Distal and Proximal Family Contextual Effects on Adolescents’ Interparental Conflict Appraisals: A Daily Diary Study

Abstract

Adolescent appraisals of interparental conflict (IPC) – perceiving IPC as threatening to their well-being or that of the family, and self-blaming attributions – are well-established pathways by which IPC confers risk for developmental disruptions and psychopathology. Recent work documents intraindividual change in IPC and appraisals that occur on a daily timescale. However, considerably less is known about how the broader family context may temper appraisals of IPC. This study provides a novel examination of the implications of distal (global ratings of family relationships in general) and proximal (fluctuations in daily family relationships) family context (family cohesion, parent-adolescent closeness, and parent-adolescent conflict) for adolescents’ propensity to form negative appraisals of daily IPC.

This sample included 144 adolescents (63% female) in two-parent families, who participated in a 21-day daily diary study. Findings indicate that intraindividual variability in adolescents’ perception of family cohesion, parent-adolescent closeness, and parent-adolescent conflict all correspond to adolescent appraisals of IPC through direct relations and moderating effects. Unique patterns emerged for boys and girls, suggesting gender differences in how adolescents incorporate the family context into their appraisals of IPC. This study expands our awareness of the importance of daily fluctuations in family relationships for adolescent risk during exposure to IPC.

The cognitive contextual framework (CCF) recognizes adolescents’ appraisals of interparental conflict (IPC) as key mediators of youth problem outcomes (Fosco, DeBoard, & Grych, 2007; Grych & Fincham, 1990). Threat appraisals refer to evaluations of IPC as undermining the child’s well-being or that of the family; self-blaming attributions refer to adolescents’ beliefs that they are responsible for causing or resolving conflicts (Atkinson et al., 2009; Grych & Fincham, 1990). The mediating role of appraisals linking IPC and youth maladjustment is well-established (e.g., Fosco & Feinberg, 2015; Gerard, Buehler, Franck, & Anderson, 2005; Grych, Harold, & Miles, 2003); other studies link IPC appraisals with adolescence-salient outcomes: substance use (Fosco & Feinberg, 2018; Tschann et al., 2002), social anxiety and poor peer relationships (Weymouth et al., 2019), and poor academic outcomes (Fosco & Bray, 2016; Harold, Aitken, & Shelton, 2007).

A second important premise of the CCF – the idea that appraisals of IPC are informed by children’s subjective perception of their broader family history, relationships, and experiences – has been much less extensively studied. Considering appraisals of IPC in isolation may provide an incomplete picture of the lived experience and the milieu of family dynamics that adolescents draw upon when evaluating threat and blame in IPC episodes (Fosco et al., 2007; Grych & Fincham, 1990). In the broader family context, warm and nurturing family relationships protect adolescents from the effects of environmental stressors (Biglan, Flay, Embry, & Sandler, 2012) and variability in warmth/nurturance or exposure to conflict in the home can confer risk (Repetti et al., 2002). Applied to the CCF, youth in families that offer warmth and support are expected to form less negative appraisals of IPC; however, in the context of adversity in other family relationships (e.g., parent-adolescent conflict), IPC may be even more distressing (Fosco et al., 2007). Preliminary studies (described below) support the hypothesis that family-level and parent-adolescent relations operate as contextual factors for IPC appraisals.

At the family-level, warm and nurturing family relationships in the form of cohesion and positive emotional expressiveness are associated with less negative appraisals of IPC. In one study, high levels of family cohesion buffered the association between destructive IPC and youth’s threat and self-blame appraisals (Lindahl & Malik, 2011). Another study found that youth in families characterized by frequent expressions of positive affect and affection as well as infrequent negative affect were less likely to report self-blame for IPC (Fosco & Grych, 2007). These studies, focusing on middle childhood outcomes, suggest that family-level functioning may contextualize IPC for youth. Less is known about these processes during adolescence.

Parent-youth relationships also have been identified as important contextual factors for IPC. Close, supportive, and warm parent-youth relationships reduce risk for threat and self-blame appraisals (DeBoard-Lucas et al., 2010; Grych et al., 2004; Lucas-Thompson & George, 2017; Selçuk et al., 2020). Harsh, rejecting, and psychologically controlling parenting increases children’s risk of forming threat and self-blame appraisals (DeBoard-Lucas et al., 2010; Selçuk et al., 2020). Interestingly, findings vary: sometimes providing support for mother (but not father)-youth relations, other times providing support for father (but not mother)-youth relations as context for IPC appraisals (e.g., Grych et al., 2004). Thus, more work is needed evaluating mother- and father-adolescent relationships as context for IPC.

Disentangling Long- and Short-Timescale Processes

This study revisits the CCF with an eye toward distinguishing between traditional, global assessments of the family context and dynamic, day-to-day fluctuations in family functioning. The vast majority of research evaluates how global ratings (i.e., those measuring general patterns or tendencies) of IPC are associated with developmental changes in adolescents’ tendency to evaluate IPC as threatening or to blame themselves for conflicts. Likewise, work evaluating family contextual factors also has relied on global assessments and between-person methods and speak primarily to developmental change processes. These studies offer key information about how dispositional IPC and appraisals correspond to probabilistic risk for youth maladjustment. Yet, an important gap remains in understanding how and why adolescents appraise day-to-day variation in IPC in terms of threat and self-blame.

Global vs. Situational Appraisals of IPC

Adding to daily diary studies of IPC (e.g., Cummings et al., 2003), recent work distinguishes between global appraisals, referring to adolescents’ general tendency to form threatening or self-blaming appraisals of IPC and situational appraisals, referring to adolescents’ appraisals of specific interparental conflict events (Authors). Situational appraisals are operationalized as a within-person process of evaluating the degree to which day-to-day fluctuations in IPC correspond to changes in threat and self-blame. Recent findings revealed that on days when IPC is elevated, adolescents experienced higher threat and self-blame than usual. In turn, daily increases in adolescents’ situational appraisals corresponded to poorer daily mood (anger, sadness, anxiety) and well-being (happiness, life satisfaction, meaning/purpose in life) (Authors). These within-person findings complement the extant literature on between-person differences in risk, yet offer novel information about how fluctuations in IPC, family context, and appraisals are experienced by adolescents and their families on a daily basis.

Distal vs. Proximal Family Context of IPC

The current study applies lessons learned from the delineation of global and situational appraisals to Grych and Fincham’s (1990) original distinction between distal and proximal contextual factors for IPC. Distal contextual factors correspond to dispositional qualities of the family (e.g., global relationship quality, parental warmth) and youth (e.g., temperament), or past experiences and exposure to IPC. Adolescents do not evaluate each conflict anew; rather, they draw upon long-standing family relationships and patterns of experience as information from which to understand new experiences (Grych & Cardoza-Fernandez, 2001). The literature reviewed earlier supports hypotheses regarding the distal family context of IPC.

