EVALUATION OF A CONFLICT RESOLUTION-PEER MEDIATION PROGRAM

This research evaluates a seven week conflict resolution peer mediation program offered at a diverse, small, private middle school ( 4th through 8th grade). Identical curriculum based measures of knowledge about conflict, self-efficacy for managing conflict, and endorsed strategies for responding to conflict were given both before and after the program. The 72 students demonstrated gains in knowledge following the intervention . Differences were not found in the other areas . An exploration of the relationship among variables revealed low correlation and shared variance. The contribution of conflict knowledge and self efficacy for managing conflict to the prediction of endorsed conflict response strategies was negligible. Student and staff perceptions regarding the amount and intensity of conflict present at the school were similar on the pretest. Posttest data indicate stafflstudent differences in perception of the amount of conflict, but not in the intensity.

The Executive Director and Headmaster of a small, private school located in the Northeast contracted with a local university clinic for the provision of conflict resolutionpeer mediat ion training at the school. A program was designed and delivered in response to that request.
Morton Deutsch (1993) stated that conflicts are an inevitable part of the human condition . It is how they are managed , or mismanaged , that determines whether they are constructive or destructive . Deutsch (I 994) conceptualizes the management of conflict as a social skill, not just a knowledge base . Like all social skills, competency is achieve through practice, which is itself subject to environmental constraints . The schoo l was interested developing and supporting this social skill.
The management of conflict has trad itionally been of interest to those working in social psychology. The Journal of Social Issues devoted an entire issue (Vol. 50, 1994) to applications of conflict management. Contributors addressed conflict at many levels : interpersonal , intergroup in school , intergroup in organizations , and international.
Milton Mapes , Jr. (1987) , Director of the National Peace Academy, argued persuasivel y for the need to create "a new science of peacemaking" (p . 138). There is a need to learn how to manage conflict, whether it be local, national , or international. To this end the Peace Academy was founded with the goals of research , education and training , and information dissemination . The American educational system allows the opportunity to educate the public, particularly the formative public, in managing conflict.
The school's request provided the impetus to respond to Mapes call for education and training . The designed program was new to both the clinic and the school and, as always , it is important to evaluate the effectiveness of new programs . Such evaluations aid program development and help impro ve upon program design (Posavac & Carey , 1992;Maher & Kruger , 1988). This research evaluates the effectiveness of the conflict resolution -peer mediation program developed for this school.
Effectiveness is but one of the dimensions suggested by Elliott , Witt, and Kratochwill ( 1991) for evaluation of classroom interventions . Other areas for consideration are acceptability , teacher/parent empowerment , and treatment integrity . Elliott et al. assumed delivery of services by classroom teachers , but their evaluation approach provides a suitable rubric for discussion of this program.

Conflict Resolution .
As stated earlier, conflict resolution is an area of interest for many groups . The potential for conflict is present whenever people interact . Fisher, Ury, and Patton (1991) made the national bestseller list with their book entitled "Getting to Yes : Negotiating Agreement Without Giving In". This manual for handling professional and personal disputes struck a cord with many people .
Conflict resolution training has tremendous potential as a preventative technique . Horowitz and Broadman (1994) recommend that schools include training of conflict management as part of the ongoing core curriculum . They feel that many youth are not receiving sufficient training through the traditional framework of family and community interactions . Training in communication and negotiation can help both children and parents in highly stressful environments . 2 School-wide conflict resolution programs have been tried in many areas of this country . The programs are designed to fit into the existing curriculum . Three alternative high schools in New York City participated in a study (Zhang , 1994) comparing the effects of cooperative learning, constructive conflict resolution , and a combination of both on measures of self esteem, social support , victimization, perceived school climate, and academic achievement . The constructive conflict resolution program had a direct positive impact on student's social support and directly and indirectly worked to decrease victimization .

