Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2010 Jan 19.
Published in final edited form as: J Res Adolesc. 2002 Sep 1;12(3):353–372. doi: 10.1111/1532-7795.00037

Adolescent Perceptions of Conflict in Interdependent and Disengaged Friendships

Shmuel Shulman 1, Brett Laursen 2
PMCID: PMC2808029  NIHMSID: NIHMS142744  PMID: 20090925

Abstract

Interdependent and disengaged friendships in a middle-class sample of suburban Israeli adolescents were examined for differences in reports of conflict behavior. A total of 194 (100 females, and 94 males) close, reciprocal friends participated in a joint problem-solving task used to categorize friendships. Interdependent friends balanced closeness and individuality by cooperating on the task, whereas disengaged friends emphasized individuality by working independently on the task. In separate interviews, these friends recounted their most important conflict from the previous week. Older adolescents (M = 17.4 years) reported more conflicts over private disrespect than did younger adolescents (M = 12.7 years), whereas younger adolescents reported more conflicts over public disrespect and undependability than did older adolescents. Differences between friendship types in conflict initiation, negative affect, and relationship impact were found among older adolescents but not younger adolescents; differences in conflict resolutions were found in both age groups. In contrast to disengaged friends, interdependent friends were better able to manage conflicts in a manner that emphasized relationship harmony over individual gain.

Few friends are able to avoid disagreements, for disputes are an inevitable part of all close relationships. The mere presence of conflict, therefore, reveals less about the quality of a relationship than does the way in which the conflict is handled (Perry, Perry, & Kennedy, 1992). Conflict management is a particularly important barometer of relationship functioning, because friends are both invested in the rewards of affiliation and free to discontinue interconnections perceived to be disadvantageous. As a result, adolescents adopt more conciliatory management strategies with friends than with nonfriends or family members (Laursen, 1993; Laursen, Finkelstein, & Betts, 2002). Additional relationship differences in conflict are presumed to follow from differences in friendship qualities, but little is known about behavioral variations in friendship and the degree to which these are manifested in disagreements (Hartup, 1992). Simply put, friendships differ on multiple dimensions and we expect these differences to be reflected in their conflict behavior. The present study was designed to address this issue, by identifying qualitative differences between adolescent friendships and extending these distinctions to interpersonal conflict behavior.

All friendships are not created equal. Several studies have confirmed qualitative differences in the organization of adolescent close peer relationships that resemble qualitative differences in family systems (Shulman, 1993; Shulman, Laursen, Kalman, & Karpovsky, 1997; Shulman, Levy-Shiff, Kedem, & Alon, 1997). In these studies, two types of relationships emerged from observations of early and midadolescent friends and late adolescent romantic partners working on a joint problem-solving task: (1) environment sensitive or interdependent, and (2) distance sensitive or disengaged. (A third type of relationship found among families, consensus sensitive or enmeshed, is rare among close peers.) Seen in this light, friendships resemble family relationships in that each is a system in which participants must negotiate mutuality and individuality. In interdependent relationships, partners strive to balance the needs of each individual. They react objectively to the free discussion of ideas and wishes. Cooperation prevails because friends look to each other for support and intimacy. In disengaged relationships, partners pursue personal goals at the expense of relationship goals. Cooperation is a means to an end in these relationships, because friends emphasize personal benefits rather than interpersonal closeness. In the absence of support and intimacy, expressions of warmth are rare.

Friendships are typically described in terms of closeness, but conflict is also an integral feature of these relationships (Laursen, Hartup, & Koplas, 1996). Conflict may provoke relationship perturbation or even disintegration, but it may also provide an opportunity to define roles, improve communication, and strengthen interconnections (Hartup, 1992, 1996). As such, disagreements are a powerful marker of relationship differences. The best evidence to support this hypothesis comes from studies of marital and family relationships. Research on married couples indicates that some prefer harmony and avoid negativity, but others are not afraid to express disagreement and appear to thrive on conflict (Gottman, 1998). Research on parent–child relationships indicates that adolescents with supportive parents prefer to compromise, whereas those with inconsistent parents tend to escalate hostilities (Rubinstein & Feldman, 1993). With the exception of studies that link negative and coercive interactions to low friendship satisfaction (e.g., Dishion, Andrews, & Crosby, 1995), research has yet to explore differences between friendship types in conflict management. However, there is evidence of individual differences such that peer conflict and negativity vary as a function of antisocial behavior, personality, and peer social status (Chung & Asher, 1996; Graziano, Jensen-Campbell, & Hair, 1996; Poulin, Dishion, & Haas, 1999). Given the prevalence of individual differences in other relationships, there is good reason to suspect similar variations in conflict management in friendships.

Conflict poses a special risk to friendships because of the voluntary nature of the affiliation (Laursen et al., 1996). Participants are free to dissolve the affiliation should exchanges fail to be mutually satisfactory, so friends must monitor their interactions to ensure that rewards and costs are shared. Sometimes the long-term needs of the dyad are contrary to the short-term desires of the participants. When such a conflict arises, friends must choose between beneficent solutions that maintain interpersonal interconnections and coercive solutions that advance individual outcomes at the expense of the relationship. It is a constant dilemma because friends are frequently in conflict; by one account, adolescents report an average of one or two disagreements with close friends every day (Laursen, 1995a). Interdependent friends should be adept at avoiding disputes and resolving those disagreements that do arise because participants tend to focus on the needs of the dyad (Shulman, 1993). Of course, interdependent friends cannot completely avoid negativity, but they can be expected to go to great lengths to minimize coercion, unpleasant affect, and detrimental outcomes. Disengaged partners, in contrast, may have difficulty compromising because of a tendency to focus on self-interests. The pursuit of personal gain at the expense of relationship amity may heighten the frequency with which conflicts arise and the tendency to manage them coercively.

