A Developmental Meta-Analysis of Peer Conflict Resolution

Share Embed


Descrição do Produto

Developmental Review 21, 423–449 (2001) doi:10.1006/drev.2000.0531, available online at http://www.idealibrary.com on

A Developmental Meta-Analysis of Peer Conflict Resolution Brett Laursen, Benjamin D. Finkelstein, and Noel Townsend Betts Florida Atlantic University A series of meta-analyses examine developmental trends in peer conflict resolution. Peer conflict resolutions are most likely to involve negotiation, with coercion and then disengagement the next most likely strategies. Patterns of conflict resolution differ with age. Coercion is common among children and disengagement is rare. Negotiation is prevalent among adolescents and young adults; the former do not differ in terms of coercion and disengagement, whereas the latter tend to avoid coercion in favor of disengagement. Conflict resolutions also vary as a function of peer relationships, assessment procedures, and reporters. Negotiation prevails in all peer relationships except those with siblings; there is more negotiation among romantic partners than among friends, and more negotiation among friends than among acquaintances. Negotiation is the overwhelming strategy of choice for those presented with hypothetical disputes, but actual conflicts tend to be resolved by coercion. Observers indicate that most conflicts involve coercive resolutions, in contrast to self-reports, which suggest that negotiation prevails. Although conclusions are qualified by the limited number of studies available, follow-up moderator analyses indicate that negotiation increases and coercion declines with age across most peer relationships, assessment procedures, and reporters such that different patterns of conflict resolution during childhood give way to the same relative ordering of strategies during young adulthood.  2001 Elsevier Science

Disagreements between peers represent a critical developmental challenge. Important social skills are acquired in interactions with agemates, one of which is the ability to amicably resolve disputes (Dunn, 1993). Children who lack this ability are at risk for maladjustment and social rejection (Newcomb, Bukowski, & Pattee, 1993). Narrative reviews of the literature suggest that conflict management skills improve with age, but they leave unspecified the manner and timing of change, and they caution that divergent developAccepted under the editorship of Dr. Grover J. Whitehurst. This investigation was completed with support from the U.S. National Institute of Child Health and Human Development (R29 HD33006) and a Johann Jacobs Foundation Young Investigator Award to the first author. Thanks are due to Margaret Ferreira, Michael Hayes, Kim Hernandez, Ernest Hodges, Julie Taylor-Ackner, John Venezia, and Amy Wilson for assisting with this project and to Erika Hoff for insightful comments on an earlier draft. We also extend our appreciation to those scholars who provided additional information about their research. Address correspondence and reprint requests to Brett Laursen, Florida Atlantic University, 2912 College Avenue, Fort Lauderdale, FL 33314. E-mail: [email protected] 423 0273-2297/01 $35.00  2001 Elsevier Science All rights reserved.

424

LAURSEN, FINKELSTEIN, AND BETTS

mental depictions emerge as a function of the reporter and the type of peer relationship (Canary, Cupach, & Messman, 1995; Hartup, 1992; Shantz, 1987). Efforts to delineate developmental trends are further complicated by discrepancies between preferred strategies for resolving conflict and actual behavior (Laursen, 1993). To overcome the limitations of conventional narrative reviews in specifying patterns of development, a meta-analysis was undertaken that describes the resolution of peer conflict from childhood through young adulthood, elaborating age-related trends as a function of relationships, reporters, and assessment procedures. The meta-analysis is designed to explicate normative developmental changes in peer conflict behavior and, in so doing, bolster intervention efforts aimed at cultivating conflict management abilities. Conflicts are ‘‘time-distributed social episodes’’ consisting of a series of discrete components that include issues, oppositions, resolutions, and outcomes (Shantz, 1987, p. 285). Conflict resolutions are of particular interest to developmental scholars because they provide a framework within which children acquire principles of justice (Ross, 1996), master the regulation of affect (Fabes & Eisenberg, 1992), and define personal autonomy (Nucci, Killen, & Smetana, 1996). An influential treatise on conflict resolution identified five conceptually distinct strategies for closing a disagreement (Vuchinich, 1990): compromise, third-party intervention, withdrawal, standoff, and submission. Empirical evidence suggests that these resolutions may be collapsed into three categories (Jensen-Campbell, Graziano, & Hair, 1996): negotiation (compromise and third-party resolution), disengagement (withdrawal and standoff ), and coercion (submission). The argument is often made that peer relationships provide children with opportunities to hone conflict resolution skills that are unavailable in other relationships. According to different developmental accounts, peers increasingly resolve disagreements through negotiation and compromise as the nature of their conflicts change and as they become more adept at avoiding coercive exchanges. Several theorists, elaborating on Piaget’s (1932/1965) assertion that peer relationships alter the child’s understanding of the social world and subsequent behavior in it, have suggested that conflict resolutions are a product of social cognitive maturity (Dunn, 1993; Selman, 1980; Smetana, 1988; Youniss & Smollar, 1985). By this account, cognitive development prompts advances in social understanding that permit self-interests to be subordinated to the interests of the partner and the relationship. Thus, social cognitive gains encourage a preference for negotiated resolutions that spurs the mastery of compromise skills. Scholars adopting an experiential perspective agree that coercion gives way to negotiation, although the purported mechanisms of developmental change differ (Hartup, 1996; Katz, Kramer, & Gottman, 1992; Laursen & Collins, 1994). By this account, tuition and practice promote the social skills necessary to resolve conflict in a manner that avoids disrupting the relationship. Thus, advances in interper-

PEER CONFLICT RESOLUTION

425

sonal skills enable more sophisticated compromise behaviors which foster a social cognitive appreciation for relationship distinctions. Neither perspective is inconsistent with the argument that developmental changes in resolutions may be traced to alterations in the nature and dynamics of peer conflict (Garvey & Shantz, 1992; Ross & Conant, 1992). In this view, age-related shifts in the resolution of peer conflict reflect the increasing salience of issues and settings that afford opportunities for compromise. The evidence supports claims of age-related shifts in the resolution of peer conflict, but discrepant data and a dearth of developmental research limit conclusions about the scope and magnitude of these changes (Canary et al., 1995; Hartup, 1992; Shantz, 1987). Considerable uncertainty surrounds potential moderators of developmental change. This meta-analytic review addresses two such moderators: relationships and research methods. Developmental changes in resolutions may be moderated by the type of relationship in which the conflict arises. There is reason to suspect that friends and romantic partners rely more on negotiation and less on coercion than acquaintances and siblings (Laursen, Hartup, & Koplas, 1996). Friends and romantic partners are invested in voluntary relationships, so their resolutions should reflect the desire to maintain rewarding interconnections; acquaintances lack investments and siblings are assured of relationship continuity, so their resolutions should evince little concern about disrupting interconnections. Advances in interpersonal understanding and social skills are often acquired first in friendships and then applied to other relationships (Ladd & Kochenderfer, 1996), so relationship differences are likely in the timing of the shift from coercion to negotiation. Distinctions between peers should diminish with age, as these gains are consolidated in all peer relationships. Developmental changes in peer conflict resolutions also may be moderated by research methods, with systematic variations anticipated as a function of assessment procedures and reporters. In terms of assessments, evidence drawn from the study of parent–child relationships suggests that hypothetical conflict scenarios evoke more constructive resolutions than those elicited by actual disputes (Smetana, Yau, & Hanson, 1991). Negotiation may be the preferred method for resolving conflict, but coercion is practiced more often. Divergent developmental depictions of the shift from coercion to negotiation are likely because assessments tap different ability domains: Appraisals of conflict behavior assay social skills, whereas responses to conflict scenarios gauge social cognitive abilities (Selman, Beardslee, Schultz, Krupa, & Podorefsky, 1986). Distinctions between actual and hypothetical events should decline with age as preferences and abilities converge on the goal of compromise. In terms of reporters, evidence drawn from studies of family processes suggests substantial discordance between participant and observer accounts of the same conflict episode (Gonzales, Cauce, & Mason, 1996). Consistent with the tendency for most people to cast their own interpersonal behavior

