School violence, social support and psychological health among Taiwanese junior high school students

Child Abuse & Neglect 37 (2013) 252–262

Contents lists available at SciVerse ScienceDirect

Child Abuse & Neglect

School violence, social support and psychological health among Taiwanese junior high school students Ji-Kang Chen a,∗ , Hsi-Sheng Wei b a b

Department of Social Work, Chinese University of Hong Kong, Sha Tin, Hong Kong Department of Social Work, National Taipei University, Taiwan

a r t i c l e

i n f o

Article history: Received 23 November 2011 Received in revised form 29 December 2012 Accepted 7 January 2013 Available online 17 February 2013 Keywords: School violence Bullying Social support Taiwan Mental health Indigenous Psychological health Chinese

a b s t r a c t Objectives: This paper examines how peer social support mediates the association between school victimization and student psychological health among junior-high students in an Asian context (Taiwan), and further examines how gender and ethnicity differ in the interrelationships of school violence, peer social support and psychological health. Methods: Data were obtained from a large-scale random sample of 1650 junior-high students (grades 7–9) in one diverse county of Taiwan. Students were given an anonymous structured questionnaire, including items regarding basic demographics and school social experiences. Results: The results of structural equation modeling analysis provided a good fit for the sample as a whole. The final model accounted for 26% of the variance in student psychological health. Overall findings showed that student psychological health is not significantly directly associated with victimization by students and student maltreatment by teachers; however, student psychological health is indirectly associated with victimization by students, mediated through peer social support. Similar findings were found for both male and female and both Han Chinese and Indigenous students. Conclusion: The findings imply that peer social support plays an important mediating role between exposure to school violence and student psychological health. The findings provide empirical evidence and information to help school practitioners and policymakers justify developing or incorporating social support into prevention and intervention strategies. The findings suggest that interventions or policies promoting social support incorporated at a national level could be effective across genders and ethnicities in Taiwan. © 2013 Elsevier Ltd. All rights reserved.

Introduction Over the past few decades, studies on the outcomes of school violence have mainly examined how exposure to school violence directly impacts student psychological health (e.g., Baldry, 2003, 2004; Chen & Wei, 2011b; Flannery, Singer, & Wester, 2004; Gladstone, Parker, & Malhi, 2006; Graham & Bellmore, 2007). Recently, a few empirical studies have shown that peer social support plays an influential mediating role between exposure to school violence and its negative outcomes (Malecki, Demaray, & Davison, 2008; Seeds, Harkness, & Quilty, 2010). However, these studies were conducted in Western countries, using small convenience samples. Little is known about how peer social support mediates outcomes of school violence in Asian cultures, and even less is known about how student gender and ethnicity influence the interrelationships of these factors in Taiwan.

∗ Corresponding author. 0145-2134/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.chiabu.2013.01.001

