & Malik, 2011), which may in turn prompt intervention in an attempt to diminish IPC severity. Interparental love, however, was not significantly associated with change in triangulation.

Parent depression also emerged as a risk factor for increases in adolescent triangulation, offering partial support for the parent hypothesis. Parent depression undermines effective conflict resolution and marital satisfaction (Kronmüller et al., 2011; Marchand & Hock, 2000), leaving adolescents to tend to the emotional needs of their depressed parent (Van Parys & Rober, 2013). In this way parental depression may set the stage for formation of cross-generational alliances between the adolescent and depressed parent (Kerig, 2005), positioning the adolescent to intervene on behalf of the depressed parent in a protective role during IPC. Conversely, parents’ angry mood did not predict significant changes in triangulation.

Adolescent difficulties with regulating emotion represented a third risk factor for increasing triangulation, supporting the adolescent hypothesis. Adolescents who typically struggle to regulate emotions may be at a greater risk for dysregulated responses to conflict episodes, such as engaging in behaviors that distract parents from conflict (Minuchin, 1974), or reciprocating negative affect expressed by parents (Katz & Hunter, 2007). Similarly, adolescents who struggle with emotion regulation are at greater risk for externalizing behaviors (Shortt et al., 2010), placing them at risk for being targeted as scapegoats for IPC (Minchin, 1974). Adolescents’ anxious mood, however, was not a significant predictor of changes in triangulation.

The Role of Parental Emotion Coaching in Reducing Risk for Triangulation

Parent reported emotion coaching emerged as both a main effect and buffer of risk for adolescent triangulation into IPC. Emotion coaching exhibited a direct association with decreases in triangulation over time in the statistical model evaluating adolescent individual risk characteristics. However, this link may be contingent on other factors considered, as emotion coaching did not predict changes in triangulation in the parent risk factor model, and moderated interparental love in the family-level model. Emotion coaching may reduce risk for triangulation above and beyond an adolescent’s ability to regulate emotions in general, potentially by providing adolescents with specific social-emotional skills to implement during IPC. However, emotion coaching may not help adolescents to navigate risk related to parent characteristics, such as parent depression.

In the family model, parents’ emotion coaching also moderated the association between interparental love and adolescent triangulation, with results differing for girls and boys. Girls’ risk for increases in triangulation at T2 was generally negatively associated with interparental love. That is, in families with low levels of interparental love, girls reported increases in triangulation over time, whereas girls in families with high levels of interparental love did not. Emotion coaching qualified this relation, such that adolescent girls who received more emotion coaching benefitted more from interparental love. However, it is worth noting that girls who experienced both low levels of interparental love and high levels of emotion coaching were the most likely to exhibit increases in triangulation over time. Perhaps for adolescent girls, emotion coaching facilitates greater emotion understanding, which may motivate their involvement when they perceive problems in the interparental relationship.

For boys, high levels of interparental love were also indicative of minimal change in triangulation over time, similar to findings for girls. However, in the context of low interparental love, parent emotion coaching was a key factor for boys’ triangulation risk. Boys in families with low levels of interparental love and low parent emotion coaching exhibited increases in triangulation over time. This finding suggests that low levels of emotion coaching may leave adolescent boys without the social-emotional awareness or skills to avoid becoming entangled in their parents’ disputes, particularly in families characterized by poor love and commitment between parents. On the other hand, boys in families with low interparental love and high emotion coaching exhibited decreases in triangulation over time, suggesting that parent emotion coaching may play a valuable role in protecting adolescent boys from being drawn into disputes between parents with particularly troubled relationships.

