Support-Seeking and Closeness Across Social Relationships : Relationship-General and Relationship-Specific Levels Analysis

The current study examined whether interpersonal behaviors and perceptions could be conceptualized and measured as relationship-general versus relationship-specific tendencies. To address this goal, we examined (1) the similarities (or concordance) in support-seeking across individuals’ social relationships and (2) how support-seeking may be related to relationship closeness at the relationship-general and relationship-specific levels. Participants were recruited from a regional university and granted course credits for participation. The final sample included 189 undergraduate students with age ranging from 18 to 21 years (M = 18.81; SD = .95), with a total of 66% females and 87% White-Americans. Results suggest that individuals expressed similar rates of support-seeking across parental, best-friend, and romantic relationships. Supporting our hypotheses, a relationship-general correlation suggests that individuals who are more likely to seek social support also perceived their social relationships as generally more intimate. Controlling for relationship-general tendencies, results also suggest that support-seeking was related to relationship closeness for each type of relationship. Theoretical and methodological implications are discussed.

Social-personality psychologists have long been interested in the question of whether people would exhibit consistent versus inconsistent behaviors across different situations (see Mischel & Shoda, 1995).Not until recently have interpersonal theorists begun to specifically describe the cross-partner similarities and differences in interpersonal behaviors and perceptions.Whereas cross-partner similarities in interpersonal behaviors and perceptions are thought to reflect underlying personal traits, cross-partner differences are the results of dyadic systems that are unique to specific relationships (Reis, Capobianco, & Tsai, 2002;Zayas, Shoda, & Ayduk, 2002).Despite the theoretical efforts made, research that targets the issue of similarities and differences in interpersonal behaviors and perceptions remains scarce.
Furthermore, coping researchers have long debated whether coping behaviors, such as support-seeking, should be conceptualized as stable behavioral tendencies according to a trait perspective (e.g., Endler & Parker, 1994) versus situational-oriented behavioral tendencies according to a transactional perspective (e.g., Lazarus & Folkman, 1984).Indeed, little is known about whether individual differences in interpersonal coping behaviors, such as seeking social support, could be decomposed as relationship-general versus relationship-specific components.
In order to provide more insights into this issue, it is important to investigate (1) similarities in the ways people seek support from their social relationships as well as their perceptions of these relationships, and (2) whether Interpersona | An International Journal on Personal Relationships interpersona.psychopen.eu| support-seeking may be related to relationship closeness at the relationship-general versus relationship-specific levels.Findings from the current study would highlight the importance of considering interpersonal behaviors and perceptions as trait-versus state-like tendencies.

Similarities of Interpersonal Behaviors Across Relationships
Interpersonal theorists suggest that individual differences in interpersonal behaviors and perceptions can be conceptualized as two major levels (Kenny, 1994;Reis et al., 2002;Zayas et al., 2002).The first level involves people's general tendency to behave or respond in a uniform way towards others, regardless of the type of relationship (relationship-general tendency).For instance, people who are more willing to seek support from parents are also more likely to seek support from their friends and romantic partners.Presumably, similarities of interpersonal behaviors across different social relationships may be underlined by an overarching, trait-like cognitive structure or mental representations (Bowlby, 1982;Reis et al., 2002).The second level involves people's tendency to behave or respond in a particular way only towards a specific partner (relationship-specific tendency).For instance, people may be more likely to seek support from their romantic partners but not parents and friends, reflecting a unique dyadic system that is not shared by either friendships or parent-child relationships.This idea is consistent with interdependence theory which suggests that interpersonal behaviors are the result of unique combinations of two partners' experiences in a relationship (Kelley & Thibaut, 1978;Zayas et al., 2002).According to these interpersonal theories, although people may exhibit consistency in their support-seeking behaviors across different relationships (relationship-general tendency), the uniqueness of each type of relationship (relationship-specific tendency) would also contribute to within-person variations in support-seeking across different social partners.
Little research attention, surprisingly, has been directed to examining the similarities of stress communication across parent versus peer domains.Fortunately, insights might be borrowed from research on youths' relationship networks.Specifically, research on youths found that perceived supportiveness and attachment styles across different types of relationships were moderately correlated (Furman, Simon, Shaffer, & Bouchey, 2002;Laursen, Furman, & Mooney, 2006).For example, youths who perceived their parents as more supportive also perceived their friends and romantic partners as more supportive (Laursen et al., 2006).Similarly, youths showed moderate convergence in their attachment mental representations across parents, friends, and romantic partner (Furman et al., 2002).Because perceived relationship quality and attachment security were theorized to be important indicators of stress communication behaviors (e.g., Florian, Mikulincer, & Bucholtz, 1995), the similarities of these constructs indeed imply that young people should exhibit similar support-seeking tendency across different social relationships.It is important to note that, however, neither perceived supportiveness nor attachment styles have directly captured stress communication, and therefore, a more direct examination of support-seeking across different relationships is needed.