Developmental differences in the management of peer conflict have been identified. Age and maturity are accompanied by a greater appreciation of the role that conflict resolutions play in maintaining friendships (Youniss & Smollar, 1985). Even young children recognize the risk that coercion poses to friendship, yet they do not become adept at averting adverse consequences until their social skills and perspective-taking abilities improve (Selman, 1980). Responses to hypothetical vignettes make it clear that older adolescents place a higher priority on negotiating with friends than do younger adolescents (Selman, Beardslee, Schultz, Krupa, & Podorefsky, 1986). Although evidence for change across adolescence in actual peer conflict behavior is lacking, a recent meta-analysis suggests that the period is marked by a gradual developmental shift in resolution tactics, as indicated by successively lower levels of coercion and higher levels of negotiation across childhood, adolescence, and young adulthood (Laursen et al., 2002). As is the case in relationships with siblings (Buhrmester & Furman, 1990), antagonism and competition between friends should decline with age, reducing the frequency of disagreements and the expression of negative affect.

The present study was designed to explore differences in the conflict behavior of interdependent and disengaged friends. Two questions were central to the investigation. The first question concerned friendship differences: Do interdependent and disengaged friends differ in conflict management behaviors? It was hypothesized that disengaged friends would report more coercive patterns of conflict management than would interdependent friends, because personal gains guide the former, whereas relationship concerns direct the latter (Shulman, 1993). Relative to interdependent friends, disengaged friends were expected to assume less responsibility for conflict, evince more negative affect, compromise less during the course of a dispute, and report more winner/loser outcomes that make the relationship worse. Because previous studies of friendship types have not considered conflict topics, no predictions were made concerning differences in the rate of conflict over specific issues.

The second question concerned age-related differences: Are there age-related differences in the conflict management behavior of adolescent friends? It was hypothesized that experience and maturity would provide adolescents with a greater appreciation of the strategies that preserve voluntary ties with friends (Laursen et al., 1996; Selman, 1980). Relative to younger adolescents, older adolescents were expected to report fewer conflicts, assume more responsibility for conflict, evince less negative affect, compromise more during the course of the dispute, and report more equal outcomes that do not negatively impact the relationship.

The proportion of interdependent friends was expected to increase with age and the proportion of disengaged friends was expected to decline. Thus, friendship type differences in conflict may interact with age-related differences in conflict and friendship types. Because previous studies have not examined the interplay of friendship types and adolescent age, no predictions were made as to whether to expect independent main effects or two-way interactions that indicate distinct age-related patterns of conflict behavior for each type of friendship.

METHOD

Pilot Survey

Two pilot studies were conducted. Participants in each study were roughly equally divided between seventh and eighth graders who were attending a public junior high school and tenth and eleventh graders who were attending a public high school in a middle-class suburb of Tel Aviv, Israel. Participants in these pilot studies were representative of the school and community from which they were drawn in terms of socioeconomics and gender ratio. The goal of these pilot studies was to modify the Interpersonal Conflict Interview (Laursen & Koplas, 1995) for use with Israeli adolescents.

In the first pilot study, 146 adolescents were invited to participate and 138 (78 females, 60 males) did so after providing their own consent and the necessary consent of their parents. Open-ended surveys determined a list of 11 issues that were the source of greatest conflict between friends. These 11 conflict issues were similar to those found in North American studies of peer conflict (for a review, see Hartup, 1992).

In the second pilot study, 276 adolescents were invited to participate and 250 (122 females, 128 males) did so after providing their own consent and the necessary consent of their parents. Participants completed questionnaires rating the frequency with which conflicts with unspecified friends arose over each of the 11 conflict issues, on a scale ranging from very infrequent (1) to very frequent (4). Factor analysis yielded a three-factor structure that explained 49.7% of the variance; item loadings for each are listed in parentheses. The first factor, public disrespect, explained 28.4% of the variance and contained four items: telling secrets (.75), public teasing (.71), gossip (.65), and being ignored during social activities (.49). The second factor, undependability, explained 11.4% of the variance and included four items: jealousy over another friend (.80), lack of closeness (.62), failure to provide assistance (.53), and failure to share or lend possessions (.43). The third factor, private disrespect, explained 9.9% of the variance and contained three items: imposing views on another (.71), competition (.68), and missing appointments (.49).

Participants

A total of 194 adolescents (100 females, 94 males) participated in the investigation. These participants were drawn from an initial pool of 270 students who completed a friendship survey. Of this total, invitations to participate in the project were extended to the 212 adolescents who met the close friendship criteria described in the following section; 194 did so after providing their own consent and the necessary consent of their parents. Students attended public junior and senior high schools in a middle-class neighborhood of Tel Aviv, Israel, and were representative of the school and community from which they were drawn with regard to their socioeconomics and gender ratio. The seventh graders (n = 120) ranged in age from 11 to 13 (M = 12.7 years) and the eleventh graders (n = 74) ranged in age from 16 to 18 (M = 17.4 years). Participation rates were similar across age groups; there were fewer eleventh grade participants than seventh grade participants because matriculation exams limited the number of eleventh-grade classrooms available for participant recruitment.

Procedure

The data were collected in three phases, each separated by approximately 1 week. During the first phase, a screening procedure identified students who were participating in a close, reciprocal friendship. Adolescents in the initial participant pool named all of their friends in an open-ended friendship survey. They also rated the closeness of each relationship on a 7-point scale ranging from low (1) to high (7). Dyads that qualified for the investigation met two criteria. First, the friendship was reciprocated; both partners identified one another as friends. Second, participants were close friends; both partners gave the relationship a closeness score of at least 6. The procedure yielded 60 pairs of seventh-grade friends (31 females, 29 males) and 37 pairs of eleventh-grade friends (19 females, 18 males). During the second phase, adolescents were separately interviewed about conflicts with this close friend. During the third phase, dyads participated in a joint problem-solving task to identify friendship typologies.