Taken together, our findings pertaining to interparental love and emotion coaching highlight three issues related to family risk for triangulation. First, for both boys and girls, high levels of interparental love corresponded to no changes in triangulation over time, underscoring the role of a loving couple relationship for family stability. Second, emotion coaching represents a valuable protective factor for boys in families with low levels of interparental love, and may speak to the importance of sensitive, engaged parenting for guiding adolescent boys within distressed families. Within such families, parents who practice emotion coaching may be taking extra precautions to ensure their sons do not become entangled in potentially distressing parental disputes. Third, processes of involvement in IPC may differ for boys and girls (e.g., Bell et al., 2001). Given that adolescent daughters may experience pressure to take on caregiving roles in distressed families (Davies & Lindsay, 2004; Kerig, 2005), emotion coaching may not provide a protective function for girls in the context of low interparental love, as they may become involved in conflict due to gendered role expectations rather than their own dysregulated emotion. In these cases, improving stability within the interparental relationship may be more effective for mitigating girls’ risk for triangulation.

Implications for Intervention

The range of risk factors for triangulation throughout the family suggest multiple viable avenues through which educational programming may curtail the progression of family triangulation to clinical levels. Notably, lower levels of interparental love combined with low emotion coaching represented a risk factor for increasing triangulation, particularly for boys. Helping parents of boys to learn and practice emotion coaching strategies may be an ideal component to be included in family-based prevention settings that address couple relationships and parenting strategies (e.g., Strengthening Families Program, Kumpfer, Molgaard, & Spoth, 1996). For girls, however, efforts to support the interparental relationship may represent an effective approach to mitigating risk for triangulation. Furthermore, parent depression emerged as a risk for increasing triangulation of both girls and boys. Efforts to improve family relationships and effective resolution of IPC may be strengthened by addressing family dynamics surrounding parents’ mental health. In particular, addressing conflict resolution challenges faced by depressed parents, as well as adolescents concerns about their family roles and responsibilities, may support families’ resolution of IPC without involvement of adolescent children.

Limitations and Future Directions

This study had several limitations. First, families in this study were predominantly White, economically advantaged, and reported high levels of positive family functioning. Future research should determine the extent to which patterns of risk found in the current study are generalizable to families facing economic and other contextual stressors. Second, auto-regressive methods represent a common but imperfect method for modeling change over time (Castro-Schilo & Grimm, 2018); future studies may provide additional insights into longitudinal patterns of triangulation by applying alternative methodologies (e.g., change scores, growth modeling). Third, separate models were constructed in order to navigate a balance between Type I and Type II error possibilities. Though this approach is valuable when evaluating potential moderation (Beauchaine et al., 2008), risk for Type I error for main effects is increased. However, we feel this risk was balanced by applying auto-regressive models with a high number of covariates.

Additionally, measurement limitations in this study warrant consideration. First, we used a general measure of adolescent triangulation into IPC that does not differentiate between specific types of triangulation (e.g., taking sides, mediation). Future studies might offer greater insights by focusing on different types of triangulation. Second, our study was limited by self-report assessments from one adolescent and one female caregiver per family. Family members engaged in triangulation tend to have discordant views of family interactions (McCauley et al., 2021). Collecting data from all family members, as well as family observation, may offer unique and valuable insights into triadic conflict processes. Third, efforts to reduce participant burden by using short measures resulted in reliance upon two measures consisting of two items, which may introduce measurement error. However, measures of psychological constructs drawing upon fewer than three items have shown to be reliable and valid (Robins et al., 2001). Fourth, two of our findings relating to family cohesion and parent depression exhibited trend-level statistical significance (p = .05), calling for caution until these findings can be replicated.

Conclusions

This study evaluated potential risk and protective factors for triangulation within multiple domains of the family. Low family cohesion, parent depression, and adolescent emotion dysregulation emerged as risk factors, while parent emotion coaching buffered the association between low interparental love and triangulation, but only for boys. Findings highlight that risk for triangulation exists throughout the family system, and also underscore the necessity of examining youth gender when evaluating protective factors in the family context.

Supplementary Material

Table 4_Supplemental
Table 3_Supplemental
Table 5_Supplemental

Figure 1. Conceptual model of family, parent, and adolescent risk factors for triangulation.

Figure 1.

Open in a new tab

Parent emotion coaching may buffer the association between risk factors and triangulation.

