Inter-Peer Group Status and School Bullying: The Case of Middle-School Students in Japan

Abstract

The aim of the current study was to investigate the relationship between school bullying and inter-peer group status among Japanese middle-school students. A total of 2294, 7th–9th grade Japanese middle-school students participated in this study. Using a self-report questionnaire, we measured their inter- and intra-peer group status, bullying behavior, victimization, coping strategy, and current victimization status. The results showed that, compared to those in the higher-status groups, students in lower-status peer groups tended to be bullied and experienced prolonged victimization. These findings suggest that a lower peer-group status predisposes students to victimization.

1. Introduction

Peer status is an important factor for students when it comes to their school life and adjustments. In previous studies, researchers characterized students’ social status among their schools or grades as “popularity” and divided it into sociometric popularity and perceived popularity. Sociometric popularity is defined as to what extent a child is accepted and liked by others [1]. Previous research found that sociomatrically popular children displayed a lower level of aggression [2], victimization [3,4], and bullying perpetration [5,6]. This means that students with higher sociometric popularity are less likely to be victims or perpetrators of bullying.

Regarding perceived popularity, the relations changed in the opposite direction. Perceived popularity entails being perceived as “cool,” prominent, or at the center of one’s peer group [7]. This construct is ordinarily measured through the peer-nomination method, wherein students are asked to rate their peers based on their behavior or traits. Although aggression is considered a socially maladaptive behavior by nature, many researchers have shown that popular students display more aggressive behaviors than their less-popular counterparts [2,8,9]. In longitudinal studies, perceived popularity predicted further overt and relational aggression [8,10]. Furthermore, these studies suggested that aggressive behavior plays an important role in the attainment and maintenance of high status. Cillessen and Mayeux [8] showed that relational aggression positively predicted subsequent perceived popularity, and Lu et al. [10] revealed that overt aggression predicted perceived popularity. Therefore, perceived popularity could promote aggressive behaviors, and this aggression could enhance and maintain one’s status. However, varying degrees of both forms of popularity are related to aggression and bullying.

Researchers have also indicated a relationship between perceived popularity and school bullying. Although students with high sociometric popularity tend not to bully others or be bullied [11], popular students tend to bully others [12,13], while unpopular students tend to be bullied [14,15]. Further, a literature review has shown that individuals engage in bullying behavior to demonstrate their “coolness” to peers [16]. These studies show that it is important to be aware of the relationship between perceived popularity, bullying, and victimization, to understand bullying in schools.

The findings concerning perceived popularity, as initially obtained in Western countries such as the United States and Europe [8], have been replicated in some areas of East Asia, such as Hong Kong [17], Taipei [18], and mainland China [10,19,20]. In other words, perceived popularity in school was found to be related to aggression, victimization, and bullying perpetration in both Western and East Asian countries. However, this relationship has not been extensively studied in the context of Japanese adolescents.

1.1. The Japanese Middle-School Context

In the Japanese middle-school system, students are typically randomly assigned to classes every school year and take courses as a group. Smaller groups of students in the same class are also assigned some chores (e.g., sweeping the classroom) by their teachers. In addition, students are expected to participate in special events such as school trips and festivals. Compared to a typical Japanese school, these special events are not held in Chinese schools [21], suggesting that Japanese schools are unique in the context of East Asia. As a result, middle-school students in Japan seem highly cohesive, and peer relationships in the classroom are important to their school life. In addition, Kanestuna [22] pointed out the difference between the Japanese and British (UK) school systems in terms of bullying. In typical British middle schools, students are assigned to classrooms according to their academic abilities and attend classes for different courses in different classrooms. Moreover, students are free to enroll in any course according to their aspirations and interests. Depending on academic performance and interests, students have relatively high mobility between classrooms, but Japanese students do not have such mobility. British students usually spend their break time in the schoolyard and can establish friendships with peers from different classes and grades. Bullying in British schools is typically perpetrated in the schoolyard by students of the same or higher grades. In contrast, in typical Japanese schools, bullying (“ijime” in Japanese) is perpetrated by classmates who are of a similar age and in the same classroom within a particular grade.

