Abstract
Relationships with parents, siblings, and friends are salient contexts to study young adolescent adjustment. Internal working models of relationships develop based on early experiences and provide a framework for how individuals navigate later relationships. Despite the enduring and impactful nature of sibling relationships, little research to date focuses on identifying patterns of relationship quality with siblings while in the presence of other relationships. The present study aimed to address this gap by (1) identifying profiles of co-occurring relationship types as defined by relational positivity and negativity, and (2) examining whether these profiles relate to internalized adjustment and prosocial behaviour. The current sample included young adolescents (n = 388) between 10 and 13 years of age (M = 10.87, SD = 0.73). Latent profile analyses yielded two distinct profiles of relational positivity and negativity: (1) high-quality (n = 244) and (2) discordant (n = 144). Compared to the high-quality profile, participants in the discordant profile demonstrated greater negativity and lower positivity across relationships, along with intense sibling relationships. Results also showed that participants with high-quality relationships were better adjusted psychosocially than those with discordant relationships. This study emphasizes the dynamic interplay of co-occurring relationships, and particularly the unique developmental role of siblings. Overall, the results indicate that future studies should consider multiple co-occurring relationships to advance our understanding of adolescent relationship networks and their influence on socio-behavioural functioning.
Introduction
Prior research highlights that relationships are critical for adolescent development and well-being, particularly those held with parents, siblings, and friends (Bukowski et al., 2019; Howe et al., 2020; Jensen et al., 2023; McHale et al., 2012). The first form of close relationships that children develop is typically with their parents or primary caregiver(s) (Ainsworth & Bell, 1970; Ainsworth et al., 1978; Bowlby, 1988). Over time, youth develop an understanding of how relationships ought to function, which informs the way in which they interact with subsequent and co-occurring relationships. The current literature highlights the importance of relationships in development, as central socialization contexts through which emotional regulation, behavioral norms, and self-concept are shaped. For example, parent-child, sibling, and friend relationships have been consistently linked to adolescents’ socio-behavioural and adaptive well-being, particularly the development of internalizing (e.g., anxiety, depressed affect, withdrawal) and externalizing (e.g., aggression, conduct problems, attentional difficulties) challenges (Alsarrani et al., 2022; Dirks et al., 2015; Laursen & Mooney, 2008). However, many of these studies either examine specific relationship dyads in isolation or exclude siblings from consideration. The purpose of this study was to directly address this limitation and identify different profiles of co-occurring relationship types as they may relate to subjective well-being and prosociality in a sample of young adolescents (Delgado et al., 2022; Thompson, 1998).
Relationships With Parents
According to the social convoy model, individuals are surrounded by a dynamic network or system of social relationships that provide support and evolve across the lifespan (Antonucci et al., 2014). This model posits that these relationships are multidimensional, and their quality is influenced by personal and situational characteristics and have important implications for well-being and adjustment (Antonucci et al., 2014). The model further conceptualizes these relationships as structured and hierarchical, varying in closeness and function, and changing over time in response to individual and contextual factors. As such, the model emphasizes that the quality and functional capacity of social ties, rather than their mere presence, are central to understanding well-being and adjustment. One of the closest relationships within this model is the parent-child relationship. From there, other close relationships remain close to the child (e.g., siblings, friends), with other less influential but existing relationships playing tertiary roles (e.g., teachers, peers, relatives).
Early theorists emphasized the importance of young children developing a “secure base,” or strong attachment to a parent or caregiver, for building self-confidence and developing an internal working model - or cognitive schema - that acts as a foundation for their relational world that may impact how they approach new close others (Bowlby & Bowlby, 2012; Bretherton & Munholland, 2008; Delgado et al., 2022; Furman et al., 2002). These theories formed the foundation for subsequent conceptualizations of relationship quality, which are defined by two orthogonal dimensions reflecting relational positivity, including closeness, trust, and warmth, as well as negativity, which includes using conflict, criticism and antagonism (Banny et al., 2011; Dryburgh et al., 2022; Furman & Buhrmester, 1985; Laursen & Mooney, 2008). High-quality adolescent-parent relationships (i.e., high positivity, low negativity) have been found to be associated with reduced internalizing and externalizing problems, which can result in better adjustment and increased resilience to stressors (Skinner et al., 2021). Conversely, parent-adolescent relationships that have higher levels of conflict have been associated with lower self-esteem, negative mood, and greater internalizing behaviours (Laursen & Mooney, 2008; McCormick & Kennedy, 1994). As children enter adolescence, past relational experiences with parents may act as their relationship framework for their siblings and friends.