Peer Mediation.
Peer mediation is one technique for the management of conflict situations . When disputing parties are unable to work out an issue they may decide to attain help with the process . Mediators aid the disputants by providing a supportive environment for smoother communication . 3 Potential school mediators either volunteer or are nominated by their teachers and/or peers to receive training on the management of conflict. Mediators are taught a structured process for the mediation sessions . They set ground rules for behavior in the session (i.e. no interruptions of the speaker, no name calling or put downs , agreement to pursue a solution, and honest answers) . If disputants are unwilling, or unable, to abide by these rules they are referred elsewhere .
During the session mediators listen carefully, frequently reflecting and restating what is presented . After the problem is explored, wants are identified and participants are encouraged to suggest solutions . It is important to note that mediators function only to aid disputants in coming to a resolution -they do not choose sides, render opinions or pronounce verdicts (Strumpf , Crawford , & Usadel, 1991).
The use of peer mediators has been strongly supported in the literature (Lane & McWhirter, 1992;Burrell & Vogl, 1990;Crary, 1992;Araki, 1990). Koch (1988) reports that upon completion of the first year of a mediated dispute resolution (MDR) program at one high school in New York City, a 50% drop in suspensions for fighting was found . He also states that a multi-ethnic high school in Hawaii experienced a reduction in fighting incidents, from 83 to 19, in its first two program years .
At the William E . Ferron Elementary School in Las Vegas , Nevada peer mediators wear orange helmets at recess to identify themselves (Cahoon, 1988) . The job of mediation is assigned by-weekly to different students in the classroom . Cahoon reports that there has been a decline in discipline referrals to the principal's office and an improvement in student problem-solving and satisfaction since the implementation of peer mediation .
In 1987 a large (806), culturally diverse (66% minority) urban middle school in the Santa Monica area began its Peer Mediation Program. A evaluation done after the second year found that 95 disputes were referred to mediators, 97% of those cases were reported as resolved . Two weeks to a month after the session most disputants indicated that the agreements reached during mediation were maintained . Students and faculty were satisfied with the program (Crary, 1992) .
The introduction of peer mediation into a school benefits not only the school, but also those students trained as peer mediators . Crary (1988) found the student mediators to be more "happy" with self and less likely to blame others on a post training measure . Improvements in academic subjects (Araki , 1990) and increases in prosocial behavior (Cahoon , 1988) on the part of mediators have been noted . Gentry and Benenson ( 1993) examined the transfer of learning to home settings .
Twenty-seven middle school students were trained and worked as peer mediators in their school. Home follow-up on these students indicated that their parents perceived a decrease in the frequency of sibling conflict along with improved communication skills (during conflict) on the part of the student. The trained students felt that they were experiencing less conflict with their siblings and those conflicts that did occur were less intense. Participants noted that siblings that were not given any school-based training in conflict management presented greater home challenges for the mediators.
Program Design .
The program is based upon the work of Deutsch (1993 ;, Dodge (1986) , and Crick and Dodge (1994). Dodge (1986) presented a five step model of social information processing . The steps are : a) encoding, b) representation, c) response search, d) response decision, and e) enactment. Each process is further refined to components . For example, the encoding process consists of sensation, perception, attention and focus . Subsequent work by Crick and Dodge (1994) reformulated this model to six, nonlinear stages that take into consideration reciprocal effects . In brief, the model proceeds as follows : a) cues are encoded , b) an interpretation is made of those cues, c) goals for that situation are clarified, d) potential responses are generated from memory or new responses are constructed in the case of novel situations , e) the response decision is made after an evaluation that is likely to include a consideration of outcomes and selfefficacy, and f) the response is made in form behavior, which is in turn a cue to be encoded by others .
Children involved in conflict may be experiencing difficulty at any point(s) in the process . It is also likely that various children will have difficulty at different points in processing social information . To accommodate these differences the program was designed to address the multiple stages of social information processing . For instance, the early units on verbal and nonverbal communication were designed to focus attention on how cues are encoded and interpreted . These units were based on exercises suggested by Shaw (1992) and Chiak and Heron (1980) . (See Appendix A for the program timeline and specific lesson plans.) Personal goal clarification is the next stage in the Crick and Dodge model. In order to assess their own goals students need a sense of why they are involved in a dispute . The program attempted to address this concern with a discussion of the sources of conflict. Material for this unit was drawn form the Schrumpf, Crawford , and Usadel (1991, pp .6-8) program . It was hoped that with increased awareness of conflict sources students would also be sensitive to the goals of all involved parties.
The generation of potential responses in the next stage . The first discussion of this area was done in conjunction with the discussion of conflict sources (week 4) . Schrumpf et al. introduced broad categories of avoidance , confrontation , and communication as response classes. Students were asked to consider their own conflict response style.
The following week dealt with the issue of perspective . It was felt that sensitivity to others perspective may influence the nature of the generated conflict responses . Student listened to a different version of a familiar fairy tale, The True Story of the Three Little Pigs by A Wolf (Scieszka, 1989). They were invited to consider the wolfs perspective of the case and how that was at odds with the pigs experience . This discussion preceded a small group exercise in which students presented both sides of a personal conflict .