1.2. “School Caste” (Inter-Peer-Group Status) and School Bullying

Although perceived popularity has not been studied in Japan, some Japanese educational critics and teachers have identified the notion of “school caste,” which seems to be similar to perceived popularity [23,24]. Over 80% of Japanese middle-school students report that they have informal peer groups in their classrooms [25,26]. School caste is defined as a status hierarchy within a peer group in a classroom [25,27]. The findings of an interview study showed that students of a high-status peer group are prominent and play a central role in daily school life (e.g., having a strong voice and assuming leadership roles during lesson, break times, and special activities) in the classroom, showing better school adjustment [27]. Furthermore, empirical studies [25,26] have measured group status in the school-caste system through “inter-peer group status,” which is defined as a peer-group status with which students are most involved in their classroom, and assesses how strongly students perceive their peer groups as playing a central role in their classroom, based on self-reporting. Indeed, these studies showed that students in high-status peer groups tend to show higher school adjustment [25,26]. School caste differs in that it represents status disparity via a group status hierarchy within the classroom, whereas perceived popularity represents a personal status hierarchy within a school or grade.

Moreover, school bullying may be attributable to the school-caste system. Investigating the relationship between inter-peer group status and school bullying, Mizuno et al. [28] conducted a cross-sectional survey study on a large sample of Japanese middle-school students. The results showed that inter-peer group status within a classroom was significantly positively correlated with bullying and negatively correlated with victimization. However, the coefficients obtained were too small (Spearman’s ρ < 0.10), which led to the conclusion that school caste only slightly correlated with school bullying.

Mizuno et al.’s [28] study had several limitations. First, the authors only measured the inter-peer group status. A previous study regarding school caste [25,26] suggested that, when exploring the effects of inter-peer group status, individual status should be controlled for by also assessing “intra-peer group status,” which is defined as how strongly students perceive their individual status as playing a central role in their peer group. Thus, the relationship between inter-peer group status and bullying must be investigated when the intra-peer group status is controlled for. Second, participants in the study were only required to report their current experiences of bullying perpetration and victimization. However, a lower status in the peer group might not only be related to victimization, but may also have some effects on coping strategies and the continuation of that victimization. For example, longitudinal studies reported that victimization predicted further victimization [29,30]. Students in a lower-status peer group might find it difficult to stop the victimization or fight back, since they have less power than their peers, which, further, leads to prolonged victimization. Hence, it is also important to include these variables to achieve a comprehensive understanding of the relationship between school caste and school bullying. Third, there is a possibility that the authors did not classify the participants accurately, as bullying and victimization were measured and analyzed according to frequency, instead of using an explicit standard. Owing to the lack of a standard, such as the one proposed by Solberg and Olweus [31], who identified someone who “previously bullied another/or was victimized more than two or three times a month” as being a bully/victim, the study might have categorized some participants inaccurately. Moreover, an overlap between bullying and victimization [32] may also make their findings less reliable. Victimized students reportedly had lower perceived popularity than the uninvolved, whereas bully–victim students were perceived to be as popular as the uninvolved [33]. Hence, we dealt with the bully–victim overlap by classifying the participants into four sub-groups to explore the difference between students who experienced only bullying (or victimization) and bully–victim students. In summary, the current study endeavored to use questionnaires to measure bullying perpetration and victimization comprehensively, and classified participants into four categories based on Solberg and Olweus’s [31] explicit standard. Hence, our study is expected to lead to a more reliable finding by addressing these concerns.

1.3. Overview of the Current Study

This study used four sub-categories of bullying—uninvolved, bully-only, victim-only, and bully-victim—to investigate the relationships between peer-group status in the classroom (i.e., inter-peer group status), bullying behavior, coping strategy, and current victimization status. According to findings from previous studies on peer status [34], we hypothesized that students in lower-status peer groups tend to experience relatively more victimization, whereas those in higher-status peer groups typically engage in more bullying perpetration. In addition, regarding the relationships between coping strategy, prolonged victimization, and group status, we hypothesized that students who are victimized in the lower-status peer group tend to not respond to or cope with victimization, and, consequently, experience prolonged victimization. Our hypotheses are presented in Figure 1.

2. Methods

2.1. Participants and Procedures

The participants were recruited from all seven public middle schools located in a rural–urban fringe area (X city) in Japan. X city, with a population of a hundred thousand, is a typical Japanese medium-sized local city. The survey was designed with the Bullying Countermeasures Liaison Council and conducted by the Board of Education of X city in July 2017, with the approval of the Principals’ Association in X city and ethical approval from the ethics review committee of the last author’s institution. In the homeroom activity, classroom teachers obtained oral informed consent from their students and provided them with the following information: the survey was anonymous, questionnaires would be collected in a sealed envelope after completion to maintain privacy, and they would experience no disadvantage related to their answers and had the right to refuse participation at any time. Prior to the survey, the Board of Education of X city informed the participants’ parents about the study. A total of 2294 Japanese students (boys = 1176 [51%] and girls = 1118) in the 7th to 9th grades (aged 12 to 15 years) participated in this study (95% of the students were from the seven schools). Participants were asked to complete the questionnaires at home and return them in a sealed envelope.