426

LAURSEN, FINKELSTEIN, AND BETTS

in a favorable light (Hinde, 1997), participants often report higher levels of negotiation and lower levels of coercion than observers. Developmental depictions of the shift from coercion to negotiation are likely to differ because reporters access different sources of information: Observers describe behavioral tangibles, whereas participants focus on personal goals and preferences (Olson, 1977). Distinctions between reporters should narrow with age as improved perspective taking skills increase the congruence of insider reports relative to outsider reports. The complexities of disentangling these competing sources of variation are such that it is not surprising that scholars have struggled with the challenge of summarizing the literature on developmental changes in the resolution of peer conflict. These difficulties suggest that the time is ripe for a meta-analysis of the topic. Unlike narrative reviews, which are constrained by the limited human capacity to assimilate and interpret large amounts of diverse data, meta-analysis offers an objective, quantitative procedure for synthesizing research, treating findings from multiple studies as a single complex data set (Hedges & Olkin, 1985). As such, it is uniquely suited to identify developmental trends across disparate data sets and to separate developmental trends from moderating influences. Two types of meta-analysis are available for the study of developmental change: direct and oblique. The procedures differ in terms of whether they contain explicit or implicit developmental data. A direct developmental meta-analysis summarizes studies that contrast two or more age groups (e.g., negotiation during childhood and adolescence). This procedure requires a substantial literature on change over time; only longitudinal and cross-sectional research qualify for inclusion. An oblique developmental meta-analysis does not involve explicit contrasts of age groups. Instead, multiple meta-analyses are conducted, each summarizing data within a distinct developmental period (e.g., negotiation and coercion during childhood). Standardized effects make possible indirect comparisons of findings from different age groups. The present investigation is typical of most developmental meta-analyses in that oblique procedures were employed because there were too few cross-sectional and longitudinal studies for direct analyses. In an oblique meta-analysis, potential moderators are isolated by conducting separate follow-up analyses to identify developmental trends among studies that share the same characteristics (e.g., self-reports of negotiation and coercion during childhood). By clarifying patterns of developmental change in peer conflict, this metaanalysis aims to advance efforts to help children with troubled peer relations. The limited success of programs designed to promote constructive conflict management (Furman & McQuaid, 1992) may be a function of undifferentiated assessments and incongruous interventions. One key to enriching these programs is a better understanding of factors that moderate normative developmental trends. The commonly held expectation that negotiation should increase with age fails to recognize that the social cognitions and social skills

PEER CONFLICT RESOLUTION

427

that promote negotiation develop according to discrete timetables. It follows that different children have different developmental deficits that require different interventions. More accurate assessments and better targeted interventions will follow from the recognition that conflict resolution skills may be acquired and applied at different rates in different peer relationships and that different methods and different reporters may present different views of individual development and adjustment. The review addresses three questions: (1) How do peers resolve conflicts? (2) Do peer conflict resolutions differ across childhood, adolescence, and young adulthood? and (3) Do the findings concerning age-related trends in peer conflict resolutions vary as a function of relationships, assessment procedures, and reporters? Rates of negotiation, coercion, and disengagement were contrasted in separate meta-analyses. Overall, high levels of negotiation and low levels of coercion were predicted for peer conflicts. Developmental differences were posited such that successive age groups should evince more negotiation and less coercion. Relationships and research methods were expected to moderate patterns of developmental change. With age, the resolution strategies of acquaintances and siblings should resemble those of romantic partners and friends, as lessons learned in close relationships are applied to other peer affiliations. Similarly, the gap between assessment procedures should decrease with age, as the practice of negotiation catches up to the optimistic level with which it is endorsed in hypothetical scenarios. Finally, self-report and observational data should grow more congruent with age, as improved social skills are reflected in more accurate interpersonal perceptions. METHOD Studies that qualified for the meta-analyses were located and reviewed in three stages. First, an abstract search identified research reports for further consideration. Second, the complete text of each research report was examined to identify those that met the criteria for the final stage of coding. Third, research reports were scored in terms of whether effect size estimates and study and sample characteristics could be ascertained. Reviewing Abstracts Computer searches of ERIC and PSYCHLIT were conducted for the years 1974 through 1998. An additional computer search of Dissertation Abstracts was conducted for 1900 through 1998. A search of key words and roots (i.e., agonism, argue, conflict, disagree, dispute, negotiate, problem solving, and resolution) generated over 70,000 abstracts. Elimination of duplicate entries reduced the total to approximately 25,000. Studies qualifying for further consideration met three criteria: (1) written in English; (2) contain original research; and (3) describe conflict among acquaintances, friends, romantic partners, or siblings. Of the abstracts identified from key words, 365 qualified

428

LAURSEN, FINKELSTEIN, AND BETTS

for the second stage of review. To assess interrater reliability, 10% of these abstracts (n ⫽ 2500) were chosen at random to be reviewed by three independent coders. Agreement on whether a study met the criteria for further consideration ranged from 87 to 96% (κ ⫽ .77 to .91). Differences among coders were resolved through discussion. Reviewing Research Reports In the second phase, two coders reviewed the complete text of the 365 research reports identified in the first phase. Studies selected for further consideration met three criteria: (1) Conflict was defined as a dyadic, interpersonal event involving overt, behavioral opposition (Shantz, 1987). Studies concerning arguments, disagreements, and quarrels were included but those limited to aggression, competition, intrapsychic events, and speech interruptions were excluded (see Collins & Laursen, 1992, for discussion). (2) Conflict among acquaintances, friends, romantic partners, and siblings could be distinguished from that in other relationships. Studies describing conflict between coworkers, spouses, and previously married couples were not included. (3) Participants included nonclinical populations ranging from early childhood through young adulthood, approximately 2 to 25 years old. Of the 365 research reports reviewed, 79 qualified for the next stage of coding. Agreement on selecting a study for further coding was 93% (κ ⫽ .86). Differences were resolved through discussion. Coding Research Reports In the third phase, three coders reviewed the 79 research reports that met the criteria for further consideration to determine whether effect size estimates and characteristics of the study and sample could be ascertained. Letters were sent to 37 investigators whose research reports lacked sufficient detail to calculate effect size estimates or to categorize the sample and method. Of the 16 who acknowledged the inquiry, 8 provided the necessary information. All reports included in the meta-analyses met two criteria: (1) central characteristics of the study and sample could be determined, including the number and age or grade level of participants, the conflict measure, and the source of the data; and (2) conflict resolution effect sizes could be estimated for the relative prevalence of negotiation, coercion, and disengagement. Negotiation describes compromise, reflecting concessions from both parties. Achieving the middle ground between two opposing positions involves sharing, turn-taking, or talking things out. Coercion describes a process whereby one party submits or capitulates to the demands of the other. Assertive tactics include commands, denials, and physical or verbal aggression. Disengagement describes dropping a conflict without achieving a resolution. Inconclusive solutions reflect withdrawals (e.g., discontinuing discussion or leaving the field) as well as standoffs (e.g., shifting the topic of speech

PEER CONFLICT RESOLUTION

429

or focus of activity). The majority of studies excluded at this stage were not empirical research reports in that they contained no quantitative data on conflict resolutions and, as a consequence, effects sizes could not be estimated. The rest of the excluded studies were divided between those that lacked sufficient detail to calculate effect sizes and those that did not consider all three conflict resolution strategies, either focusing exclusively on one or two ¨ sterman et al., 1997) or conflating strategies (e.g., Killen & Turiel, 1991; O two or more strategies (e.g., Selman et al., 1986; Slomkowski & Dunn, 1992). Effect sizes could not be estimated for studies that focused on a single conflict resolution strategy. Although effect sizes could be estimated for studies that were limited to two conflict resolution strategies, the manner in which strategies were typically combined or excluded (i.e., disengagement was usually either omitted or counted as a form of negotiation) systematically inflated these effect size estimates and made them conceptually and empirically distinct from estimates based on studies that included all three strategies. Third-party resolutions were not possible in some studies and were not considered in others, so this category was omitted. Of the 79 research reports coded, 31 were included in the meta-analyses. Interrater reliability on whether a study met the inclusion criteria ranged from 92 to 100% (κ ⫽ .85 to 1.00). Differences were resolved through discussion. Table 1 describes the study and sample characteristics of research reports included in the meta-analyses. It is not a complete summary of each study, but rather an overview of the data contributed to the meta-analyses. Selection and coding procedures altered the form or appearance of some data from that in the original report. In some studies, missing cases (i.e., Canary & Cupach, 1988; Johnson, Johnson, Dudley, & Acikgoz, 1994; Richardson, Hammock, Lubben, & Mickler, 1989; Rose & Asher, 1999) and attrition (i.e., Rafaelli, 1992) reduced the number of participants in the meta-analyses. In other studies, clinical subsamples (i.e., Lieber, 1994), intervention groups (i.e., Johnson et al., 1994; Johnson, Johnson, Dudley, Mitchell, & Fredrickson, 1997), and participants who described nonpeer relationships (i.e., Hammock, Richardson, Pilkington, & Utley, 1990; Vitaro & Pelletier, 1991) were excluded from the meta-analyses. Care was taken to ensure that each sample was represented only once in a meta-analysis. To this end, different investigations with the same participants and methods were identified, and the report with the most subjects was selected for inclusion. One investigation (i.e., Graziano, Jensen-Campbell, & Hair, 1996) contained two separate studies, each with distinct participants and different procedures. These unique studies were treated as separate research reports. Classification of Research Reports for Meta-Analyses Three independent coders classified research reports on five dimensions: (1) participant age group, (2) conflict measure, (3) peer relationship,