Sibling and Friend Relationships
Similar to parents, the social convoy model would suggest that sibling relationships are also considered to have a significant impact on children and adolescents. Siblings tend to be the first type of similar-age relationships that children develop and, in many ways, provide an avenue for practicing socialization and providing support outside of primary caregivers (Jensen et al., 2023). While this area of research is particularly understudied relative to parent-child relationships, the literature indicates that siblings help one another to develop skills (i.e., conflict resolution, play, social cues, confidence) and further construct internal working models that can be later strengthened in the context of friendships (Howe et al., 2020; Jensen et al., 2023; McHale et al., 2012). Unlike friendships, sibling relationship quality may be directly influenced by familial relationships, though the precise mechanisms remain obscure and may interact with several confounding factors (e.g., family structures including birth order, gender of dyad) (Howe et al., 2020; Jensen et al., 2023).
Closer and high-quality sibling relationships are linked to lower rates of depression, greater resilience, better adjustment, and conflict resolution (Howe et al., 2020; Jensen et al., 2023; McHale et al., 2012; Milevsky & Levitt, 2005). Additionally, positive relationships with siblings may act as a buffer for negative internalizing and externalizing outcomes arising from other stressors (i.e., bullying, mistreatment) (Fry et al., 2021; Milevsky & Levitt, 2005; Persram et al., 2023). Contrastingly, low quality (i.e., low positivity and high negativity) or affect intense (i.e., high negativity and high positivity) relationships with siblings are related to riskier and antisocial behaviours, and poorer conflict resolution (Jensen et al., 2023; McHale et al., 2012). Nevertheless, understanding the quality of sibling relationships within the broader family system continues to produce mixed findings as the structural characteristics of sibling relationships provide different social experiences across families, such that similar patterns of warmth or conflict may support development in some contexts (e.g., stressful events) while contributing to difficulties in others (Jensen et al., 2023; Milevsky & Levitt, 2005).
Over time, adolescents’ social networks evolve and expand, with friendships being added to the existing foundation of familial relationships within the social convoy. During adolescence, friendships become an important addition to the overall relational system and can be approached in line with previously developed internal working models based on familial relationships (Delgado et al., 2022; Groh et al., 2017; Milevsky & Levitt, 2005; Thompson, 1998). Structurally, friendships differ from sibling and parent relationships as friendships are voluntary relationships with individuals at or around the same age, that are built on positivity, trust, and mutual reciprocity (Bukowski et al., 2019). Like sibling relationships, friendships aid in social learning and skill building, emotional support, and companionship (Bagwell & Bukowski, 2018). High-quality friendships are related to greater subjective well-being and self-esteem, and reduced loneliness and depression (Alsarrani et al., 2022; Schwartz-Mette et al., 2020). Conversely, low quality or unstable friendships are typically associated with poor internalizing and externalizing difficulties, as well as poorer conflict resolution (Lodder et al., 2017; Schwartz-Mette et al., 2020; Waldrip et al., 2008). Greater emphasis on certain relationships can be especially present in adolescence as friendships may now be perceived as more valued, prompting the development of autonomy and identity, and providing a sense of belonging (Laursen & Bukowski, 1997; Schwartz-Mette et al., 2020).
The function of sibling relationships and friendships may appear similar; however, adolescents may place higher value on incorporating prosocial behaviour and meaningful conflict resolution with friendships, whereas siblings may be considered more constant and permanent relational figures, having previously been described as the longest lasting social relationship (Jensen et al., 2023). Moreover, high-quality friendships during young adolescence may help to compensate for negativity across close relationships with family and friends (Rubin et al., 2004). However, there is limited and mixed work to date on specifically understanding internal working models for sibling relationships with these other co-occurring relationships, and a gap in understanding what this means for relationship development and adjustment in adolescence.
