Effects of Conflict Resolution Training Integrated into a Kindergarten Curriculum

Child Development, May/June 2000, Volume 71, Number 3, Pages 772 784 Effects of Conflict Resolution Training Integrated into a Kindergarten Curriculum Laurie Stevahn, David W. Johnson, Roger T. Johnson, Katie Oberle, and Leslie Wahl The effectiveness of a conflict resolution training program was examined in an American midwestern suburban elementary school. Participants were 80 kindergartners randomly assigned to an experimental or control condition in morning or afternoon time blocks. Children in the experimental condition received 9 hr of conflict resolution training integrated into a curriculum unit on friendship taught daily for 4 consecutive weeks. Children in the control condition were taught the identical friendship unit for the same period of time without conflict resolution training. Teachers rotated equally across conditions. Significant differences between trained and untrained children occurred in their knowledge and retention of the conflict resolution procedure, willingness and ability to use the procedure in conflict situations, and conceptual understanding of friendship. INTRODUCTION Conflict is posited as an essential impetus to change, adaptation, and development by almost every major theory that addresses fundamental issues in social and cognitive development (Shantz, 1987). Freud (Hall, 1954), for example, posited that children are in continuous conflict with adults and other children over the incompatibility between the child s drives (the wish ) and society s demands and rules (the ought ). Erikson (1959) elaborated on Freud s theorizing by positing that every child is in conflict with the representatives of society (parents, teachers, older children), with crises of psychosocial development (such as whether or not to trust), and with modes of adapting (to hold on or to let go ). Piaget (1928, 1932) posited that children face continuous conflict because their current sensorimotor and mental schemas do not fit with the demands of objects and events. Children try to resolve the conflicts with assimilation and accommodation, which ultimately leads to a reduction of egocentrism. Such interpersonal conflicts create intrapersonal cognitive conflict that results in not only cognitive development but also the ability to co-operate with others. Ethological theorists have pointed out that conflict is the social context in which social structures such as dominance hierarchies emerge in groups and general adaptation to the environment takes place (Weigel, 1984). Yet for all of this theoretical emphasis, little research has been conducted on how young children engage in conflict and whether they can learn how to manage conflicts constructively. The overall purposes of this study are to determine how 5- and 6-year-old children attending kindergarten engage in conflict and whether they can be trained to manage conflicts constructively. The current evidence indicates that young children tend to engage in conflicts frequently (5 to 8 per hr; Shantz, 1987). The conflicts tend to be of relatively short duration (an average of 31 s) and the majority tend to be settled by the children themselves with quite clear win/lose outcomes. Selman (1980, 1981) concluded that young children either engage in (1) impulsive physical behavior to get what they want or to avoid harm or (2) unilateral actions based on control or appeasement of the other person (stopping interaction or physical aggressiveness). He posits that reciprocation based on trades and exchanges takes place when children are beyond the age of 8 and that children may recognize that mutually satisfying agreements are possible when they are beyond the age of 14. The first purpose of this study is to examine the extent to which kindergarten children understand what a conflict is and the procedures they use to resolve their conflicts. To resolve conflicts constructively, individuals must be able to take into account both their own interests and the interests of others involved. Engaging in integrative negotiations, for example, requires that a person focus on maximizing joint outcomes and finding a solution that satisfies the interests of all involved parties. There is a controversy over whether kindergarten children (ages 5 and 6) have the cognitive capacity to do so. Developmental psychologists such as Selman (1980, 1981) predict that kindergarten children are not developmentally able to take the perspective of others in a way that results in searching for mutually satisfying agreements and therefore will not be able to learn and use an integrative negotiation 2000 by the Society for Research in Child Development, Inc. All rights reserved. 0009-3920/2000/7103-0021

Stevahn et al. 773 procedure. Engaging in integrative negotiations, furthermore, requires making a comparison of benefits resulting from any proposed agreement. Nicholls (1978, 1984), Boggiano and Ruble (1986), and Ruble, Boggiano, Feldman, and Lochi (1980) have all provided evidence that children less than 7 or 8 years old are unable to make comparisons with others. The inability of children to compare themselves and their outcomes with others may interfere with their understanding of relative standing that characterizes resolving conflicts of interests. Almost all of this research, however, is observational research that has focused on the frequency and nature of conflicts that naturally occur between young children primarily in preschool and home settings and especially in friendship relations (Shantz, 1987). This research, therefore, does not address the issue of training. Alternatively, social psychologists predict that if the training is conducted appropriately, kindergarten children will be able to learn integrative negotiation procedures to resolve their conflicts constructively. Johnson, Johnson, and Dudley (1992) and Johnson, Johnson, Dudley, and Acikgoz (1994), for example, included first- through sixth-grade students (7- through 13-year-old children) in their conflict resolution training studies and concluded that the younger children learned the integrative negotiation procedure as well as the sixth graders. Almost nothing is known, however, about the effects of conflict resolution training on 4-, 5-, or 6-year-olds in classroom settings (such as preschool, structured daycare, or kindergarten). The second purpose of this study is to determine whether kindergarten children