Support-Seeking and Relationship Closeness
Past relationship theories and studies have consistently demonstrated that support-seeking or self-disclosure is closely related to relationship closeness (Chow & Buhrmester, 2011;Chow, Ruhl, & Buhrmester, 2013;Reis & Shaver, 1988).Therefore, investigating the link between support-seeking and relationship closeness in the context of social networks should be an important step towards understanding these constructs as relationship-general versus relationship-specific components.Specifically, theorists suggest that interpersonal perceptions may also be conceptualized at the relationship-general and relationship-specific levels (Kenny, 1994;Reis et al., 2002;Zayas et al., 2002).For example, although people may be more inclined to perceive that their social relationships are more supportive in general, within-person differences in perceived supportiveness across relationships should simultaneously exist due to the unique dyadic system of each type of relationship.
Based on this line of reasoning, results from past studies that investigated the link between support-seeking and relationship closeness should be reinterpreted cautiously.For example, studies found that people's greater tendency to seek support were related to more intimate relationship quality (Chow & Buhrmester, 2011;Laurenceau, Barrett, & Pietromonaco, 1998).If individual differences in interpersonal behaviors (e.g., support-seeking) and perceptions (e.g., perceived intimacy) could be decomposed as two different levels, the findings of previous studies would have confounded between the relationship-general and relationship-specific tendencies.That is, existing studies have failed to detangle the possibility that the associations between support-seeking and intimacy may be attributable to (1) a relationship-general tendency in which people are generally more likely to perceive their social relationships as intimate and seek support from their social partners and/or (2) a unique dyadic system that exists only in their romantic relationships.According to past theories (e.g., Reis et al., 2002), relationship quality and support-seeking could indeed be related at both relationship-general and relationship-specific levels.

The Current Study
The current study aimed to examine: (1) similarities (or concordance) in the ways people seek support from their social relationships as well as their perceptions of these relationships, and (2) whether support-seeking may be related to relationship closeness at the relationship-general versus relationship-specific levels.Based on past theories and research, it was hypothesized that people's relationship-general support-seeking would be related to their overall perceived social relationships closeness and support-seeking from each type of relationship would be related to the corresponding relationship closeness, holding their relationship-general tendency constant.
averaged across father and mother to form the "parents" composite.Internal consistencies for the support-seeking scales were satisfactory, αs range = .79to .83.
Relationship closeness -Four parallel versions of the relationship closeness 15-item subscale from the Network of Relationships Inventory (NRI; Furman & Buhrmester, 1985) were used to measure participants' relationship closeness with father, mother, best friend, and romantic partner.For example, a closeness item was "How happy are you with your relationship with your (father/mother/best friend/romantic partner)?"Respondents rated how much/often each feature occurred in their relationship on a scale from 1 (Never or hardly at all) to 5 (Always or extremely much).Items were averaged to form the relationship closeness scores.For the present purpose, scores were averaged across father and mother to form the "parents" composite.Internal consistencies were all satisfactory, αs range = .92to .94.