Proximal contextual factors refer to more recent or concurrent experiences that may influence youth appraisals of IPC events, such as children’s mood at the time conflict occurs or their expectations for the intensity or outcome of an unfolding conflict episode (Grych & Fincham, 1990). Consistent with this idea, we draw on recent findings that family-level and parent-adolescent relationships fluctuate in a meaningful way from day to day (Fosco et al., 2019; Fosco & Lydon-Staley, 2020; Lippold, Davis, Lawson, et al., 2016). Thus, the state of family relationships on days when IPC occurs, relative to the family’s typical relationship quality, may serve as important proximal context from which adolescents appraise IPC.

The proximal family context may offer cues to adolescents regarding the potential for IPC to undermine family well-being, or shape attributions of blame. For example, adolescents may perceive IPC as particularly threatening if it is accompanied by declines in family cohesion, or if one or both parents also became angry or hostile with their adolescent. Conversely, if family cohesion or parent-adolescent conflict remain unchanged on days of IPC, adolescents may appraise IPC less negatively. The absence of research on proximal family context reflects a key gap in the literature that would offer direct guidance to practitioners who often may seek to advise parents about ways to mitigate the effects of IPC for their child.

Adolescent Gender Differences in Models of IPC

Whether IPC confers different risks to boys and girls has been an important discussion for some time (e.g., Davies & Lindsay, 2004). Findings have been largely inconsistent with regard to gender differences in the impact of IPC on boys’ and girls’ outcomes, especially in studies predicting global maladjustment (van Eldik et al., 2020). van Eldik and colleagues (2020) state: “developmental and gender differences may be more apparent in terms of children’s specific responses than in terms of broadband adjustment problems” (p. 7); others have called for tests of gender moderation of within-person appraisals of IPC (Fosco & Lydon-Staley, 2019). Whether specific instances of IPC confer differential risk to boys and girls remains unclear.

Gender differences in situational appraisals of IPC may require careful unpacking to detect. In an early daily diary study of IPC, no gender differences were found for children’s emotional responses to daily IPC (Cummings et al., 2002; 2003); however, recent analytic methods that disentangle within-person and between-person effects (Bolger & Laurenceau, 2013) may yield new information. Other daily diary studies of family relationships document gender differences. For example, girls tend to experience more negative affect during family interactions than boys do, particularly during the 9^th and 10^th grades (Larson et al., 1996). Moreover, girls exhibit larger-magnitude emotional responses to daily stressors (Lippold, Davis, McHale, et al., 2016) and to interpersonal stressors (i.e., conflict with mothers, fathers, and/or friends) (Flook, 2011) than boys do. This evidence suggests that the proximal daily contexts in which adolescents appraise IPC may differ for boys and girls. Guided by prior studies that reveal gender differences in responses to family functioning on a daily timescale, we include adolescent gender as a moderator to account for the possibility that fluctuations in family relations may contextualize situational appraisals of IPC differently for boys and girls.

The Current Study

In this study, we disentangled distal and proximal aspects of family context to gain new insights into their contributions to adolescents’ daily threat and self-blame appraisals of IPC. We focused on family-level cohesion, mother-adolescent (MA) closeness and conflict, and father-adolescent (FA) closeness and conflict as promising contextual factors, based on theory and prior work. Consistent with propositions of the CCF, we evaluated two potential ways that family context functions in this model: a) as a moderator of the IPC-appraisal link by buffering or amplifying the risk of forming negative appraisals of IPC, or b) by directly shaping the meaning IPC has for adolescents, evidenced by a direct relation between contextual factors and adolescent appraisals (Grych & Fincham, 1990).

Daily diary data, combined with global reports of family-level and parent-adolescent functioning, were leveraged to disentangle between-person and within-person functioning (Hamaker et al., 2007). Guided by the CCF, we operationalized the distal family context as adolescents’ reports of global family functioning to capture their “psychological construal” of general (or trait-like) patterns of family relations that may inform their evaluations of conflict (Grych & Fincham, 1990). Proximal family context was operationalized as within-person deviations from their own central tendency across days, likened to a “state” (Hamaker et al., 2007). Thus, it was possible to ascertain whether fluctuations in family functioning, on days when IPC was higher than usual, might modify the degree to which IPC was appraised negatively (moderation) or function as an additive effect for adolescent appraisals. Prior (between-person/distal family context) studies lend support to both hypotheses; however, due to the dearth of within-person studies on this topic, we did not have specific hypotheses as to whether moderation or main effect findings would emerge. We hypothesized the following:

Distal family context effects.

We expected that distal family functioning would moderate within-person relations among daily IPC and situational appraisals, such that adolescents in families characterized by high cohesion or close parent-adolescent relationships would evaluate IPC less negatively than those in families with poorer relationship quality. Likewise, adolescents in families with more parent-adolescent conflict were expected to exhibit a larger increase in threat or self-blame in response to daily instances of IPC.

Proximal family context effects.

We evaluated whether proximal family relations operated as moderators of within-person associations of IPC and appraisals or as main effects in within-person models. Generally, we expected that family cohesion and parent-adolescent closeness would have protective effects, and that parent-adolescent conflict would be a risk factor for more negative appraisals of IPC.

Finally, we evaluated adolescent gender differences in these models to answer recent calls for such work (van Eldik et al., 2020). Given inconsistencies in prior research on IPC, no specific hypotheses were posed. However, daily diary family research suggests that girls may be more sensitive to proximal family contextual factors than boys.

Method

Data for the current study were collected between August, 2015 and November, 2016 as part of the {Study Title}, wherein parents and adolescents completed up to 21 daily reports of family functioning, mood, and well-being. This study {title} was approved by the {blind} Institutional review board. Adolescents reported on IPC, appraisals, and family relationships.

Participants

The current sample (same as in {Authors}) was drawn from 150 families of 9^th and 10^th grade adolescents recruited through high schools in Pennsylvania and through family referrals, to take part in a multiple time-scale experience sampling study of family relationships, mood, and well-being. Families were eligible for participation if they met six criteria: (1) two-caregiver family status, (2) adolescents lived in one household continuously, (3) internet access and means to complete daily surveys at home, (4) English fluency, (5) the participating adolescent was in 9^th or 10^th grade, and (6) one parent and adolescent agreed to participate (via consent, assent, respectively). Because of a focus on adolescent relationships with their mothers and fathers, 7 families were dropped because they did not conform to this family structure (e.g., parent and aunt caregivers).