6
In the sixth week began with a presentation of how communication can escalate or de-escalate conflict situations (Kriedler , 1984). Students were given a handout with rules for fair fighting (Schmidt & Friedman, 1985, p. 8) and asked to rewrite a brief conflict scene to reflect those rules .
The exercise was followed by a return to the discussion of potential conflict responses . At this point specific strategies were introduced to the students . These strategies were drawn from the work of Shaw (1992 , Lesson 33 :3) and Schmidt and Friedman ( 1985, p. 9). Students were asked to consider the appropriateness of strategies to situations . They were also asked to project potential outcomes . Crick and Dodge (1994) suggested that the response decision is based , in part , on a consideration of outcomes and self-efficacy. The program sought to make the link between strategies and outcome overt .
The final stage of the model is behavior. The final week of the program focused on response in a conflict situation . All the units were tied together in a small group exercise. The groups were given different conflict scenarios to work through . They were asked to identify the conflict, determine it's likely source , and discuss the perspective of the parties involved . The next step was to generate possible ways of managing the conflict and discuss the potential consequences of each. Groups then had to decide which strategy they would endorse , or behavior they would enact.

Context of the Study
The contracting school was a privately funded middle school for grades 4 through 8. The mission of the school was to prepare low-income and minority students demonstrating average or above-average ability for college preparatory high school programs . In keeping with this agenda , 92% of the students received some scholarship aid from the school.
Students at this school came from many different neighborhoods , both within the city and from the surrounding communities , and brought with them a diversity of experiences . A number of academic classes included hearing-impaired students from another school . Students attending the private school were separated from old friends and habits. They had to adapt to a different and, in many cases, a more vigorous academic routine . In order to support the students in these transitions, the school attempted to create a close-knit , caring community .
Teachers interacted freely with the students , eating at a common lunch time and remaining after school to provide additional help. Weekly community meetings for the entire school were held to share information and talents , to acknowledge accomplishments, and to celebrate occasions . Older students were encouraged to help younger students and it is not unusual for teachers and students to discuss homework after school hours by telephone .
This school works hard to foster friendship and affiliation among its diverse student body . To that end they felt it desirable for students to receive some formal training on managing conflict. Of immediate concern to staff was the "tension" among .
certain of the students attending the school. For example, a new, upper level student was experiencing difficulty in gaining acceptance by peers . Two lower grade students were involved in a power struggle over friends and free time activities . Another student refused to cooperate during group projects (the school utilized many cooperative learning activities) . For these reasons the school requested that the conflict resolution -peer mediation program be designed and implemented during the '93-'94 school year. 8 The school requested that all students receive training in conflict resolution as part of their health curriculum. They also requested that two students from each class/grade level be trained as peer mediators . This research addresses the evaluation of a schoolwide conflict resolution -peer mediation program that was delivered to this small, private school in the spring of 1994.
The purpose of this conflict resolution -mediation program was to train students in techniques for managing interpersonal conflict. It was predicted that students would : l) increase their knowledge about conflict resolution, 2) increase the variety and effectiveness of the conflict resolution strategies endorsed, and 3) show improved student self-efficacy for conflict resolution .
The variable of greatest interest to the school was the daily conflict resolution strategies used by students. The relationships of self-efficacy, knowledge , and endorsed strategies will be explored . Of particular interest is the contributions of knowledge and self efficacy to strategies employed .