2.2. Measurements

Inter- and intra-peer group status. The protocol for instruction and status measurement was based on that described by Mizuno and Ota [25]. First, students were asked to recall a group that they were most-closely affiliated with, and then evaluate their inter- and intra-group status. Inter-group status was measured using the following question: “Do you think that your peer group plays a central role in your class?” Intra-group status was measured using items such as “Do you think that you play a central role in your peer group?” All items were rated on a five-point Likert scale (from 1 = “not at all” to 5 = “strongly agree”). Higher scores represented a higher group status in their classroom and higher individual status in their peer group. Participants who were not affiliated with any peer group in their class were coded as missing data (7.4%).

Victimization. In research, the definition of bullying includes an aggressor’s intention to attack and engage in repeated aggressive behavior, and the unbalanced power relationship between the aggressor and the victim [35]. However, in Japan, the Ministry of Education, Culture, Sports, Science and Technology (MEXT) defines bullying as “any physically or psychological behavior (including behavior committed via Internet) committed by one student toward another student who attends the same school (including primary school, junior high school, senior high school, secondary school and special education school) or with a certain relationship, which causes the victim to feel physical pain and/or psychological distress” [36]. Since the victim’s perception of bullying is emphasized in Japan, we did not provide any definition of school bullying to the participants. The participants were required to report their experiences of being bullied since the beginning of the school year (April) using Kato et al.’s [37] eight-item scale.

Table 1. Frequency of victimization and bullying behavior.

Items	Never	Only Once	Once a Month	Once a Week	A Few Times a Week
Excluded or ignored	89.1%	7.4%	2.0%	0.9%	0.7%
Had one’s belongings hidden or stolen	91.0%	5.9%	1.9%	0.7%	0.4%
Talked about behind one’s back	73.4%	12.5%	7.2%	4.4%	2.5%
Kicked or punched	91.2%	4.9%	1.9%	1.1%	1.0%
Called names	90.4%	4.5%	2.3%	1.6%	1.2%
Harassed on cyberspace	98.3%	1.2%	0.3%	0.1%	0.1%
Attacked jokingly	88.9%	5.5%	2.5%	1.6%	1.6%
Sexually harassed	98.9%	0.7%	0.1%	0.1%	0.1%
Exclude or ignore	82.9%	8.3%	2.8%	3.3%	2.7%
Hide or steal someone’s belongings	84.4%	7.2%	3.2%	2.8%	2.4%
Talk about someone behind their back	69.5%	10.3%	6.4%	6.3%	7.4%
Kick or punch	88.1%	4.4%	2.5%	2.5%	2.4%
Call someone names	85.3%	4.5%	3.1%	2.9%	4.2%
Harass on cyberspace	93.9%	3.2%	1.7%	0.5%	0.7%
Attack jokingly	85.7%	5.2%	3.3%	2.2%	3.6%
Sexually harass	96.3%	1.9%	0.6%	0.6%	0.7%

Note. The 8 items above are victimization items and 8 items below are bullying items.

lists the victimization items. All items were rated on a five-point Likert scale (1 = “never,” 2 = “only once,” 3 = “once a month,” 4 = “once a week,” and 5 = “several times a week”).

Bullying. Participants were also asked to report their experiences of perpetrating bullying since the beginning of the school year using eight bullying items from Kato et al.’s scale [37]. The items assessing bullying are listed in Table 1. All items were rated on a five-point Likert scale (1 = “never,” 2 = “only once,” 3 = “once a month,” 4 = “once a week,” and 5 = “several times a week”). There was a clear criterion based on which participants could be identified as either experiencing school bullying or not [31,38]; namely, a student bullying or being bullied by others more than once a week would be considered a bully or victim. Based on this criterion, students who responded “once a week” or “a few times a week” to any of the items were regarded as either victims or perpetrators involved in school bullying.

Coping strategy. Participants who reported being victimized even once were asked to respond to six additional questions separately (“asking perpetrator to stop bullying,” “fighting back,” “seeking help from parents,” “seeking help from teachers,” “seeking help from friends,” and “doing nothing”). We required students to answer each question by rating items on a two-point scale (0 = “no,” 1 = “yes”).