2.6- to 6-year-olds 6th to 8th grades

48 162

23.5-year-olds College students

140 342 3.3- to 5.3-year-olds

College students

124

53

2- to 5-year-olds College students

166 263

7.5- to 10.1-year-olds College students 11- to 14-year-olds 4- to 7-year-olds

92 488 192

Blackford (1993) Canary and Cupach (1988) Caplan, Bennetto, and Weissberg (1991) DeHart, Duffy, Kucharczak, Ghazanfari, and Johnson (1999) Eisenberg and Garvey (1981) Graziano, Jensen-Campbell, and Hair (1996, study 1) Graziano, Jensen-Campbell, and Hair (1996, study 2) Haferkamp (1991–1992) Hammock, Richardson, Pilkington, and Utley (1990) Hartup, Laursen, Stewart, and Eastenson (1988) Iskander, Laursen, Finkelstein, and Fredrickson (1995) Jensen-Campbell, Graziano, and Hair (1996)

Age or grade

26

N

Study

Hypothetical

Hypothetical

Actual

Actual Hypothetical

Actual

Actual Hypothetical

Actual

Hypothetical Actual Hypothetical

Conflict measure

All peer groups

Acquaintances, friends

Acquaintances, friends

Friends Friends, siblings

Acquaintances

Acquaintances All peer groups

Friends, siblings

Acquaintances Acquaintances Acquaintances, friends

Peer relationship

TABLE 1 Study and Sample Characteristics of Research Reports Included in Meta-Analyses

Self-report

Self-report

Observer

Self-report Self-report

Observer

Observer Self-report

Observer

Self-report Self-report Self-report

Source of data

430 LAURSEN, FINKELSTEIN, AND BETTS

9- to 11-year-olds 15.2- to 17.8-year-olds 11- to 17-year-olds 4.2-year-olds 11th grade 15.2-year-olds College students 5th to 9th grades 19.4-year-olds 4th to 5th grades College students 9th to 12th grades 20-year-olds 22.4-year-olds 1st to 2nd grades 11- to 15-year-olds 21.6-year-olds

666 100 90 160 443 57 96 264

6th to 9th grades

60 74 212 96 15 50 64 137 79 147

1st to 6th grades

86

Actual Hypothetical Actual

Hypothetical Actual Actual Hypothetical Hypothetical

Actual Actual Hypothetical Actual Hypothetical Actual Hypothetical Actual Hypothetical

Hypothetical

Hypothetical

Acquaintances Acquaintances, friends Acquaintances

Friends Acquaintances Acquaintances Romantic partners Acquaintances

Friends All peer groups Acquaintances Acquaintances Acquaintances Siblings Romantic partners Friends, siblings Friends, romantic partners

Acquaintances

Acquaintances

Note. N ⫽ participants included in effect size estimates. All peer groups ⫽ acquaintances, friends, siblings, and romantic partners.

Johnson, Johnson, Dudley, and Acikgoz (1994) Johnson, Johnson, Dudley, Mitchell, and Fredrickson (1997) Joshi (1999) Laursen and Koplas (1995) Levya and Furth (1986) Lieber (1994) McFarland and Culp (1992) Montemayor and Hanson (1985) Pistole (1989) Raffaelli (1992) Richardson, Hammock, Lubben, and Mickler (1989) Rose and Asher (1999) Sillars (1980) Stein, Bernas, and Calicchia (1997) Teismann and Mosher (1978) Trubisky, Ting-Toomey, and Lin (1991) Vitaro and Pelletier (1991) Whitesell and Harter (1996) Witteman (1992) Observer Self-report Self-report

Self-report Self-report Observer Observer Self-report

Self-report Self-report Self-report Observer Self-report Self-report Self-report Self-report Self-report

Self-report

Self-report

PEER CONFLICT RESOLUTION

431

432

LAURSEN, FINKELSTEIN, AND BETTS

(4) source of data, and (5) participant demographics. Interrater reliability for coding study and sample characteristics ranged from 88 to 100% (κ ⫽ .73 to 1.00). Differences were resolved through discussion. Participant age group. Studies were classified into three categories on the basis of participant age or school grade: childhood, adolescence, or young adulthood. A total of 10 reports involved children1 (2- to 10-year-olds or preschool to 5th grade), 10 reports included adolescents (11- to 18-year-olds or 6th to 12th grades), and 11 reports described young adults (19- to 25year-olds or college undergraduates). Conflict measure. Two types of conflict were identified. Actual conflicts describe the resolution of real disagreements. Hypothetical conflicts describe proposed resolutions to vignettes or imaginary events. A total of 15 reports involved actual conflicts and 16 reports concerned hypothetical conflicts. Peer relationship. Peer relationships were divided into four categories: acquaintances, friends, romantic partners, and siblings. Acquaintances included dormitory roommates, nonfriends, and classmates not otherwise specified as friends. Two reports identified friends empirically (e.g., reciprocal nominees in sociometric interviews) but otherwise participants defined and identified friends and romantic partners according to their own criteria. A total of 14 reports involved acquaintances, 3 reports described friends, 2 reports included romantic partners, and 1 report concerned siblings. In addition, 4 reports described acquaintances and friends; 3 reports concerned friends and siblings; 1 report involved friends and romantic partners; and 3 reports included acquaintances, friends, romantic partners, and siblings. Source of data. Data collection procedures were coded into two categories: observer reports and self-reports. A total of 8 studies described data obtained from observations. A total of 23 studies described self-report data obtained from interviews and questionnaires. Participant demographics. A total of 19 studies reported that the majority of participants were North Americans of European ancestry, 1 study reported that most participants were not North American, and 11 studies lacked sufficient information to characterize the sample. Socioeconomic status (SES) classifications revealed that the majority of participants were middle SES in 10 studies and lower SES in 1 study. No SES data were reported in 20 studies. All 31 studies included males and females, but many failed to differenti1 The childhood age group spans at least three distinct developmental periods: toddlerhood, early childhood, and middle childhood. As such, it may obscure developmental changes in conflict resolution. Unfortunately, the number of studies within each of the potential subcategories of childhood was small (k ⫽ 2 to 6) and the distribution of these studies varied systematically by moderator variables. Using this broad category, we were able to keep the number of studies in the childhood age group similar to the number of studies in the adolescence and young adulthood age groups, thus ensuring that each age group had a similar distribution of studies across moderator variables.

PEER CONFLICT RESOLUTION

433

ate resolutions according to gender. Insufficient data precluded separate analyses of conflict resolutions as a function of participant demographics. Statistical Analyses Effect size estimates. The Pearson product–moment correlation coefficient r may be used to describe nonexperimental data in a meta-analysis (Hedges & Olkin, 1985). Effect size estimates of r were calculated directly from means and standard deviations or indirectly from F, t, and χ 2 values (Schwarzer, 1989), where r represents the standardized difference between two resolution strategies derived from a trichotomized continuous variable (i.e., peer conflict resolutions may be divided into separate proportion scores for negotiation, coercion, and disengagement that total 100%). Cohen (1992) suggests that correlational data be interpreted in terms of small (r ⫽ .1), medium (r ⫽ .3), and large (r ⫽ .5 or greater) effects. For unspecified results that did not reach statistical significance, effect sizes were conservatively estimated as 0. Table 2 presents total peer conflict resolution effect size estimates for each research report in the meta-analyses. Effect size estimates were combined to obtain a mean unweighted 2 population effect (Hedges & Olkin, 1985). Prior to combining estimates, r values from individual research reports were converted to Z scores with Fisher’s r to Z transformation. An average Z score was calculated then transformed back to r. Homogeneity analyses determined whether research reports provided consistent estimates of population effects. The homogeneity statistic Q resembles the χ 2 test with k ⫺ 1 degrees of freedom, where k represents the number of effect sizes. Lack of statistical significance on homogeneity tests suggests that effect size estimates provided by different studies are similar in magnitude and direction. Statistically significant results indicate divergence among studies in effect size estimates (i.e., heterogeneity), which may be the product of outliers, moderators, or highly variable phenomena. Plan of analyses. Multiple meta-analyses test the prediction that peers practice negotiation and avoid coercion when resolving disputes. Each consists of three conflict resolution contrasts: (1) negotiation with coercion, (2) negotiation with disengagement, and (3) coercion with disengagement. Contrasts describing population effects are made first for overall peer conflict and then separately for each age group 3 (childhood, adolescence, and young 2 Population effects varied systematically across age groups, conflict measures, peer relationships, and sources of data. Unweighted effects are presented because the number of studies within each of these categories is more evenly balanced than the number of participants. Thus, unweighted population effects are less likely than weighted population effects to be biased by specific samples or methods. This approach is not without disadvantages, however, because individual participants make unequal contributions to unweighted population effects. 3 Supplemental analyses divided the childhood age group into preschool (2- to 5-year-olds; k ⫽ 5, N ⫽ 314) and middle childhood (6- to 10-year-olds; k ⫽ 6, N ⫽ 1,001) age groups. These findings should be interpreted with caution because the preschool studies consisted