Instruments

Conflict interview

A modified version of the Interpersonal Conflict Interview (Laursen & Koplas, 1995) identified all disagreements that had arisen between friends during the preceding week, and assayed specific characteristics of each. Conflict was defined in terms of behavioral opposition (Shantz, 1987, p. 284): “You and your friend had a difference of opinion; you objected to something he or she said or did, or your friend objected to something you said or did; or you and your friend had a quarrel or an argument.” First, participants rated each of the 11 conflict issues on a 4-point scale ranging from very infrequent (1) to very frequent (4). Conflict frequency represented the summed average of scores from each of these issues. Second, participants were asked to identify, from this list of 11 issues, the most important conflict from the past week that arose with the close friend identified by the screening procedure. Both friends completed the same instrument but did not necessarily identify the same most important conflict. Follow-up questions addressed five components of this most important conflict. Open-ended responses to specific questions were transcribed and two coders later scored these responses. Kappa coefficients of interrater reliability, based on dual coding of 70 interviews, are given in parentheses.

  1. Initiation (.94). Who started the conflict? Responses were coded into one of two categories: (1) assumed responsibility for the conflict and (2) did not assume responsibility for the conflict.

  2. Negative affect. How angry were you during the conflict? Participants rated disagreements on a 4-point scale ranging from a little angry (1) to very angry (4).

  3. Resolution (.84). How was the conflict resolved? Responses were coded into one of four categories: (1) power assertion, which involved one party acceding to the demands of the other; (2) compromise, which entailed negotiation; (3) withdrawal, which encompassed disengagement; and (4) third-party resolution, which described solutions proposed or imposed by a nonparticipant.

  4. Outcome (.85). What was the outcome of the conflict? Responses were classified into one of three categories: (1) Winner/loser, in which 1 participant attained all of his or her original conflict goals; (2) equality, in which both participants achieved similar levels of their original conflict goals; and (3) no outcome, in which neither participant attained any of their original conflict goals.

  5. Relationship impact (.89). Did the conflict affect your relationship? Responses were coded into one of three categories: (1) made the relationship worse, (2) no impact on the relationship, and (3) made the relationship better.

Previous studies have revealed adequate internal reliability and moderate test–retest stability for the Interpersonal Conflict Interview and for a questionnaire version of this instrument (Laursen, 1995b; Laursen & Koplas, 1995). Depictions of conflict derived from these measures have been linked to adaptive features of adolescent development such as self-esteem and peer social competence (Collins, Laursen, Mortensen, Luebker, & Ferreira, 1997).

Problem-solving task

A modified version of the Card Sort Problem-Solving Procedure (Reiss, 1981) provided a behavioral measure of interpersonal interaction processes from which friendships were classified (Shulman, 1993). Dyads were presented with separate sets of 16 cards, each containing a row of letters varying in order and length. Participants were instructed to use any criteria to sort the cards into any number of piles and that any solution was acceptable except for a sort of one pile (of all the cards) or 16 piles (each card separately). Two dimensions were available for ordering the cards: the number of letters (e.g., CPQRLT, CPQRLRLT, CPQRLRLRLT) and the pattern or order of the letters (e.g., CPQRLT, RLTRLT, CPQCPQ).

The task was divided into two phases. First, adolescents sorted their own cards separately, without speaking. Second, adolescents again sorted their own cards separately, but communication was permitted. No instructions on dyadic agreement were provided, so joint or individual sorting decisions were possible and partners decided for themselves whether to cooperate or work independently. The task is somewhat analogous to academic study assignments in which teamwork is neither prohibited nor encouraged, and individuals are free to devise their own study procedures.

Problem-solving behavior was coded in terms of configuration and coordination. For each dimension, dyads were categorized as high or low on the basis of a priori cutoff scores. These fixed criteria were derived from the original Card Sort Problem-Solving Procedure (Reiss, 1981) and subsequently validated in several studies of early (11- to 13-year-olds) and midadolescent (14- to 16-year-olds) close peer and late adolescent (17- to 19-year-olds) romantic relationships (Shulman, 1993; Shulman, Laursen, et al., 1997; Shulman, Levy-Shiff, et al., 1997). Each set of cards includes a single optimal solution based on the length or pattern of letters. Both sorts are each separately scored according to an empirically derived formula that reflects the extent to which the solution is similar or different from the optimal sort. Sort scores range from 0.00 (dissimilar) to 1.00 (similar).

Configuration reflects improvement or deterioration in the quality of the card sort from the first trial to the second trial. It describes the extent to which friends contributed to each other’s card sort when given the opportunity to interact. High configuration denotes positive scores (greater than 0) that indicate that interaction improved problem solving (i.e., the second sort resembled the optimal solution more than the first sort). Low configuration denotes negative scores (less than 0) that indicate that interaction hindered problem solving (i.e., the second sort resembled the optimal solution less than the first sort).

Coordination reflects (1) solution similarity (ranging from unique [0.00] to identical [1.00]) and (2) the difference (in SD units) in the time taken to complete the second sort. Large similarity scores indicate a substantial degree of overlap in solution sorts. Small difference scores indicate that participants finished trials at the same time. High coordination (sort similarity greater than 68% and sort SD less than 4 s) indicates that friends worked together and solved the problem in a similar manner. Low coordination (sort similarity less than 68% and sort SD greater than 4 s) indicates that friends did not cooperate.