can learn a procedure for resolving conflicts by seeking agreements that maximize mutual gain. Learning how to use an integrative negotiation procedure to resolve conflicts constructively involves more than simply being able to memorize the steps of a procedure. A Guttman-type scale designating levels of mastery of a conflict resolution procedure may be constructed: 0 Did not learn the procedure 1 Can state the steps of the procedure as a way they would in general resolve a conflict 2 Can retain their knowledge of the procedure over time 3 Can role play accurately by using the steps to resolve a conflict with a peer 4 Can appropriately apply the steps by explaining how to resolve a specific conflict with another person 5 Can use the procedure spontaneously with peers to resolve actual conflicts 6 Can transfer the use of the procedure to helping peers use it appropriately to resolve their conflicts The third purpose of this study is to determine at what level the kindergarten children learn procedures for resolving conflicts constructively. The higher the level of children s mastery of the conflict resolution procedures, the stronger the evidence that 5- and 6-yearold children attending kindergarten can in fact learn how to resolve conflicts in an integrative way. Previous research on conflict resolution programs has relied almost entirely on questionnaires and interviews (Johnson & Johnson, 1996). Few experimental studies have used behavioral measures. In the Johnson et al. (1992) study, students were placed in a simulated conflict situation 4 to 5 months after the training had ended and were videotaped. Johnson et al. (1994) also videotaped elementary school children resolving a conflict with a peer in a simulation before and after the conflict resolution training. In addition, 4 months after the training had ended, students who received the training were observed all day for 10 days to document the frequency with which they used the conflict resolution procedure in naturally occurring conflicts. Other than these two studies, behavioral measures of children s use of conflict resolution procedures have not been used in the research. The fourth purpose of this study is to provide behavioral measures of children s mastery of the conflict resolution procedures taught. Although there is a great deal of discussion about the need for conflict resolution programs, they rarely are implemented and rarely get institutionalized into school life. Institutionalization may depend on two issues. One is whether conflict training is an add-on program separate and apart from required curricula or is integrated into existing curricula. Competing priorities make adding yet another program to what many perceive as an already overcrowded curriculum challenging (if not impossible). The more conflict resolution training is integrated into the existing curricula, the more likely it is to be institutionalized. Another issue is impact of conflict resolution training on academic learning after it has been integrated into existing curricula. Innovations in schools have tended to disappear unless they have shown positive effects on student achievement (Fullan with Stiegelbauer, 1991; Johnson, 1970). There is evidence that the curriculum-integrated approach to teaching conflict resolution is effective and will increase academic achievement in English literature and social studies in middle and high school settings (Stevahn, Johnson, Johnson, Green, & Laginski, 1997; Stevahn, Johnson,

774 Child Development Johnson, Laginski, & O Coin, 1996; Stevahn, Johnson, Johnson, & Real, 1996; Stevahn, Johnson, Johnson, & Schultz, 1998). There are no empirical data on impact of the curriculum-integrated approach to conflict resolution training on learning in elementary school settings. This study provides some data by integrating conflict resolution training into a kindergarten thematic interdisciplinary unit of instruction on friendship. Previous research on conflict resolution training programs has by and large lacked methodological rigor (Johnson & Johnson, 1996). Most studies lacked control groups, did not assign students randomly to conditions, and relied predominately on anecdotal and self-report survey measures (e.g., Burrell & Vogl, 1990; Metis Associates, 1990; Opotow, 1991; Peterson & Peterson, 1990). Most studies, furthermore, were conducted in isolation from systematic, ongoing programs of research and therefore were not replicated or extended (e.g., Araki, 1990; Bradley, 1989). These problems both decrease confidence in conclusions and limit generalizability of results. In this study a training program directly based on the theory of integrative negotiations was examined (Johnson & Johnson, 1995), a control condition was included, participants were randomly assigned to conditions, teachers were rotated across conditions, and a range of selfreport and behavioral premeasures and postmeasures were used. METHOD Participants. Participants were 80 kindergartners in four classes in a suburban, kindergartan through fifth grade public elementary school in Edina, Minnesota. These children constituted the entire kindergarten population in the school. Thirty-nine of the children were in two morning classes (20 and 19, respectively) and 41 were in two afternoon classes (21 and 20, respectively). The children who attended kindergarten in the morning were randomly assigned across classes to an experimental (12 boys and 8 girls) or control (13 boys and 6 girls) condition to study an instructional unit on friendship. Those who attended kindergarten in the afternoon also were randomly assigned across classes to an experimental (8 boys and 11 girls) or control (15 boys and 7 girls) condition. All children were from middle-class backgrounds and were heterogeneous in academic achievement. The randomization resulted in both gifted and special needs children being assigned to each condition. The total population of the kindergarten through fifth grade elementary school was 476 students. Although some of the kindergartners had older brothers or sisters enrolled in the school, most of the kindergartners had limited interaction with them and other older students during the day because of the daily kindergarten schedule and the location of the kindergarten rooms in the school. Most kindergartners arrived directly in their classroom from home and left directly from their classroom to go home. Most did not eat lunch at school or participate