European Journal of Psychology and Educational Research

: This study explored the impact of the peer support programme on pupils at Japanese technical high school for over 6 years. A total of 268 pupils (an intervention group 112; a control group 114) were invited for the study and the pupils of the intervention group were given peer support training and they joined its supporting activities. All the pupils were assessed three times by adopting IRS, RSES10 and GHQ28. The results showed that the programme seemed to give positive influences on the peer supporters in terms of IRS, RSES10 and GHQ28. Also, both group members were classified into low-scoring groups and high-scoring groups, depending on their scores at the first assessment. Then, the results of analysis showed that the pupils from both low & high-scoring groups significantly improved their scores in IRS. In RSES10 and GHQ28, the pupils from the low-scoring groups improved their scores, but those from the high-scoring groups did not improve much. As a conclusion, even in a technical high school where the majority of pupils was male pupils (over 90%), the peer support programme seemed to give positive influences on the peer supporters in terms of IRS, RSES10 and GHQ28.


Introduction
Peer support is an approach that builds on the helpfulness and altruism characteristic of friendship by extending it beyond friendship to the wider peer group. Cowie and Wallace (2000) argued that through peer support interventions, young people can learn to deal with conflict and to relate to one another in a more constructive, non-violent way. The peer support approaches have been continually developed over the past 50 years in Western Nations. In Japan, the peer support approaches were introduced to school education in the early 1990's when several children's issues had been exacerbated such as bullying, youth suicide, class disruption and school absenteeism (Ministry of Education, Culture, Sports, Science and Technology, 2001). The peer support approach was applied as a part of preventive methods against these issues, and since the year 2000 has gained popularity amongst schoolteachers and researchers. In Japan, the peer support approaches have been facilitated by several educational groups, and the Japan Peer Support Association (JPSA) was the largest and most influential organization across, which have had great influences on the theoretical developments of the peer support and its practices.
In the study of peer support, which was applied in educational settings, there were several research issues which were described in previous studies (e.g., Houlston & Smith, 2009). One of the issues was that there were not much quality longitudinal studies, which were designed to examine the impact of peer support programmes between an intervention group and a control group (Baginsky, 2004). There were several longitudinal studies, but most of them were partially based on the qualitative data which were derived from participant's self-reports to show their feelings about the impact of the peer support.
Another issue was that there were not many studies regarding male peer supporters' practices. Due to the nature of peer support activities, such as caring for others, male pupils tend not to be involved into the activities as a peer supporter. Even if male pupils joined the peer support activities, their dropout rate was generally high comparing to those of female pupils.
It seems that male pupils did not want to be regarded as a "teacher's pet" by other male pupils. Also, male pupils might think that joining peer support activities was a threat to their sense of masculinity (Olweus & Endresen, 1998). Hence, the peer support programme seemed not to be adapted well in male-dominated school environments. Considering these issues, this study was designed to examine the impact of the peer support programme on mainly male pupils while adopting a long-term peer support practice. Methodology

Objectives
The research objectives were as follows; 1) To evaluate the impact of the peer support programme on the peer supporters comparing to the control group in terms of a long-term practice.
2) To evaluate the impact of the peer support programme at the male-dominated school by utilizing existing school activities as a part of peer support activities.