The card sort procedure is a well-established family assessment tool, predicting clinical referrals, psychopathology, and treatment outcomes (Reiss, Costell, Jones, & Berkman, 1980; Shulman & Klein, 1983; Shulman & Zohar, 1991). It is also a valid and reliable index of adolescent close peer relationships. Evidence suggests that the card sort procedure describes stable patterns of adolescent friendships: Over a 2-month period, 30 out of 34 adolescent friendship dyads retained the same classification (Korin, 1992). Strong associations, r = .56, have been found between card sort appraisals and observer ratings of cooperation for 43 adolescent friendships (Alon, 1996). Among adolescent friends and romantic partners, typologies identified by the procedure have been linked to self-reports of relationship conceptions and attitudes, as well as to perceptions of relationship closeness and control (Shulman, 1993; Shulman, Laursen, et al., 1997; Shulman, Levy-Shiff, et al., 1997).

The Reiss (1981) classification procedure yields three relationship types: interdependent, disengaged, and enmeshed. Interdependent dyads are high on coordination and high on configuration. Disengaged dyads are low on coordination and either low or high on configuration; dyads low on coordination are considered disengaged regardless of their configuration score, because improvements in problem solving cannot be attributed to cooperation between partners. Enmeshed dyads are high on coordination and low on configuration.

All dyads in the present study were classified as either interdependent or disengaged; no dyads met the criteria for enmeshed friendships.1 Among seventh graders, 70% were classified as disengaged friends (23 females, 19 males) and 30% were classified as interdependent friends (8 females, 10 males). Among eleventh graders, 46% were classified as disengaged friends (12 females, 5 males) and 54% were classified as interdependent friends (7 females, 13 males). Friendship types differed across age groups, χ2(1, N = 97) = 5.52, p < .01. Most younger adolescent friendships were disengaged, whereas older adolescents were involved in a similar proportion of interdependent and disengaged friendships. There were no statistically significant gender differences in the distribution of friendship types.

RESULTS

Each set of analyses addressed two central questions: (1) Do interdependent and disengaged friends differ in conflict management behaviors? (2) Are there age-related differences in the conflict management behavior of adolescent friends? Conflict variables based on rating scale data (i.e., conflict frequency and affective intensity) were examined with separate analyses of variance (ANOVA) and t tests. Conflict variables based on frequency data (i.e., conflict initiation, resolution, outcome, and relationship impact) were examined with separate log-linear and χ2 analyses.

Two sets of preliminary analyses explored the potential contribution of gender and reporter. The first concerned reporter. In analyses of rating score data, each member of the dyad was entered into an ANOVA as a repeated-measures independent variable. No statistically significant main effects or interactions emerged involving reporter. In χ2 analyses of frequency data, reports that contrasted each member of the dyad revealed no statistically significant effects for reporter on any conflict component variable, except conflict initiation. All conflict components were significantly correlated, p < .05, across reporters, r and Cramer’s V ranged from .25 to .35, despite the fact that some friends did not describe the same conflict. Dyad scores could not be aggregated because of potential partner differences in the target conflict, so analyses were conducted instead on scores from one randomly selected participant per dyad. Additional analyses (not presented herein) were repeated using reports from the other member of the dyad. The same pattern of statistically significant results emerged.

The second set of preliminary analyses concerned gender. Analyses of rating score data included gender as a between-subjects variable. No statistically significant main effects or higher order interactions emerged involving gender on any conflict component variable, except a two-way interaction between gender and age on negative affect. Among older adolescents, females (M = 2.73, SD = .94) reported greater levels of conflict anger than did males (M = 1.93, SD = .85). Among younger adolescents, there were no statistically significant differences between females (M = 2.39, SD = 1.12) and males (M = 2.57, SD = 1.17). Chi-square analyses of frequency data revealed no statistically significant effects for gender on any conflict component variable. Because no consistent gender differences emerged from previous research and none were expected in this study, gender was not included in the analyses described in the text.

Conflict Frequency

A repeated-measures ANOVA was conducted with two between-subjects independent variables—age (younger and older) and friendship type (interdependent and disengaged)—and one within-subjects independent variable—of conflict issue (public disrespect, undependability, and private disrespect). Conflict frequency was the dependent variable. Results revealed a significant main effect for conflict issue, F(2, 94) = 16.95, p < .001, and a two-way interaction between age and conflict issue, F(2, 94) = 6.22, p < .001. To explore this two-way interaction, three separate follow-up t tests identified statistically significant differences, p < .05, between older and younger adolescents on the frequency of each conflict issue. Younger adolescents rated conflicts over public disrespect and undependability to be more frequent than did older adolescents, whereas older adolescents rated conflicts over private disrespect to be more frequent than did younger adolescents (see Table 1).

TABLE 1.

Frequency Ratings of Conflict Issues by Adolescent Age Groups

Conflict Issue Younger Adolescents M (SD) Older Adolescents M (SD)
Public disrespect 1.73a (.49) 1.57b (.39)
Undependability 1.89a (.43) 1.58b (.47)
Private disrespect 1.97b (.54) 2.15a (.62)
Open in a new tab

Note. N = 60 younger adolescents and 37 older adolescents. Different subscripts indicate statistically significant differences, p < .05, between age groups in follow-up t tests. Issues were rated on a scale ranging from 1 (very infrequent) to 4 (very frequent).

Conflict Initiation

A 2 (age) × 2 (friendship type) × 2 (conflict initiation: assumed responsibility and did not assume responsibility) log-linear analysis revealed a three-way interaction among age, friendship type, and conflict initiation, χ2(1, N = 97) = 4.58, p < .05. To elaborate these differences, 2 (friendship) × 2 (conflict initiation) follow-up χ2 analyses were conducted separately for each age group. Among older adolescents, interdependent friends were more likely to take responsibility for initiating a conflict than not, whereas the reverse was true of disengaged friends, χ2(1, N = 37) = 4.18, p < .05. Among younger adolescents, there were no statistically significant differences in the initiation of conflict across friendship types (see Table 2).