Relational Congruence and Incongruence Across Relationships
While each of parent-child, sibling, and friend relationships can be seen as distinct, separate relationships, it is crucial to recognize their co-occurring and interdependent nature. From a family systems and socio-ecological perspective, individuals are embedded within multiple, overlapping relational contexts that operate simultaneously and influence one another, such that quality of relationships are best understood as part of a broader relational environment rather than in isolation (Bowen, 1978; Bronfenbrenner, 1979). Similarly, the social convoy model emphasizes that individuals’ networks of concurrent relationships can vary in structure and quality, and that these configurations may carry distinct implications for psychosocial adjustment (Antonucci et al., 2014). Therefore, it is possible that these relationships function similarly; that is, they maintain relational congruence (Bowlby & Bowlby, 2012; Bretherton & Munholland, 2008; Delgado et al., 2022). For instance, positive and high-quality relationships with parents may spill over into other congruent high-quality relationships with siblings and friends (Dishion et al., 2019; Oh et al., 2021). Importantly, congruent high-quality relationships (i.e., mother, father, friends) are related to better psychosocial adjustment (i.e., fewer emotional difficulties and behavioural challenges including aggressive and rule breaking behaviours; Laursen & Mooney, 2008). Conversely, there may also be relational incongruence, such that qualities of relationships may not remain consistent between all dyadic relationships (Oh et al., 2021; Sherman et al., 2006). McHale et al. (2012) describe some sibling relationships as having a combination of positivity and negativity that reflect an ‘affect-intense’ style of relationship, which is defined as having high levels of both relational positivity and negativity. As such, certain relationships may contribute to incongruence across relationships. These patterns highlight that relational congruence is not universal and that different relationship contexts may play distinct roles within an individual’s broader relational ecology.
Incorporating person-centered analyses to classifying dyadic relationships through their quality and congruence has become a prominent method of understanding types and outcomes of relationship structures. Work by Oh et al. (2021) determined cross-cultural profiles of relationships in adolescents as they relate to socio-behavioural functioning. They broadly identified four profiles or types of relationships, including high-quality (high positivity, low negativity), moderate quality (moderate negativity and positivity), low quality (lower positivity, higher negativity), and discordant (high levels of both positivity and negativity) in relationships across youth reports of maternal, paternal, and friendship quality. Moreover, adolescents in the discordant profile were nominated by peers as being less prosocial (e.g., less helpful or fair), and displaying more conflict compared to those in the high-quality profiles. These findings underscore the nuance in relationship networks while also pointing to the need to further understand and classify relationships as they relate to socio-behavioural outcomes. The importance of relationship congruence and quality between adolescent and sibling relationships has yet to be explored and are not typically included in studies on networks of relationships in adolescents, providing a key gap in the existing literature.
Present Study
The purpose of this study is twofold. First, we sought to address prior limitations in examining relationship quality between adolescents and their siblings to identify subpopulations of youth who hold similar perceptions of relationships with each of their mothers, fathers, siblings and friends. Drawing on person-centered analyses, namely latent profile analysis, we aimed to better understand the consistency between each of the relationships that are close to adolescents’ lives (Laursen & Hoff, 2006). This approach moves beyond traditional variable-centered approaches that assess relational interactions and allows us to empirically test whether adolescents differ meaningfully in the configuration of relationship quality across these concurrent relational contexts, consistent with the social convoy model. Specifically, this study examined the degree to which youth held congruent or incongruent perceptions of these critical relationships, and subsequently, what predicted membership within these relational profiles. Second, we sought to determine how these profiles vary on measures of socio-behavioural outcomes, namely internalizing difficulties (i.e., depressed affect, anxiety, avoidance), aggression, victimization, and prosocial behaviour (i.e., justice, care). By linking distinct configurations of relationship quality to socio-behavioural functioning, we examined whether the structure of adolescents’ concurrent relationship networks is related to adjustment beyond overall relationship positivity alone. Given that the function of sibships and friendships are believed to be similar during adolescence, we expected that the number of profiles would comparable to other studies’ findings on relationships between parents and friends (e.g., 2–4 profiles; Hiatt et al., 2015; Oh et al., 2021; Tang et al., 2023). Further, we also hypothesized that those in the high-quality profile (i.e., high positivity, low negativity) would report better socio-behavioural outcomes compared to those in other identified profiles.