Acknowledgments

Support was provided by the Karl R. and Diane Wendle Fink Early Career Professorship for the Study of Families awarded to Gregory Fosco. Dr. Fosco was supported by the Penn State Social Science Research Institute. Devin McCauley was supported by the Prevention and Methodology Training Program (T32 DA017629; MPIs: J. Maggs & S. Lanza) with funding from the National Institute on Drug Abuse. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding agencies.

We would like to acknowledge the insightful feedback of Susan McHale on multiple drafts of this paper. We also gratefully acknowledge the contributions of Mengya Xia, Hio Wa Mak, Keiana Mayfield, Emily LoBraico, and Amanda Ramos for their assistance in collecting and preparing the data, and to the participating schools and families that made this project possible.

Footnotes

The authors have no competing interests to declare.

References

  1. Beauchaine TP, Neuhaus E, Brenner SL, & Gatzke-Kopp L (2008). Ten good reasons to consider biological processes in prevention and intervention research. Development and Psychopathology, 20, 745–774. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bell LG, Bell DC, & Nakata Y (2001). Triangulation and adolescent development in the US and Japan. Family Process, 40, 173–186. [DOI] [PubMed] [Google Scholar]
  3. Bloom BL (1985). A factor analysis of self-report measures of family functioning. Family Process, 24, 225–239. [DOI] [PubMed] [Google Scholar]
  4. Bowen M (1978). Family therapy in clinical practice. New York, NY: Jason Aronson, Inc. [Google Scholar]
  5. Braiker HB, & Kelley HH (1979). Conflict in the development of close relationships. In Burgess RL & Huston TL (Eds.), Social exchange in developing relationships (pp. 135–168). New York, NY: Academic Press, Inc. [Google Scholar]
  6. Buchanan CM, & Waizenhofer R (2001). The impact of interparental conflict on adolescent children: Considerations of family systems and family structure. In Booth A, Crouter AC, & Clements M (Eds.), Couples in conflict (pp. 149–160). Mahwah, NJ: Erlbaum. [Google Scholar]
  7. Buehler C, Franck KL, & Cook EC (2009). Adolescents' triangulation in marital conflict and peer relations. Journal of Research on Adolescence, 19, 669–689. [Google Scholar]
  8. Camisasca E, Miragoli S, & Di Blasio P (2019). Children’s triangulation during inter-parental conflict: Which role for maternal and paternal parenting stress? Journal of Child and Family Studies, 28, 1623–1634. [Google Scholar]
  9. Castro-Schilo L, & Grimm KJ (2018). Using residualized change versus difference scores for longitudinal research. Journal of Social and Personal Relationships, 35, 32–58. [Google Scholar]
  10. Cox MJ, & Paley B (1997). Families as systems. Annual Review of Psychology, 48, 243–267. [DOI] [PubMed] [Google Scholar]
  11. Cox MJ, Paley B, & Harter K (2001). Interparental conflict and parent-child relationships. In Grych JH & Fincham FD (Eds.), Interparental conflict and child development (pp. 249–272). New York: Cambridge University Press. [Google Scholar]
  12. Cummings EM, Ballard M, El-Sheikh M, & Lake M (1991). Resolution and children's responses to interadult anger. Developmental Psychology, 27, 462–470. [Google Scholar]
  13. Davies PT, Coe JL, Martin MJ, Sturge-Apple ML, & Cummings EM (2015). The developmental costs and benefits of children’s involvement in interparental conflict. Developmental Psychology, 51, 1026–1047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Davies PT, & Lindsay LL (2004). Interparental conflict and adolescent adjustment: Why does gender moderate early adolescent vulnerability?. Journal of Family Psychology, 18, 160–170. [DOI] [PubMed] [Google Scholar]
  15. Driver JL, & Gottman JM (2004). Daily marital interactions and positive affect during marital conflict among newlywed couples. Family Process, 43, 301–314. [DOI] [PubMed] [Google Scholar]