434

LAURSEN, FINKELSTEIN, AND BETTS

adulthood), each conflict measure (actual and hypothetical), each peer relationship (acquaintances, friends, romantic partners, and siblings), and each source of data (self-reports and observations). A series of follow-up metaanalyses were planned to explore developmental moderators. In contrasts describing moderator effects, population effects are disaggregated, and effects for peer relationships, conflict measures, and data sources are examined within the childhood, adolescence, and young adulthood age groups. The hypothesis that peers prefer negotiation and shun coercion yields two predictions concerning the relative magnitude of resolution strategies. The direction of an effect (positive or negative) is a measure of support for the hypothesis. First, positive effects in contrasts involving negotiation indicate greater levels of negotiation than disengagement or coercion. Negative effects in these contrasts indicate less negotiation than disengagement or coercion. Second, positive effects in contrasts involving coercion indicate lower levels of coercion than negotiation or disengagement. Negative effects in these contrasts indicate greater coercion than negotiation or disengagement. RESULTS Table 3 describes a total of 36 population effects from peer conflict resolution meta-analyses. Table 4 describes a total of 63 moderator effects from follow-up developmental meta-analyses. The number of studies available for some moderator analyses was limited, so these results should be interpreted with caution. Confidence intervals that do not contain 0 denote statistically significant effects. Unless otherwise indicated, effects were not homogeneous. Population Effects Overall peer conflict. Significant population effects emerged for all peer conflict contrasts. Peers resolve conflicts more often with negotiation than with coercion or disengagement and more often with coercion than with disengagement. Age group. Significant population effects emerged for all age group contrasts, except for the contrast between coercion and disengagement during adolescence. Children resolve conflicts more often with coercion than with negotiation or disengagement and more often with negotiation than with disengagement. Adolescents resolve conflicts more often with negotiation than with coercion or disengagement; there is no difference in the prevalence of coercion and disengagement. Young adults resolve conflicts more often with negotiation than with coercion or disengagement and more often with disengagement than with coercion. Thus, there are progressive age-related increases in negotiation and declines in coercion; negotiation becomes the preprimarily of observations of actual conflicts with acquaintances, whereas the middle childhood studies tended to be self-reports of hypothetical disagreements with friends.

435

PEER CONFLICT RESOLUTION

TABLE 2 Effect Size Estimates of Peer Conflict Resolution by Research Reports Study Blackford (1993) Canary and Cupach (1988) Caplan, Bennetto, and Weissberg (1991) DeHart, Duffy, Kucharczak, Ghazanfari, and Johnson (1999) Eisenberg and Garvey (1981) Graziano, Jensen-Campbell, and Hair (1996, study 1) Graziano, Jensen-Campbell, and Hair (1996, study 2) Haferkamp (1991–1992) Hammock, Richardson, Pilkington, and Utley (1990) Hartup, Laursen, Stewart, and Eastenson (1988) Iskander, Laursen, Finkelstein, and Fredrickson (1995) Jensen-Campbell, Graziano, and Hair (1996) Johnson, Johnson, Dudley, and Acikgoz (1994) Johnson, Johnson, Dudley, Mitchell, and Fredrickson (1997) Joshi (1999) Laursen and Koplas (1995) Levya and Furth (1986) Lieber (1994) McFarland and Culp (1992) Montemayor and Hanson (1985) Pistole (1989) Raffaelli (1992) Richardson, Hammock, Lubben, and Mickler (1989) Rose and Asher (1999) Sillars (1980) Stein, Bernas, and Calicchia (1997) Teismann and Mosher (1978) Trubisky, Ting-Toomey, and Lin (1991) Vitaro and Pelletier (1991) Whitesell and Harter (1996) Witteman (1992)

Negotiation with coercion

Negotiation with disengagement

Coercion with disengagement

.35 ⫺.26 ⫺.23

.25 ⫺.63 .41

.12 .46 ⫺.41

⫺.83

⫺.55

⫺.68

⫺.43 .66

.36 .58

⫺.62 .19

.86

.79

.33

.85 .29

.79 .33

.22 ⫺.08

⫺.50

⫺.31

⫺.33

.80

.59

.71

.78

.56

⫺.51

⫺.60

⫺.53

.16

.29

.26

.35

⫺.10 ⫺.39 .71 ⫺.93 .65 ⫺.27 .47 ⫺.39 .43

.17 ⫺.41 .80 .50 ⫺.59 ⫺.40 .51 ⫺.53 .56

⫺.27 ⫺.02 ⫺.42 ⫺.94 .84 .15 ⫺.13 .20 ⫺.17

.71 ⫺.29 .46

.81 ⫺.16 .35

⫺.22 ⫺.14 .13

.53 .62

.51 .44

.02 .29

.04 ⫺.01 ⫺.17

.32 .25 .29

⫺.22 ⫺.31 ⫺.42

Note. r ⫽ effect size estimate. Positive r values in contrasts of negotiation and coercion, and in contrasts of negotiation and disengagement indicate greater levels of negotiation. Positive r values in contrasts of coercion and disengagement indicate greater levels of disengagement.

.16 ⫺.22 .21 .44 ⫺.18 .45 .12 .25 .52 ⫺.06 ⫺.18 .28

Overall peer conflict Age group Childhood Adolescence Young adulthood

Conflict measure Actual Hypothetical

Peer relationship Acquaintances Friends Romantic partners Siblings

Source of data Observer Self-report

.19)

to to to to

.15) .28) .56) .00)

(⫺.25 to ⫺.10) (.25 to .30)

(.08 (.21 (.47 (⫺.12

(⫺.22 to ⫺.14) (.42 to .48)

(⫺.27 to ⫺.17) (.15 to .27) (.41 to .47)

(.14 to

(95% Confidence interval)

.28 .23

.20 .31 .49 ⫺.02

.06 .40

.19 .09 .41

.24

r

to to to to

.23) .34) .54) .05)

(.21 to .35) (.20 to .26)

(.17 (.27 (.45 (⫺.08

(.02 to .11) (.37 to .43)

(.14 to .25) (.03 to .15) (.38 to .44)

(.22 to .27)

(95% Confidence interval)

Negotiation with disengagement

⫺.40 .05

⫺.07 ⫺.12 ⫺.12 ⫺.17

⫺.21 .06

to to to to

⫺.04) ⫺.08) ⫺.07) ⫺.11) (⫺.46 to ⫺.33) (.02 to .08)

(⫺.11 (⫺.16 (⫺.18 (⫺.23

(⫺.25 to ⫺.17) (.02 to .09)

(⫺.36 to ⫺.26) (⫺.03 to .09) (.01 to .09)

(⫺.10 to ⫺.05)

⫺.08 ⫺.31 .03 .05

(95% Confidence interval)

r

Coercion with disengagement

8 23

21 14 6 7

15 16

10 10 11

31

k

691 4,301

3,169 2,513 1,081 995

1,952 3,040

1,283 1,101 2,608

4,992

N

Note. r ⫽ unweighted population effect size; k ⫽ total number of independent samples; N ⫽ total number of independent participants. Positive r values in contrasts involving negotiation indicate greater levels of negotiation than coercion or disengagement. Positive r values in contrasts involving coercion indicate lower levels of coercion than negotiation or disengagement.