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

253

Using a large-scale random sample from Taiwan, this study expands the literature on school victimization by examining how peer social support mediates the association between exposure to school violence and student psychological health in an Asian context. This inquiry also examines how gender and ethnicity differ in the interrelationships of school violence, peer social support and psychological health among Taiwanese students. Taiwan is one example of an Asian country in which long-standing cultural values and outside influences must be considered simultaneously when examining these interrelationships. It has experienced democratization, rapid industrialization and urbanization, a growing population of immigrants from other Asian countries, and an expanding economy over the past four decades (e.g., Chen & Astor, 2010, 2012). Yet Taiwanese are still primarily guided by values stemming from a blend of Confucianism, Taoism, and Buddhism. Recently, a handful of published studies have indicated that school violence is severe in Taiwan. For example, a report indicated 26.9% of Taiwanese students were exposed to physical, emotional and sexual types of maltreatment by teachers, even though teacher aggression against students is strictly prohibited and banned in Taiwan (Chen & Wei, 2011a, 2011b). Other national data indicate that 59.7% of Taiwanese students have been involved in school peer violence (e.g., Chen & Astor, 2009a, 2009b). These studies suggested immediate attention should be paid to design an effective intervention to alleviate school violence (e.g., Chen & Astor, 2009a, 2009b; Chen & Wei, 2011a, 2011b). Definition School violence in this study is defined as any behavior intended to harm students psychologically or physically while in school (Benbenishty & Astor, 2005). The term “violence” includes physical violence (e.g., beating), verbal and social violence (e.g., cursing and social exclusion), and threatening behavior (e.g., extortion) (see Benbenishty & Astor, 2005 for a critical discussion). Based on this definition, the two most common types of school victimization discussed in current literature are: student victimization by students (e.g., Benbenishty & Astor, 2005; Chen & Wei, 2011b) and student maltreatment by teachers (e.g., Chen & Wei, 2011a, 2011b). This study examines these two types of school victimization. Social support has been defined as assistance, help or information leading individuals to believe they are loved, cared for and valued and that they share mutual obligation with members of a network (Cobb, 1976; Cohen & Wills, 1985; Seeman, 1996). It includes emotional support (e.g., trust and empathy), instrumental support (e.g., time and money), appraisal support (e.g., evaluative feedback) and informational support (e.g., knowledge and skills) (Malecki & Demaray, 2003; Tardy, 1985). In addition, social support can be categorized into actual use of social support and perception of social support (Malecki & Demaray, 2002). Studies have indicated that perception of peer support has greater influence on adolescent mental health than does actual received social support, whether from peers or from other sources, such as parents and school adults (Papalia, Olds, & Feldman, 2009). Thus, this study examines students’ perception of peer social support. Literature review Theories related to trauma or violence posit that exposure to school violence can lead to student psychological distress, such as depression and anxiety (Baldry, 2003, 2004; Chen & Wei, 2011b; Flannery et al., 2004; Gladstone et al., 2006; Graham & Bellmore, 2007). However, why and how school victimization leads to these negative psychological outcomes is less well explained (Malecki et al., 2008; Seeds et al., 2010). Furthermore, existing empirical studies on the link between school victimization and distress are mixed; some indicate a strong association (e.g., Baldry, 2003, 2004), while others show a weak or insignificant relationship (e.g., Chen & Wei, 2011b). These findings raise the question of whether the levels of the psychological outcomes of school violence depend on how certain psychosocial mechanisms react to violence exposure (Chen & Wei, 2011b). The literature suggests that social support can be a potential psychosocial mechanism in the link between exposure to school violence and poor student mental health (e.g., Cowie & Olafsson, 2000; Kochenderfer-Ladd & Skinner, 2002; Naylor & Cowie, 1999; Williams, Connolly, Pepler, & Craig, 2005). Two models have been widely examined in the literature related to this topic: the stress buffer model and the main effect model (e.g., Cohen & Wills, 1985; Davison and Demaray, 2007; Demaray & Malecki, 2003; Kaniasty & Norris, 1993; Prelow, Mosher, & Bowman, 2006). The stress buffer model examines how social support moderates the negative outcomes of exposure to school violence, hypothesizing that high level of support can reduce negative outcomes. By contrast, the main effect model examines how high levels of social support promote mental health, regardless of the level of victimization. This model suggests that all students can benefit from social support even if they are not experiencing school violence. However, the empirical findings about these models appear contradictory; some support the stress buffer model, some the main effect model, and others neither (e.g., Chen & Wei, 2011b; Cohen & Wills, 1985; Cowie & Olafsson, 2000; Davison and Demaray, 2007; Kaniasty & Norris, 1993; Kochenderfer-Ladd & Skinner, 2002; Naylor & Cowie, 1999; Williams et al., 2005). Social support as mediator Alternatively, the social support deterioration model theorizes that trauma and stressful events can both directly and indirectly, through perceived social support, influence psychological well-being (Kaniasty & Norris, 1993; Norris & Kaniasty, 1996; Prati & Pietrantoni, 2010; Prelow et al., 2006; Punamaki, Komproe, Qouta, El-Masri, & de Jong, 2005). Specifically,