Participant and Data Collection
Participants were invited from a state technical high school which was located in a large city in Aichi Prefecture, Japan. The school has about 840 pupils in total and 93% of them were male pupils. As the intervention group, annuary twenty pupils were invited and most of them were members of the school health committee. Scales For the evaluation of the peer support programme, the following scales were adopted. Interpersonal Relationship Scale (IRS) is a 6-item scale for assessing self-assertion, self-understanding, open-mindedness, understanding of others, sensibility and reliability. The items are rated on a 5-points Likert scale and the larger numbers indicates more skillful. The scale was originally developed for assessing impact of Structured Group Encounter. Interpersonal Relationship Scale yielded relatively high reliability for the samples (α=0.83).
Rosenberg Self Esteem Scale (RSES10) is a 10-item scale for evaluating individual self-esteem. The items are rated on a 5point Likert format, and the larger numbers indicates one's higher self-esteem. RSES10 yielded acceptable reliability for the samples (α=0.75) Also, the clean-up committee members were selected as a control group. In Japan, a new academic year traditionally starts in April, therefore the pupils of the intervention group started joining the peer support training and supporting activities from April. For the evaluation of the peer supporters (the intervention group) and the control group, assessments were conducted three times per year; before the peer support programme starts (April), after 7 months (November) and after 10 months (the end of January). The three scales, which were Interpersonal Relationship Scale (IRS), Rosenberg Self Esteem Scale (RSES10) and General Health Questionnaire (GHQ28), were applied for the evaluations in both the intervention group (peer supporters) and the control group. At the end of the final training session (at the end of January), the peer supporters were asked to complete a self-evaluation questionnaire, which provided some space for free descriptive answers about their experiences regarding the activities.

Framework of the Peer Support Activities
The structure of the peer support programme of Japan Peer Support Association is composed of a cycle of 4 phases, which are 1) Training, 2) Planning, 3) Support activities, and 4) Group supervision (Evaluation), and this is generally called the Peer Support Programme in Japan (Morikawa et al., 2002).
In this study, the peer support programme started annually from April, and its training, planning and support activities were carried out while considering the contents of the school events. Group supervision (Evaluation) was held in November, then, the training, planning and support activities were continued until the February. Falling backward (Feeling of reliability) Dec. Listening VI: Five steps for solving problems (Giving a helping hand to)

Analyzing of Data
A series of analysis of variance was conducted with using IBM SPSS Statistics 25, to compare the values between the intervention group (peer supporters) and the control group. Also, in terms of the results of the first assessment, the participants (both the intervention group and the control group) were divided into two groups (low/high-scoring groups) for further evaluation. Since the participants were selected from young adolescents, this seemed to maximize individual differences in their scores. With careful consideration of their individual differences in social and emotional developments, the participants were divided into two groups (low/high-scoring groups) after the first assessment. If their scores were lower than the median in the first assessment, they were classified in the low-scoring group. Similarly, if their scores were the higher than median, they had been classified as high-scoring group. In GHQ28, the lower score means healthier in terms of mental states. Thus, participants who scored lower than median, had been classified as high-scoring groups, which means to be received positive influences. Eventually, each group was classified as follows: (IRS) low-scoring group ≦ 20 < high-scoring group, (RSES10) low-scoring group ≦ 30 < high-scoring group and (GHQ28) low-scoring group ≧ 6 ＞ highscoring group. Also, this study applied three different scales for the assessment, which may involve common method bias. Thus, Harman's single factor test was used to check the common method bias, and the result (26.579) indicated that there were no significant common methods biases across the Interpersonal Relationship Scale (IRS), Rosenberg Self Esteem Scale (RSES10) and General Health Questionnaire (GHQ28).

Low-Scoring Group by Year
In IRS, there was an interaction in year 2005 (p< .01) and there were significant main effects of time in every year. In RSES10, there was a significant interaction in 2005 and significant main effects of time in all year except 2008. Also, among the intervention group, there were significant increases in all years except 2008 and 2011. On the contrary, in the control group, there were significant increases in 2011 (after 7 months, p< .05; after 10 months, p< .01).    8.2 1.6 6.9 4.3 6.3 4.7 A×B 1.52 †p< .10; *p< .05; **p< .01; ***p< .001 1：0 month 2 ： After 7months 3： After 10 months IRS; Low-scoring group ≦ 20 < High-scoring group RSES10; Low-scoring group ≦ 30 < High-scoring group GHQ28; Low-scoring group ≧ 6 ＞ High-scoring group