The analytic sample for this study included 144 adolescents (91 female, 53 male) between 13 and 16 years old (M_Age = 14.62, SD_Age = 0.83) who were identified (via parent report) as White (84.0%), African American/Black (2.1%), Native American/American Indian (0.7%), Asian (4.2%), Hispanic/Latino (0.7%), Multiracial (6.9%), or missing (0.7%). Female caregivers were identified as the adolescent’s mother (96.53%), stepmother (2.78%), or foster mother (0.69%). Male caregivers were identified as the adolescent’s father (81.94%), stepfather (15.28%), foster father (0.69%), or other male caregiver (1.39%). Couples were married (n = 132), or cohabiting (n = 12). Parents reported living together for an average of 17.6 years (SD = 6.7). Participating parents’ education spanned graduate or professional training (23.6%), college degree (28.5%), associate’s degree or some college (29.9%), high school degree or equivalent (14.6%), to less than a high school degree (2.8%), with 0.7% missing this information. Caregivers also reported their partner’s education, from graduate or professional training (19.4%), college degree (20.1%), associate’s degree or some college (28.5%), high school degree or equivalent (23.6%), or less than a high school degree (6.9%), with 1.4% missing data. Family income ranged from ‘less than $10,000’ to ‘$125,000 and over’ (Median = ‘$70,000 – $79,999’).

Procedure

Families were recruited through emails sent to parents from school principals. Interested parents accessed a study web page describing the purpose and design of the study, and provided consent to participate and contact information. Research staff reviewed family eligibility, then sent adolescents a description of the study and an opportunity to assent or decline participation. Following assent, adolescent and parents completed baseline surveys, after which 21-day daily diary protocols were initiated. Person-specific links to daily questionnaires were emailed separately to parents and adolescents at 7:00 PM each night, followed by a reminder (text message or phone call). Participants were instructed to complete the daily survey before going to bed, although access links remained open until 9:00 AM the next morning. In rare cases where participants completed surveys the following morning, they were instructed to report on the prior day. Daily questionnaires took approximately 5 minutes to complete each night. The N = 144 families analyzed here provided 2657 daily reports on between 11 to 21 days (M_Days = 19.03 (90.62%), SD_Days = 2.44). Parents and adolescents were compensated with gift cards to Amazon.com or Wal-Mart (based on preference) at each stage: $25 for the baseline assessment and up to $75 for completing the 21 daily surveys.

Measures

Our empirical analysis makes use of adolescents’ daily reports about IPC, threat, self-blame, family cohesion, and parent-adolescent relationship quality to capture within-person processes consistent with proximal family context hypotheses. Reliability for daily measures were calculated in two ways: R_c to evaluate the reliability of within-person change across days (Bolger & Laurenceau, 2013) and R_1F to evaluate reliability of between-person differences, while accounting for nesting of occasions within persons (Cranford et al., 2006). Distal family context was assessed global measures in the baseline survey to capture general parent-adolescent relations and family cohesion. Reliability was calculated using Cronbach’s alpha. For all metrics, higher values reflect better reliability, with values exceeding .70 viewed as good. Scale means, SDs and reliability are reported in Table 1. All daily items are reported in Supplemental Table 1.

Table 1.

Descriptive Statistics of Study Variables.

		Correlations
		(within-person below diagonal, between-person above diagonal)
		1	2	3	4	5	6	7	8
1	IPC	---	.16	.14	−.50*	−.19	−.50*	.30	.52*
2	Threat	.15	---	.98*	−.04	−.05	−.07	.02	.06
3	Blame	.12	.47	---	−.04	−.06	−.05	.03	.05
4	Cohesion	−.19	−.04	−.09	---	.43*	.64*	−.26*	−.37*
5	MA Close	−.13	−.06	−.11	.37	---	.49	−.34*	−.16
6	FA Close	−.13	−.06	−.13	.11	.39	---	−.27*	−.50*
7	MA Con	.16	.06	.13	−.17	−.36	−.15	---	.50*
8	FA Con	.17	.07	.17	−.16	−.16	−.22	.26	---
	Daily diary (within-person) variables
	Mean	0.98	−2.14	−2.15	7.90	8.48	7.65	0.98	0.95
	SD	2.02	2.41	2.37	2.42	2.13	2.83	1.99	2.01
	R1F	0.86	0.94	0.95	0.92	0.94	0.97	0.78	0.84
	Rc	0.83	0.83	0.81	0.76	0.87	0.84	0.78	0.78
	Global (between-person) variables
	Mean	1.01	−2.18	−2.18	4.23	4.22	3.78	1.73	1.69
	SD	1.49	2.09	2.11	0.67	0.68	0.96	0.76	0.84
	α	--	--	--	.82	.92	.96	.86	.91

Note. Within-person correlations (below diagonal) reflect a sample average of within-person correlations, no tests of statistical significance are available. Between-person correlations (above diagonal) are calculated using individual mean score for IPC and appraisals, and baseline assessments for distal family context variables.

^*denotes between-person correlations that were statistically significant (p < .05).

Interparental Conflict.

Adolescents rated that day’s level of conflict between caregivers, using a slider scaled 0 (“Not at All”) to 10 (“A Lot”) in 0.1 increments. As evidence of validity, average IPC across days has been found to be correlated with parent reports (blinded), and with established baseline measures of IPC (r = .58, p < .01).

Situational Threat and Blame Appraisals.

To capture within-person variation, adolescents were asked to rate their threat and self-blame on each day, compared to how they usually feel from −5 (Less than Usual) to 5 (More than Usual), with the slider anchored at 0 (Neither More nor Less than Usual), in 0.1 increments. Adolescents responded to two items about situational threat and three items measuring situational self-blame.

Proximal family context.

Adolescents reported on family-level and parent-child relationship quality each day. Family cohesion was measured using three items selected from the short version of the Family Environment Scale (Bloom, 1985) to fit with a daily timescale. Prior work has shown that the average score across days is correlated with the standard Family Environment Scale (r = .68, p < .01; Authors). Adolescents rated four items assessing parent-adolescent closeness, repeated for MA and FA closeness. Across-day averages were correlated with baseline measures of closeness (MA: r = .60 p < .01; FA: r = .77, p < .01). Adolescents also rated two items for parent-adolescent conflict, repeated for each parent. Average daily scores were correlated with baseline measures of conflict (MA: r = .52, p < .01; FA: r = .58, p < .01).