Methods
Participants
Our sample was generated from a larger study (e.g., Persram et al., 2021; Wood et al., 2017, N = 430) on early adolescent relationships. Among the total sample, 9% had no siblings and 14% of parents completing the demographics questionnaire did not report information regarding the number of siblings. Approximately 7% of the total sample had one sibling, 45% had two siblings, 21% had three siblings, and 4% had four or more siblings. For this study, our analytic sample consisted of 388 fifth- (n = 178) and sixth-grade (n = 210) students (N = 205 boys and 183 girls) with at least one sibling from 19 classrooms in three mixed-gender public schools located in a large city in Quebec, Canada. Schools were from the same regional school board and selected based on school administrators’ agreements to participate. School principals were contacted by the research team to determine their interest in participating in this research. English was typically the language used for instruction and between children during peer interactions. All early adolescents were included in the larger study, and only those who received permission to be involved in the study were included. The minimum participation rate was 80% in each classroom.
The ages of these students ranged from 10–13 years (M = 10.89 years, SD = 0.73) and included lower- and upper- middle socioeconomic class students. The marital status of parents were intact heterosexual marriages (57.5%) while the remaining ranged between divorced or single-parent families. All participants reported having cisgender heterosexual parents. Parental education for mothers (M = 14.33 years, SD = 2.87) and fathers (M = 14.62 years, SD = .2.26) was comparable, Mean father income was $64,778.48 (SD = $29,601.49) and mean mother income was $39,489.44 (SD = $26,262.37).
Measures
Relationship Quality. The Network of Relationships Inventory Social Provisions Version (NRI-SPV; Furman & Buhrmester, 1985) was used to assess relationship quality across all four dyads (mother, father, sibling, friend) to develop scales of positivity and negativity. This 36-item questionnaire asks participants to rate the extent to which their mother, father, closest-age sibling, and their closest friend in school on subscales of positivity (affection, reliable alliance, enhancement of worth, intimacy, instrumental help, companionship, and nurturance of others) and negativity (conflict, antagonism, punishment, and criticism). Each item is rated on a Likert scale with key anchor points at 1 (almost never), 3 (sometimes), and 5 (almost always). Subscales were computed by averaging the items within its respective measure only if participants rated at least two of the three items. Relational positivity and negativity scores were calculated by aggregating the subscales that represented each dimension for each of the four relationships. Mean scores were not computed for participants missing more than two subscales. Cronbach’s alpha within this sample revealed good to excellent internal consistency across scales of positivity (α = .90–.95) and negativity (α = .83–.89).
Self-Rated Anxiety. Participants rated three items intended to measure feelings of anxiety on a 5-point Likert scale ranging from 1 (never true) to 5 (always true) (Panarello & Bukowski, 2021; Wood et al., 2017). These items were “I am nervous or tense” “I worry a lot” and “I get stressed a lot”. The reliability of the self-rated anxiety measure was considered good for the present study (α = .79).
Self-Rated Depressed Affect. Participants were instructed to rate how much each of the three items intended to measure their self-rated depressed affect described them on a 5-point Likert scale ranging from 1 (never) to 5 (almost always) (Persram et al., 2023; Wood et al., 2017). These items were “I feel lonely” “I feel sad” and “I feel unhappy.” Higher total scores indicated greater self-rated depressed affect. Cronbach’s alpha within this sample was acceptable (α = .76).
Peer Nominations of Socio-Behavioural Outcomes. Participants completed peer assessments, where they nominated who in their class fit characteristics that were consistent with measures of relational aggression (e.g., “Someone who talks bad about others behind their backs to hurt them”), physical aggression (e.g., “Someone who gets involved in physical fights”), victimization (e.g., “Others do mean things to him/her”), avoidance (e.g., “Someone who would rather play alone than with others”), anxiety (e.g., “Someone who is nervous or tense”), depressed affect (e.g., “Someone who is sad”), care (“Someone who helps other people with their problems), and justice (e.g., “Someone who makes sure that everyone is treated equally). Four sets of measures of socio-behavioural outcomes were tested: (a) internalizing difficulties (i.e., depressed affect, anxiety, avoidance), (b) aggression perpetration, (c) victimization, and (d) prosocial behaviour (i.e., justice, care). Each participant received a mean score indicating how often they were nominated by their same-gender classroom peers on each of the behavioural characteristics. These scores were then adjusted to account for class size (see Velásquez et al., 2013), and mean scores were calculated. Cronbach’s alpha ranged from fair to excellent: (a) internalizing difficulties (.83), (b) aggression perpetration (.88), (c) victimization (.87), and (d) prosocial behaviour (.85)
Procedure
Ethics approval was granted by Toronto Metropolitan University Research Ethics Board (#2024-236). An information letter and a parental consent form was sent home to parents who signed and returned it to the classroom teacher indicating whether they consented or not to having the child participate in the project. In addition, parents received a demographics questionnaire in which they provided information related to the family structure. Participant assent was also obtained before moving forward with the questionnaires after receiving parental consent. Participants were given a small reward of school supplies for returning the consent form regardless of whether parents gave permission for them to be in the study and those who completed the study received a t-shirt.