in playground activities with other students. The two kindergarten rooms in the school also formed a self-contained wing at one end of the building. The wing had its own entrance, which was used predominantly by the kindergartners and their parents. Independent variable. This study used a pre-post, experimental-control group design. The independent variable was the presence versus the absence of conflict resolution training integrated into a kindergarten curriculum unit. In the experimental condition, the children were taught a conflict resolution procedure as an integrated component of a kindergarten curriculum unit on friendship. The curriculum unit on friendship lasted 4 consecutive weeks and contained 18 lessons (one 30-min lesson per day) for a total of 9 hr of instruction. The conflict resolution training integrated daily into the unit in the experimental condition consisted of teaching children (1) what is and is not a conflict and (2) a six-step integrative negotiation procedure to resolve conflicts constructively (Johnson & Johnson, 1995). The six-step negotiation procedure was (1) recognizing that a conflict exists and expressing a desire to resolve it constructively ( Stop. We have a conflict. Let s work it out. ), (2) stating what you want and giving your underlying reasons ( I want... because... ), (3) expressing how you feel ( I feel mad or sad. ), (4) communicating your understanding of what the other person wants and why ( You want... because... ), (5) inventing three or more optional solutions that maximize mutual gain ( Some ideas are... ), and (6) reaching an agreement by selecting and shaking hands on one of the options ( We agree on... ). A visual icon was created to correspond to each step and was posted during instruction (see Figure 1). The conflict resolution procedure was presented as an inherent part of building and maintaining a friendship. Children in the experimental condition engaged in a range of activities designed to help them learn and practice the integrative negotiation procedure as part of the thematic instructional unit on friendship. The children recited the negotiation steps, correctly sequenced sets of cards that contained each step, colored the visual icons that represented each step, watched the teacher demonstrate how to use the steps, and practiced using the procedure themselves to resolve

Stevahn et al. 775 Figure 1. Integrative negotiation procedure. conflicts embedded in the friendship unit. Picture books on friendship, for example, contained conflicts between characters. Children in pairs used the negotiation steps (with guidance from the teacher) to role play resolutions to those conflicts. Other unit activities were aimed at teaching friendship concepts such as sharing, caring, helping, listening, playing and working together, taking turns, and being fair. Those activities consisted of tasks to be completed in pairs and the materials were arranged so that children would be able to practice the negotiation procedure. In one activity, for example, children individually copied a rhyme about friendship on a valentine and then were paired to color their valentines. Each pair was given one box of crayons containing only one red crayon. The children used the negotiation steps (with guidance from the teacher) to resolve their conflict over both wanting to use the red crayon simultaneously. Children used the negotiation steps to practice resolving both storybook and task-activity conflicts a total of 15 times. In the control condition, children were taught the identical curriculum unit on friendship for the same number of sessions and hours. They used the same instructional materials and activities to learn the same friendship concepts but no conflict resolution procedures were included. There was an experimental and control group in both the morning and afternoon time blocks. Dependent measures. Before the study began, the participating children were interviewed to determine their knowledge and understanding of the concept conflict. The What Conflict Means To Me interview consisted of asking each child, Have you ever heard of the word conflict? If the child responded yes, he or she was asked, What does it mean? How would you explain what conflict is to someone who does not know? Each oral response was written verbatim and scored for knowledge of conflict (1 point for knowing the word conflict, 1 point for accurately defining it, 0 points for no knowledge). After the interview, all children were given the definition of conflict and a standard concrete example. The study was then conducted and seven dependent variables were measured and scored independently by two coders. Interrater agreement was 91% or higher on the dependent scores. The first dependent variable was the degree to which children learned the negotiation procedure. It was measured by using the How I Manage Conflict interview administered immediately before the study began and at the end of the study. Each child was asked: When you have a conflict with someone, what do you do step-by-step to solve it? Each oral response was written verbatim and scored for the presence of the six steps of the negotiation procedure (1 point per step). The second dependent variable was retention of the negotiation procedure. The How I Manage Conflict interview was readministered in May, 10 weeks after the study ended. Oral responses were written verbatim and scored identically to previous coding. The third dependent variable was willingness and ability to apply the negotiation procedure to conflict situations like those that occur in school. The Negotiation Conflict Scenario interview was administered immediately before the study began and after the study ended. This measure involved reading to each child a brief scenario about the child wanting to use a classroom computer but being blocked from doing so by a classmate. Each child was asked to explain what he or she would do to solve the conflict. Oral responses were written verbatim and categorized in three ways. The first categorization was the Conflict Management Scale, in which behaviors for dealing with conflict are arranged on a 12-point continuum from most destructive (physical aggression) to most constructive (full integrative negotiation). The second categorization was the Conflict Strategies Theory Scale derived