Distal family context.

Distal family cohesion was measured using five items from the shortened Family Environment Scale (Bloom, 1985). Respondents rated how true each of the items had been in the past month, on a scale from 1 (Almost Never) to 5 (Almost Always). Distal parent-adolescent closeness was measured using 10 items from the Inventory of Parent and Peer Relationships (IPPA; Armsden & Greenberg, 1987), rated from 1 (Completely Untrue) to 5 (Completely True). Adolescents reported on each item separately for each parent. Distal parent-adolescent conflict was assessed with 5 items measuring the frequency of parent’s hostile behaviors toward them, from the Behavioral Affective Rating Scale (Conger, 1989). Response options ranged from 1 (Almost Never) to 5 (Almost Always).

Analysis Plan

Multilevel models were used to evaluate our hypotheses because they are particularly suited to address the nested nature of the intensive repeated measures (21 days nested within persons; Snijders & Bokser, 2012). To disentangle distal (between-person) and proximal (within-person) features of family functioning, we parameterized time-invariant (between-person) and time-varying (within-person) aspects of IPC, family cohesion, parent-adolescent closeness, and parent-adolescent conflict (see Bolger & Laurenceau, 2013). Between-person variables include usual IPC, calculated as the individual mean across the 21 days, and distal family context, using global measures of family relations at baseline. These variables were grand-mean centered such that positive values indicated higher average levels relative to others in the sample. Within-person variables, referred to as day’s IPC and proximal family context, were calculated as a daily deviation score from individuals’ mean scores. Values reflect days in which adolescents experienced higher (positive values) or lower (negative values) levels of IPC, cohesion, closeness, or parent-adolescent conflict than usual.

We conducted five sets of multilevel models, evaluating family cohesion, MA closeness, FA closeness, MA conflict, and FA conflict as contextual factors of adolescent situational threat and self-blame. A general example of the model is presented below.

At level 1 (day-level variables) the equation was constructed as:

$${Appraisal}_{it}={\beta }_{0i}+{\beta }_{1i}Day\text{'}s{IPC}_{it}+{\beta }_{2i}Prox{Fam}_{it}+{\beta }_{3i}{Time}_{it}+{\beta }_{4i}Day\text{'}sIPC\ast Prox{Fam}_{it}+e_{it}$$ (1)

where Appraisal_it reflects threat or self-blame for person i on day t; β_0i indicates the expected threat or self-blame in the middle of the study (time was centered at day 10.5) for an individual experiencing an average level of IPC, family context, and the interaction between day’s IPC and proximal family context; β_1i indicates the association between day’s IPC and appraisals; β_2i indicates the association between proximal family context and appraisals (consistent with a direct contextual role for appraisals), β_3i indicates the effect of time in study on appraisals in order to account for time as a third variable (see Bolger & Laurenceau, 2013); and β_4i indicates the day-level interaction of day’s IPC and proximal family context. Finally, e_it are day-specific residuals that were allowed to autocorrelate (AR1).

Person-specific intercepts and associations from the Level 1 model were specified at Level 2 (family-level variables) as:

$${\beta }_{0i}={\gamma }_{00}+{\gamma }_{01}Usual{IPC}_i+{\gamma }_{02}Dist{Fam}_i+{\gamma }_{03}{Gender}_i+{\gamma }_{04}UsualIPC\ast {Gender}_i+{\gamma }_{05}DistFam\ast {Gender}_i+u_{0i}$$ (2a)

$${\beta }_{1i}={\gamma }_{10}+{\gamma }_{11}Dist{Fam}_i+{\gamma }_{12}{Sex}_i+{\gamma }_{13}DistFam\ast {Gender}_i+u_{1i}$$ (2b)

$${\beta }_{2i}={\gamma }_{20}+{\gamma }_{21}{Sex}_i+u_{2i}$$ (2c)

$${\beta }_{3i}={\gamma }_{30}$$ (2d)

$${\beta }_{4i}={\gamma }_{40}+{\gamma }_{41}{Sex}_i$$ (2e)

where the γs are sample-level parameters and the u’s are residual between-family differences that may be correlated, but are uncorrelated with e_it. As shown in equations 2a–c, between family associations for usual IPC, distal family context, and adolescent gender with appraisals are indicated by γ₀₁, γ₀₂, and γ₀₃ respectively. Presented in the second step of the analyses, we also tested 2-way interactions of usual IPC*gender (γ₀₄) and distal family context*gender (γ₀₅).

Cross-level interactions were calculated for distal family context and adolescent gender as moderators of daily effects. Two-way interactions tested whether distal family context (γ₁₁) or adolescent gender (γ₁₂) moderated the relation between day’s IPC and appraisals (Eq. 2b). Additionally, we tested whether the main effect of proximal context on appraisals differed for boys and girls (γ₂₁) (Eq. 2c). Three-way interactions were used to evaluate whether the distal family context and gender together qualified the association between day’s IPC and appraisals (γ₁₃), or whether gender qualified the day-level interaction for day’s IPC and proximal family context (γ₄₁). Statistically significant interactions were followed-up by examining simple slopes at high (+1 SD) and low (−1 SD) levels of the moderator. To facilitate interpretation of the intercept as levels of appraisals for the average person, adolescent gender was sample-mean centered. Analyses were conducted using the nlme package in R version 3.6.2 (Pinheiro et al., 2020) using maximum likelihood estimation. As a note, alternative analyses were run using individuals’ average rating across daily reports as distal contextual factors rather than the baseline measures reported here; the substantive findings were nearly identical.

Results

Table 1 presents the descriptive statistics and correlations for study variables. We then tested a series of multilevel models predicting adolescent threat and self-blame appraisals, focusing on positive family relations – family cohesion, and mother- and father-adolescent closeness, followed by two sets of models evaluating mother- and father-adolescent conflict as contextual factors.

Family Cohesion.

Models evaluating family cohesion are presented in the first column of Table 2. On days when IPC was higher than usual, adolescents reported higher threat (γ₁₀ = .15) and self-blame (γ₁₀ = .10). Additionally, on days when proximal family cohesion was higher than usual, adolescents experienced lower self-blame than usual (γ₂₀ = −.07); however, proximal family cohesion was not associated directly with changes in threat appraisals. In both models, the three-way interactions for day’s IPC*proximal cohesion*gender (γ₄₁) were statistically significant predicting threat (γ₄₁= .05) and self-blame (γ₄₁= .05).

Table 2.