r

Contrast

Negotiation with coercion

TABLE 3 Peer Conflict Resolution Meta-Analyses: Population Effects

436 LAURSEN, FINKELSTEIN, AND BETTS

PEER CONFLICT RESOLUTION

437

vailing resolution strategy in adolescence but coercion does not fall below disengagement until young adulthood.4 Conflict measure. Significant population effects emerged for all conflict measure contrasts. Actual conflicts are resolved more often with coercion than with negotiation or disengagement and more often with negotiation than with disengagement. Hypothetical conflicts are resolved more often with negotiation than with coercion or disengagement, and more often with disengagement than with coercion. Thus, responses to hypothetical conflicts, where negotiation prevails and coercion is avoided, bear little relation to actual conflict behavior, where coercion prevails and disengagement is avoided. Peer relationship. Significant population effects emerged for all peer relationship contrasts except those involving negotiation among siblings. Acquaintances, friends, and romantic partners resolve conflicts more often with negotiation than with coercion or disengagement and more often with coercion than with disengagement. Siblings resolve conflicts more often with coercion than with disengagement; the prevalence of negotiation does not differ from that of coercion or disengagement. Thus, negotiation is prevalent in all peer relationships, except those with siblings; there is more negotiation among romantic partners than among friends and more negotiation among friends than among acquaintances. Rates of coercion are not lower than those of disengagement in any peer relationship. Source of data. Significant population effects emerged for all source of data contrasts. Observer reports reveal that peers resolve conflicts more often with coercion than with negotiation or disengagement and more often with negotiation than with disengagement. Self-reports reveal that peers resolve conflicts more often with negotiation than with coercion or disengagement and more often with disengagement than with coercion. Thus, reporters disagree about the relative prevalence of peer conflict resolutions: Self-reports indicate high rates of negotiation and low rates of coercion, but observer reports indicate high rates of coercion and low rates of disengagement. Moderator Effects Conflict measure by age group. Significant moderator effects emerged for actual conflicts in all age groups, except for the contrast between negotiation and disengagement during childhood. Children resolve actual conflicts more often with coercion than with negotiation or disengagement; there is no difference in the prevalence of negotiation and disengagement. Adolescents resolve actual conflicts more often with disengagement than with negotiation or coercion and more often with coercion than with negotiation. The effect 4 In separate analyses of the preschool and middle childhood age groups, significant population effects emerged for all contrasts. Levels of coercion were lower and levels of negotiation and disengagement were higher for younger children relative to older children.

.39 .43 .51

⫺.35 .33 .36

.05 .11 .60

Childhood Adolescence Young adulthood

Childhood Adolescence Young adulthood

Childhood Adolescence Young adulthood

(⫺.02 to (.04 to (.56 to

.11) .18) .64)

(⫺.43 to ⫺.27) (.27 to .39) (.32 to .40)

.45) .49) .55)

(⫺.64 to ⫺.50) (⫺.24 to ⫺.06) (.29 to .40)

⫺.57 ⫺.15 .35

Childhood Adolescence Young adulthood

(.34 to (.36 to (.47 to

(95% Confidence interval)

r

Age group

Negotiation with coercion (95% Confidence interval)

(⫺.02 to .18) (⫺.35 to ⫺.17) (.24 to .35)

(.29 to (.25 to (.45 to

.41) .39) .53)

(.05 to (.14 to (.22 to

.22) .27) .31)

.22 .16 .56

(.15 to (.09 to (.51 to

.28) .23) .60)

Peer relationship: Friends

.13 .21 .27

Peer relationship: Acquaintances

.35 .32 .49

Conflict measure: Hypothetical

.08 ⫺.26 .30

Conflict measure: Actual

r

Negotiation with disengagement

⫺.19 ⫺.23 .09

⫺.33 ⫺.01 .16

.23 ⫺.03 .02

⫺.60 .12 .09

r

.30) .05) .07)

(⫺.25 to ⫺.12) (⫺.29 to ⫺.16) (.02 to .15)

(⫺.41 to ⫺.25) (⫺.07 to .05) (.11 to .20)

(.17 to (⫺.11 to (⫺.03 to

(⫺.66 to ⫺.53) (.02 to .21) (.03 to .15)

(95% Confidence interval)

Coercion with disengagement

TABLE 4 Peer Conflict Resolution Meta-Analyses: Moderator Effects

5 5 4

7 8 6

4 6 6

6 4 5

k

867 754 892

517 970 1,682

892 656 1,492

391 445 1,116

N

438 LAURSEN, FINKELSTEIN, AND BETTS

— .02 .33

⫺.64 — .74

.30 .18 .35

Childhood Adolescence Young adulthood

Childhood Adolescence Young adulthood

Childhood Adolescence Young adulthood

.10) .41)

.53) .60)

(.25 to (.12 to (.31 to

.36) .24) .39)

(⫺.70 to ⫺.57) — (.68 to .79)

— (⫺.07 to (.24 to

— (.36 to (.49 to

— (.27 to (.50 to .45) .60)

— (⫺.18 to ⫺.01) (.39 to .53)

(⫺.05 to — (.61 to .73)

.17)

.32 .06 .34

(.26 to (.00 to (.30 to

.37) .13) .37)

Source of data: Self-report

.06 — .67

Source of data: Observer

— ⫺.09 .46

Peer relationship: Siblings

— .36 .55

.13 .02 .03

⫺.65 — .17

— ⫺.04 ⫺.18

— ⫺.35 ⫺.01

(.07 to (⫺.05 to (⫺.02 to

.20) .08) .07)

(⫺.71 to ⫺.58) — (.06 to .29)

— (⫺.13 to .04) (⫺.27 to ⫺.09)

— (⫺.44 to ⫺.26) (⫺.08 to .07)

5 9 9

5 1 2

1 4 2

0 2 4

966 1,011 2,324

317 90 284

26 517 452

0 374 707

Note. r ⫽ unweighted population effect size; k ⫽ total number of independent samples; N ⫽ total number of independent participants. Positive r values in contrasts involving negotiation indicate greater levels of negotiation than coercion or disengagement. Positive r values in contrasts involving coercion indicate lower levels of coercion than negotiation or disengagement.

— .45 .55

Childhood Adolescence Young adulthood

Peer relationship: Romantic partners

PEER CONFLICT RESOLUTION

439

440

LAURSEN, FINKELSTEIN, AND BETTS

of adolescent coercion with disengagement was homogeneous. Young adults resolve actual conflicts more often with negotiation than with coercion or disengagement and more often with disengagement than with coercion. Significant moderator effects emerged for hypothetical conflicts in all age groups, except for contrasts between coercion and disengagement during adolescence and young adulthood. Children resolve hypothetical conflicts more often with negotiation than with coercion or disengagement and more often with disengagement than with coercion. Adolescents and young adults resolve hypothetical conflicts more often with negotiation than with coercion or disengagement; there is no difference in the prevalence of coercion and disengagement. Taken together, age-related shifts in actual conflicts, supplanting coercion first with disengagement and then with negotiation, bring behavior in real disputes closer to depictions derived from hypothetical conflicts.5 Peer relationship by age group. Significant moderator effects emerged for acquaintances in all age groups, except for the contrast between coercion and disengagement during adolescence. Children resolve conflicts with acquaintances more often with coercion than with negotiation or disengagement and more often with negotiation than with disengagement. Adolescents resolve conflicts with acquaintances more often with negotiation than with coercion or disengagement; there is no difference in the prevalence of coercion and disengagement. Young adults resolve conflict with acquaintances more often with negotiation than with coercion or disengagement and more often with disengagement than with coercion. Significant moderator effects emerged for friends in all age groups, except for the contrast between negotiation and coercion during childhood. Children resolve conflicts with friends more often with negotiation and coercion than with disengagement; there is no difference in the prevalence of negotiation and coercion. Adolescents resolve conflicts with friends more often with negotiation than with coercion or disengagement and more often with coercion than with disengagement. Young adults resolve conflicts with friends more often with negotiation than with coercion or disengagement and more often with disengagement than with coercion. Significant moderator effects emerged for romantic partners in the adolescent and young adult age groups, except for the contrast between coercion and disengagement during young adulthood. No studies of romantic relationships were available to determine conflict resolutions during childhood, un5 In separate analyses of the preschool and middle childhood age groups, significant moderator effects emerged for all contrasts of actual conflict, except middle childhood negotiation with disengagement and middle childhood coercion with disengagement. In actual conflicts, levels of coercion were lower and levels of negotiation and disengagement were higher for older children relative to younger children. There were significant moderator effects for all middle childhood contrasts of hypothetical conflict, but there were too few studies of hypothetical conflict available to compare the conflict resolution strategies of preschool children.