Discussion
The primary focus of the study was to examine how the peer support programme give the impact on the male-dominated school in terms of a long-term practice. As a whole, the results of 6 years data showed that there were significant main effects of time in IRS (after 7 months, p< .001; after 10 months, p< .001), RSES10 (after 7 months, p< .001; after 10 months, p< . 001) and GHQ10 (after 7months, p< .001; after 10 months, p< .001). This seemed to indicate that there were significant positive effects in all the scales among the intervention group (peer supporters), and this implied that the peer support programme effectively worked and gave positive influences on the male-dominated high school pupils. However, among the control group, there were also some significant effects in IRS (after 10 months p< .001) and in RSES10 (after 7 months p< .05; after 10 months p< .05). In order to clarify the reason for this, it is necessary to investigate more details by considering some factors. For instance, the experiences of the peer support training and activities may alter the influence of the peer supporters depending on their personal development rate. As mentioned earlier, since the participants were the young adolescents, this seemed to maximize individual differences in their scores. Thus, before the peer support programme started, all the participants were assessed with three scales, and they were classified into two groups (the lowscoring group and the high-scoring group). On comparison between low and high-scoring groups, the low-scoring groups in both the intervention and the control groups had significant increases surprisingly in all the scales (IRS, RSES10 and GHQ10). On the contrary, the high-scoring groups in both the intervention and the control groups did not show any significant increases. But, only in IRS, there was an increasing tendency among the intervention group. These results implied that the participants in the low-scoring group were easily influenced by the peer support programme and positive climates of the school. All in all, the intervention group slightly showed better improvement against the control group in all the scales.
As shown in Table 6, unexpectedly each scale had very weak positive correlations with other scales. Generally, children who often mis-behaved, tended to have lower scores in interpersonal relationship, individual self-esteem and mental health conditions (Aoki, 2007;Makino & Tagami, 1998;Matsuura et al., 2020;Miyauchi, 2013;Okada et al., 2010;Olweus & Endresen, 1998;Oshio, 2001). Therefore, it was expected that each scale would show relatively positive relationships with other scales, but it did not. Since the intervention group (of low-scoring group) in all the scales, showed significant improvements, further investigation is required to fully understand the results of the correlations.
Also, from the results of the table 7 (low-scoring data by year), in IRS, there were significant improvements at the early stage (after 7 month) among the intervention group in most years. This implied that the peer support programme gave positive influences on the participants in terms of interpersonal relationships. Likewise, among the control group, there were significant improvements in 2009 and 2011 in IRS. This implied that not the peer support programme, but some other factors may give positive influences on the members of the control group. Some educational activities in school, generally feature the school's educational functions to enhance children's interpersonal relationships, self-esteem and prosocial behaviours (Taruki & Ishikuma, 2006). In this sense, the peer support programme seemed to play a role in reinforcing the school's educational functions. Also, the programme completed its first cycle in July (Training -Planning -Supporting activities -Group supervision) and continued to run another cycle of programme. Therefore, one of the possible reasons was that meaningful synergies between educational activities in the school and the peer support programme seemed to help both the intervention and the control group to improve their personal abilities and skills.
In RSES10, there were significant increases among the intervention group in year 2005year , 2007year , 2009year and 2010year . In 2008 there was no significant increase, but this does not mean that the peer support programme did not work well. In fact, in 2008, the participants of the intervention group solidified their foundation of the group, and their peer support activities were dramatically developed. Nevertheless, the results showed there were no significant improvements in all the scales. When assessing children's improvement in self-esteem, it seems necessary to consider not only their raising scores but also other issues. As Cowie and Wallace (2000) mentioned, when peer supporters start deepening their understanding of themselves and others, peer supporters temporarily tend to decrease their self-esteem. This is because their self-concepts seem to be agitated in the process of deepening their understanding (Kato et al., 2018;Nakagaki, 2011). Then, after a while, generally their self-esteem increases again. In this vein, their temporary decrease in their self-esteem seems to be part of the process of their emotional development, thus, this seemed to happen when the assessments were conducted. From the above, the peer support programme seemed to give positive influences to their self-esteem. Interestingly, however, in the year 2011, there was a significant improvement not in the intervention group, but in the control group. Several pupils who belonged to the mechanical department, applied for the national high school design idea competition, and won a grand prize, which gave them an opportunity to commercialise their ideas (works). Eventually, this was a big news to the whole school and this seemed to give a positive influence to many other pupils as well. This may be one of the reasons for a significant increase in their self-esteem in 2011.