Participants
The participants were the 81 students and 10 staff members of an urban middle school. Approximately 46% of the student body were African American, 25% Caucasian, 11% were Hispanic, 14% Southeast Asian, 2% Native American, and 2% of mixed heritage .
The school had students in fourth through eighth grade . There was one class on each of the grade levels. The gender and number of students by grade level was as follows : fourth grade -16 students (9 male and 7 female), fifth grade -16 students (9 male and 7 female), sixth grade -18 students (8 male and 10 female), seventh grade -15 students (10 male and 5 female), and eighth grade -16 students (7 male and 9 female). In total, there were 43 male and 38 female students in the school.
The exact ages of the students was not available. However, it should be noted that this school recruits students of average or above average ability and does not retain students . In spite of extensive help given to students who struggle with the curriculum demands, some students do choose to return to their local school(s) . For this reason it is likely that the majority of students were close to the standard age for each grade . The approximate age for fourth grade is 9 yr., fifth grader is 10 yr., sixth grader is 11 yr., seventh grader is 12 yr., and eighth grader 12 yr.
The staff consisted of 8 full time teachers, 1 teaching intern, and 2 administrative people who did some teaching. Three of these people were male and the remaining 8 were female. The majority of the staff is Caucasian. Two members are African American.
Data were collected on those students and staff present at school on the days of pre-and post-testing . This resulted in 72 complete sets of student questionnaires . There were 5 pretest and 11 post test staff responses .

Measures
Pre-and post-training questionnaire The conflict resolution portion of the program was assessed by means of a pre-and post-training questionnaire (see Appendixes Band C) based on the curriculum . The questionnaire consisted of the Conflict Resolution Scale (Schmidt & Friedman, 1990) that measures knowledge about conflict and its resolution . This 21 question true/false measure was designed in conjunction with the Creative Conflict Solving Program For Kids (Schmidt & Friedman, 1990). No reliability or validity data are available for the measure . This is followed by eight multiple choice questions designed specifically for this study . The first 4 questions address self-efficacy in conflict situations . Students are asked to respond either A) this is true , B) this is sometimes true , or C) this is not true , to statements regarding their ability to manage conflicts in general, at school , with other students, and with staff members . The remaining 4 questions look at perceived level and frequency of conflict at the school and the perceived importance of the Peer Mediation -Conflict Resolution program . No reliability or validity data exist for this measure .
The final measure was the Conflict Response Behavioral Scale (Kreidler , 1984) which is a self report of strategies used. Fifteen conflict strategies are listed and students are asked to check whether they always, sometimes, or never use each of the tactics . This scale is designed as a curriculum based measure of Kreidler's (1984) Creative Conflict Resolution program and lacks reliability and validity data . Mediation assessment A written questionnaire (see Appendix E) was designed by the researcher to evaluate mediator's knowledge of basic program components . There are no reliability or validity data for this curriculum-based measure . It consists of 5 brief answer questions regarding the role and function of a peer mediator. Students must achieve a score of 80% or better to become peer mediators .
A live role-play scenario was used to assess students' ability to mediate . The students were divided into 2 groups of 4 students to enact the role play. Two of the students acted as disputants and 2 peer mediators conducted the session within each group . Upon completion , the students switched places for a second session . The former disputants worked as mediators , and the former mediators enacted a different dispute .
These role-play scenarios were evaluated by 2 independent judges who were knowledgeable about the process and goals of peer mediation. . Performance on this measure was rated either oass/fail.

Procedures
At the time of the researcher 's involvement in the program , the school had already invested in the Schrumpf, Crawford , and Usadel (1991) Peer Mediation: Conflict Resolution in Schools package for training the peer mediators . The program was designed for use by middle and high school students. It contains copies of all necessary forms, one program guide for training, and student manuals. This package was reviewed by Benson and Benson (1993) . They found the program to be "comprehensive and wellsequenced" (p. 429) and recommended its use .
The peer mediation training took place over the course of 4 weekly sessions. Each training session lasted 2 hours , so the mediators were trained for a total of 8 hours . Each mediator received a program workbook (Schrumpf et al., 1991), which was covered in the course of training .
In the first session mediators were given an overview of conflict sources and responses . Students then learned about the characteristics and role of peer mediators .
The remainder of the session was spent learning and practicing the communication skills of active listening, summarizing, and clarify.
The second session focused on the peer mediation process . The 6 mediation steps from the Schrumpf et al. (1991) program (i.e., open the session, gather information, focus on common interests, create options, evaluate options and choose a solution, and write the agreement and close) were briefly described and a case example presented. The session then discussed each step individually in considerable detail.
At the third meeting students used workbook activities to practice gathering information and focusing on common interests . Groups of three students worked to role play conflict situations through the first three mediation steps. The students were asked to critique their performances and offer suggestions to each other. Once comfortable with that part of the process discussion turned to ways of encouraging disputants to create options for themselves . The session ended with the writing of practice agreements .
The final session consisted of a post-test (see Appendix E) and the judged roleplay scenarios previously mentioned under mediation assessment. All eight of the mediators met the requirements to become official school mediators .
The timeline and 7 lesson plans for the conflict resolution portion of the program are presented in Appendix A. (The reader is referred to the Program Design section of this paper for a discussion of th e elements .) Training was delivered once a week in all grades by the same trainer over the course of 7 weekly sessions . Each session lasted an hour. The trainer spent 50 minutes in each class room over the course a day. The program components were delivered as specified by the program timeline. The pre-test was administered during the first session and the post-test during the last session .