776 Child Development from dual-concerns theory (Johnson & F. Johnson, 1997; Johnson & R. Johnson, 1995). Participant responses were classified on a 5-point continuum (forcing, withdrawing, smoothing, compromising, negotiating for mutual gain). In the third categorization, each response was scored for the presence of the six steps of the integrative negotiation procedure (1 point per step). In all three categorizations, 0 points were assigned to I don t know responses. The fourth dependent variable was a behavioral measure of actual use of the negotiation procedure in a simulated conflict with a classmate. The Conflict Simulation Measure was administered immediately following the friendship unit. All participants were randomly paired and each pair was given a different conflict (such as both wanting the only copy of a popular picture book, both wanting to use the only calculator, or both wanting the only colorful bookmark available in a stack of bookmarks). Each pair was asked to solve their conflict and was videotaped doing so. A pair score was calculated by counting the number of integrative negotiation steps the two participants used to resolve the conflict (scores ranged from 0 for no steps used by either child to 12 for all six steps used by both children). The fifth dependent variable was application of the integrative negotiation procedure to resolve actual conflicts that occurred between participants in the classroom. The kindergarten teachers used the Negotiation Frequency Tally to count the number of times both trained and untrained children attempted to use integrative negotiation to resolve conflicts that occurred during a 30-min free-play period scheduled each day. The measure was used daily during the 4 weeks during which the study was conducted. The sixth dependent variable was willingness and ability to help others use integrative negotiation. It was measured by using the Mediation Conflict Scenario interview administered immediately before and after the study. This measure involved reading to each child a brief scenario about watching two students who were lining up for the bus arguing over their place in line (one student thought the other cut in front). Each child was asked to explain what he or she would do to help the two students solve their conflict. Oral responses were written verbatim and categorized in two ways. The first categorization was the Mediation Constructiveness Scale (adapted from the Conflict Management Scale) in which behaviors for helping others manage conflict are arranged on a 14- point continuum from most destructive (encouraging physical aggression between disputants) to most constructive (assisting disputants to use the integrative negotiation procedure). The second categorization was the Mediation Conflict Strategies Theory Scale (adapted from the Conflict Strategies Theory Scale) in which responses are classified on a 6-point continuum. In the third categorization, each response was scored for the presence of the steps in the mediation process described in Teaching Students To Be Peacemakers (Johnson & Johnson, 1995). Those steps are (1) stop hostility between disputants, (2) ask disputants if they would like a mediator, (3) gain commitment from disputants for mediation, (4) facilitate integrative negotiation between disputants, and (5) formalize the disputants negotiated agreement. One point was assigned for each step present. In all three categorizations, 0 points were assigned to I don t know responses. The seventh dependent variable was mastery of the friendship concepts taught (children s conceptual understanding of friendship). It was measured by using the Friendship Interview administered individually at the end of the study. Each child was asked: What does it mean to be a friend? What do friends do? Oral responses were written verbatim. Each response was scored for the total number of friendship concepts listed (1 point for each separate and different concept). Procedure. The study was conducted in February of 1997 in a midwestern, suburban, kindergarten through fifth grade public elementary school. The study comprised four phases: pretesting, training, posttesting, and administering retention measures. In January of 1997 permission was obtained to conduct the study and a 4-week kindergarten curriculum unit on friendship was planned to contain the conflict training. The unit thematically was designed to coincide with Valentine s Day. Before the study began, all of the kindergartners individually were administered four premeasures: What Conflict Means To Me, How I Manage Conflict, the Negotiation Conflict Scenario, and the Mediation Conflict Scenario. The children randomly were assigned to experimental or control conditions (a morning and afternoon group for each condition, respectively). Training in the experimental condition consisted of teaching kindergartners the six-step negotiation procedure integrated into 18 lessons on friendship (1 per day for 4 weeks). Kindergartners in the control condition were taught the identical friendship lessons for the same period of time by using the same materials and activities but without the negotiation procedure. The morning and afternoon groups that constituted the experimental condition received identical instruction each day for consistency in implementing that treatment. The morning and afternoon groups that constituted the control condition also received the same

Stevahn et al. 777 daily instruction for consistency in implementing the control treatment. Experimental and control groups met simultaneously in the morning or afternoon, respectively, in separate classrooms for approximately 30 min each day for a total of 9 hr of instruction. Instruction was conducted by the two kindergarten teachers in the school who rotated across experimental and control conditions in both morning and afternoon groups every 2 days. The rotation resulted in each teacher conducting an equal number of lessons in all experimental and control groups. A staff development consultant who assisted in the morning rotated across both teachers equally, which resulted in co-teaching an equal number of lessons in the morning experimental and control groups. The teachers and groups also rotated across the two kindergarten rooms (which were adjacent to each other) so that both teachers taught in the two rooms an equal number of times and experimental and control groups received instruction in both rooms an equal number of times. In other words, all of the kindergartners were in their own kindergarten room and the other kindergarten