There were many studies, which explored the impact of peer support programmes on children's self-esteem. Some studies did not show significant positive effects (Badura et al., 2000;Sanada, 2020;Sasaki, 2005;Sato, 2018;Takahashi & Kurihara, 2006), but the others did show significant effects (Ellis et al., 2005;Houlston & Smith, 2009;Ikegaya & Kasai, 2003;Kondo & Sakai, 2007;María Soledad, 2007). One of the possible reasons why several studies did not show an increase in positive effects, maybe that the peer support programmes focused heavily on the training aspect, and peer supporters did not have much opportunities involved in supporting activities, which gave them opportunities to do actual practices in daily life. Also, another possible reason seemed that there were several cases which provided the peer support trainings with more than 300 pupils and its trainings might be inferior in quality (Takahashi & Kurihara, 2006). While, in the studies showing positive effects, most these studies provided training with a relatively small group (from 14 to 38 peer supporters). Also, in some cases, the selection of peer supporters was conducted in a democratic manner, such as voting by pupils (María Soledad, 2007), and peer support programmes were supported by the whole school policies and systems, which helped peer supporters to be actively involved into their supporting activities (Ellis et al., 2005;Houlston & Smith, 2009;María Soledad, 2007). In this vein, it is presumed that when peer supporters were not involved in the whole process of peer support programmes (cycle of the peer support activities) in an appropriate manner, the studies less likely showed positive results of peer support programmes.
In this study, the results of the year 2005 (the first year of 6 years practice) showed an increasing tendency in self-esteem scores, which had a similar increase rate with Houlston and Smith's study (2009). Since the first year (2005), this study annually kept showing significant improvements on peer supporters' self-esteem scores, and this is assumed that the cycle of the peer support activity worked well to give positive influence on peer supporter's self-esteem. Self-esteem appears to be enhanced when they received positive feedbacks and/or evaluations by others (Endo et al., 1992). Therefore, teachers who took the lead in the peer support programmes, needed to take enough time for evaluation (Group supervision) of peer supporters to receive positive feedback and to go through the cycle of peer support activities multiple times.
From the table 7, in GHQ28, there were significant main effects of both the intervention and the control groups and time in 2005. Also, there were significant improvements among the intervention group every year except 2009. This result means that the peer support programme gave positive influence on pupil's mental sanity in terms of GHQ28. In regards to the 2009 results, the influenzas epidemic in Japan had significant impact on the school from November to February. For this reason, there was a possibility for pupils to be also affected by this issue.
In this study, various supporting activities, such as volunteer work, visiting nursing homes to conduct special classes, and school cultural festivals, were conducted after the training sessions. These supporting activities were designed to mainly support the elderly and young children, and the participants were actively involved before July. Basically, the participants were able to actively utilize their skills which were gained from the training sessions, and their sense of accomplishment and satisfaction seemed to contribute to their improvements of their scores in the scales. After the end of all the training sessions and supporting activities, the participants answered a questionnaire about their personal supporting experiences. From the table 8, there were only 8 cases which the participant worked for mediating quarrels, and there were no supporting activities which were related to "Study support (learning support)". The majority of the participants answered, "Presentation at cultural festival", "Greeting", "Care for classmates", "Lend an ear to others", "Activities of school health committee", "Cleaning activities", "Volunteer work", etc, and these supporting activities were comparatively simple and easily embedded in daily school life. In a sense, supporting activities, which were related to "Mediating quarrel" and "Study support (learning support)", seemed to be considered as difficult supporting activities.
Also, in order to run the peer support programme in an effective manner, each training session was arranged and scheduled in accordance with the school events. For instance, in the session 3 "Trust walk & Assertion" (see the table 2), the participants learned and gained the skills such as "guidance method" and "escorting", then they joined a local volunteer support activity to care for younger children. In the session 7 & 8, the participants learned about "questioning which suits the situation", and "the difference between resecure and support". After they gained appropriate skills and knowledge, which helped them to care for others, then, they joined AED course and sports festivals. In the session 9, through the personally assessments, the participants deepened their self-understanding, and in session 10, participants learned about the importance of coping with stress by themselves. Then, they joined the health campaign at the school cultural festivals, which enlightened knowledge of health issues as a peer educator. On the health campaign, there were some opportunities for general participants to join their own stress check and/or personality assessments. Before the health campaign started, the peer supporters were already able to gain relative skills and knowledge through the training and its simulations, thus they were able to provide quality supports with others in an effective manner. By arranging the training sessions in this way, the activities of the school events were able to be used as the supporting activities of the peer support programme. This idea tended to make it easy for peer supporters to plan the supporting activities, and also enhanced the quality of the school events.
In addition, a school nurse in charge of the school health committee, successfully managed to run the peer support programme, and this made it easy for the peer supporters (participants) to build a good relationship of trust with a school nurse. The school nurse endeavoured to effectively run the cycle of the peer support programmes, and this seemed to contribute to high participation rate in training sessions and less dropout. This eventually seemed to encourage the peer supporters to be involved in a wide range of activities and to gain skills effectively. As shown, existing school events and activities were utilized as opportunities for the peer supporters to apply their gained skills. This unique methodology smoothly embedded the spirit of the peer support programme within the whole school system. This implied a possible beneficial effect of the peer support programme to deal with various issues in school, such as school absenteeism, bullying and suicide.