Evaluating Cohesion and Parent-Adolescent Closeness as Proximal and Distal Contextual Factors for Appraisals

	Family Cohesion		MA Closeness		FA Closeness
	Threat	Blame	Threat	Blame	Threat	Blame
Main Effects	Est (SE) p	Est (SE) p	Est (SE) p	Est (SE) p	Est (SE) p	Est (SE) p
Intercept (γ00)	−2.18 (.17) .000	−2.17 (.18) .000	−2.19 (.17) .000	−2.18 (.18) .000	−2.20 (.18) .000	−2.22 (.18) .000
Day’s IPC (γ10)	.15 (.04) .000	.10 (.04) .004	.17 (.04) .000	.11 (.03) .001	.18 (.04) .000	.10 (.03) .002
Usual IPC (γ01)	.36 (.13) .006	.27 (.14) .046	.31 (.12) .009	.23 (.12) .056	.32 (.13) .017	.24 (.14) .077
Prox Fam (γ20)	−.03 (.02) .148	−.07 (.03) .008	−.06 (.02) .003	−.14 (.03) .000	−.06 (.03) .041	−.13 (.03) .000
Dist Fam (γ02)	.32 (.29) .271	.22 (.30) .465	−.01 (.26) .964	.00 (.26) .996	.08 (.21) .701	.06 (.21) .784
Gender (γ03)	.35 (.36) .340	.31 (.37) .404	.25 (.36) .489	.23 (.37) .535	.21 (.36) .563	.18 (.37) .623
Time (γ30)	.02 (.00) .00	.02 (.00) .000	.02 (.00) .000	.02 (.00) .000	.02 (.00) .000	.02 (.00) .000
2-way interactions
Day’s IPC*Prox Fam (γ40)	−.03 (.01) .014	.00 (.01) .928	−.02 (.01). 169	.02 (.01) .072	.03 (.01) .021	.00 (.01) .980
Day’s IPC*Dist Fam (γ11)	−.07 (.05) .179	.01 (.05) .773	−.03 (.05) .530	.00 (.04) .938	−.04 (.04) .327	.02 (.03) .613
Day’s IPC*Gen(γ12)	.02 (.08) .843	.02 (.08) .761	−.02 (.08) .807	.02 (.07) .783	.03 (.08) .673	.04 (.07) .569
Usual IPC*Gen (γ04)	.21 (.26) .431	.24 (.27) .391	.12 (.23) .600	.22 (.23) .337	.24 (.26) .360	.32 (.27) .234
Prox Fam*Gen (γ21)	−.03 (.05) .607	−.05 (.06) .365	−.12 (.04) .003	−.07 (.07) .255	.04 (.06) .470	.01 (.06) .924
Dist Fam*Gen (γ05)	.62 (.63) .325	.49 (.64) .452	.23 (.53) .670	.56 (.52) .284	.45 (.43) .294	.51 (.44) .249
3-way interactions
Day’s IPC*Prox Fam*Gen (γ41)	.05 (.02) .033	.05 (.02) .012	−.02 (.02) .492	.02 (.02) .475	.01 (.02) .672	.04 (.02) .033
Day’s IPC*Dist Fam*Gen(γ13)	−.03 (.10) .791	−.13 (.09) .184	−.07 (.10) .500	−.06 (.09) .492	.01 (.08) .934	−.08 (.07) .239
Random Effects
Icpt σu0	2.05 [1.82, 2.32]	2.08 [1.85, 2.35]	2.06 [1.82, 2.32]	2.08 [1.85, 2.34]	2.05 [1.82, 2.32]	2.09 [1.85, 2.35]
Day’s IPC σu1	.30 [.24, .37]	.27 [.21, .35]	.30 [.24, .37]	.26 [.20, .33]	.31 [.25, .39]	.26 [.20, .33]
Prox Fam σu2	.12 [.07, .20]	.18 [.13, .24]	.03 [.00, 4.69]	.23 [.17, .30]	.16 [.11, .23]	.23 [.18, .29]
Corr: Icpt, Day’s IPC ru0,u1	−.39 [−.60, −.12]	−.18 [−.42, .09]	−.41 [−.62, −.13]	−.18 [−.47, .15]	−.40 [−.60, −.13]	−.20 [−.44, .07]
Corr: Icpt, Prox Fam ru0,u2	.40 [−.04, .72]	.27 [−.07, .55]	.05 [−.30, .39]	.36 [.04, .61]	.06 [−.18, .30]	.21 [−.08, .47]
Corr: Day’s IPC, Prox Fam ru1,u2	−.15 [−.52, .27]	.09 [−.32, .47]	−.02 [−.16, .12]	−.05 [−.55, .48]	.14 [−.22, .47]	−.07 [−.29, .17]

Notes: Est=estimate; SE=standard error; p = p value; Dist Fam= Distal Family Context; Prox Fam = Proximal Family Context; Coh=cohesion; IPC = Interparental Conflict; Gen = Gender.

Probes of these three-way interactions were conducted separately for threat and self-blame and are plotted in Figure 1. We first examined the simple slope for the within-person association between day’s IPC and situational threat, across high and low levels of proximal cohesion, plotted separately for boys and girls. For adolescent girls, day’s IPC was associated with threat on days when cohesion was lower than usual (b = .20, p = .00); but was not associated with threat on high-cohesion days. For boys, day’s IPC was associated with threat at both high (b = .17, p = .02) and low (b = .16, p = .02) levels of proximal cohesion.

Figure 1. Illustrative Plots for Proximal Family Cohesion as Context for Appraisals of IPC.

Note. 1a. Plots of Day’s IPC*Proximal Cohesion*Gender Predicting Situational Threat and Blame Appraisals, by Adolescent Gender.

♦ Indicates sample mean

Predicting situational blame (also see Figure 1), the pattern of results for girls was similar to those above. Specifically, on days when proximal cohesion was lower than usual, day’s IPC more strongly corresponded to increases in self-blame (b = .12, p = .01); but on days when proximal cohesion was higher than usual, this relation was not significant. For boys, the pattern of results was different. Specifically, on days when proximal cohesion was high, day’s IPC was positively associated with self-blame (b = .16, p = .02); on days when proximal cohesion was low, the day’s IPC was not associated with self-blame.

MA Closeness.

Results from models testing MA closeness are reported in the second column of Table 2. As before, on days when IPC was higher than usual, adolescents reported higher levels of threat (γ₁₀ = .17) and self-blame (γ₁₀ = .11). Over and above IPC, on days when adolescents felt closer to their mothers, they reported decreases in threat (γ₂₀ = −.06) and in self-blame (γ₂₀ = −.14). However, the main effect of proximal MA closeness on situational threat was qualified by adolescent gender (γ₂₁ = −.12), indicating the association between daily MA closeness and situational threat was only significant for boys (b = −.14, p = .00).