Results
To determine whether there was an increase in student knowledge about conflict resolution a related measures 1 test was used to analyze the pre-and post-training student questionnaires. This meant that each student acted as his or her own control. Tallies were made of the number of correct responses on the Conflict Resolution Scale. It was expected that students would have a higher number of correct answers on the measure subsequent to training, thus reflecting an increase in knowledge about conflict and its resolution .
The predicted increase in the variety and effectiveness of conflict resolution strategies endorsed by students was evaluated by changes on the Response to Conflict Scale. The Response to Conflict Scale was scored Always =1, Sometimes =2, and Never = 3 and summed to create a continuous measure . It was expected that the training would be reflected in increased endorsement in the use of communication (items 3, 4, 6, 8, 11, and 12) along with a reduction in the use of aggressive and/or passive strategies (items 1, 2, 5, 7, 9, 10, 13, 14, and 15). Reverse coding was used so that a low score reflects movement toward the increased use of communication strategies . A related measures 1 test was used to analyze the pre-and post-training questionnaires .
To address the question of improved self-efficacy for conflict resolution a comparison was made of pre-and post-training scores on the self-efficacy questions . A related measures 1 test was used to compare the scores. A higher score reflects greater confidence in one's ability to manage conflict. On questions 1 and 4 the answer "This is true" was scored 3, the answer "this is sometimes true" was scored 2, and the answer "this is not true" was scored 1. On questions 2 and 3 the scoring was reversed with "This is true" scored 1 and "This is not true" scored 3.
In order to explore the relationship among the measures of self-efficacy, knowledge about conflict, and endorsed conflict resolution strategies correlation coefficients were determined for each of the combinations at both data collection points. Results of the knowledge and self-efficacy scales were used as the independent variables and the score on the endorsed strategies scale was the dependent variable in a multiple regression analysis to determine important correlation's between knowledge and selfefficacy and strategies for successful resolution. 14 Comparisons of pre and post-test scores on the measures of knowledge about conflict, preferred responses to conflict, and self-efficacy in resolving conflicts were made by means of dependent groups 1 tests . Table 1   The comparison of pre and post-test scores on the Conflict Resolution Scale was found to be statistically significant 1 (71) = 4.48, p < .05, indicating that students' knowledge about conflict and its resolution increased . Differences between pre and post scores were not found to be significant on the Responses to Conflict Scale, 1(71) = -1.35, u>.05, nor on the self-efficacy measure , t(71) = -0.85, u>.05.
It was discovered that the post test contained only one of the four self efficacy questions from the pretest measure (i.e., I'm very successful in resolving conflicts that happen to me) . To determine whether the value of a single response differed significantly from the mean of all four responses, a dependent t-test was used to compare these conditions on the pretest. A comparison found the difference to be significant, 1 (71) = -3.35, ll < .05. This indicates that the response to a single question is not representative of the mean response to four questions .
A standard multiple regression was used to explore the relationship among the measures of self-efficacy, knowledge about conflict, and endorsed conflict resolution strategies . The variable of greatest interest to the school was the strategies employed by students during conflict -hence response to conflict was used as the dependent variable in the regressions .
The first multiple regression used information from the pretest measure .
Responses to conflict was the dependent variable . Knowledge and self efficacy, as measured by the single question, were the independent variables . The correlation matrix indicates that knowledge has a correlation of -.107 with response to conflict. Self efficacy has an even weaker correlation of -.096 with response to conflict . The correlation between the two independent variables is -.125.
The independent variables, knowledge and self efficacy, accounted for only 2.3% of the variance in responses to conflict . The beta weights for both knowledge and self .
A second regression was done on the pretest data utilizing the mean response for the four self efficacy questions instead of the single response . Knowledge and self efficacy correlated at -.225 . The correlation between self efficacy and response to conflict was found to be . 01. As stated above, knowledge has a correlated of -.107 with response to conflict. 16 The proportion of variance in endorsed responses to conflict accounted for by the best linear combination of these two variables was 1.16%. With beta weights of -.014 for self efficacy and -.11 for knowledge, neither was found to make a significant contribution to predicting response to conflict.
Data from the post test measure were analyzed with the same procedure . The correlation of knowledge and self efficacy to response strategies were found to be -.126 and -.132 , respectively. The independent variables (knowledge and self efficacy) correlated .143 with each other.
Knowledge and self efficacy accounted for only 2. 9% of the variance in response to conflict . Self efficacy had a beta weight of -.116 and knowledge had a beta weight of -. 109. Neither variable was found to be significant.