Conclusion
Even in a male-dominated high school, which are generally considered to be difficult to run peer support activities, the peer support programme was effective in providing a positive influence on the peer supporters. Also, in terms of interpersonal relationships (IRS), both low and high-scoring groups showed their positive improvements. In terms of both self-esteem (RSES10) and mental health (GHQ28), the participants who had low scores, improved their scores, however, the participant who had high scores, did not show their improvements much. This study also demonstrated that even male pupils who tend to feel uncomfortable being involved into peer support, could be actively engaged in the peer support activities, with careful considerations.

Recommendations
One of the main contributions of this study is demonstrating the outline of practical examples in its training and support activities. In terms of organizing the peer support programme, several practical keys are suggested as below.

Combining Peer Support Activities and the Existing School Activities
One of the recommendations is to combine peer support activities and the existing school activities as a part of the peer support programme. As mentioned, the student's health care committee members were guided to work as the peer supporter and they successfully played their role. Likewise, it might be highly possible that other school activities, such as club activities, are used as the peer support programme, depending on school conditions. This easily encourages male pupils, who do not want to be labeled as a teacher's pet from other pupils, to join the programme, and also seems to play a role in enhancing the school's educational functions and the quality of the school events.

Well-Organised Planning
Another recommendation is to focus on well-organised planning for productive activities. In planning meetings, peer supporters need to plan their own supporting activities. This seems to make them feel less pressure and they may smoothly join and/or be involved into activities. Peer supporters are also advised to contact the teachers when they encounter with the serious cases. Then, by depending on the teacher's decision, appropriate peer supporters need to be allocated to tackle the serious issues. Having effective plannings helps to make sure that the peer supporters will experience their own productive activities.

Limitations
There are several limitations of this study. One of the primary limitations of the study was that some parts of the result showed positive impacts among the control group too. The second limitation was the validity of material selection for the study. The third limitation was that the author struggled to stabilize the frequency and timing of the training and support.