TABLE 2.

Conflict Initiation × Friendship Types among Younger and Older Adolescents

Disengaged n (%) Interdependent n (%)
Younger adolescents
 Assumed responsibility 20 (47.7) 11 (61.1)
 Did not assume responsibility 22 (52.3) 7 (38.9)
Older adolescents*
 Assumed responsibility 7 (41.1) 15 (75.0)
 Did not assume responsibility 10 (58.8) 5 (25.0)
Open in a new tab

Note. N = 59 disengaged dyads and 38 interdependent dyads. Separate Friendship Type × Conflict Initiation χ2 analyses were conducted for younger and older adolescents.

*

Differences in χ2 within age groups, p < .05.

Conflict Negative Affect

An ANOVA was conducted with two between-subjects independent variables—age and friendship type. Negative affect was the dependent variable. A two-way interaction emerged between age and friendship type, F(1, 95) = 13.64, p < .001. Follow-up t tests, p < .05, conducted separately for each age group, contrasted interdependent and disengaged friends on negative affect (see Table 3). Among older adolescents, disengaged friends reported greater levels of conflict anger than interdependent friends. Among younger adolescents, there were no statistically significant differences between disengaged friends and did interdependent friends on negative affect.

TABLE 3.

Conflict Negative Affect × Friendship Types among Younger and Older Adolescents

Disengaged M (SD) Interdependent M (SD)
Younger adolescents 2.46 (.85) 2.47 (.75)
Older adolescents 3.00a (.89) 1.84b (.76)
Open in a new tab

Note. N = 59 disengaged dyads and 38 interdependent dyads. Different subscripts indicate statistically significant differences, p < .05, between age groups in follow-up t tests. Negative affect was rated on a scale ranging from 1 (a little angry) to 4 (very angry).

Conflict Resolution

A 2 (age) × 2 (friendship type) × 4 (conflict resolution: power assertion, compromise, withdrawal, and third-party resolution) log-linear analysis revealed a two-way interaction between friendship type and conflict resolution, χ2(3, N = 97) = 9.04, p < .05. There were greater levels of power assertion and third-party resolutions among disengaged friends than among interdependent friends, whereas compromise and withdrawal were more prevalent among interdependent friends than among disengaged friends (see Table 4).2

TABLE 4.

Conflict Resolution × Friendship Types

Disengaged n (%) Interdependent n (%)
Power assertion 12 (20.3) 2 (5.3)
Compromise 29 (49.2) 26 (68.4)
Withdrawal 6 (10.2) 7 (18.4)
Third-party resolution 12 (20.3) 3 (7.9)
Open in a new tab

Note. N = 59 disengaged dyads and 38 interdependent dyads.

Conflict Outcome

A 2 (age) × 2 (friendship type) × 3 (conflict outcome: winner/loser, equality, and no outcome) log-linear analysis failed to reveal statistically significant differences in the distribution of these conflict behaviors. Regardless of age and friendship type, most adolescents reported equal outcomes in conflicts with close friends.

Conflict Relationship Impact

A 2 (age) × 2 (friendship type) × 3 (conflict relationship impact: made it better, no impact, and made it worse) log-linear analysis revealed a three-way interaction among age, friendship type, and conflict relationship impact, χ2(2, N = 97) = 7.52, p < .05. To elaborate these differences, 2 (friendship) × 3 (conflict relationship impact) follow-up χ2 analyses were conducted separately for each age group. Among older adolescents, no relationship impact was reported more often by interdependent friends than by disengaged friends, whereas both positive and negative relationship impacts were reported more often by disengaged friends than by interdependent friends, χ2(2, N = 37) = 6.47, p < .05. Among younger adolescents, there were no statistically significant differences in conflict relationship impact across friendship types (see Table 5).

TABLE 5.

Conflict Relationship Impact × Friendship Types among Younger and Older Adolescents

Disengaged n (%) Interdependent n (%)
Younger adolescents
 Made relationship better 10 (23.8) 4 (22.2)
 No impact on relationship 20 (47.6) 9 (50.0)
 Made relationship worse 12 (28.6) 5 (27.8)
Older adolescents*
 Made relationship better 4 (23.5) 1 (5.0)
 No impact on relationship 5 (29.4) 14 (70.0)
 Made relationship worse 8 (47.1) 5 (25.0)
Open in a new tab

Note. N = 59 disengaged dyads and 38 interdependent dyads. Separate Friendship Type × Conflict Relationship impact χ2 analyses were conducted for younger adolescents (n = 60 dyads) and older adolescents (n = 37).

*

Differences in χ2 within age groups, p < .05.

DISCUSSION

The results indicate that age tends to moderate friendship differences in conflict behavior. Among older adolescents, distinct patterns of conflict management emerged for interdependent and disengaged friends. Relative to disengaged friends, interdependent friends tended to assume responsibility, avoid anger, and report fewer negative consequences for the relationship. In both age groups, interdependent friends reported less coercion and withdrawal, and more compromise resolutions than did disengaged friends. Interdependent friends acted in a manner that displayed more concern for the well-being of the relationship than concern for personal gain. These friendships were typically unaffected by the vicissitudes of disagreement. In contrast, disengaged friends allowed conflict to escalate, perhaps because participants were more interested in prevailing than in achieving a mutually satisfactory solution. Perceptions of these friendships tended to vacillate, with the relationship reportedly improving or worsening according to the behavior displayed during the most recent important dispute.