Current victimization status. Those who reported being victimized were asked to answer the following question using a three-point scale (1 = “resolved,” 2 = “somewhat resolved,” and 3 = “never resolved”): “Have your victimization problems been resolved now?”

2.3. Data Analysis

We analyzed the data using HAD ver. 16.05, a Microsoft Excel macro program developed by Shimizu [39]. To explore the relationship between school caste and school bullying, we performed logistic-regression analyses while controlling for sex, grade, and intra-group status simultaneously, in addition to estimating clustered robust standard errors to adjust for dependence between classrooms. We used pairwise deletion to deal with the missing values. Before conducting regression analyses, we checked for multicollinearity and outliers. Based on variance-inflation factors (VIFs) and the results of Grubbs’s tests calculated by R, the regression models indicated no multicollinearity (VIFs < 2.00), and inter- and intra-peer groups did not include outliers (p > 0.05).

3. Results

3.1. Response Rates and Descriptive Statistics

The raw response rate scores for all items pertaining to victimization and bullying are presented in Table 1. Few participants reported being bullied or victimized. Based on the established criterion [31,38], 11.1% (n = 253) of the students were regarded as perpetrators of bullying, and 14.7% (n = 336) were regarded as victims.

For coping strategy and current victimization status, only the participants categorized as victims were included in the analysis. The frequency of the six coping-strategy items was as follows: When students were bullied, 30.2% (n = 106) of them asked the perpetrator to stop the bullying and 24.3% (n = 85) fought back. Few students sought help from parents (19.1%, n = 67), teachers (19.4%, n = 68), and friends (23.9%, n = 84) when or after being bullied. Moreover, 37.6% (n = 118) of the students reported “doing nothing.” In addition, the response rate for current victimization status indicated that victimization had been somewhat resolved overall: “resolved” = 26.8% (n = 88), “somewhat resolved” = 41.5% (n = 136), and “never resolved” = 31.7% (n = 104).

The descriptive statistics for inter-peer group status and intra-peer group status were as follows: M = 3.07 (SD = 1.05) and M = 2.95 (SD = 0.98), respectively. These results are similar to those of a previous study [25].

3.2. Categorization of Participants into Subgroups of School Bullying

The percentage of involvement in bullying was significantly related to being a victim (χ² (1, n = 2248) = 290.53, p < 0.001, φ = 0.36). As mentioned before, four sub-categories relating to distinct experiences of bullying were set to investigate the relationship between inter-peer group status and school bullying more accurately. These were the bully-only category (5.6%, n = 125), the victim-only category (9.0%, n = 202), the uninvolved category (79.8%, n = 1794), comprising individuals who were neither bullies nor victims, and the bully–victim category (5.6%, n = 127), which included both bullies and victims. Most of the participants fell into the uninvolved category.

3.3. Inter-Peer Group Status and Sub-Categories of School Bullying

A multinomial logistic regression analysis was conducted to explore the relationships between inter-peer group status and the four sub-categories of school bullying. The dependent variable was the four sub-categories of school bullying, and the uninvolved category was used as a reference category (

Table 2. Multinomial logistic regression for the four sub-categories of school bullying (n = 2042).

Reference: Uninvolved	Bully Only			Victim Only			Bully–Victim
	OR	95% CI		OR	95% CI		OR	95% CI
Sex a	1.03	0.73	1.43	0.60 ***	0.43	0.84	0.42 ***	0.26	0.66
Grade a	0.98	0.72	1.35	0.66 **	0.54	0.81	0.87	0.66	1.13
Intra-peer group status a	0.94	0.7	1.25	0.91	0.75	1.12	1.04	0.78	1.37
Inter-peer group status	1.02	0.78	1.34	0.77 *	0.63	0.95	0.90	0.69	1.19

Note. OR = odds ratio, CI = confidence interval, ^a = control variable. Grade was coded as a numeric variable as follows: 1 = 7th, 2 = 8th, 3 = 9th. Sex was coded as follows: 0 = boy, 1 = girl. *** p < 0.001; ** p < 0.01; * p < 0.05.

). The results indicated that, while inter-peer group status was negatively associated with the victim-only category (OR = 0.77, p = 0.02, 95% CI [0.63, 0.95]), it was not significantly associated with the bully-only category (OR = 1.02, p = 0.89, 95% CI [0.78, 1.32]) or the bully–victim category (OR = 0.90, p = 0.48, 95% CI [0.69, 1.19]). Intra-peer group status was not significantly associated with any of the three bullying sub-categories. Regarding demographic variables, boys and lower-grade students tended to be victimized, and there were more boys in the bully–victim category than girls.