Plan of Analysis
Primary analyses were conducted using Mplus version 8.6 (Muthen & Muthen, 1998). A series of latent profile analyses (LPA) were performed with the number of classes ranging from 1 to 6 using the robust maximum likelihood estimator. A person-centered approach to LPA was selected because it enables the identification of unobserved subgroups based on shared response patterns across multiple relational indicators (i.e., quality, congruence), allowing for the examination of whether distinct relationship profiles are differentially associated with study outcomes (Kaniusonyte & Laursen, 2021; Laursen & Hoff, 2006). The indicators in these models were the observed scores for relational positivity and negativity across the four relationships. Each LPA was estimated with 5000 random starting values and 1000 iterations. Means and variances were estimated freely within each profile. Evaluation of the model with the best number of profiles was determined using several indices, including the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), and the sample-size adjusted BIC (ABIC), where lower values in successive models indicated a better fit (Ferguson et al., 2020). In addition, the adjusted Lo-Mendell-Rubin likelihood ratio test (LMR) compared the model with k profiles with the model with k – 1 profiles, with nonsignificant values indicating that the model with k – 1 is preferred (Lo et al., 2001). Entropy was also examined as an index of classification accuracy, with higher values closer to 1 indicating greater certainty in profile assignment; however, entropy was not used as a primary criterion for model selection. Models that contained profiles with less than 5% of the sample membership were rejected (Ferguson et al., 2020). Using the R3STEP procedure in Mplus, the effect of covariates (i.e., gender, age, grade) was analyzed within the LPA. Finally, we tested differences between the profiles on several socio-behavioural outcomes. To accomplish this, we evaluated profile differences on measures of peer-assessed aggression, avoidance, and prosociality as well as self-reported internalizing difficulties using the auxiliary Bolck-Croon-Hagenaars (BCH) command in Mplus. This method allows to test significant differences in socio-behavioural outcomes as a function of profile membership (Bakk & Vermunt, 2016; Bolck et al., 2004).
Across the 36 items of the NRI, item-level missingness ranged from 2.09% to 16.74%. Missingness was addressed using multiple imputation, in which 100 datasets were imputed at the item level. Using this imputed dataset, items were aggregated by subscale and subsequently used to compute composite scores of relational positivity and negativity for each of the four relationships (i.e., mothers, father, siblings, friends). Following imputation and aggregation, residual missingness as a result of insufficient data for imputation was limited to composite scores of relationship positivity and negativity related to sibling relationships and friendships (0.55%). A Little’s MCAR test showed that our imputed data was missing completely at random, χ2(20) = 24.10, p = 0.24. When performing LPA, remaining missing data for composite scores was handled in Mplus using the Full Information Maximum Likelihood (FIML) estimation (Muthen & Muthen, 1998).
Results
Descriptive Statistics for Relationship Quality
Descriptive Statistics and Correlations Among Relationship Quality Variables
Note. *p < .05, **p < .01, ***p < .001
Latent Profile Analysis
LPA Model Selection Indices
Scree Plot depicting AIC, BIC, and ABIC values
(a) Relationship positivity and negativity for 2-profile solution (b) Relationship positivity and negativity for the 3-profile solution
Within-Profile Differences in Relationship Quality Across Relationship Dyads
Note. Significant differences in relationship quality (i.e., significant difference between relational positivity and negativity) for each relationship dyad in the high-quality and discordant profiles, respectively (p < .05).
Between-Profile Differences in Relationship Quality
Note. Significant differences for each relationship dyad between high-quality and discordant profiles (p < .05).
Following the identification of the 2-profile model, we included age, gender (boy, girl), and number of siblings as covariates to test whether these variables predicted profile membership using the R3STEP procedure. Using the high-quality profile as the reference class, no covariates were found to be statistically significant predictors of membership in the discordant profile. An exploratory LPA was further performed to understand how the presence of siblings may influence profile membership (see Supplemental Materials [Table S2]). Interestingly, profile membership was not found to be solely driven by the quality of sibling relationships, suggesting that siblings may not be the primary source of these differences in relationship quality, but may be the clearest display of discordance and high conflict. For instance, discordance in relationships was still present with and without siblings, but siblings may serve as a lens through which adolescents’ discordant relational functioning become more apparent through more affect-intense interactions.