Mutuality and individuality, central themes in the organization of close relationships, are often invoked to explain processes of adolescent adaptation. According to systems theorists, family members must strike a balance between cohesion and personal growth; those who accommodate these competing requirements demonstrate greater competence than do those who exclusively emphasize mutuality or individuality (Minuchin, 1974). The same demands confront adolescent friends. Participants in interdependent relationships appear to moderate their personal objectives to attain a solution that preserves mutuality. Participants in disengaged relationships manage conflicts in such a way that personal objectives appear to override relationship harmony and undermine mutuality.

It has long been argued that peer relationships serve as a foundation for the child’s understanding of the social world (Piaget, 1932/1965). Interpersonal conflict is of particular interest to developmental scholars because it provides an important context in which children attain and refine social skills, acquire principles of justice, master the regulation of affect, and define personal autonomy (Puttalaz & Sheppard, 1992). These benefits are more apt to derive from disagreements with friends than from those with other peers, in part because of the sheer volume of conflicts that involve friends and in part because of the significance of the relationship and the disagreements that arise within it (Hartup, 1992). Furthermore, the mutuality characteristic of balanced friendships encourages participants to raise and negotiate disagreements in the best interest of the relationship (Wynne, 1984). By specifying developmental differences between friendships in patterns of conflict management, this investigation offers new insight into processes whereby peer relationships may influence individual social adjustment.

Three previous studies of friendship quality and interpersonal conflict offer some insight into the results obtained in the present study. The first (Shulman, Elicker, & Sroufe, 1994) contrasted the disagreements of preadolescents who shared the same attachment history. Friends with secure attachment histories tended to resolve conflict in a manner that enabled a quick resumption of play. Friends with ambivalent attachment histories disengaged from conflict and avoided one another for a day or two, then resumed play without resolving the dispute. The second study (Hartup, Laursen, Stewart, & Eastenson, 1988) contrasted preschool friends who differed in their level of interpersonal interaction. Children who spent more than 25% of their free play time together managed conflicts in a more conciliatory manner than those who spent 10% to 25% of their free time together. Specifically, the conflicts of engaged friends contained less negative affect and power assertion, and more equitable outcomes than did those of detached friends. The third study (Dishion et al., 1995) examined rates of negative engagement (i.e., negative verbal and nonverbal, verbal attack, coerce, and physical aggression) in adolescent friendships as a function of self-reports of relationship satisfaction and teacher, self-, and interviewer reports of antisocial behavior. Dyadic relationship satisfaction was linked to observed negativity, whereas antisocial behavior was not. Although these findings must be interpreted with caution because of the varied participant ages, when taken together with the results of the present study, they point to the conclusion that differences between friends in conflict management reflect salient distinctions between relationships in the manner in which friendships are organized and in the experiences they provide.

In several instances in the present study, age moderated differences between adolescent friendships in conflict behavior. Because we offered no predictions about the potential interaction of age and friendship type, these findings must be interpreted with caution. That said, the consistent pattern of results raises the possibility that different types of friendships elicit different developmental trajectories. Contemporary discussions of social exchange theory (Laursen & Hartup, 2002) and attachment theory (Furman & Wehner, 1994) agree that the quality of initial interactions with close peers has a strong influence on the quality of subsequent interactions with that peer and on the quality of subsequent relationships with other peers. Different types of friendships differ in their selection and influence processes, which, in turn, may foster different individual and dyadic pathways for interdependent friends and disengaged friends. Thus, age may be expected to interact with friendship style as interdependent friends gain more positive experiences over time that contribute to improved conflict management tactics. Disengaged friends, in contrast, may lack these opportunities of growth.

The fact that interdependent and disengaged friends did not differ in the reported frequency of conflict should not come as a surprise. Conflict rate is an imprecise measure of relationship quality. The consequences of a disagreement are mediated by features of the relationship; trust and closeness help to determine whether conflict is functional or dysfunctional (Cooper, 1988). It was surprising, however, not to find differences between friendships on conflict outcomes. Perhaps outcomes are so important to these voluntary relationships that, regardless of friendship style, equality is expected if the relationship is to continue. This may be especially true of the close friendships that were the focus of the present study. Conflicts may also differ such that outcomes are less free to vary across friendship types in important disagreements as compared with those that are not important (Laursen & Koplas, 1995).

This study was conducted with a sample of middle-class Israeli adolescents. Israel, in general, and its suburban middle class, in particular, represent a Western culture with values similar to those found in other industrial societies. Unlike a previous study that employed similar methods to identify friendship types (Shulman, Laursen, et al., 1997), there were no gender differences in the present study, a fact that we attribute more to the unique design of this study than to cultural differences in the friendships of boys and girls. The present study concerned reciprocated friendships in which both partners perceived the relationship to be very close. Gender differences may be less pronounced under these circumstances, because of ceiling or floor effects on behaviors. In the absence of additional research that extends these findings to less intimate friendships, these conclusions should be applied only to the closest of friendships.

A few additional caveats should be mentioned. Context exerts a powerful influence over the behavior of friends (Krappmann, 1996). Voluntary affiliations are especially vulnerable to the demands of the setting because their interconnections are tenuous and unstable. Adolescent friendships are more stable than are childhood friendships and perhaps less susceptible to contextual pressures (Hartup, 1992), but questions about ecological validity remain nevertheless. The card sort procedure is a well-established assessment tool for use in family relationships, and accumulating evidence supports the distinctions it draws between friendships. Differences between interdependent and disengaged friends have been documented in self-reports of participant attitudes toward their relationship and perceptions of closeness (Shulman, 1993; Shulman, Laursen, et al., 1997; Shulman, Levy-Shiff, et al., 1997), but associations between friends’ problem solving and observed behavior outside of the laboratory have yet to be established. Confidence in the present findings is bolstered by their similarity to those from related studies and, together, suggest that friendship differences in conflict behavior transcend variations in specific interaction processes. However, additional moderators should be considered, such as the length of the relationship and the characteristics of the individual participants. Finally, it was not possible to parse participant perceptions of conflict because some friends described different conflicts; future studies should consider the role of shared and independent perceptions of friendship behavior, a process best done by ensuring that all dyads experience and describe the same conflicts.