room an equal number of times for instruction. Similarly, each teacher taught in her own kindergarten room and in the other kindergarten room an equal number of times. This room rotation was established because the randomization of kindergartners to experimental or control conditions across classrooms resulted in children from both rooms in each group. All of the groups that resulted from the randomization (both experimental and control) were called friendship groups. All of the children were told that these cross-class groups would help them make new friends with other kindergartners in the school. During the 4 weeks during which the study was conducted, the two kindergarten teachers used the Negotiation Frequency Tally during a daily free-play period to count attempts by kindergartners to use the integrative negotiation procedure to resolve actual conflicts. Teachers noted whether those attempting to use the procedure were trained or untrained. The free-play period lasted approximately 30 min each day. Children in experimental and control groups were mixed during the free-play period. The teachers, therefore, did not use the six-step integrative negotiation procedure to assist students in resolving their conflicts during that time (or any other time during the day when trained and untrained children were mixed) to ensure that treatments would not be contaminated. Instead, the teachers assisted students by using classroom management methods used before the study. At the end of the study, all kindergartners individually were administered the postmeasures. After all of the postmeasures were given, the conditions were combined and all the kindergartners were taught one lesson on the six-step integrative negotiation procedure. Ten weeks after the study ended, the How I Manage Conflict measure was readministered individually to all kindergartners to assess retention. Analysis. A two-way analysis of variance was conducted on experimental versus control conditions and morning versus afternoon groups. Because there were no significant differences at the.01 level between the morning and afternoon groups on the dependent variables, the two groups were combined. Differences between the experimental (trained) and control (untrained) conditions were then analyzed by using unpaired t tests. Totals were calculated on the Negotiation Frequency Tally. RESULTS An analysis was conducted to determine whether there were significant differences between the morning and the afternoon classes. There were no significant differences between the experimental morning and afternoon groups or the control morning and afternoon groups on any of the dependent variables. They were, therefore, combined into one experimental and one control condition. Before the study began, the children were interviewed to determine their understanding of the concept conflict. No significant difference was found between conditions. Only two children in the experimental condition and three in the control condition had heard of the word conflict. None of those children accurately defined the term. Table 1 shows that before training, there was no significant difference between conditions; none of the children included any of the negotiation steps in their descriptions of how they resolved conflict. After Table 1 Mean Scores on Dependent Measures Measure Trained Untrained t test Probability Negotiation steps pretest 0 0 Negotiation steps posttest 4.53 (1.83).05 (.22) 15.58 (77).001 Negotiation steps retention 4.24 (1.89) 2.22 (2.30) 4.22 (76).001 Conflict simulation postmeasure 9.69 (1.19) 1.33 (.70) 23.54 (29).001 Friendship concept recall posttest 5.47 (2.16) 3.66 (1.80) 4.07 (77).001 Note: Standard deviations are in parentheses.

778 Child Development Table 2 Mean Scores on the Negotiation Conflict Scenario Interview Measure (Computer Conflict) Pretest Posttest Classification M SD n M SD n Conflict management scale Trained 5.61 1.63 39 7.95 2.71 38 Untrained 5.27 1.75 40 5.63 1.88 41 Conflict strategies theory scale Trained 1.15.49 39 2.82 1.83 38 Untrained 1.55.93 40 1.34.73 41 Integrative negotiation steps Trained 0 0 39 1.63 2.40 38 Untrained 0 0 40 0 0 41 Note: Pretest, trained versus untrained: conflict management scale, t(77).892, p.375; conflict strategies theory scale, t(77) 2.356, p.021; integrative negotiation steps, the mean score for both trained and untrained was 0 and therefore we did not conduct a t test on the comparison. Posttest, trained versus untrained: conflict management scale, t(77) 4.43, p.001; conflict strategies theory scale, t(77) 4.77, p.001; integrative negotiation steps, the mean score for untrained was 0 and therefore we did not conduct a t test on the comparison. training, trained children included significantly more negotiation steps in their descriptions than untrained children, t(77) 15.58, p.001. Ten weeks after the study ended, trained children still included significantly more negotiation steps in their descriptions of how they resolved conflict compared with untrained children, t(76) 4.22, p.001. To measure children s willingness and ability to apply the negotiation procedure to conflicts representative of those that occur in school, children were interviewed as to how they would resolve a conflict over the use of a computer. Children s responses were analyzed in three ways (see Table 2). For the Conflict Management Scale analysis, there was no significant difference between trained and untrained children before training. After training, however, the trained children used significantly more constructive strategies to resolve the conflict than the untrained children, t(77) 4.43, p.001. For the Conflict Strategies Theory Scale analysis, there was no significant difference between trained and untrained children on the pretest. On the posttest, trained children used significantly more constructive strategies than untrained children, t(77) 4.77, p.001. The computer conflict scenario responses also were analyzed to determine the frequency with which children used each step of the negotiation procedure. On the pretest, none of the children used any of the steps. On the posttest, trained children used more steps than untrained children. We did not conduct a t test on this comparison because the mean score for the untrained students was 0. Thirty-seven percent of the trained children, however, used one or more steps of the negotiation procedure (see Table 3). Two behavioral measures were taken of