FA Closeness.

Analyses of FA closeness are reported in the third column of Table 2. As with prior models, on days when IPC was higher than usual, adolescents reported increased threat (γ₁₀ = .18) and increased self-blame (γ₁₀ = .10). Accounting for days’ IPC, on days when adolescents felt closer to their fathers, they reported decreased threat (γ₂₀ = −.06) and decreased self-blame (γ₂₀ = −.13). Proximal closeness with fathers qualified the within-person association between IPC and threat appraisals (γ₄₀= .03). Probing this interaction revealed that the magnitude of the within-person association between day’s IPC and situational threat was slightly higher on days of higher FA closeness than usual (b = .21, p = .00), compared to days of lower FA closeness than usual (b = .15, p = .00).

Turning to analyses of self-blame, the three-way interaction for day’s IPC, proximal FA closeness, and adolescent gender was statistically significant (γ₄₁= .04). For girls, on days when they felt less close to their fathers than usual, the day’s IPC was more strongly associated with blame (b = .11, p = .01); however, on days of higher than usual father-adolescent closeness, daily variation in IPC did not correspond to self-blame (b = .07, p = .14). For boys, the pattern was different. On days when boys felt closer to their fathers than usual, the within-person association between day’s IPC and blame was stronger (b = .16, p = .01); but on days of lower than usual FA closeness, day’s IPC was not associated with self-blame (b = .10; p = .09). Plots of this interaction were nearly identical to Figure 1b, and thus are not presented for space.

MA Conflict.

Presented in Table 3, we evaluated MA conflict as a contextual factor for adolescent appraisals of IPC. As in other models, on days when IPC was higher than usual, adolescents reported elevated threat (γ₁₀ = .16) and self-blame (γ₁₀ = .07). Further, on days when adolescents experienced higher conflict with their mothers than usual, they reported higher threat (γ₂₀ = .06) and self-blame (γ₂₀ = .15) than usual. None of the two-way interactions were statistically significant. However, the day’s IPC*proximal MA conflict*gender three-way interactions were significant for both threat (γ₄₁= −.06) and self-blame (γ₄₁= −.04). For girls, on days of low MA conflict, day’s IPC was not significantly associated with increases in threat appraisals (b = .10, p = .19) and was not associated with self-blame (b = .03, p = .49). However, on days when MA conflict was higher than usual, day’s IPC was more strongly associated with girls’ threat appraisals (b = .19, p =. 01) and self-blaming attributions (b = .09, p = .03). For boys, the association between day’s IPC and threat (b = .21, p = .06) and self-blame (b = .12, p = .02) was stronger on days of lower MA conflict than usual. However, when MA conflict was higher than usual, IPC was not associated with boys’ threat (b = .14, p = .22) or self-blame (b = .08, p = .14), but these appraisals were generally elevated. These plots are presented in Figure 2a.

Table 3.

Evaluating Parent-Adolescent Conflict as Proximal and Distal Contextual Factors for Appraisals

	MA Conflict		FA Conflict
	Threat	Blame	Threat	Blame
Main Effects	Est (SE) p	Est (SE) p	Est (SE) p	Est (SE) p
Intercept (γ00)	−2.19 (.17) .000	−2.20 (.18) .000	−2.19 (.17) .000	−2.19 (.18) .000
Day’s IPC (γ10)	.16 (.04) .000	.07 (.03) .014	.16 (.04) .000	.06 (.03) .014
Usual IPC (γ01)	.33 (.12) .007	.20 (.12) .104	.29 (.13) .030	.21 (.14) .136
Prox Fam (γ20)	.06 (.02) .009	.15 (.02) .000	.06 (.03) .034	.15 (.03) .000
Dist Fam (γ02)	−.16 (.24) .515	−.05 (.25) .848	−.14 (.25) .562	−.06 (.25) .798
Gender (γ03)	.27 (.36) .461	.23 (.36) .534	.31 (.36) .393	.25 (.37) .499
Time (γ30)	.02 (.00) .000	.02 (.00) .000	.02 (.00) .000	.02 (.00) .000
2-way interactions
Day’s IPC*Prox Fam (γ40)	.01 (.01) .184	.00 (.01) .389	.01 (.01) .179	.02 (.01) .047
Day’s IPC*Dist Fam (γ11)	.06 (.05) .216	.01 (.04) .875	.04 (.04) .358	.04 (.03). 143
Day’s IPC*Gen(γ12)	.03 (.08) .695	.04 (.06) .504	−.01 (.08) .944	.03 (.06) .620
Usual IPC*Gen (γ04)	.13 (.23) .583	.28 (.24) .258	.14 (.26) .593	.25 (.27) .348
Prox Fam*Gen (γ21)	.02 (.05) .664	.06 (.05) .198	.09 (.05) .090	.06 (.06) .348
Dist Fam*Gen (γ05)	−.31 (.52) .555	−.66 (.05) .212	−.20 (.48) .673	−.27 (.49) .581
3-way interactions
Day’s IPC*Prox Fam*Gen (γ41)	−.06 (.02) .000	−.04 (.02) .022	−.04 (.02) .047	−.02 (.02) .204
Day’s IPC*Dist Fam*Gen (γ13)	.07 (.10) .511	−.07 (.08) .418	.02 (.08) .801	.01 (.06) .820
Random Effects
Intercept σu0	2.06 [1.83, 2.33]	2.10 [1.86, 2.37]	2.06 [1.82, 2.32]	2.09 [1.85, 2.36]
Day’s IPC σu1	.30 [.24, .37]	.22 [.16, .29]	.30 [.24, .37]	.18 [.13, .25]
Prox Fam σu2	.14 [.10, .19]	.16 [.13, .21]	.16 [.12, .22]	.22 [.18, .28]
Corr: Intercept, Day’s IPC ru0,u1	−.43 [−.64, −.16]	−.30 [−.54, −.01]	−.43 [−.63, −.17]	−.28 [−.55, .05]
Corr: Intercept, Prox Fam ru0,u2	.09 [−.38, .53]	.06 [−.28, .17]	.06 [−.20, .31]	−.17 [−.43, .11]
Corr: Day’s IPC, Prox Fam ru1,u2	.19 [−.35, .64]	.20 [−.17, .52]	.17 [−.23, .52]	.33 [−.17, .69]

Notes: Est=estimate; SE=standard error; p = p value; Dist Fam= Distal Family Context; Prox Fam = Proximal Family Context; IPC = Interparental Conflict; MA = Mother-Adolescent; FA = Father-Adolescent Gen = Gender.