PEER CONFLICT RESOLUTION

441

doubtedly because most children are not involved in romantic relationships. Adolescents resolve conflicts with romantic partners more often with negotiation than with coercion or disengagement and more often with coercion than with disengagement. Young adults resolve conflicts with romantic partners more often with negotiation than with coercion or disengagement; there is no difference in the prevalence of coercion and disengagement. Significant moderator effects emerged for siblings in the adolescent and young adult age groups, except in contrasts involving coercion during adolescence. Too few studies of sibling relationships were available to determine conflict resolutions in childhood. Adolescents resolve conflicts with siblings more often with disengagement than with negotiation; the prevalence of coercion does not differ from that of negotiation or disengagement. Young adults resolve conflicts with siblings more often with negotiation than with coercion or disengagement and more often with coercion than with disengagement. The effect of young adult coercion with disengagement was homogeneous. Taken together, age-related increases in negotiation are accompanied by a decline in coercion in all relationships except those with siblings.6 Source of data by age group. Significant moderator effects emerged for observer reports in the child and young adult age groups, except in the contrast between negotiation and disengagement. Too few studies with observer reports were available to determine conflict resolutions in adolescence. Observers indicate that children resolve conflicts more often with coercion than with negotiation or disengagement; there is no difference in the prevalence of negotiation and disengagement. Observers indicate that young adults resolve conflicts more often with negotiation than with coercion or disengagement and more often with disengagement than with coercion. Significant moderator effects emerged for self-reports in all age groups, except in the contrast between negotiation and disengagement during adolescence and in the contrast between coercion and disengagement during adolescence and young adulthood. Self-reports indicate that children resolve conflicts more often with negotiation than with coercion or disengagement and more often with disengagement than with coercion. Self-reports indicate that adolescents resolve conflicts more often with negotiation than with coercion; the prevalence of disengagement does not differ from that of negotiation and coercion. Self-reports indicate that young adults resolve conflicts more often with negotiation than with coercion or disengagement; there is no difference in the prevalence of coercion and disengagement. 6 In separate analyses of the preschool and middle childhood age groups, significant moderator effects emerged for all contrasts of preschool acquaintances and middle childhood friends; there were no statistically significant differences between the conflict resolution strategies of preschool friends and middle childhood acquaintances. For both friends and acquaintances, levels of coercion were lower and levels of negotiation and disengagement were higher for older children relative to younger children.

442

LAURSEN, FINKELSTEIN, AND BETTS

Taken together, discrepancies between reporters diminish with age, as increases in negotiation and declines in coercion bring observer reports in line with self-reports.7 DISCUSSION This meta-analytic review specified normative patterns of peer conflict, identified developmental alterations in conflict resolutions, and elaborated findings for age related trends in terms of relationships and research methods. The results confirmed the hypothesis that negotiation is the predominant means of resolving conflict between peers, but they did not support the assertion that peers refrain from coercion. Overall levels of coercion are somewhat greater than those of disengagement. Even so, the resolution of peer conflicts stand in sharp contrast to the resolution of parent–child conflicts, where coercion predominates and negotiation is atypical (Laursen, 1993). Peers share power, so negotiation ensures equivalent outcomes for both participants. In contrast, parents hold more power than children, so coercion ensures favorable results for the puissant. Peers do not eschew coercion completely, but there is support for the view that participants make an effort to avoid behaviors that may disrupt future interactions. Developmental variations emerged to temper these generalizations. Resolution strategies vary across childhood, adolescence, and young adulthood. The relative importance of coercion diminishes with age as that of negotiation and disengagement increase. Children tend to resolve disputes with coercion and refrain from disengagement. Adolescents favor negotiation, and practice disengagement and coercion in equal parts. Young adults evince high levels of negotiation and low levels of coercion. Elaborating on previous reviews which suggested that coercion gives way to negotiation (Laursen et al., 1996; Newcomb & Bagwell, 1995), these findings are important in that they describe incremental advances in the use of disengagement during successive age periods. Much has been made of developmental increases in negotiation, but the parallel emergence of disengagement suggests that improvements in conflict resolution skills also may include learning to walk away from a dispute. Qualifications must accompany conclusions about developmental change. Findings for conflict measures confirmed hypothesized differences between hypothetical and actual conflict resolutions: Coercion is favored in actual conflicts whereas negotiation prevails in hypothetical disagreements. Find7 In separate analyses of the preschool and middle childhood age groups, significant moderator effects emerged for all contrasts of observer reports, except middle childhood negotiation with disengagement. Observers indicate that levels of coercion were lower and levels of negotiation and disengagement were higher for older children relative to younger children. Significant moderator effects emerged for all middle childhood contrasts of self-reports, but too few self-report studies were available to compare the conflict resolution strategies of preschool children.

PEER CONFLICT RESOLUTION

443

ings also supported assertions that resolutions to real and hypothetical conflicts grow more similar with age: Increasing negotiation and declining coercion in actual conflicts diminish the discrepancy between assessment procedures. The difficulties of linking professed preferences to behavioral practices in peer conflict have been enumerated elsewhere (Hay, ZahnWaxler, Cummings, & Iannotti, 1992), but these results are a pointed reminder to those who study interpersonal conflict that hypothetical scenarios elicit levels of negotiation that are much higher than those found in actual disputes. Concerns about the validity of responses to hypothetical events are likely to be raised in light of the weak correspondence between measures. Those who favor the procedure could argue that hypothetical scenarios prompt similar embellishments from all respondents or that hypothetical scenarios amplify individual differences on dimensions that separate the well adjusted from the less well adjusted. Those less favorably inclined toward the procedure could argue that hypothetical accounts confound differences in interpersonal skills with differences in interpersonal perceptions, such that children and adolescents who accurately ascribe actual resolutions to hypothetical events may be indistinguishable from those with a proclivity to think and act coercively. One thing is certain: Hypothetical accounts imply that negotiation is normative, which exaggerates the level of interpersonal competence typically displayed by children and adolescents during real disputes. This offers an important lesson on the topic of social adjustment because although troubled youth negotiate less than their classmates, the predominant use of coercion during this period is not, in and of itself, atypical or maladaptive (Putallaz & Sheppard, 1992). Interventions must acknowledge these interpersonal realities, striving to improve negotiation skills without conveying unrealistic expectations about their implementation frequency. As predicted, patterns of conflict resolution vary across peer relationships. Negotiation is the preferred means of resolving disputes in voluntary peer relationships, yet rates among romantic partners exceeded those among friends, and rates among friends surpassed those among acquaintances. Assumptions that patterns of conflict resolution in peer relationships grow more similar with age were partially confirmed. From childhood through young adulthood, negotiation increases and coercion declines for all peers except siblings. The prevalence of negotiation among peers reflects the realities of sharing power. Yet within these parameters, there are differences between relationships in rates of negotiation that are consistent with differences in the degree to which affiliations are interdependent and voluntary (Laursen et al., 1996). Interdependence characterizes friendships and romantic relationships. Participants are invested in continuing exchanges, so they rely on negotiation to preserve interconnections. Acquaintances, in contrast, lack interdependence. Participants have no history of rewarding interactions, so there is little incentive to negotiate for the sake of amity. Sibling relationships are interdependent, but unlike other peer relationships, the affiliation is oblig-