254

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

the model emphasizes that stressors or trauma events can involve dramatic change in available social support sources and potentially diminish the perception of social support, which then leads to poor psychological outcomes (Berrera, 1986). School violence has been recognized as a major school stressor, a type of trauma that challenges students’ mental health (Graham & Bellmore, 2007). Some researchers argue that once students experience school violence perpetrated by school peers and/or teachers, their sense of insecurity about interpersonal relationships and distrust of available resources may lead to perceptions of less peer social support. Subsequently, the risk of suffering psychological distress is greatly increased (Malecki et al., 2008; Seeds et al., 2010). To date, the social support deterioration model has been widely applied to explain psychological distress among people who experience natural disaster, family violence, and racial discrimination (e.g., Kaniasty & Norris, 1993; Norris & Kaniasty, 1996; Prati & Pietrantoni, 2010; Prelow et al., 2006; Punamaki et al., 2005). However, relatively few studies have employed this model to examine school victimization, and most of those studies used small convenience samples from Western countries (Malecki et al., 2008; Seeds et al., 2010). There is a lack of empirical data showing how this theoretical model applies to the outcomes of school violence in Chinese/Asian contexts. In order to fill the research gap, the current study examines how the social support deterioration model is relevant to the context of school violence in Taiwan. It is beneficial to theory, policy and practice to examine this kind of topic in Chinese/Asian societies, such as Taiwan, since culture is an important factor affecting perception of social support (Kim, Sherman, & Taylor, 2008). For example, Asian culture has been categorized as collectivistic, with individual interests subordinate to group norms (Chen, 2000; Triandis, 1995). Once students are exposed to school violence, they may perceive that seeking social support would burden their social networks, cause personal problems for others and disrupt group harmony (Kim et al., 2008). This may erode their perception of social support and, in turn, negatively impact their psychological health (Kim et al., 2008). If this is true, school practitioners and researchers will have empirical evidence to justify increased use of social support as a core component of school violence intervention, regardless of culture or country (Demaray & Malecki, 2006). Theoretical model for this study Based on previous literature and theory, this study proposes that students exposed to school violence (i.e., student victimization by students and student maltreatment by teachers) report lower levels of perception of peer social support and psychological health. Most importantly, this study proposes that perception of peer social support mediates the effects of exposure to school violence on student psychological health. Previous studies indicate that girls experience less school victimization, perceive more social support, but report worse psychological well-being than do boys (Chen & Wei, 2011a, 2011b; Cumsille & Epstein, 1994; Guay, Billete, & Marchand, 2006; Kendler, Myers, & Prescott, 2005; Norris, Foster, & Weisshaar, 2002; Reevy & Maslach, 2001; Sharir, Tanasescu, Turbow, & Maman, 2007). It has thus been argued that the interrelationships of school violence, social support and psychological distress may differ with gender (Guay et al., 2006). Furthermore, there are two major ethnic groups in Taiwan: Han Chinese and Indigenous (Cheng & Jacob, 2008; Chuang & Huang, 2007; Yeh & Chiang, 2005). Like their counterparts in Australia, New Zealand and United States, Indigenous peoples in Taiwan are a minority suffering from discrimination and oppression, and possessing relatively few educational, economic, political and social resources (Cheng & Jacob, 2008; Chuang & Huang, 2007; Yeh & Chiang, 2005). This inequality may influence students’ level of exposure to school violence, perception of social support, and mental health, which raises the question of how these interrelationships differ with ethnicity (Chen & Wei, 2011b; Cheng & Jacob, 2008; Chuang & Huang, 2007; Yeh & Chiang, 2005). The current study examines the theoretical model, and how gender and ethnicity differences impact the relationships of exposure to school violence, perception of peer social support and psychological health. The current study expects the model to fit both male and female as well as both Han Chinese and Indigenous students, but certain paths within the model may differ. Methods The data used in this study were based on a large-scale project entitled “Yilan County Youth Life Conditions Survey” (Wei, Chen, & Chen, 2009). Yilan, with a diverse population of about 0.5 million, is located in northeastern Taiwan and is one of the regions with a large number of Indigenous residents (Wei et al., 2009). The survey was conducted throughout Yilan County among more than 1600 junior-high students (grades 7–9), with a response rate of around 90%. The students were given a structured questionnaire in classrooms under the guidance of professionally trained survey monitors. The questionnaire included over 150 items regarding basic demographics and personal, family and school experiences, and took approximately 30 min to complete. Respondents were assured of anonymity, were encouraged to respond truthfully and were free to withdraw from the study at any time and for any reason. This study was supported by the Yilan County Government, whose internal review board approved and supervised the questionnaires, procedures, informed consent forms and ethical concerns (Wei et al., 2009). The sample was designed to represent all Yilan County 7th–9th-grade students. The probability sampling method was a multi-stage cluster random sample, and the strata were the twelve geographical districts of Yilan County. In the first stage, schools were randomly selected from the sampling frame according to the appropriate strata. In the next stage, two classes