3.4. Inter-Peer Group Status and Coping Strategy

Six binominal logistic regression analyses were conducted to explore the relationships between inter-peer group status and coping strategy. The six “coping strategy” items were used as dependent variables in each model (

Table 3. Binominal logistic regressions for the six coping strategies (n = 266).

	Asking Perpetrator to Stop Bullying			Fighting Back			Seeking Help from Parents
	OR	95% CI		OR	95% CI		OR	95% CI
Sex a	0.61	0.34	1.09	0.33 **	0.17	0.66	3.72 ***	1.94	7.13
Grade a	0.71	0.50	1.00	1.49 *	1.03	2.15	0.85	0.57	1.27
Intra-peer group status a	0.90	0.70	1.16	1.05	0.76	1.45	0.75	0.53	1.07
Inter-peer group status	1.07	0.82	1.39	1.23	0.92	1.66	1.13	0.83	1.55
	Seeking help from teachers			Seeking help from friends			Doing nothing
	OR	95% CI		OR	95% CI		OR	95% CI
Sex a	1.63	0.81	3.27	3.56 ***	2.08	6.10	0.82	0.51	1.33
Grade a	0.82	0.60	1.13	0.73	0.51	1.07	1.18	0.88	1.59
Intra-peer group status a	1.01	0.68	1.50	1.05	0.75	1.46	0.96	0.72	1.27
Inter-peer group status	0.86	0.59	1.26	0.82	0.60	1.11	1.00	0.77	1.32

Note. OR = odds ratio, CI = confidence interval, ^a = control variable. Grade was coded as a numeric variable as follows: 1 = 7th, 2 = 8th, 3 = 9th. Sex was coded as follows: 0 = boy, 1 = girl. *** p < 0.001; ** p < 0.01; * p < 0.05.

). Inter-peer group status and intra-group status were not significantly associated with any of the reaction categories. Regarding demographic variables, boys and higher-grade students tended to fight back, while girls tended to seek help from parents and friends.

3.5. Inter-Peer Group Status and Current Victimization Status (never Resolved, Resolved, and Somewhat Resolved)

A multinomial logistic regression analysis was conducted to explore the relationships between inter-peer group status and recent victimization status (

Table 4. Multinomial logistic regression for the three current victimization status (n = 277).

Reference: Never Resolved	Resolved			Somewhat Resolved
	OR	95% CI		OR	95% CI
Sex a	0.30 ***	0.03	0.41	0.43 **	0.15	1.81
Grade a	1.84 **	0.15	0.61	1.27	0.25	0.76
Intra-peer group status a	1.24	1.00	1.89	1.20	0.88	1.62
Inter-peer group status	1.38 *	1.18	2.86	1.19	0.89	1.81

Note. OR = odds ratio, CI = confidence interval, ^a = control variable. Grade was coded as a numeric variable as follows: 1 = 7th, 2 = 8th, 3 = 9th. Sex was coded as follows: 0 = boy, 1 = girl. *** p < 0.001 ** p < 0.01 * p < 0.05.

). The dependent variable was the three categories of current victimization status, and the “never resolved” category was used as a reference category. Compared with “never resolved” as the reference category, the logistic regression analysis revealed that inter-peer group status predicted resolution to victimization (OR = 1.38, p = 0.047, 95% CI [1.18, 2.86]). However, intra-peer group status did not significantly predict the current victimization status. Regarding demographic variables, girls and higher-grade students tended to resolve victimization situations.

4. Discussion

4.1. Main Findings

The main purpose of this study was to examine the associations between inter-peer group status and school bullying among Japanese middle-school students. We comprehensively examined the relationship between school caste and school bullying using an explicit criterion to classify students’ involvement in school bullying, and included their reaction to victimization, current victimization status, and intra-group status. The results showed that inter-peer group status was negatively related to unilateral victimization. Further, no significant relationship was observed between inter-peer group status and the remaining categories, and no significant relationship was observed between intra-peer group status and any of the categories. Inter-peer group status was also positively related to fighting back when confronted with aggressive behaviors and successful resolution of the victimization problem, but was not related to any of the remaining reactions. Moreover, no significant relationship was found between intra-peer group status and any of the coping strategies.