  16. Feinberg ME (2003). The internal structure and ecological context of coparenting: A framework for research and intervention. Parenting: Science and Practice, 3, 95–131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fosco GM, & Bray BC (2016). Profiles of cognitive appraisals and triangulation into interparental conflict: Implications for adolescent adjustment. Journal of Family Psychology, 30, 533–542. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fosco GM, Caruthers AS, & Dishion TJ (2012). A six-year predictive test of adolescent family relationship quality and effortful control pathways to emerging adult social and emotional health. Journal of Family Psychology, 26, 565–575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Fosco GM, & Grych JH (2008). Emotional, cognitive, and family systems mediators of children's adjustment to interparental conflict. Journal of Family Psychology, 22, 843–854. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Fosco GM, & Grych JH (2010). Adolescent triangulation into parental conflicts: Longitudinal implications for appraisals and adolescent-parent relations. Journal of Marriage and Family, 72, 254–266. [Google Scholar]
  21. Fosco GM, Lippold M, & Feinberg ME (2014). Interparental boundary problems, parent–adolescent hostility, and adolescent–parent hostility: A family process model for adolescent aggression problems. Couple and Family Psychology: Research and Practice, 3, 141–155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Garside RB, & Klimes-Dougan B (2002). Socialization of discrete negative emotions: Gender differences and links with psychological distress. Sex Roles, 47, 115–128. [Google Scholar]
  23. Gerard JM, Buehler C, Franck KL, & Anderson O (2005). In the eyes of the beholder: The functional roles of perceived appraisals associated with interparental conflict. Journal of Family Psychology, 19, 376–384. [DOI] [PubMed] [Google Scholar]
  24. Gottman JM, Coan J, Carrere S, & Swanson C (1998). Predicting marital happiness and stability from newlywed interactions. Journal of Marriage and the Family, 60, 5–22. [Google Scholar]
  25. Grych JH, Raynor SR, & Fosco GM (2004). Family processes that shape the impact of interparental conflict on adolescents. Development and Psychopathology, 16, 649–665. [DOI] [PubMed] [Google Scholar]
  26. Grych JH, Seid M, & Fincham FD (1992). Assessing marital conflict from the child's perspective: The children's perception of interparental conflict scale. Child Development, 63, 558–572. [DOI] [PubMed] [Google Scholar]
  27. Hankin BL, Fraley RC, Lahey BB, & Waldman ID (2005). Is depression best viewed as a continuum or discrete category? a taxometric analysis of childhood and adolescent depression in a population-based sample. Journal of Abnormal Psychology, 114, 96–110. [DOI] [PubMed] [Google Scholar]
  28. Jenkins JM, Smith MA, & Graham PJ (1989). Coping with parental quarrels. Journal of the American Academy of Child & Adolescent Psychiatry, 28, 182–189. [DOI] [PubMed] [Google Scholar]
  29. Katz LF, & Hunter EC (2007). Maternal meta-emotion philosophy and adolescent depressive symptomatology. Social Development, 16, 343–360. [Google Scholar]
  30. Katz LF, Maliken AC, & Stettler NM (2012). Parental meta-emotion philosophy: A review of research and theoretical framework. Child Development Perspectives, 6, 417–422. [Google Scholar]
  31. Kaufman EA, Xia M, Fosco G, Yaptangco M, Skidmore CR, & Crowell SE (2015). The difficulties in emotion regulation scale short form (DERS-SF): Validation and replication in adolescent and adult samples. Journal of Psychopathology and Behavioral Assessment, 38, 1–13. [Google Scholar]
  32. Kerig PK (2005). Revisiting the construct of boundary dissolution: A multidimensional perspective. Journal of Emotional Abuse, 5, 5–42. [Google Scholar]