children s actual use of the negotiation procedure. Children participated in a simulation immediately following the study. They randomly were assigned to pairs and given a conflict to resolve. Table 1 shows that trained pairs used significantly more of the integrative negotiation steps to resolve their conflicts than untrained pairs, t(29) 23.54, p.001. Teachers also systematically observed children during daily 30-min free-play periods to measure how often they would spontaneously use the negotiation procedure in conflicts with each other. Table 4 shows that a total of 16 attempts to use negotiation occurred over a 4-week time period. All of the attempts were made by trained children. None of the untrained children attempted to use negotiation to resolve conflicts that occurred during the free-play time. The children also were interviewed about their willingness and ability to help others use integrative negotiation (see Table 5 and Table 6). The children were asked to describe what they would do to help two classmates resolve a conflict about cutting in line. Before training began, there was no significant difference between conditions for the Mediation Constructiveness Scale. Following the training, trained children used significantly more constructive interventions to help others resolve conflict than untrained children, t(77) 3.70, p.001. For the Mediation Conflict Strategies Theory Scale analysis, there was no significant difference between trained and untrained children on the pretest. On the posttest, trained children used significantly more constructive intervention strategies than untrained children, t(77) 2.92, p.0046. The cutting in line conflict scenario responses also were analyzed to determine the frequency with which children used mediation steps. On the pretest, there was no significant difference between trained and untrained children in the number of mediation steps used. On the posttest, trained children used significantly more steps than untrained children, t(77) 3.67 p.001. Finally, the children s mastery of the concepts taught in the friendship unit was measured. Children were interviewed as to their conceptual understanding of friendship (see Table 1). At the end of the study, trained children compared to untrained recalled significantly more of the friendship concepts taught than did the untrained children, t(77) 4.07, p.001. Notably, 19 of the 38 children in the trained condition

Stevahn et al. 779 Table 3 Responses on the Negotiation Conflict Scenario Interview Measure (Computer Conflict) Pretest Posttest Classification Trained n 39 Untrained n 40 Trained n 38 Untrained n 41 Conflict management scale Blank (don t know) 0 2 (5%) 0 3 (7%) Physical aggression 0 0 0 0 Verbal threat 0 1 (2.5%) 0 0 Unsatisfactory withdrawal 0 0 0 0 Tell the teacher 13 (33%) 10 (25%) 5 (13%) 5 (12%) Command/request 1 (2%) 3 (7.5%) 1 (3%) 3 (7%) Satisfactory withdrawal 6 (15%) 14 (35%) 7 (18%) 14 (35%) Invoke norms 19 (50%) 10 (25%) 9 (24%) 16 (39%) Propose alternatives 0 0 2 (5%) 0 Negotiation: intent expressed 0 0 0 0 Negotiation: 1 or 2 steps 0 0 3 (8%) 0 Negotiation: 3 to 5 steps 0 0 6 (16%) 0 Negotiation: 6 steps 0 0 5 (13%) 0 Conflict strategies theory scale Blank (don t know) 0 2 (5%) 0 3 (7%) Forcing33 (85%) 24 (60%) 15 (40%) 24 (59%) Withdrawing0 4 (10%) 3 (8%) 6 (15%) Smoothing6 (15%) 10 (25%) 4 (10%) 8 (19%) Compromising0 0 2 (5%) 0 Negotiating 0 0 14 (37%) 0 Integrative negotiation steps 0 steps 39 (100%) 40 (100%) 24 (63%) 41 (100%) 1 step 0 0 1 (3%) 0 2 steps 0 0 2 (5%) 0 3 steps 0 0 1 (3%) 0 4 steps 0 0 1 (3%) 0 5 steps 0 0 4 (10%) 0 6 steps 0 0 5 (13%) 0 Note: Response percentages are in parentheses. stated that an important aspect of friendship was solving conflicts. DISCUSSION There are numerous reasons why young children should be taught procedures for managing their conflicts constructively. The first is that if they are not Table 4 Time Week Negotiation Frequency Tally during Daily Free-Play Attempts by Trained Children to Use Negotiation Attempts by Untrained Children to Use Negotiation 1 10 0 2 6 0 3 0 0 4 0 0 Total 16 0 taught constructive procedures for resolving conflicts, children develop in unhealthy ways that are destructive not only to themselves but also to people around them and society as a whole. Eron and Tolan (see Murray, 1998), for example, conducted a study of violence among young children that included nearly 5,000 students (40% African American, 40% Latino, 16% European American) and lasted over 7 years. They concluded that (1) children often learn violent behaviors as early as kindergarten; (2) without intervention, the violence tends to persist and to escalate; (3) by the time children are 8 years old, violent habits are almost impossible to break; and (4) intervention programs should begin as early as possible. Another reason why young children should be taught conflict resolution procedures is that healthy development requires it. Conflict with peers and adults is a central developmental process that leads to social, cognitive, and psychological development as long as it is managed constructively. The minimal requirements for managing conflicts constructively are to un-

780 Child Development Table 5 Mean Scores on the Mediation Conflict Scenario Interview Measure (Cutting in Line Conflict) Pretest Posttest Classification M SD n M SD n Mediation constructiveness scale Trained 6.67 2.42 39 9.63 3.09 38 Untrained 7.22 2.38 40 7.37 2.32 41 Mediation conflict strategies theory scale Trained 1.61 1.44 39 2.97 2.28 38 Untrained 1.45 1.28 40 1.71 1.52 41 Mediation steps Trained.13.34 39.63.82 38 Untrained.22.42 40.12.33 41 Note: Pretest, trained versus untrained: mediation constructiveness scale, t(77) 1.034, p.3045; mediation conflict strategies theory scale, t(77).539, p.5914; mediation steps, t(77) 1.121, p.2657. Posttest, trained versus untrained: mediation constructiveness scale, t(77) 3.70, p.001; mediation conflict strategies theory scale, t(77) 2.92, p.0046; mediation steps, t(77) 3.67, p.001. derstand what is and is not a conflict and to master a procedure that resolves conflicts in a way that leaves both parties satisfied with the agreement and improves their relationship (Deutsch, 1973; Johnson & Johnson, 1995). In this study, before the conflict resolution training began, the kindergarten children were interviewed to determine (1) their knowledge and understanding of the word conflict and (2) the procedures they used to resolve