Discussion
This research was undertaken to evaluate a new conflict resolution -peer mediation program instituted at a small, private middle school. The conflict resolution was delivered to all students over the course of seven weeks . Eight peer mediators were chosen by the school and trained with the Schrumpf et al. ( 1991) program. The peer mediation program was not instituted early enough in the year to generate data on the number and disposition of referrals . The school intends to collect and analyze data on this . .
program m upcommg years .
The conflict resolution portion of the program will be discussed in terms of the dimensions suggested by Elliott et al. (1991). To review, the dimensions were effectiveness, acceptability, parent/teacher empowerment, and treatment integrity .

Effectiveness
It was predicted that students would demonstrate an increase in knowledge about conflict resolution and endorse more effective strategies for resolving conflicts . It was also predicted that students would reflect improved self-efficacy for conflict resolution .
Only one of these predictions was supported by this research. Students did demonstrate significant gains in the area of knowledge about conflict.
The headmaster of this school requested the program in response to perceived tension among the diverse student body. School staff was primarily interested in the behavior of students involved in conflict. Of the measures employed by this study the Response to Conflict Scale came the closest to evaluating what students do when in conflict. Unfortunately, pre and post tests were not significantly different, suggesting little change in behavior .
The program consisted of seven weekly sessions, each lasting about 50 minutes .
The students did have the opportunity to practice conflict management skills during the sessions, but the transfer of these techniques to the student's own, emotionally-charged conflicts might require more extensive practice than the time constraints allowed .
The transfer of techniques could have been assisted by teacher participation in the training sessions . Since the teachers were not part of the training , they could not be expected to model the communication techniques and process of the program. It is likely that they continued with the past practice of referring disputants to the office.
Self efficacy for conflict resolution did not change significantly. Students did demonstrate improved knowledge about conflict on the post test . It is hoped that , with time and practice , the students will put this knowledge to work and feel confident in their ability to manage conflicts .
An additional concern for the potential effectiveness of the program stems from the timing of program delivery. The training sessions did not start until the beginning of May, and finished by mid June . There was only several days of school remaining . This was not an ideal situation for the development and support of skills. It would have better to start the program in late fall to afford students ample time for practicing the techniques .

Acceptability
This program was developed from a number of existing conflict management sources . It was reviewed by the headmaster and curriculum development coordinator prior to its implementation in the school and changes were made at their suggestion . The program was presented to a parents' advisory board and received their approval. This program was acceptab le to the aforementioned groups .
Unfortunately , two important stakeholder groups did not have input into the program design . Classroom teachers and students were not consulted in program development. Separate student and teacher focus groups to discuss the nature of the school conflicts would have been a valuable addition to program development and may have increased their investment in the treatment.
The other acceptability issue has to do with the choice of students to become peer mediators . Two of the mediators appear to have been chosen for reasons other then their innate ability to work with people . For instance , one of the selected students was frequently referred to the office for fighting. The other student had few friends. These students would have been appropriate nominations in an established, successful peer mediation program. However, this was a new program and it had yet to gain acceptance with the student body. At this early stage it would have been better to train students who were respected by their peers .
Teacher/Parent Empowerment Unfortunately , this program did little to empower teachers and/or parents. The program was delivered to students by an outside agent at a time when teachers were not in the classroom. Teachers had not been consulted in program development and did not know the program content. Teacher involvement in both the development and delivery of the conflict resolution program may have facilitated the transfer of training to daily situations.
In a similar manner, parents were also uninformed of the program's content. This was a missed opportunity . A recommendation for future programs is to have a parent workshop in conjunction with the school program. That way families will be learning the same skills together and the likelihood of generalization is increased . Horowitz and Broadman ( 1994) recommended the institution of community based workshops . A workshop offered to parents would be a step in that direction .