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

255

were randomly selected from each grade in the selected schools. The sample included all students in the selected classes (Wei et al., 2009). This study examined 1650 students from 12 junior high schools. Around half the students were boys (51.4%), 48.2% were girls, and 0.4% did not indicate gender. Over three-fourths of students were Han Chinese (78.8%), 17.5% were Indigenous, and 3.7% did not indicate ethnicity. The grade level distribution was: 34.7%, 7th grade; 29.8%, 8th grade; and 35.5%, 9th grade. Measurement Demographics. The student demographic variables (e.g., gender and grade level) were self-reported. Student maltreatment by teachers (˛ = .70). This latent variable involved three items asking students how many times they had experienced maltreatment by teachers in school during this semester. The three items in this scale were selected and adapted from a version of the California School Climate and Safety Survey (CSCSS) (e.g., Benbenishty & Astor, 2005; Chen & Wei, 2011a, 2011b; Furlong et al., 2005; Wei & Chen, 2012; Wei, Williams, Chen, & Chang, 2010). The original items in the CSCSS were translated from English to Chinese to form a standardized version, and a standard back-translation procedure was employed in order to ensure accuracy. A five-point Likert scale was provided (1 = never, 2 = 1–2 times, 3 = 3–4 times, 4 = 5–6 times and 5 = 7 times or more). Previous studies showed that the CSCSS demonstrated high validity and reliability (Chen & Wei, 2011a, 2011b; Furlong et al., 2003, 2005). The three items were: “Teachers hit, beat, slapped or kicked you”; “Teachers cursed you”; and “Teachers mocked, insulted or humiliated you”. The score for this variable was the sum of the ratings for these three items. School victimization by students (˛ = .76). This latent variable asked students if they had experienced victimization by their peers in school in this semester. The five items in this scale were selected from a version of the CSCSS (e.g., Benbenishty & Astor, 2005; Chen & Wei, 2011a, 2011b; Furlong et al., 2005; Wei & Chen, 2012; Wei et al., 2010). The original items in the adapted version of the CSCSS were translated from English to Chinese to form a standardized version, and a standard back-translation procedure was employed in order to ensure accuracy. A five-point Likert scale was provided (1 = never, 2 = 1–2 times, 3 = 3–4 times, 4 = 5–6 times and 5 = 7 times or more). These items related to being hit, beaten or kicked; being cursed; being socially excluded; and being threatened or having rumors spread. The score for this variable was the sum of the ratings for these five items. Peer social support (˛ = .90). This variable involved twelve items asking students about their perception of social support from school peers. These items were selected from a Chinese version of the Children and Adolescent Social Support Scale (CASSS) (Chen, 2005; Malecki & Demaray, 2002; Malecki, Demaray, & Elliott, 2000). CASSS has been used to measure 3rd–12th-grade students’ perception of social support (Malecki & Demaray, 2002; Malecki et al., 2000). The Chinese version of CASSS has been shown to have high reliability and good factorial validity (Chen, 2005; Malecki & Demaray, 2002; Malecki et al., 2000). The results of exploratory factor analysis suggested the scale was unidimensional. The twelve items were randomly parceled into three indicators (i.e., Support 1, Support 2 and Support 3) in order to build up robust latent structures (Alhija & Wisenbaker, 2006; Little, Lindenberger, & Nesselroade, 1999; Marsh, Hau, Balla, & Grayson, 1998), and the score of each indicator was the sum of the items included. Support 1 included four items related to peers’ friendly treatment and peer assistance (˛ = .82, factor loading = .91). Support 2 included four items related to peers’ sharing, suggestion and advice, and activity participation and invitation (˛ = .84, factor loading = .92). Support 3 included four items related to peers’ appraisal and appreciation, and spending time together (˛ = .86, factor loading = .90). Psychological health (˛ = .84). This latent variable involved six items asking students about their general mental health situation. These six items in this scale were selected from the 12-item Chinese version of the General Health Questionnaire-12 (GHQ-12) (e.g., Chan, 1993; Ye, 2009). The GHQ-12 has been used to measure general psychological health in many cultures and countries (e.g., Chan, 1993; Ye, 2009), and the Chinese version of GHQ has been shown to have high reliability and good factorial validity (Ye, 2009). Some studies have identified the GHQ-12 as comprising two meaningful dimensions (Ye, 2009), which are mostly defined by negatively worded versus positively worded items. In this study, positively worded items were used because previous studies suggested negatively worded effects in the Chinese version of GHQ-12 (e.g., Ye, 2009). These six items included, for example, “being able to concentrate on what you are doing”, “being able to enjoy your normal day-today activities” and “been feeling reasonably happy, all things considered”. A five-point Likert scale was provided (1 = never, 2 = seldom, 3 = sometimes, 4 = often and 5 = all the time). The score for this variable was the sum of the ratings for these six items. Analysis plan The primary analysis method in this study is latent-variable structural equations modeling (SEM) with maximum likelihood (ML) estimation using the AMOS program. SEM is a statistical methodology that takes a confirmatory approach to the analysis of structural theory, describing relationships among endogenous factors (Bentler, 1988). The structure/hypothesized model can be tested statistically in a simultaneous analysis of an entire system of variables to determine the extent to which