  33. Kronmüller KT, Backenstrass M, Victor D, Postelnicu I, Schenkenbach C, Joest K, Fiedler P, & Mundt C (2011). Quality of marital relationship and depression: Results of a 10-year prospective follow-up study. Journal of Affective Disorders, 128, 64–71. [DOI] [PubMed] [Google Scholar]
  34. Kumpfer KL, Molgaard V, & Spoth R (1996). The Strengthening families program for the prevention of delinquency and drug use. In McMahon RJ & Peters RD & McMahon RJ (Eds.), Preventing childhood disorders, substance abuse, and delinquency (pp. 241–267). Thousand Oaks, CA: Sage Publications, Inc. [Google Scholar]
  35. Lindahl KM (1998). Triadic family observational coding: The use of a global coding system with a multi-ethnic sample. In Annual Convention of the Association for Advancement of Behavior Therapy, Washington, DC. [Google Scholar]
  36. Lindahl KM, & Malik NM (2011). Marital conflict typology and children's appraisals: The moderating role of family cohesion. Journal of Family Psychology, 25, 194–201. [DOI] [PubMed] [Google Scholar]
  37. Marchand JF, & Hock E (2000). Avoidance and attacking conflict-resolution strategies among married couples: Relations to depressive symptoms and marital satisfaction. Family Relations, 49, 201–206. [Google Scholar]
  38. McCauley DM, Sloan CJ, Xia M, & Fosco GM (2021). Same family, divergent realities: How triangulation preserves parents’ illusory harmony while adolescents navigate interparental conflicts. Journal of Family Psychology, 35, 128–137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Miller RB, & Wright DW (1995). Detecting and correcting attrition bias in longitudinal family research. Journal of Marriage and the Family, 921–929. [Google Scholar]
  40. Minuchin S (1974). Families and family therapy. Cambridge, MA: Harvard University Press. [Google Scholar]
  41. Olson DH, Waldvogel L, & Schlieff M (2019). Circumplex model of marital and family systems: An update. Journal of Family Theory & Review, 11, 199–211. [Google Scholar]
  42. Radloff LS (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. [Google Scholar]
  43. Robins RW, Hendin HM, & Trzesniewski KH (2001). Measuring global self-esteem: Construct validation of a single-item measure and the Rosenberg Self-Esteem Scale. Personality and Social Psychology Bulletin, 27, 151–161. [Google Scholar]
  44. Schermerhorn AC, Cummings EM, DeCarlo CA, & Davies PT (2007). Children's influence in the marital relationship. Journal of Family Psychology, 21, 259–269. [DOI] [PubMed] [Google Scholar]
  45. Shortt JW, Stoolmiller M, Smith-Shine JN, Mark Eddy J, & Sheeber L (2010). Maternal emotion coaching, adolescent anger regulation, and siblings’ externalizing symptoms. Journal of Child Psychology and Psychiatry, 51, 799–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Terry PC, Lane AM, & Fogarty GJ (2003). Construct validity of the profile of mood states-adolescents for use with adults. Psychology of Sport and Exercise, 4, 129–139. [Google Scholar]
  47. Van Parys H, & Rober P (2013). Trying to comfort the parent: A qualitative study of children dealing with parental depression. Journal of Marital and Family Therapy, 39, 330–345. [DOI] [PubMed] [Google Scholar]
  48. van Eldik WM, de Haan AD, Parry LQ, Davies PT, Luijk MP, Arends LR, & Prinzie P (2020). The interparental relationship: Meta-analytic associations with children’s maladjustment and responses to interparental conflict. Psychological Bulletin, 146, 553–594. [DOI] [PubMed] [Google Scholar]
  49. Vassilopoulos SP, & Banerjee R (2008). Interpretations and judgments regarding positive and negative social scenarios in childhood social anxiety. Behaviour Research and Therapy, 46(7), 870–876. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Table 4_Supplemental
NIHMS1737310-supplement-Table_4_Supplemental.docx (19.5KB, docx)
Table 3_Supplemental
NIHMS1737310-supplement-Table_3_Supplemental.pdf (30.9KB, pdf)
Table 5_Supplemental
NIHMS1737310-supplement-Table_5_Supplemental.docx (18.4KB, docx)

RESOURCES