Similarities (or Concordance) Across Relationships
Descriptive statistics of the study variables are presented in Table 1.Our first research goal was to examine the similarities of support-seeking and relationship closeness across relationship types.We addressed this goal by first investigating the correlations between different relationship types (see Table 1).As expected, individuals exhibited high similarities in support-seeking from parents, best friend, and romantic partner (rs ranged from .39 to .46).We further examined similarities across all three types of relationships through an intraclass correlation.The intraclass correlation for support-seeking was .42.This finding suggested that individuals exhibited a high level of consistency in support-seeking tendencies across different social relationships, with shared variance of about 42%.It is important to note that all three types of relationships shared merely half of the variance, implying substantial within-individual variations in support-seeking.Then, analyses also revealed that individuals exhibited moderate rates of similarity in their perceived relationship closeness across parents, best friend, and romantic partner (rs ranged from .26 to .40).The intraclass correlation for relationship closeness was .31,suggesting that individuals showed a moderate level of consistency in their perceptions of closeness with social relationships, with shared variance of about 31%.

Association Between Support-Seeking and Relationship Quality
Our second research goal was to examine whether support-seeking may be related to relationship closeness at the relationship-general and -specific levels.To address this goal, we adopted a Confirmatory Factor Analysis (CFA) approach (see Figure 1) which allowed us to partition each measured (manifest) variable into three com-Interpersona 2013, Vol.7(1), 88-96 doi:10.5964/ijpr.v7i1.114ponents: (1) variance that is common to all relationships, (2) variance that is unique to a specific relationship, and (3) measurement error.For example, measured support-seeking with parents may reflect individuals' tendency to seek support from relationships, tendency to seek support only from parents, and measurement error.Similarly, measured relationship closeness with parents may reflect individuals' general closeness with all relationships, closeness with parents, and measurement error.For the CFA model, we specified two latent variables, relationship-general support-seeking and relationship quality, to account for the relationship-general similarities across the three types of relationships.The residuals, therefore, would account for relationship-specific variance and measurement error.The latent variables were allowed to covary to estimate the relationship-general link between support-seeking and relationship quality.We then examined whether the model fit would improve significantly when the residuals within each type of relationship were allowed to covary (e.g., support-seeking with parents and relationship closeness with parents).Significant improvement of the model fit or significant correlations between relationship-specific support-seeking and relationship closeness would support the notion that the association between these constructs may reflect both relationshipgeneral and relationship-specific components.
The model without correlated residuals did not fit the data well, χ Support Seeking support-seeking and relationship quality should be conceived at two different levels: relationship-general and relationship-specific.Correlation coefficients are presented in Figure 1.As expected, the relationship-general correlation showed that individuals who were generally more likely to seek social support also tended to perceive their social relationships as more intimate.As expected, relationship-specific correlations showed that for each type of relationship, greater support-seeking was related to higher intimacy.

Discussion
The current study highlights the importance of considering interpersonal behaviors and perceptions at the relationship-general and relationship-specific levels.Indeed, individuals' support-seeking behaviors from parents, best friend, and romantic partner were found to be consistently related.These findings are consistent with past studies that found similarities in youths' perceptions of relationship quality and attachment mental representations across different types of relationships (Furman et al., 2002;Laursen et al., 2006).The moderate similarities in supportseeking imply that there might be an overarching personality trait or mental structure that underlines the systematic similarities.
One potential mechanism would be attachment mental representation.According attachment theory, people internalize their interpersonal experiences, including those with parents, friends, and romantic partners, and develop overarching mental representations of self-versus others (Bowlby, 1982;Furman et al., 2002).These mental representations are thought to systematically guide how people perceive their social relationships as well as the interpersonal behaviors that are directed towards their social partners, including support-seeking behaviors (Florian et al., 1995).Indeed, research has found that general attachment security is related to support-seeking behaviors in the context of parents, friends, and romantic relationships (Florian et al., 1995).Integrating current findings and attachment research, one remaining question, however, is that when attachment security or other broader personality trait (e.g., neuroticism) is accounted for, should similarities in stress communication styles across relationships still be observed?In other words, if mental representations or trait structures are the underlying mechanisms that explain the similarities, the interrelatedness across different social relationships should fade away when these constructs statistically controlled for.Although the current study did not directly address this question, it has provided important insights into this issue and a foundation for future research.
It is important to note that, however, the similarities for support-seeking are moderately high at best.The imperfect correlations suggest that there are within-person variations in how individuals communicate stress with their social partners.If assumed underlying personality traits do not lead to perfect consistencies in interpersonal behaviors towards different partners, what are the factors that may explain such variability?Examining the associations between relationship quality and support-seeking should provide important insights into this question.First, results demonstrated that although people are more inclined to have similar attitudes/reactions concerning perceived intimacy across different relationships, within-individual variations across relationships also existed.Second, model comparison results demonstrated that linking perceived relationship closeness and support-seeking at the relationship-general level did not account for sufficient variability in both constructs.The significant model fit improvement by correlating the relationship-specific variables suggested that people have unique patterns of communicating stress with each type of relationship, depending on the specific perceived relationship quality.These findings are consistent with past interpersonal theories that argue that each type of relationship provides a unique "situational factor" that interpersonal traits may be expressed (Zayas et al., 2002).Conceivably, perceived quality of each type of relationship may provide a unique dyadic context that only allows the expression of supportseeking behaviors that are above and beyond the relationship-general tendency.
The current study added important, yet provocative, ideas to past relationship studies that investigated the association between support-seeking and relationship closeness (Chow & Buhrmester, 2011;Reis & Shaver, 1988).
The current study found that individual differences in interpersonal behaviors (e.g., support-seeking) and perceptions (e.g., perceived intimacy) could be decomposed as two different levels.This implies that existing studies on the association between support-seeking and relationship closeness have failed to detangle the possibility that the association between support-seeking and intimacy may be attributable to whether people are generally more likely to perceive their social relationships as intimate and seek support from their social partners and/or whether the association only exists in a specific relationship.Thus, it is important for future research on social relationships to employ a technique similar to the current study in order to partition the correlation between interpersonal behavior and perception into relationship-general and relationship-specific levels.