conflicts. The results indicated that only 5 of the 80 children (2 in the experimental condition and 3 in the control condition) had heard of the word conflict and none could accurately define it. On the pretests, no children included any integrative negotiation steps in their explanations of how they managed conflict or how they would manage a typical conflict involving the use of a computer. The predominant strategies used tended to be telling the teacher, withdrawing from the conflict, making threats, and invoking social norms such as fairness, sharing, and taking turns. On the role-playing postmeasure, all untrained children resolved the conflict by one child suggesting a solution and the other child immediately conceding. These results corroborate the findings of previous researchers (Selman, 1980, 1981; Shantz, 1987), who found that kindergarten-age children used forcing and withdrawal as their major strategies for managing conflicts. Is the lack of understanding and competence due to the developmental limitations of children of kindergarten age, or is it due to a lack of training in how to manage conflict constructively? A number of developmental psychologists (e.g., Nicholls, 1984; Selman, 1980, 1981) argue for the former explanation whereas a number of social psychologists (e.g., Johnson & Johnson, 1996) argue for the latter. The resolution of this controversy is somewhat complex because a procedure can be learned at different levels. Simple memorization of a negotiation procedure, for example, may have a minimal effect on healthy development. Thus, a Guttman-type scale was developed to measure the level of learning, understanding, and internalization of the integrative negotiation procedure taught. The level of learning of the procedure may indicate whether the developmental level or lack of training position is more valid. The first level of mastery was the volunteering of the procedure as the way to resolve conflict. After training, 47% of the trained children included all six negotiation steps in their explanations of how they managed conflict and an additional 21% included four or five steps. In contrast, 95% of the untrained children included zero steps in their explanations. Thus, the training was successful in teaching kindergarten children the integrative negotiation procedure in a relatively short period of time as part of an academic unit on friendship. The second level of mastery was to retain the knowledge of the procedure weeks after the training had ended. Even 10 weeks after the study ended, trained children still retained more of the steps than untrained children. Thirty-two percent of those who received training remembered all six steps and an additional 38% remembered four or five steps. In contrast, only 12% of the untrained students knew all six steps and only an additional 20% knew four or five steps. This finding is noteworthy because after the study ended (1) all of the kindergartners were taught one lesson on the steps of the integrative negotiation procedure, (2) the steps were posted on the walls in both kindergarten rooms, and (3) the kindergarten teachers used the steps in their classrooms to help children resolve actual conflicts throughout the remaining months of the school year. This high rate of retention may have resulted from the trained kindergartners using the integrative negotiation procedure to manage day-to-day conflicts with classmates and peers. The third level of mastery of the integrative negotiation procedure was to accurately role play the procedure in resolving a conflict with a classmate. The results of this study indicated that when placed in a conflict simulation situation, every pair of trained students applied the integrative negotiation procedure to resolve their conflict. Of the 16 trained pairs, 10 used all six of the negotiation steps, 4 pairs used

Stevahn et al. 781 Table 6 Responses on the Mediation Conflict Scenario Interview Measure (Cutting in Line Conflict) Pretest Posttest Classification Trained n 39 Untrained n 40 Trained n 38 Untrained n 41 Mediation constructiveness scale Blank (don t know) 2 (5%) 2 (5%) 0 2 (5%) Promotes physical/verbal aggression 0 0 0 0 No intervention: observes 0 0 0 0 No intervention: walks away 1 (3%) 0 2 (5%) 0 Labels behavior as inappropriate 0 1 (2.5%) 0 1 (2%) Tells the teacher 8 (20%) 4 (10%) 1 (3%) 2 (5%) Commands/orders disputants to stop 10 (26%) 9 (22.5%) 3 (8%) 9 (22%) Advises to forgive and forget 3 (8%) 2 (5%) 1 (3%) 3 (7%) Arbitrates: listens to both sides 3 (8%) 4 (10%) 3 (8%) 8 (20%) Advises to invoke norms 8 (20%) 15 (37.5%) 13 (34%) 11 (27%) Proposes alternative solutions 4 (10%) 3 (7.5%) 5 (13%) 5 (12%) Mediation: gets an adult mediator 0 0 0 0 Mediation: gets a peer mediator 0 0 0 0 Mediation: gives solutions 0 0 1 (3%) 0 Mediation: facilitates full negotiation 0 0 9 (23%) 0 Mediation conflict strategies theory scale Blank (don t know) 2 (5%) 2 (5%) 0 2 (5%) Facilitating forcing 29 (74%) 32 (80%) 20 (53%) 30 (74%) No intervention 1 (3%) 0 2 (5%) 0 Facilitating withdrawal 0 1 (2.5%) 0 1 (2%) Facilitating smoothing 3 (8%) 2 (5%) 1 (3%) 3 (7%) Facilitating compromise 4 (10%) 3 (7.5%) 5 (13%) 5 (12%) Facilitating negotiation 0 0 10 (26%) 0 Mediation steps 0 steps 34 (87%) 31 (77.5%) 22 (58%) 36 (88%) 1 step 5 (13%) 9 (22.5%) 8 (21%) 5 (12%) 2 steps 0 0 8 (21%) 0 3 steps 0 0 0 0 4 steps 0 0 0 0 5 steps 0 0 0 0 Note: Response percentages are in parentheses. five steps, and 1 pair used four steps. All but 1 pair generated at least two or more alternative solutions from which they mutually chose one for their final resolution. In contrast, none of the 15 untrained pairs used the negotiation procedure to deal with their conflicts. Although the children in 1 untrained pair did express their feelings and children in 3 untrained pairs did state their wants, doing so was not part of a systematic procedure for resolving the conflicts. Instead, in all 15 untrained pairs, one child suggested a solution and the other child immediately conceded. Given that the children learned, retained, and could accurately perform the integrative negotiation steps, the fourth level of mastery was whether they could actually use the procedure in a conflict situation similar to those they face in their own lives. Before training, the predominant strategies used by all of the children to resolve the computer conflict scenario were telling the teacher, withdrawing from the conflict (letting the other person have the computer without confronting the issue), or invoking social norms such as fairness, sharing, and taking turns. After training, 37% of the trained children used all or part of the integrative negotiation procedure to solve the conflict; none of the untrained children used integrative negotiation to deal with the conflict. Untrained children instead predominately continued to deal with the conflict by telling the teacher, withdrawing, or invoking social norms. The ability of trained students to use the procedure in an actual conflict increases the confidence schools can have in seeking to teach young children how to resolve their conflicts constructively. The fifth level of mastering the integrative negotiation procedure was to spontaneously use the procedure to resolve an actual conflict with a peer. The kindergarten teachers recorded any uses of the integrative negotiation procedure they observed during the daily