To conclude, adolescent friends managed conflict consistent with relationship distinctions based on dimensions derived from family systems theory. Interdependent friends behaved in a manner that balanced short-term individual outcomes against long-term relationship consequences, whereas disengaged friends tended to emphasize individual benefits at the expense of relationship outcomes. The long-term consequences of these differences remain to be determined. Although it may be the case that some friends thrive on contention, it may also be that children in disengaged friendships do not appreciate that the coercive behavior typical of disengaged friendships could be counterproductive in relationships with other peers.

Acknowledgments

The second author received support for the preparation of this manuscript from the U.S. National Institute of Child Health and Human Development (R29-HD33006).

Footnotes

1

Although there are important conceptual differences between the use of absolute and relative cutoff scores, in the present study the distinction was of little practical significance. Using the overall median score (.64) instead of the a priori cutoff score (.68) on coordination sort similarity resulted in a net change in the friendship classification of 1 dyad, which produced no appreciable differences in the overall pattern of results. Using age group medians (seventh grade = .54, eleventh grade = .67) on coordination sort similarity, 48% of seventh graders (14 females, 15 males) and 46% of eleventh graders (12 females, 5 males) were classified as disengaged friends, 45% of seventh graders (16 females, 11 males) and 54% of eleventh graders were classified as interdependent friends, and 7% of seventh graders (1 females, 3 males) and no eleventh graders were classified as enmeshed friends. Chi-square analyses failed to reveal statistically significant age or gender differences in this distribution of friendship types.

2

Identical analyses were conducted using the age group medians to classify friendship types; consensus sensitive or enmeshed dyads were excluded from these analyses due to small cell sizes. Similar results were obtained for all conflict components except conflict resolution. A 2 (age) × 2 (conflict type) × 4 (conflict resolution) log-linear analysis revealed a two-way interaction between age and conflict resolution, χ2(3, N = 93) = 11.62, p .01. Younger adolescents reported greater levels of power assertion (23%) and withdrawal (18%) than did older adolescents (3% and 8%, respectively), whereas older adolescents (73%) reported greater levels of compromise than did younger adolescents (43%). There were no differences between younger adolescents (16%) and older adolescents (16%) on levels of third-party resolution.

Contributor Information

Shmuel Shulman, Bar Ilan University.

Brett Laursen, Florida Atlantic University.