444

LAURSEN, FINKELSTEIN, AND BETTS

atory rather than voluntary. Conflict resolutions have no bearing on relationship continuity, so participants are not motivated to avoid relationship disruption. Differences between peer relationships in the relative frequency of resolution strategies and their diminution with age provide important new evidence suggesting that friends play a central role in the development of conflict management skills. Together with the findings concerning variations in patterns of change as a function of assessment procedures, there is support for the argument that social cognitive advances promoting negotiation emerge from experiences with friends and prompt new strategies for managing conflict that are initially implemented and refined in this relationship (Dunn, 1993; Selman, 1980; Smetana, 1988; Youniss & Smollar, 1985). As predicted, patterns of conflict resolution vary across sources of data. Observers indicate that most conflicts are resolved with coercion; self-reports suggest that negotiation prevails. Observers indicate that peers shun disengagement; self-reports suggest that it is coercion that is avoided. As anticipated, patterns of conflict resolution grow commensurate with age. Observed increases in negotiation are accompanied by declines in coercion that minimize differences between reporters. Apparently, descriptions of conflict resolutions are susceptible to perceptual and response biases. Self-reports reveal greater levels of negotiation and disengagement than observer reports, raising the possibility that individuals may be tempted to provide socially desirable responses (Canary et al., 1995). Different scales of reference also may contribute to differences between reporters. Observers rely on absolute behavioral criteria to define resolution strategies, whereas participants rate the episode relative to personal goals and past relationship experiences (Olson, 1977). Then, too, the time frame available to reporters may contribute to differences in conflict resolutions. Observers interpret closure as resolution, but a participant planning to revisit a dispute later views a temporary cessation of hostilities as disengagement. Finally, perceptual differences may be exacerbated by methodological constraints. Observations often take place in closed laboratory settings, where disengagement is not much of an option and where disincentives for coercion are removed (Collins & Laursen, 1992). Although it was not possible to distinguish between these sources of influence, the present study is important in that it demonstrates that the gulf between self-reports and observer reports of interpersonal conflict narrow with age. Practically speaking, this means that scholars may expect insider reports and outsider reports to yield similar depictions of conflict among adults, but that the methods ought not be considered interchangeable in studies of children and adolescents. This investigation is not the final word on peer conflict resolution. The present study employed operational definitions that sharpened distinctions between resolution strategies by eliminating studies with categories that crossed dimensional boundaries. As a consequence, the number of research reports meeting the qualifying criteria was limited. Caution is warranted in

PEER CONFLICT RESOLUTION

445

interpreting the results because of the small number of studies available for some of the moderator analyses and because aggregate age-groups were constructed that may have obscured subtle developmental differences. Although supplemental analyses suggested less coercion and more negotiation and disengagement in the middle childhood group relative to the preschool age group, insufficient data precluded the replication of these trends with all moderator variables. A lack of data also prevented the consideration of two important moderators. First, previous work indicates that gender (Miller, Danaher, & Forbes, 1986) and ethnicity (Killen & Sueyoshi, 1995) shape conflict resolutions, which leaves open the possibility that participant demographic characteristics moderate patterns of developmental change. Second, previous research indicates that conflict topics shape conflict resolutions (Killen & Turiel, 1991) and change with age (Youniss, 1980). Thus, there is reason to suspect that conflict topics may moderate developmental changes in peer conflict resolution strategies, but experience cautions against ascribing causality to this correlational evidence: Recent research indicates that conflict topics are not responsible for differences between parent–child and friend relationships in conflict resolutions, even though they are independently linked to both relationships and resolutions (Adams & Laursen, 2001). Simply put, more studies are needed before firm conclusions can be advanced concerning higher order interactions among moderator variables. As with any meta-analysis, the file drawer problem must be considered. Unpublished studies may contain findings that would alter the conclusions of this review. Typically, the file drawer problem overestimates effect sizes because nonsignificant results tend not to be published (Hedges & Olkin, 1985). In the current meta-analyses, the file drawer problem poses the greatest threat to contrasts of coercion with disengagement, where an average of 21 null population effects and 21 null moderator effects would render findings nonsignificant. By comparison, contrasts of negotiation with coercion require an average of 59 null population effects and 38 null moderator effects, and contrasts of negotiation with disengagement require an average of 60 null population effects and 26 null moderator effects to change the statistical significance of the findings. Thus, despite the small number of studies available for the meta-analyses, fail-safe figures suggest that most of the effects are quite robust, especially those concerning negotiation. We conclude that successful strategies for managing conflict develop gradually with age, primarily through experience in close peer relationships. Sharing power and preserving interconnections are strong incentives to practice negotiation and avoid coercion. Age-related changes are consistent with assertions that social-cognitive maturity promotes dispute management skills, but vicissitudes across relationships and research methods in the findings that describe developmental trajectories underscore the need to delineate the ontogenetic basis of transformations in peer conflict resolution.

446

LAURSEN, FINKELSTEIN, AND BETTS

REFERENCES* Adams, R., & Laursen, B. (2001). The organization and dynamics of adolescent conflict with parents and friends. Journal of Marriage and the Family, 63, 97–110. *Blackford, I. T. (1993). Validation of the social problem solving inventory (SPIN). Unpublished doctoral dissertation, Fuller Theological Seminary, School of Psychology, Pasadena, CA. *Canary, D. J., & Cupach, W. R. (1988). Relational and episodic characteristics associated with conflict tactics. Journal of Social and Personal Relationships, 5, 305–325. Canary, D. J., Cupach, W. R., & Messman, S. J. (1995). Relationship conflict: Conflict in parent–child, friendship, and romantic relationships. Thousand Oaks, CA: Sage. *Caplan, M., Bennetto, L., & Weissberg, R. P. (1991). The role of interpersonal context in the assessment of social problem-solving skills. Journal of Applied Developmental Psychology, 12, 103–114. Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. *DeHart, G. B., Duffy, L., Kucharczak, K., Ghazanfari, S., & Johnson, S. (1999, April). Conflict between siblings and friends in early and middle childhood. Paper presented at the meeting of the Society for Research in Child Development, Albuquerque, NM. Dunn, J. (1993). Young children’s close relationships: Beyond attachment. Newbury Park, CA: Sage. *Eisenberg, A. R., & Garvey, C. (1981). Children’s use of verbal strategies in resolving conflicts. Discourse Processes, 4, 149–170. Fabes, R., & Eisenberg, N. (1992). Young children’s coping with interpersonal stress. Child Development, 63, 116–128. Furman, W., & McQuaid, E. L. (1992). Intervention programs for the management of conflict. In C. U. Shantz & W. W. Hartup (Eds.), Conflict in child and adolescent development (pp. 402–429). New York: Cambridge Univ. Press. Gonzales, N. A., Cauce, A. M., & Mason, C. A. (1996). Interobserver agreement in the assessment of parental behavior and parent-adolescent conflict: African American mothers, daughters, and independent observers. Child Development, 67, 1483–1498. *Graziano, W. G., Jensen-Campbell, L. A., & Hair, E. C. (1996). Perceiving interpersonal conflict and reacting to it: The case for agreeableness. Journal of Personality and Social Psychology, 70, 820–835. *Haferkamp, C. J. (1991–1992). Orientations to conflict: Gender, attributions, resolution strategies, and self-monitoring. Current Psychology: Research and Reviews, 10, 227–240. *Hammock, G. S., Richardson, D. R., Pilkington, C. J., & Utley, M. (1990). Measurement of conflict in social relationships. Personality Processes and Individual Differences, 11, 577–583. Hartup, W. W. (1992). Conflict and friendship relations. In C. U. Shantz & W. W. Hartup (Eds.), Conflict in child and adolescent development (pp. 186–215). New York: Cambridge Univ. Press. Hartup, W. W. (1996). Cooperation, close relationships, and cognitive development. In W. M. Bukowski, A. F. Newcomb, & W. W. Hartup (Eds.), They company they keep: Friendship in childhood and adolescence (pp. 213–237). New York: Cambridge Univ. Press. *Hartup, W. W., Laursen, B., Stewart, M. I., & Eastenson, A. (1988). Conflict and the friendship relations of young children. Child Development, 59, 1590–1600. Hay, D. F., Zahn-Waxler, C., Cummings, E. M., & Iannotti, R. J. (1992). Young children’s * Studies included in the meta-analyses.