Limitations and Future Research
One major limitation of the current study was that although it was shown that support-seeking could be conceptualized at relationship-general and relationship-specific levels, the findings revolved around a "trait" approach to coping behaviors and captured only the typical ways in which people communicate stress across different relationships (general) and within each type of relationship (specific).The current study provides limited information about temporal and contextual variations in interpersonal coping behaviors (Reis et al., 2002).Although we concur with some theorists that coping behaviors can be trait-like behavioral tendencies (e.g., Endler & Parker, 1994), we also recognize the importance of conceptualizing support-seeking from a "transactional" perspective (Lazarus & Folkman, 1984).That is, individuals may constantly modify their support-seeking strategies depending on the nature of the stressors and circumstances, even with the same partner.Thus, it is important for future research to modify the current paradigm so that it can be extended to capture both trait-and state-like support-seeking behaviors.
Second, although the current study emphasizes the different ways in which people involve their partner in their coping process, one major limitation of the current study is that we did not examine the role of the partner.Indeed, it is possible that the ways individuals cope with stress are partially explained by their partners' ways of providing support.Previous research on friend dyads suggests that there are distinctive patterns of associations between friends' support-seeking and support-giving behaviors (Chow & Buhrmester, 2011).For instance, individuals seek more support from their friends when they are sensitive and responsive.Thus, future research should examine whether people's support-seeking behaviors are systematically associated with their partners' support-giving behaviors at both relationship-general versus relationship-specific levels.
Another limitation of the current study was that it solely relied on participants' self-report data about their social relationships.Future research should consider including multiple reporters' perspectives on target participants' support-seeking behaviors as well as their relationship quality.Furthermore, future research may also use a laboratory observation paradigm to capture actual support-seeking interactions with different partners.Nevertheless, results from the current study provide an important foundation towards these future directions.

Conclusion
Despite the limitations, the current study makes two major contributions to the existing social relationships research.
First, the current findings provided a novel way of conceptualizing interpersonal behaviors and perceptions, suggesting that it is important to put individual differences in these constructs within the broader context of relationship networks.Second, findings from the current study also warrant that researchers should be cautious when interpreting results that are interpersonally-oriented in that the source of variance may stem from relationship-general and/or relationship-specific levels.

Figure 1 .
Figure 1.A CFA approach to model relationship-general versus relationship-specific support-seeking and relationship closeness.Curved lines with doubled arrows are correlations and straight lines with single arrows are structural paths (factor loadings).Standardized coefficients are presented.All coefficients are significant at p < .01.

Table 1
Descriptive Statistics and Correlations Among Corresponding Variables in the Three Relationships **p < .01.