Profile Differences in Socio-Behavioural Outcomes
Test of Equality of Means Across the Two Latent Profiles
Note. Statistically significant differences between profiles (p < .05) indicated with asterisks.
Exploratory Analysis Using Traditional Variable-Centered Approaches
To further support our analytical approach, we performed separate hierarchical linear regressions, a common variable-centered approach, to assess whether there were additive and interactive effects of perceptions of relationship quality across relationships on each of the socio-behavioural outcomes. We first assessed these associations with relationship positivity and found that only father positivity was associated with less self-reported anxiety (β = −.18, p = .02) and depressed affect (β = −.18, p = .02). After accounting for all possible 2-way and 3-way interactions, the 4-way interaction between relationship positivity across each of the relationships was not significant for either anxiety (β = .06, p = .49) or depressed affect (β = .09, p = .31). For the peer-assessed outcomes, mother positivity was significantly related to greater justice (β = .22, p = .002) and care (β = .21, p = .002), but less avoidance (β = −.17, p = .01) and depressed affect (β = −.17, p = .02). Sibling (β = .14, p = .03) and friend positivity were independently associated with greater care (β = .14, p = .03). Neither of the 4-way interactions between relationship positivity and each of the outcomes were statistically significant (all ps > .11).
We then examined these associations with relationship negativity and found no significant additive or interactive effects on self-reported anxiety and depressed affect (all ps > 13). Similarly, we did not find statistically significant additive or interactive effects of relationship negativity on any of the peer-assessed socio-behavioural outcomes (all ps > .05). These findings suggest that the influence of relationship negativity may not be adequately captured by linear additive or interaction terms and may instead depend on general patterns of experiences in relationships. Thus, the lack of interactive effects using variable-centered approaches further supports the need to consider the relational profiles identified here (i.e., high-quality and discordant) using person-centered LPAs, which allow for the identification of distinct configurations of relational experiences that may differentially relate to adjustment outcomes. Variable-centered approaches in this context may miss specific interactions across the four simultaneous relationships (i.e., 4-way interactions).
Discussion
The present study examined how relationship quality between young adolescents and their mother, father, closest-age sibling, and friends may be classified into distinct profiles, with each profile being associated with key socio-behavioural outcomes. By using a person-centered approach in this study, we highlight the features that comprise these relationships as aspects related to the individual and not entirely as separate drivers that impact overall changes in youth development (Laursen & Hoff, 2006). As such, we discuss our findings within the context of the subpopulations of youth identified in our analyses rather than with a traditional additive and interactive relational lens.
In this study, latent profile analyses revealed two distinct profiles. The first profile was labelled as high-quality, characterized by consistent high relationship positivity and low relationship negativity across each of the four relationships, while the second profile was termed discordant, as evidenced by similarly high positivity and higher negativity than the high-quality profile. It is important to note that the discordant profile does not demonstrate low quality relationships, rather, it highlights relationships with high positivity and higher degrees of negativity, tension, or conflict. Though the discordant profile highlighted a distinct affect-intensity in sibling relationships (i.e., high positivity and negativity), the overall profile showed general greater negativity across relationships when compared to the high-quality profile. While high-quality and discordant profiles have been similarly found in prior research (Oh et al., 2021), this study adds to the literature by identifying relationship profiles that include the unique contribution of siblings.
Across both profiles, mother, father, and friend positivity and negativity characterized high-quality relationships, suggesting congruence across these networks. In line with the internal working model and social convoy model (Antonucci et al., 2014; Bowlby & Bowlby, 2012; Bretherton & Munholland, 2008), when primary familial relationships are built on high-quality interactions and maintain high positivity and low negativity, a prosocial internal working model for relationships may be developed that has a spillover effect to additional members (i.e., friends) across the overall relational system. While siblings had the highest level of negativity compared to other relationships within the high-quality profile, the difference between positivity and negativity in this relationship remained congruent with the overall relationship network, demonstrating an overall congruent relational system. This may be due to the nature of sibling relationships, which encompass both strong positive and negative interactions (Howe et al., 2020; Jensen et al., 2023; McHale et al., 2012).