256

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

Table 1 Means and standard deviations of each scale by gender and by ethnicity. Overall

Victimization by studentsa Maltreatment by teachersa School peer social supportb Psychological healthb

Gender

6.73 (2.86) 3.80 (1.75) 41.63 (10.73) 19.32 (5.01)

Ethnicity

Male

Female

Han

Indigenous

7.10 (3.20) 4.04 (1.98) 39.64 (11.00) 18.88 (5.33)

6.34 (2.39) 3.53 (1.36) 43.75 (10.05) 19.80 (4.62)

6.73 (2.84) 3.82 (1.80) 41.84 (10.77) 19.39 (5.04)

6.61 (2.88) 3.67 (1.51) 40.96 (10.32) 19.15 (4.90)

Note: Standard deviations are in parentheses. a Scale: 1 = never, 2 = 1–2 times, 3 = 3–4 times, 4 = 5–6 times and 5 = 7 times or more. b Scale: from 1 = never to 5 = all the time.

it is consistent with the data. Confirmative factor analysis (CFAs) was first conducted to ensure the measurement model was a good fit. Following the CFAs, the final SEM model, with the full dataset of all junior-high students, was tested. Next, comparative analyses were conducted in order to determine whether patterns of relationships and mediating effects were different between male and female students and between Han Chinese and Indigenous students. In this comparative analysis, all factor loadings, paths, and covariances were constrained to be equal simultaneously to the same model across ethnicities and genders. Then, the model was tested by releasing path constraints one at a time in order to find out if releasing equality constraints would significantly improve the fit. There are various indicators of the goodness of fit for a specific model. The chi-square coefficient is used to assess the size of discrepancies between the relationships in the original data matrix from those implied by the model. A low chisquare measure reveals non-significant discrepancies, meaning that the data “fit” the theoretical model. However, due to the sensitivity of the chi-square coefficient to sample size, it is not preferred for large samples such as the one in this study. Researchers have addressed the chi-square limitation by developing goodness-of-fit indices that take a more pragmatic approach to the evaluation process. More commonly used fit indices includes Bentler and Bonnett’s (1980) Normed Fit Index (NFI), Bollen’s (1989) Incremental Fit Index (IFI), and Bentler’s (1990) Compared Fit Index (CFI). Typically, these three fit indices consider a model to be a good fit when the value is above .90 (Bentler, 1992), and a superior fit when it is close to .95 (Hu & Bentler, 1999). A common misfit measure, the Root Mean Square Error (RMSEA), was also reported in the SEM analysis. The RMSEA considers a mediocre fit to range from .08 to .10, and a good fit to be below .06 (Hu & Bentler, 1999). Due to the large sample size of this study, statistical significance was set at ˛ < .01. Results Descriptive statistics Table 1 shows descriptive statistics (mean and standard deviation) broken down by gender and ethnicity for the study variables. The correlations between all variables in this study are shown in Table 2. The results show that ethnicity is not significant correlated with any variables in this study (r = −.02 to −.03, p > .01). T-tests were further conducted in order to investigate possible differences on the main variables based on ethnicities. No significant differences between Han Chinese and Indigenous students were found in victimization by students [t(1557) = .656, p > .05], maltreatment by teachers [t(1564) = 1.213, p > .05], school peer social support[t(1524) = 1.233, p > .05], and psychological health [t(1503) = .697, p > .05]. The two types of school violence (student victimization by students and student maltreatment by teachers) were positively correlated (r = .39, p < .01), and both of them were marginally correlated with peer social support (for victimization by students: r = −.18, p < .001; for maltreatment by teachers: r = −.07, p < .01) and with psychological health (for victimization by students: r = −.13, p < .001; for maltreatment by teachers: r = −.10, p < .01). Peer social support and psychological health were positively correlated (r = .45, p < .01). Table 2 Intercorrelations between variables in this study.