Figure 2. Illustrative Plots for Proximal Parent-Adolescent Conflict as Context for Appraisals of IPC.

a. Plots of Day’s IPC*Proximal MA Conflict*Gender Predicting Situational Threat and Blame Appraisals, by Adolescent Gender.

b. Plots of Day’s IPC*Proximal FA Conflict*Gender Predicting Situational Threat Appraisals, by Adolescent Gender.

FA Conflict.

In the FA conflict model, on days when IPC was higher than usual, adolescents reported increased threat (γ₁₀ = .16) and increased self-blame (γ₁₀ = .06). In addition, days of elevated proximal FA conflict also were associated with increases in threat appraisals (γ₂₀ = −06) and increases in self-blame (γ₂₀= .15). For threat, the three-way interaction for day’s IPC, proximal FA conflict, and adolescent gender was statistically significant (γ₄₁= −.04). For girls, day’s IPC was significantly associated with increases in threat across levels of proximal FA conflict, yet was stronger at high (b = .19, p = .00) relative to low (b = .13, p = .01) levels of FA conflict. For boys, the pattern was similar yet reversed. Day’s IPC was more strongly associated with increases in threat at low (b = .17, p = .01) relative to high (b =.14, p = .04) proximal FA conflict (Figure 2b). However, in relation to self-blame, proximal FA conflict (γ₄₀ = .02) moderated the effects of day’s IPC. On days of higher than usual IPC and FA conflict, adolescents experienced increases in self-blame (b = .08, p = .01); whereas, on days of lower than usual FA conflict, the day’s IPC was not associated with self-blame (b = .05, p = .18).

Discussion

Daily experiences of interparental conflict have been linked to adolescents’ situational threat and self-blame; in turn, these appraisals correspond to diminished well-being and elevated angry, anxious, and depressed mood (Authors). The current study expands on this work by examining how daily intraindividual variability in broader family relationships – conceptualized as family-level cohesion, parent-adolescent closeness, and parent-adolescent conflict – may contextualize daily IPC and appraisals. Applying within-person analytic methods to data from a 21-day daily diary study, we provide novel evidence for the role of proximal family relationships in predicting adolescent’s situational threat and self-blame.

As a whole, the findings in this study offer consistent and compelling support for the notion that daily family relationships shape the ways in which adolescents appraise fluctuations in IPC. On days when families are more cohesive and when adolescents feel closer to their parents than usual, adolescents experience less threat and self-blame. Similarly, on days of elevated parent-adolescent conflict, adolescent appraisals of IPC also are elevated. This general finding was present in all of the analyses; however, in several cases, tests of moderation qualified the nature of these daily contextual effects. Guided by the CCF, we hypothesized that family contextual factors would effectively buffer (e.g., family cohesion) or amplify (e.g., MA conflict) the relation between day’s IPC and situational appraisals.

The within-person methods and daily timescale used in this study allowed for comparisons of families to themselves across days and identified robust implications of proximal family relations for adolescent appraisals of IPC. Across analyses, proximal family context, but not distal family context, was a consistent predictor of within-person variation in threat and self-blame appraisals. This suggests that adolescents are particularly attuned to proximal family dynamics when forming interpretations of IPC episodes. However, nuance in our results called for a deeper interpretation of the differences in these processes for girls and boys.

Considerations for Adolescent Girls and Boys

Responding to prior calls for the study of youth gender differences in models of IPC (Fosco & Lydon-Staley, 2019; van Eldik et al., 2020), we considered whether gender might moderate effects. An emergent theme in our analyses was that the broader family context appeared to function in different ways for boys and girls. Before discussing specific findings, it is helpful to first identify the general patterns in our data where gender moderation was found.

For girls, when indications of other relationships were positive (e.g., cohesion is high), or low in conflict (e.g., low MA conflict), girls were less prone to appraise daily IPC negatively. This pattern across our analyses suggests that girls draw on multiple indicators of the family when they formulate their situational appraisals of IPC events. This interpretation is consistent with Davies and Lindsay’s (2004) application of the gender intensification hypothesis (Hill & Lynch, 1983) to guide understanding of gender differences in risk from IPC. During adolescence there may particularly strong socialization pressures to conform to traditional gender roles. From this view, girls may develop a greater propensity toward communion and interpersonal concern for family relationships that can amplify risk from IPC for internalizing problems (Davies & Lindsay, 2004). In light of the current findings, perhaps girls’ strengths in communion help explain how family relationships contextualize their appraisals of IPC. Based on our findings, it seems that girls draw on daily variation in family cohesion and their relationships with parents to guide their evaluations of the threat or blame associated with IPC.

Unlike girls, it seems that boys still attend to IPC, even in the presence of protective (e.g., cohesion) or absence of vulnerability (e.g., MA conflict) factors. However, on days when protective factors were low, or vulnerability was elevated, boys exhibited elevated appraisals, even if IPC was low. It is as though boys, rather than drawing on the broader family context to inform evaluations of IPC events, appraise multiple facets of the family in terms of self-blame or threat (particularly for self-blame). This suggests that boys may have a less-differentiated appraisal process in the family, and that, even on days when IPC was low, boys may feel responsible for problems occurring in the family. These findings also fit with hypotheses that boys are encouraged to develop a sense of agency and an interest in themselves as individuals (Davies & Lindsay, 2004). Boys’ socialized agency may lead them to interpret family events egocentrically, rather than consideration of other family members’ actions. From an intervention standpoint, boys in families with elevated IPC, or elevated parent-adolescent conflict, may benefit from cognitive interventions that help them avoid over-attributing blame to themselves. With these ideas in mind, we turn this discussion to specific family contextual factors of IPC.

Family Cohesion

Analyses examining the role of family cohesion revealed that proximal, but not distal, cohesion was a significant contextual factor for IPC. Three-way interactions indicated that daily fluctuations in cohesion modified the association between day’s IPC and adolescent appraisals differently for boys and girls. For girls, when proximal cohesion was high, days’ IPC was not associated with changes in their threat or self-blame; however, on days of low cohesion, girls experienced more threat and self-blame in response to increases in day’s IPC. For boys, day’s IPC predicted situational threat regardless of daily variation in cohesion. However, differences were evident for self-blame. On high-cohesion days, daily IPC was associated with increases in self-blame; on low-cohesion days, boys’ self-blame was elevated, regardless of day’s IPC.