Our findings also showed that in the discordant profile, adolescents experienced more negativity with their mothers, fathers, siblings, and friends and less positivity with their mothers, fathers, and siblings relative to the high-quality profile. Notably, friendship positivity remained comparably high across profiles, distinguishing friendships from familial relationships. Interestingly, sibling relationships in the discordant profile specifically highlighted the point of incongruence within this network of relationships by exhibiting both high positivity and negativity. Relationships with siblings may be special in that they involve both high positivity and high negativity and offer an environment to express themselves in an unfiltered manner and for practicing conflict resolution. Unlike friendships, which are voluntary and built on reciprocity and positivity, sibling relationships are involuntary, longer lasting, and embedded within a shared family environment (Jensen et al., 2023; McHale et al., 2012). Thus, sibling relationships within discordant familial networks may provide opportunities to engage in more intense conflict styles (Jensen et al., 2023; McHale et al., 2012), as there may also be less at stake for youth due to this relationship’s enduring nature. Therefore, when considering the internal working model of relationships, it can be theorized that when a higher degree of conflict, antagonism, punishment, and criticism exist within primary relationships (i.e., parents), this negativity may extend into sibling relationships, resulting in incongruent or affect-intense relationships within family systems. Notably, within the discordant profile, friendships appeared to retain relatively high levels of positivity despite increased negativity, suggesting possible spillover from sibling negativity into other familial relationships, and a distinct relational pattern for friendships compared to familial relationships. This may indicate that youth maintain positive features of friendships even in the context of broader relational discord.. This finding highlights the separate and purposeful nature of maintaining high-quality extrafamilial relationships, such as friendships, during young adolescence (Bagwell & Bukowski, 2018), which may help to achieve balance or compensate for negativity across their relationship network (Rubin et al., 2004).
Additionally, there is strong evidence in the literature for the association between relationship quality and socio-behavioural outcomes (e.g., Alsarrani et al., 2022; Laursen & Mooney, 2008; McHale et al., 2012). In our study, we extend this work to now include sibling relationships as another key relational context that is associated with youth development. Our findings revealed significant differences across profiles that were largely in support of our hypotheses. Specifically, young adolescents in the discordant profile had higher self-rated internalizing difficulties (i.e., depressed affect, anxiety) and peer-assessed victimization, whereas those in the high-quality profile had exhibited more prosocial behaviours (i.e., justice, care). In line with previous work, there is clear evidence for high-quality relationships with parents, siblings, and friends, though often not considered simultaneously, as being critical to the adjustment and development of youth (Alsarrani et al., 2022; Jensen et al., 2023; Laursen & Mooney, 2008; Lodder et al., 2017; Schwartz-Mette et al., 2020; Waldrip et al., 2008). Taken together, these findings suggest that discordance in close relationships may place young adolescents at heightened risk for emotional distress and reduced prosocial behaviours, whereas consistently high-quality relationships may promote overall adjustment.
Interestingly, there were no profile differences in aggression or avoidance, or relationships that were mostly negative within this sample. This is likely due to overall low base rates on these indices, the demographic make-up of the sample, or an overall well-adjusted sample of young adolescents. In the case of the latter, one would expect that a conflictual profile, characterized by high negativity and low positivity, would experience more adverse outcomes. Participants in this sample largely reported positively valenced relationships, which may speak to substantive qualitative changes adolescents experience in their relationships that may minimize the conflicted nature of relationships that exist in childhood.
Additionally, while prior literature notes that positive relationships between siblings relate to lower rates of internalizing and externalizing difficulties resulting in better adjustment and conflict resolution (Jensen et al., 2023), young adolescents with affect-intense sibling relationships may not experience these positive benefits to the same degree. Similarly, greater negativity in sibling relationships is associated with increased negative externalizing outcomes (Buist & Vermande, 2014; Dirks et al., 2015), therefore, more experiences of positivity with siblings may work to protect or buffer against these negative outcomes (Milevsky & Levitt, 2005).
Implications
This study presents unique implications for theory, research, and practice. The findings highlight the value of understanding adolescent relationships as part of a broader network of relationships, rather than focusing on individual relationships in isolation. By identifying high-quality and discordant relationship profiles across parents, siblings, and friends, the findings advance how we understand the interactions and influence of relational systems on socio-behavioural outcomes. In particular, the unique role of sibling relationships, often overlooked by research and marked by both strong positivity and strong negativity (Jensen et al., 2023; McHale et al., 2012), further suggests the importance of examining how adolescents experience and manage relationship intensity, conflict, and leverage prosocial behaviours.