Regarding demographic variables in relation with school bullying, coping strategy, and current victimization status, boys tended to experience school bullying and did not seek help. Moreover, lower-grade students were more involved in school bullying than higher-grade students. These results are in line with those of previous studies, which reported that boys and lower-grade students participated more in bullying [40] and sought less help than girls [41].

4.2. Peer-Group Status and School Bullying

Our results indicated that students in lower-status peer groups experienced victimization more often and had a lesser chance of resolving victimization than those in higher-status peer groups, although inter- and intra-group status did not show a significant relationship with coping strategy. Specifically, students with a lower peer-group status experienced more victimization, and their victimization tended to be prolonged. This is consistent with previous studies [34], which found a negative association between perceived popularity and victimization. However, our results demonstrate that students’ inter-peer group status was related to their experiences of victimization and not their own status in their group. Moreover, unlike previous studies on perceived popularity [8], the current study, which used inter- and intra-peer group status as an indicator for status, did not show a significant link with bullying perpetration.

As for the relationship between victimization and lower peer-group status in a given classroom, our study suggests that being affiliated with a lower-status peer group was related to being bullied and prolonged victimization. This is inconsistent with the results of previous studies on perceived popularity. First, intra-peer group status, as an indication of one’s individual status within a group, was not related to bullying perpetration, victimization, coping strategy, or current victimization status. Our results showed that, in Japanese middle schools, the peer-group status of students is a more important predictor of being bullied than their individual status within the peer group. This finding suggests contextual differences in classroom systems and peer relationships in the classroom between Japan and other countries. Second, inter-peer group status was not significantly related to bullying perpetration; whereas prior studies on perceived popularity found a positive relation with bullying perpetration [12], and such behavior was used to attain and maintain a high status [8]. However, our results implied that, in Japanese middle schools, bullying perpetration did not seem to foster or maintain membership in a high-status group in a given classroom. A previous Japanese study by Suzuki [27] found that students in a high-status peer group behaved aggressively toward those in lower-status peer groups. However, students in the high-status group and even those in the lower-status group did not regard their behavior as bullying but instead, considered it a jocular form of communication or simply funny, illustrating the concept of a school caste. This finding suggests that belonging to a lower-status group may predispose students to victimization by a high-status group through aggressive communication. Nonetheless, the study did not reveal the link between group status and engaging in bullying perpetration; thus, it remains unclear which groups of students are more likely to engage in bullying perpetration. This might indicate that students in a lower-status group tend to bully students in a same-status group; whereas all students, including those in the lower- and higher-status groups, are bullies. However, further study needs to confirm these possibilities by exploring who exactly is involved in bullying, for instance.

4.3. Limitations of the Study

This study has some limitations in relation to the methodology adopted. First, prior studies used peer nomination, which is the method of assessing variables of interest based on the number of nominations by others, to assess peer status and bullying behavior [7], while we used a self-report questionnaire. Hence, the data might be influenced by social-desirability bias and participants’ biased appraisal of their own behavior. Although Japanese researchers have used peer nomination extensively before, using this method has been prohibited since the 1990s, due to widespread criticism of the ethical issues associated with many Japanese press organizations and academics (see Mizuno and Tang [42]). Future research could replicate our results by using other novel methods to confirm whether these findings are characteristics specific to the Japanese classroom. Second, our findings were based on a cross-sectional survey; thus, future studies using a longitudinal survey are needed to explore the causal relationships between the variables. Finally, our study did not explore the factors that contribute to a reduction in bullying behavior. Although bullying is a problem behavior, the perpetrators need support, similarly to the victims. Thus, it is important to incorporate this perspective in future research to prevent school bullying.

4.4. Implications of the Study

Although some earlier Japanese studies suggested an association between school caste and school bullying [23,24], no empirical evidence has been found for this hypothesis until now. In addition, although Mizuno et al. [28] analyzed the relationship between school caste and school bullying, some limitations made their study less valid because it did not establish an explicit standard [31] for bullying, nor did it consider the overlap between bullying perpetrators and victims. The current study attempted, for the first time, to explore the association between school caste and school bullying using a four sub-category structure, while controlling for individuals’ status within peer groups. Moreover, the relationship between school caste and school bullying was further investigated by assessing coping strategy and current victimization status. The current study showed that, in the Japanese-classroom context, the status of the group that students belong to may play a more important role in victimization than the student’s individual status within their own group. We believe that these empirical results, which show that students in a lower peer group in their classroom may be at a risk of being bullied, will provide novel insight into the students victimized by school bullying in Japan.