Treatment Integrity
The program was planned and delivered by the same, outside agent in five separate classrooms. All of the units were delivered to each of the classes. However, the composition of each class was different. In some classrooms the students interacted freely with the material. They were not self-conscious in their role-play attempts and focused on the subject matter.
Other classes were more hesitant to take risks. This was particularly true of the 10 boys and 5 girls in seventh grade . Several of the boys tended to exaggerate their role-play lines, rendering the scene comic instead of informative. At times the group was willing to discuss the conflicts they encounter , but at other times they seemed more interested in an upcoming field trip .
The program was presented to each group , but there was considerable variability in how the classes interacted with the material. There was variability across groups but there was also variability within groups across the seven weeks. Absent students not only effected the makeup of a class for that day, but also were behind the following week. One youngster missed the first two sessions and was lost on week 3.
The treatment was delivered as planned, but it would be misleading to suggest that each class received the same program . The interpretations and questions presented by each class was unique . The personal conflicts of fourth graders differ considerably from those of eighth graders , and those differences led to subtly different conflict resolution programs .

Summary
The program under consideration sought to teach students skills for managing school conflicts . The results of this evaluation indicate only minimal success in that endeavor. Students did increase knowledge , but self efficacy and self reported behavior appear unchanged. This is not an uncommon finding in the evaluation of short term programs. Ladd and Mize (1983) reviewed a number of social skills training programs . They noted that increases in knowledge of appropriate behavior did not necessarily translate into action and recommended the strengthening of components to enhance performance and maintenance/ generalization .
Similarly, Nastasi and DeZolt (1994) discovered in their review of the substance abuse literature that increases in drug related knowledge had low correspondence with behavior change . They also recommended cognitive-behavioral methods to induce the desired behavior change .
This program did attempt to allow students the opportunities to think through and act out conflict situations, but the opportunities were isolated to once a week with an outside party . Increased practice opportunities and teacher attendance at the training sessions may have facilitated the transfer of skills.
This results of this research indicate some changes for future programs . The first suggested change is the incorporation of student and teacher input in the design, delivery, and evaluation stages . A better consideration of these important stakeholder groups may lead to improvements in treatment acceptability and teacher/student empowerment. Such improvements are likely to facilitate the transfer of training.
Evaluation design would be strengthened by the incorporation of behavioral observations . Teachers could be requested to collect data before and after the intervention . Input from teacher and student groups could be used to design a checklist of common conflicts and how they are handled . Such a list may provide a more objective measure of change.  (Shaw, 1992, (Schrumpf, Crawford, & Usadel, 1991, pp. 6-8) Responses to Conflict Discussion (Schrumpf et al, 1991, pp.  If necessary , go back and talk about the conflict situations that were not discussed last week . Students listen to "The True Story of the Three Little Pigs" 5 min. by A Wolf (Scieszka, 1989) .
Talk about perspectivehow the same thing looks different 5 min. from another angle . Students break into small groups with their conflict stories . Two 10 min. sets of footprints , facing each other , are on the floor . Students present both "sides" of their conflicts using the footprints . Reconvene as a large group to discuss how it felt to play the 5 min.
other person. Did that change how they viewed the conflict? Suggest that they practice perspective taking throughout the week . de-escalate conflict. Read through "Fighting Fair" and discuss. Break into small groups to analyze the conflict situations presented at the bottom of page 8. Students rewrite the scenes to reflect fair fighting. May be presented as skits to other students . Review previously discussed strategies (week 4) and Schmidt & Friedman handout (1985, p.9) Talk about when each strategy may be appropriate and its likely outcome . Note that different situations require different approaches. Ask students to list the strategies they use most frequently and consider what the consequences are. 10 min. 20 min. 5 min. covered by the program. Introduce role playing exercise. Each group will receive a different problem situation to 20 min. work through. Students will identify the conflict, decide its likely source, and discuss the perspectives of the parties involved. They will then decide how to manage the conflict. The small groups will present their dilemmas and either role-play or talk through their solutions. Identical to pre-test questionnaire. Read the following sentences and circle the answer that is best for you.

1.
I am very success ful in resolving conflicts that happen to me .
A. This is true B. This is sometimes true C. This is not true 2.
When I get into a disagreement with another student at school , I am not sure what to do .
A. This is true .