These findings support both the internal working model and social convoy model, highlighting how consistent relational patterns within the family unit may extend outward to peer relationships (Antonucci et al., 2014; Bowlby & Bowlby, 2012; Bretherton & Munholland, 2008). However, the persistence of high-quality friendships in the context of discordant familial relationships suggests that adolescents may also actively seek out and maintain positive peer connections as a form of compensation within higher conflict atmospheres at home. Therefore, this study adds an important step by highlighting the critical value of nonfamilial relationships for youth, particularly during a developmental period where there is a shift in priorities and greater emphasis on identity formation outside of the family system. The study of adolescent relationships tends to miss the additional component that interactions with siblings bring to relationship development, particularly as we consider socio-behavioural development and adjustment. Taking a holistic perspective, individuals are influenced by the sum of their network of relationships, with unique contributions from each individual dyad.
From a practical perspective, these results suggest that interventions targeting adolescent well-being should consider the broader relational system, particularly the nuanced role of sibling relationships. In particular, these sibling relationships may act as practice for exploring conflict styles and conflict resolution that may stand in isolation compared to other dyadic relationships for adolescents. Affect-intense sibling dynamics may signal some underlying familial discord and be linked to greater internalizing difficulties and peer victimization. Supporting adolescents in navigating these relationships, while reinforcing prosocial peer connections, may help mitigate risk and promote resilience and adaptability.
Limitations & Future Directions
While this study provides valuable insights into relational profiles of young adolescents and their related socio-behavioural outcomes, there are several limitations to be acknowledged. Prior literature demonstrates that there may be several reasons for differences in relationship quality, particularly within family systems such as divorce, birth order, and parenting styles (Kaniusonyte & Laursen, 2021; McHale et al., 2012
Conclusion
This study offers a more holistic understanding of young adolescent relational experiences by identifying distinct relationship profiles across parents, siblings, and friends, and linking these to socio-behavioural outcomes. This is the first study to identify high-quality and discordant relationships in adolescent relationship profiles that include sibling dyads, which underscores the importance of considering not just individual relationships, but the broader configuration and quality of relational networks in young adolescence. Notably, the role of siblings emerged as both distinct and influential, offering insight into how adolescents practice and navigate both conflict and connection within their familial environment.
In sum, our findings indicate that early relational experiences within the family shape the tone of other relationships, including friendships. Yet, the persistence of high-quality friendships even in the context of discordant familial relationships highlights adolescents’ active agency in seeking out supportive connections. These patterns have meaningful implications for intervention efforts, which should take into account the dynamic interplay of co-occurring relationships, particularly the unique developmental role of siblings. Ultimately, this research advances our understanding of adolescent relationship networks and their impact on socio-behavioural functioning. It also points toward important avenues for future inquiry, including the value of longitudinal, context-rich investigations that explore how these relational profiles change and impact development over time.
Supplemental Material
Supplemental Material - Typologies of Young Adolescents’ Close Relationships With Parents, Siblings, and Friends: Predictors and Outcomes
Supplemental Material for Typologies of Young Adolescents’ Close Relationships With Parents, Siblings, and Friends: Predictors and Outcomes by Aaron Palachi, Samantha Gauvreau, Gerly Ngyema Nang, Nicole S. J. Dryburgh, William M. Bukowski, Ryan J. Persram in Journal of Social and Personal Relationships
Footnotes
Acknowledgments
This research was supported by a grant awarded to 5th author by the Fonds de recherche du Québec: Société et Culture. We wish to thank the schools, teachers, principals, parents, and participants who supported this work. We gratefully acknowledge support from those who supported the data collection, especially Gordon Rosenoff, Felicia Meyer, and Jonathan Santo.
Ethical Considerations
Ethics approval was granted by the Toronto Metropolitan University Research Ethics Board for secondary analysis on June 19th, 2024 (REB#2024-236).
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant awarded to 5th author by the Fonds de recherche du Québec: Société et Culture.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data that supports the findings of this study are available from the corresponding author upon reasonable request.
Patient Consent Statement
Parental consent was obtained when data was originally collected. The consent included permission for researchers to later conduct secondary analyses like the present study.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