PEER CONFLICT RESOLUTION

447

views about conflict with peers: A comparison of the daughters and sons of depressed and well women. Journal of Child Psychology and Psychiatry, 33, 669–683. Hedges, L., & Olkin, I. (1985). Statistical methods for meta-analysis. San Diego: Academic Press. Hinde, R. (1997). Relationships: A dialectical perspective. Hove, UK: Psychology Press. *Iskander, N., Laursen, B., Finkelstein, B., & Fredrickson, L. (1995). Conflict resolution among preschool children: The appeal of negotiation in hypothetical disputes. Early Education and Development, 6, 359–376. *Jensen-Campbell, L. A., Graziano, W. G., & Hair, E. C. (1996). Personality and relationships as moderators of interpersonal conflict in adolescence. Merrill-Palmer Quarterly, 42, 148–164. *Johnson, D. W., Johnson, R., Dudley, B., & Acikgoz, K. (1994). Effects of conflict resolution training on elementary school students. Journal of Social Psychology, 134, 803–817. *Johnson, D. W., Johnson, R., Dudley, B., Mitchell, J., & Fredrickson, J. (1997). The impact of conflict resolution training on middle school students. Journal of Social Psychology, 137, 11–21. *Joshi, A. (1999, April). Emotions and strategies: Children’s reports of emotions in conflict situations and their conflict resolution strategies. Paper presented at the meeting of the Society for Research in Child Development, Albuquerque, NM. Katz, L. F., Kramer, L., & Gottman, J. M. (1992). Conflict and emotions in marital, sibling, and peer relationships. In C. U. Shantz & W. W. Hartup (Eds.), Conflict in child and adolescent development (pp. 122–149). New York: Cambridge Univ. Press. Killen, M., & Sueyoshi, L. (1995). Conflict resolution in Japanese social interactions. Early Education and Development, 6, 317–334. Killen, M., & Turiel, E. (1991). Conflict resolution in preschool social interactions. Early Education and Development, 2, 240–255. Ladd, G. W., & Kochenderfer, B. J. (1996). Linkages between friendship and adjustment during early school transitions. In W. M. Bukowski, A. F. Newcomb, & W. W. Hartup (Eds.), They company they keep: Friendship in childhood and adolescence (pp. 322– 345). New York: Cambridge Univ. Press. Laursen, B. (1993). Conflict management among close peers. In B. Laursen (Ed.), New Directions for Child Development. Close friendships in adolescence (No. 60, pp. 39–54). San Francisco: Jossey–Bass. Laursen, B., & Collins, W. A. (1994). Interpersonal conflict during adolescence. Psychological Bulletin, 115, 197–209. Laursen, B., Hartup, W. W., & Koplas, A. L. (1996). Towards understanding peer conflict. Merrill-Palmer Quarterly, 42, 76–102. *Laursen, B., & Koplas, A. L. (1995). What’s important about important conflicts? Adolescents perceptions of daily disagreements. Merrill-Palmer Quarterly, 41, 536–553. *Levya, F. A., & Furth, H. G. (1986). Compromise formation in social conflicts: The influence of age, issue, and interpersonal context. Journal of Youth and Adolescence, 15, 441–452. *Lieber, J. (1994). Conflict and its resolution in preschoolers with and without disabilities. Early Education and Development, 5, 5–17. *McFarland, W. P., & Culp, W. H. (1992). Interpersonal skill training for effective conflict resolution. The School Counselor, 39, 304–310. Miller, P. M., Danaher, D. L., & Forbes, D. (1986). Sex-related strategies for coping with interpersonal conflict in children aged five and seven. Developmental Psychology, 22, 543–548.

448

LAURSEN, FINKELSTEIN, AND BETTS

*Montemayor, R., & Hanson, E. (1985). A naturalistic view of conflict between adolescents and their parents and siblings. Journal of Early Adolescence, 5, 23–30. Newcomb, A. F., & Bagwell, C. L. (1995). Children’s friendship relations: A meta-analytic review. Psychological Bulletin, 117, 306–347. Newcomb, A. F., Bukowski, W. M., & Pattee, L. (1993). Children’s peer relations: A metaanalytic review of popular, rejected, neglected, controversial, and average sociometric status. Psychological Bulletin, 113, 99–128. Nucci, L. P., Killen, M., & Smetana, J. G. (1996). Autonomy and the personal: Negotiation and social reciprocity in adult-child social exchanges. In M. Killen (Ed.), New Directions for Child Development. Children’s autonomy, social competence, and interactions with adults and other children: Exploring connections and consequences (No. 73, pp. 7–24). San Francisco: Jossey–Bass. Olson, D. H. (1977). Insiders’ and outsiders’ views of relationships: Research studies. In G. Levinger & H. L. Raush (Eds.), Close relationships: Perspectives on the meaning of intimacy (pp. 115–135). Amherst: Univ. of Massachusetts Press. ¨ sterman, K., Bjo¨rkqvist, K., Lagerspetz, K. M. J., Landau, S. F., Fraczek, A., & PastoO relli, C. (1997). Sex differences in styles of conflict resolution: A developmental and cross-cultural study with data from Finland, Israel, Italy, and Poland. In D. P. Fry & K. Bjo¨rkqvist (Eds.), Cultural variations in conflict resolution: Alternatives to violence (pp. 182–197). Mahwah, NJ: Erlbaum. Piaget, J. (1965). The moral judgment of the child. New York: Free Press. [Original work published in 1932] *Pistole, M. C. (1989). Attachment in adult romantic relationships: Style of conflict resolution and relationship satisfaction. Journal of Social and Personal Relationships, 6, 505–510. Putallaz, M., & Sheppard, B. H. (1992). Conflict management and social competence. In C. U. Shantz & W. W. Hartup (Eds.), Conflict in child and adolescent development (pp. 330–355). New York: Cambridge Univ. Press. *Raffaelli, M. (1992). Sibling conflict in early adolescence. Journal of Marriage and the Family, 54, 652–663. *Richardson, D. R., Hammock, G. S., Lubben, T., & Mickler, S. (1989). The relationship between love attitudes and conflict responses. Journal of Social and Clinical Psychology, 8, 430–441. *Rose, A. J., & Asher, S. R. (1999). Children’s goals and strategies in response to conflicts within a friendship. Developmental Psychology, 35, 69–79. Ross, H. S. (1996). Negotiating principles of entitlement in sibling property disputes. Developmental Psychology, 32, 90–101. Ross, H. S., & Conant, C. L. (1992). The social structure of early conflict: Interaction, relationships, and alliances. In C. U. Shantz & W. W. Hartup (Eds.), Conflict in child and adolescent development (pp. 153–185). New York: Cambridge Univ. Press. Schwarzer, R. (1989). Meta-analysis programs. Freie University of Berlin. Selman, R. L. (1980). The growth of interpersonal understanding. New York: Academic Press. Selman, R. L., Beardslee, W., Schultz, L. H., Krupa, M., & Podorefsky, D. (1986). Assessing adolescent interpersonal negotiation strategies: Toward the integration of structural and functional models. Developmental Psychology, 22, 450–459. Shantz, C. U. (1987). Conflicts between children. Child Development, 58, 283–305. *Sillars, A. L. (1980). Attributions and communication in roommate conflicts. Communication Monographs, 47, 180–200. Slomkowski, C. L., & Dunn, J. (1992). Arguments and relationships within the family: Differ-

PEER CONFLICT RESOLUTION

449

ences in young children’s disputes with mothers and sibling. Developmental Psychology, 28, 919–924. Smetana, J. G. (1988). Concepts of self and social convention: Adolescents’ and parents’ reasoning about hypothetical and actual family conflicts. In M. R. Gunnar & W. A. Collins (Eds.), The Minnesota Symposia on Child Psychology: Vol. 21. Development during the transition to adolescence (pp. 79–122). Hillsdale, NJ: Erlbaum. *Stein, N. L., Bernas, R. S., & Calicchia, D. J. (1997). Conflict talk: Understanding and resolving arguments. In T. Givon (Ed.), Conversation: Cognitive, communicative, and social perspectives (pp. 233–268). Philadelphia: Benjamin. *Teismann, M. W., & Mosher, D. L. (1978). Jealous conflict in dating couples. Psychological Reports, 42, 1211–1216. *Trubisky, P., Ting-Toomey, S., & Lin, S. L. (1991). The influence of individualism–collectivism and self-monitoring on conflict styles. International Journal of Intercultural Relations, 15, 65–84. *Vitaro, F., & Pelletier, D. (1991). Assessment of children’s social problem-solving skills in hypothetical and actual conflict situations. Journal of Abnormal Child Psychology, 19, 505–518. Vuchinich, S. (1990). The sequential organization of closing in verbal family conflict. In A. D. Grimshaw (Ed.), Conflict talk: Sociolinguistic investigations of arguments in conversations (pp. 118–138). New York: Cambridge Univ. Press. *Whitesell, N. R., & Harter, S. (1996). The interpersonal context of emotion: Anger with close friends and classmates. Child Development, 67, 1345–1359. *Witteman, H. (1992). Analyzing interpersonal conflict: Nature of awareness, type of initiating event, situational perceptions, and management styles. Western Journal of Communication, 56, 248–280. Youniss, J. (1980). Parents and peers in social development: A Piaget–Sullivan perspective. Chicago: Univ. of Chicago Press. Youniss, J., & Smollar, J. (1985). Adolescent relations with mothers, fathers, and friends. Chicago: Univ. of Chicago Press. Received: February 11, 2000; published online October 2, 2001

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.