782 Child Development free-play period. Trained children at their own initiative attempted to use the negotiation procedure to deal with conflicts 16 times, all during the first 2 weeks of the study. The absence of attempts in weeks 3 and 4 may have resulted for many reasons, such as conflicts rarely occurring, trained children interacting primarily with untrained classmates, or trained children not transferring the procedure to the conflicts occurring during free-play time. Teachers did not assist trained students in the use of the negotiation procedure because trained and untrained students were mixed during the free-play time and assisting students would have contaminated the treatments. The sixth level of mastery of the integrative negotiation procedure was to voluntarily help two peers who are having a conflict use the procedure to resolve their conflict constructively. Before training on the cutting in line conflict scenario, the children predominately indicated that they would intervene by getting the teacher, ordering the disputants to stop, or advising the disputants to invoke social norms. After training, however, 26% of the children who received training indicated that they would mediate the conflict by using the negotiation procedure to help the disputants reach agreement on a solution. None of the untrained children indicated that they would use the negotiation procedure to mediate a resolution between the disputants. These results are noteworthy given that children were not taught a mediation procedure. The results of this study provide strong evidence that the children not only learned the integrative negotiation procedure but also tended to internalize it. The learning, retention, and ability to perform the integrative negotiation steps indicated a thorough mastery of the integrative negotiation procedure. Both the voluntary application of the procedure to conflict scenarios and the spontaneous use of the procedure during free-play time indicated some transfer. These results indicate that 5- and 6-year-old kindergarten children can in fact learn the procedure and many of them can transfer it to a variety of situations. If educators wish to train kindergarten children how to manage their conflicts constructively, it is certainly possible. On the pretests, not a single child included any negotiation steps in their explanations of how they managed conflicts. Untrained children still did not know the negotiation steps at the end of the study. These findings are important because they indicate that, without training, students may never learn negotiation procedures, even if they are in an environment where negotiation procedures are taught and used by others. The importance of training young children in how to manage conflicts constructively cannot be overemphasized. Children often engage in conflicts with peers and adults. Healthy social, cognitive, and psychological development depend on children resolving their conflicts constructively. The results of this study indicate that with specific training children can learn to do so; without specific training children may fail to do so. Although teaching young children how to manage their conflicts constructively has important implications for social, cognitive, and psychological development, the time it takes to do so among the competing priorities of teachers is problematic. Almost all conflict training programs have been add-on, stand-alone programs. Historically, most add-on school programs that have not been linked to academic achievement have not been widely adopted or maintained over time (Fullan with Stiegelbauer, 1991; Johnson, 1970, 1979). Integrating conflict training into academic coursework in ways that enhance achievement, therefore, increases the probability that conflict programs will be adopted and institutionalized into the daily life of classrooms and schools. The results of this study indicate that children who were taught conflict resolution as an integrated component of the thematic instructional unit on friendship scored higher on the achievement measure given at the end of the unit than did children who studied the unit without conflict resolution training. This finding corroborates previous studies where the Teaching Students To Be Peacemakers (Johnson & Johnson, 1995) program was integrated into English literature and social studies units at the middle and high school levels in Canada and the United States (Stevahn et al., 1997; Stevahn, Johnson, Johnson, Laginski, et al., 1996; Stevahn, Johnson, Johnson, & Real, 1996; Stevahn et al., 1998). Practicing negotiation in the context of the academic curriculum apparantly makes the curriculum more meaningful and memorable, perhaps because it involves students in higher order reasoning about the subject matter. The results of this study, therefore, increase the probability that conflict resolution will be adopted by teachers and schools who can see the advantage of simultaneously teaching children to manage conflicts constructively and increasing their academic learning. In summary, this study added important new data on the effectiveness of school-based conflict resolution programs in a number of ways. First, most studies on conflict training programs have used paperand-pencil, self-report measures (Johnson & Johnson, 1996). Behavioral measures in the research on conflict resolution training are lacking. This study, by using two behavioral measures, provides new evidence as to the effectiveness of conflict resolution training pro-