References

  1. Alon E. Intimacy in adolescent romantic relationships: Individual and systemic perspectives. Bar Ilan University; Israel: 1996. Unpublished master’s thesis. [Google Scholar]
  2. Buhrmester D, Furman W. Perceptions of sibling relationships during middle childhood and adolescence. Child Development. 1990;61:1387–1398. doi: 10.1111/j.1467-8624.1990.tb02869.x. [DOI] [PubMed] [Google Scholar]
  3. Chung T, Asher SR. Children’s goals and strategies in peer conflict situations. Merrill-Palmer Quarterly. 1996;42:125–147. [Google Scholar]
  4. Collins WA, Laursen B, Mortensen N, Luebker C, Ferreira M. Conflict processes and transitions in parent and peer relationships: Implications for autonomy and regulation. Journal of Adolescent Research. 1997;12:178–198. doi: 10.1177/0743554897122003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cooper CR. The role of conflict in adolescent–parent relationships. In: Gunnar MR, Collins WA, editors. The Minnesota Symposia on Child Psychology: Vol. 21. Development during the transition to adolescence. Hillsdale, NJ: Erlbaum; 1988. pp. 181–187. [Google Scholar]
  6. Dishion TJ, Andrews DW, Crosby L. Antisocial boys and their friends in early adolescence: Relationship characteristics, quality, and interactional processes. Child Development. 1995;66:139–151. doi: 10.1111/j.1467-8624.1995.tb00861.x. [DOI] [PubMed] [Google Scholar]
  7. Furman W, Wehner EA. Romantic views: Toward a theory of adolescent romantic relationships. In: Montemayor R, Adams GR, Gullota GP, editors. Relationships during adolescence. Thousand Oaks, CA: Sage; 1994. pp. 168–195. [Google Scholar]
  8. Gottman JM. Psychology and the study of marital processes. Annual Review of Psychology. 1998;49:169–197. doi: 10.1146/annurev.psych.49.1.169. [DOI] [PubMed] [Google Scholar]
  9. Graziano WG, Jensen-Campbell LA, Hair EC. Perceiving interpersonal conflict and reacting to it: The case for agreeableness. Journal of Personality and Social Psychology. 1996;70:820–835. doi: 10.1037//0022-3514.70.4.820. [DOI] [PubMed] [Google Scholar]
  10. Hartup WW. Conflict and friendship relations. In: Shantz CU, Hartup WW, editors. Conflict in child and adolescent development. New York: Cambridge University Press; 1992. pp. 186–215. [Google Scholar]
  11. Hartup WW. The company they keep: Friendships and their developmental significance. Child Development. 1996;67:1–13. [PubMed] [Google Scholar]
  12. Hartup WW, Laursen B, Stewart MI, Eastenson A. Conflict and the friendship relations of young children. Child Development. 1988;59:1590–1600. doi: 10.1111/j.1467-8624.1988.tb03686.x. [DOI] [PubMed] [Google Scholar]
  13. Korin N. Typology of close friendships and friendship reasoning in early adolescence. Tel Aviv University; Israel: 1992. Unpublished master’s thesis. [Google Scholar]
  14. Krappmann L. Amicitia, drujba, shin-yu, philia, Freundschaft, friendship: On the cultural diversity of a human relationship. In: Bukowski WM, Newcomb AF, Hartup WW, editors. The company they keep: Friendship in childhood and adolescence. New York: Cambridge University Press; 1996. pp. 1–40. [Google Scholar]
  15. Laursen B. Conflict management among close peers. In: Laursen B, editor. New directions for child development: No. 60. Close friendships in adolescence. San Francisco: Jossey-Bass; 1993. pp. 39–54. [DOI] [PubMed] [Google Scholar]
  16. Laursen B. Conflict and social interaction in adolescent relationships. Journal of Research on Adolescence. 1995a;5:55–70. [Google Scholar]
  17. Laursen B. Variations in adolescent conflict and social interaction associated with maternal employment and family structure. International Journal of Behavioral Development. 1995b;18:151–164. [Google Scholar]
  18. Laursen B, Finkelstein BD, Betts NT. A developmental meta-analysis of peer conflict resolution. Developmental Review. 2002;21:423–449. [Google Scholar]
  19. Laursen B, Hartup WW. The origins of reciprocity and social exchange in friendships. In: Laursen B, Graziano W, editors. New Directions for Child and Adolescent Development: No. 95. Social exchange in development. San Francisco: Jossey-Bass; 2002. [DOI] [PubMed] [Google Scholar]
  20. Laursen B, Hartup WW, Koplas AL. Towards understanding peer conflict. Merrill-Palmer Quarterly. 1996;42:76–102. [Google Scholar]
  21. Laursen B, Koplas AL. What’s important about important conflicts? Adolescents’ perceptions of daily disagreements. Merrill-Palmer Quarterly. 1995;41:536–553. [Google Scholar]
  22. Minuchin S. Families and family therapy. Cambridge, MA: Harvard University Press; 1974. [Google Scholar]
  23. Perry DG, Perry LC, Kennedy E. Conflict and the development of antisocial behavior. In: Shantz CU, Hartup WW, editors. Conflict in child and adolescent development. New York: Cambridge University Press; 1992. pp. 301–329. [Google Scholar]
  24. Piaget J. The moral judgment of the child. New York: Free Press; 1965. Original work published 1932. [Google Scholar]
  25. Poulin F, Dishion TJ, Haas E. The peer influence paradox: Friendship quality and deviancy training within male adolescent friendships. Merrill-Palmer Quarterly. 1999;45:42–61. [Google Scholar]
  26. Puttalaz M, Sheppard BH. Conflict management and social competence. In: Shantz CU, Hartup WW, editors. Conflict in child and adolescent development. New York: Cambridge University Press; 1992. pp. 330–355. [Google Scholar]
  27. Reiss D. The family’s construction of reality. Cambridge, MA: Harvard University Press; 1981. [Google Scholar]
  28. Reiss D, Costell R, Jones C, Berkman H. The family meets the hospital: A laboratory forecast of the encounter. Archives of General Psychiatry. 1980;37:141–154. doi: 10.1001/archpsyc.1980.01780150031003. [DOI] [PubMed] [Google Scholar]
  29. Rubinstein JL, Feldman SS. Conflict resolution behavior in adolescent boys: Antecedents and adaptational correlates. Journal of Research on Adolescence. 1993;3:41–66. [Google Scholar]
  30. Selman RL. The growth of interpersonal understanding: Developmental and clinical analyses. New York: Academic Press; 1980. [Google Scholar]
  31. Selman RL, Beardslee W, Schultz LH, Krupa M, Podorefsky D. Assessing adolescent interpersonal negotiation strategies: Toward the integration of structural and functional models. Developmental Psychology. 1986;22:340–349. [Google Scholar]
  32. Shantz CU. Conflict between children. Child Development. 1987;58:283–305. [Google Scholar]
  33. Shulman S. Early and middle adolescent close friendship: Typology and friendship reasoning. In: Laursen B, editor. New Directions for Child Development: No. 60. Close friendships in adolescence. San Francisco: Jossey-Bass; 1993. pp. 55–71. [DOI] [PubMed] [Google Scholar]
  34. Shulman S, Elicker J, Sroufe LA. Stages of friendship growth in preadolescence as related to attachment history. Journal of Social and Personal Relationships. 1994;11:341–361. [Google Scholar]
  35. Shulman S, Klein MM. Distance-sensitive and consensus-sensitive families: The effect on adolescent referral for psychotherapy. American Journal of Family Therapy. 1983;11:45–58. [Google Scholar]
  36. Shulman S, Laursen B, Kalman Z, Karpovsky S. Adolescent intimacy: Revisited. Journal of Youth and Adolescence. 1997;26:597–607. doi: 10.1023/A:1024586006966. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Shulman S, Levy-Shiff R, Kedem P, Alon E. Intimate relationships among adolescent romantic partners and same-sex friends: Individual and systemic perspectives. In: Shulman S, Collins WA, editors. New Directions for Child Development: No. 78. Romantic relationships in adolescence: Developmental perspectives. San Francisco: Jossey-Bass; 1997. [DOI] [PubMed] [Google Scholar]
  38. Shulman S, Zohar D. Family type and behavior problems of three-year-olds. American Journal of Family Therapy. 1991;19:266–276. [Google Scholar]
  39. Wynne LC. The epigenesis of relational systems: A model for understanding family development. Family Process. 1984;23:297–318. doi: 10.1111/j.1545-5300.1984.00297.x. [DOI] [PubMed] [Google Scholar]
  40. Youniss J, Smollar J. Adolescent relations with mothers, fathers, and friends. Chicago: University of Chicago Press; 1985. [Google Scholar]

RESOURCES