Taken together, day-to-day variation in family cohesion appears to be an important consideration in understanding the impact of daily IPC on adolescent appraisals. Little work has previously evaluated family cohesion as a contextual factor; however, our findings regarding moderation effects complement those of Lindahl and Malik (2011). At the daily timescale, our results offer guidance to families on days when IPC occurs to preserve family cohesion as much as possible. Optimal family functioning is found when IPC is low and cohesion is high. However, promoting daily cohesion in families may buffer girls from IPC events; for boys, cohesion may still reduce risk for self-blame when IPC is low. Even when interparental conflicts occur, parents’ efforts to compartmentalize disagreements and promote family-wide cohesion can minimize the impact of IPC on adolescent functioning.

Parent-Adolescent Closeness

Closeness with parents appeared to operate differently in MA and FA relationships. Our findings supported MA closeness as primarily an additive protective factor. On days of high MA closeness, adolescents experienced less self-blame, unqualified by moderators and accounting for day’s IPC. Moreover, on days when boys felt closer to their mothers, they also found IPC less threatening; however, the MA closeness-threat link was not evident for girls.

FA closeness was a moderator of the IPC-threat relation, but only in magnitude. On days of closeness to fathers, youth were slightly more prone to appraise IPC as threatening than on days when they felt less close to their fathers. In relation to self-blame, a three-way interaction for FA closeness emerged, in a pattern similar to family cohesion. On days of high FA closeness, girls were less prone to blame themselves for IPC, whereas boys were more responsive to IPC.

Although nuanced, these findings emphasize the importance of daily fluctuations in parent-adolescent closeness, as is reported elsewhere (Fosco et al., 2019; Lippold et al., 2016) and as a contextual factor for IPC. The findings that MA closeness operated as an additive protective factor is largely consistent with past work (DeBoard-Lucas et al., 2010; Grych et al., 2004; McDonald et al., 2012); however, the within-person methods used in this study suggest that interventions may profitably advise mothers to intentionally foster closeness with their adolescents on days when IPC occurs.

Our findings regarding FA closeness raise important questions as to why, in some cases, FA closeness was associated with increases in adolescents’ tendency to appraise IPC events negatively. Regarding findings for threat, it may be that adolescents are more likely to appraise IPC as threatening on days with higher FA closeness, due to concerns that the IPC episode may jeopardize their relationship with their father, or perhaps elicit fear that IPC will present a loyalty conflict (Grych et al., 2004). However, findings regarding FA closeness and self-blame mirror gender differences described earlier, supporting ways in which girls may benefit from FA closeness and boys may be more likely to evaluate IPC from an egocentric point of view. Overall, future work replicating these results are critical in understanding these mixed results.

Parent-Adolescent Conflict

Vulnerability models – in this study on the role of parent-adolescent conflict in the effects of IPC – also offer insight into how risk is conferred. Turning first to MA conflict models, distinct patterns emerged for boys and girls and were consistent across models predicting threat and self-blame. For girls, day’s IPC was more strongly associated with threat and self-blame on days of elevated MA conflict than on days when MA conflict was low. This suggests that girls are particularly attuned to conflict with their mothers when evaluating the significance and causes of IPC. For boys, day’s IPC corresponded to threat and blame in the absence of MA conflict, but when MA conflict was high, boys experienced elevated appraisals, irrespective of IPC. Thus, it seems that IPC and MA conflict both inform boys’ appraisals.

FA conflict was less nuanced in its role: adolescents in conflictual relationships with their fathers were more vulnerable to daily IPC, regardless of adolescent gender. On days of elevated FA conflict, adolescents consistently experienced elevated threat appraisals. However, FA conflict amplified the within-person relation between daily IPC and self-blame. These findings point to the importance of mitigating FA conflict for adolescents coping with IPC.

Findings in this study underscore the impact of parent-adolescent conflict in shaping adolescents’ evaluations of IPC. This expands on existing work documenting correlations between harsh parenting and appraisals (DeBoard-Lucas et al., 2010; Selçuk et al., 2020), and demonstrates within-person associations between fluctuations in parent-adolescent conflict and adolescents’ active processing of IPC on a daily timescale, underscoring the importance of parents’ compartmentalization of IPC. On days when IPC occurs, parents are more likely to engage in conflict or harsh parenting with their children (Almeida et al., 1999; Kouros et al., 2014). Integrating across these studies, on days when both IPC and parent-adolescent conflict occur, adolescents are at pronounced risk for perceiving conflict as threatening or blaming themselves for parental discord. Thus, the cumulative literature converges to highlight the importance of intervention strategies that support adolescent well-being by aiding parents’ awareness and providing alternatives to emotional transmission following IPC episodes.

Limitations and Future Directions

This study, although advancing the study of family contextual factors in relation to adolescents’ appraisals of IPC, has important limitations. First, this study focused on family contextual factors, but was not exhaustive. Other family factors, such as family chaos or sibling relationships also may contextualize daily findings (Lucas-Thompson & George, 2017). Further, other factors may contribute to the ways in which adolescents appraise IPC, such as temperament (Davies et al., 2016) or ecological factors such as peer relationships, experiences in school, and community or neighborhood characteristics (Fosco et al., 2007; McCauley, Weymouth, Feinberg, & Fosco, 2019), which warrant future consideration. Second, although adolescent gender accounted for differences in our findings, future studies that evaluate potential underlying mechanisms (e.g., orientation toward agency or communion) for boys and girls would advance this work. Additionally, given the novelty of our gendered findings, replication is needed. Third, future work also should consider issues of diversity in terms of SES, race/ethnicity, and sexual orientation in replicating the current findings. Fourth, the modest sample size required separate tests of family contextual factors and may be vulnerable to Type I error; studies with larger samples would also be able to evaluate multiple contextual factors in the same analytic model. Fifth, we were limited to adolescent self-report data; replication using multi-method and multi-informant designs would further bolster confidence in these findings.

Conclusion

This study provides a first test of proximal family contextual factors in relation to adolescents’ daily appraisals of IPC. Building on prior work establishing threat and self-blame appraisals as key pathways of risk for a wide range of problem outcomes, our results provide robust evidence for the roles of family cohesion, parent-adolescent closeness, and parent-adolescent conflict as contributors to adolescents’ threat and self-blame appraisals of IPC. Leveraging 21-day daily diary methods, it was possible to evaluate within-person effects that offer valuable insight, unencumbered by potential third-variable concerns, into daily processes that may reduce the problematic impact of interparental conflict on adolescent adjustment. Adolescent boys and girls exhibited distinct patterns of risk in this sample, highlighting the need for more work on gender differences to better understand these results.