1. Gender 2. Ethnicity 3. Student victimization by students 4. Student maltreatment by teachers 5. School peer social support 6. Psychological health

1

2

–

.05 –

Note: Gender is coded 0 = male, 1 = female; ethnicity is coded 0 = Han, 1 = Indigenous. * p < .01

3 −.13* −.02 –

4 −.15* −.03 .39* –

5 .19* −.03 −.18* −.07* –

6 .09* −.02 −.13* −.10* .45* –

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

257

Victimization by Students

-.25*

-.01 School Peer Social Support R2= .05

.52*

.08

.50*

Psychological Health R2= .26

-.08

Maltreatment by Teachers

Fig. 1. Structural equation modeling of direct and mediating effects on junior-high students’ psychological health.

Overall model The results of the analysis, based on the total sample, provided a good fit to the data [2 (113, N = 1650) = 629.38, p < .001, and with NFI = .95, IFI = .96, CFI = .96, and RMSEA = .053]. This suggested that the model is a good one. Fig. 1 illustrates the paths in this model. Fig. 1 shows that student psychological health is not directly associated with victimization by students (ˇ = −.01). However, psychological health is indirectly associated with victimization by students, mediated through perception of peer social support. Contrary to our prediction, student maltreatment by teachers has an insignificant association with psychological health (ˇ = −.08), and a weakly indirect association mediated through peer support. Overall, perception of peer social support was the best predictor of student psychological health (ˇ = .50). Taken together, all of these variables accounted for approximately one-fourth of the explained variance for student psychological health (R2 = .26). Gender-comparison analysis In this analysis, factor loadings, paths, and covariances were constrained to be equal, in order to fit the covariance matrices of the male and female subgroups simultaneously to the same model. The analysis provided a good fit to the data [2 (245, N: males = 848, females = 795) = 921.705, p < .001 and with NFI = .92, IFI = .94, CFI = .94, and RMSEA = .04]. Next, the model was tested to determine if releasing equality constraints on the paths could significantly improve the fit. After releasing path constraints one at a time, it was determined that the release of the constraints between victimization by students and perception of peer social support, between student maltreatment by teachers and perception of peer social support, as well as between perception of peer social support and psychological health, yielded a significantly better fit. Thus, the results for the final model with the three constraints released together were [2 (242, N: males = 848, females = 795) = 897.88, p < .001 and with NFI = .92, IFI = .94, CFI = .94, and RMSEA = .04]. Fig. 2 represents the results of this analysis. Differences between genders existed on several paths. The association of victimization by students and student maltreatment by teachers with perception of peer social support was stronger for females (ˇ = −.43 and ˇ = .22, respectively) than for males (ˇ = −.13 and ˇ = .06, respectively). For both genders, perception of peer social support had the strongest association with psychological health; however, the association was stronger for females (ˇ = .57) than for males (ˇ = .44). Finally, the overall model explained 20% of the variance in student psychological health for males (R2 = .20) and 34% for females (R2 = .34). Ethnicity analysis The ethnicity analysis was conducted in similar fashion. First, factor loadings, paths, and covariances were constrained to be equal, in order to fit the covariance matrices of the Han Chinese and Indigenous subgroups simultaneously to the same model. The analysis provided a good fit to the data [2 (245, N: Han = 1031, Indigenous = 288) = 876.84, p < .001 and with NFI = .93, IFI = .95, CFI = .95, and RMSEA = .04].

258

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

Victimization by Students -.13* -.43*

School Peer Social Support R2= .01 .12

.40* .52*

-.01 -.01 .44* .57*

Psychological Health R2= .20 .34

-.08 -.08

.06 .22*

Maltreatment by Teachers

Fig. 2. Structural equation modeling of direct and mediating effects on male and female junior-high students’ psychological health. The coefficients in regular print and those in bold italics, represent, respectively, the results for the male and the female samples.

Next, path constraints were released, one at a time. It was determined that releasing the constraint between victimization by students and perception of peer social support yielded a significantly better fit to the model. Thus, the results for the final model, with the one constraint released, were [2 (244, N: Han = 1031, Indigenous = 288) = 868.108, p < .001 and with NFI = .93, IFI = .95, CFI = .95, and RMSEA = .04]. Fig. 3 represents the results of the ethnicity analysis. Fig. 3 also shows that the beta coefficients for both the Han Chinese and Indigenous models were remarkably similar. However, the association between victimization by students and perception of peer social support was stronger for Han Chinese students (ˇ = −.28) than for Indigenous students (ˇ = −.08). The overall model explained a similar proportion of variance on student psychological health (for Han, R2 = .26; for Indigenous, R2 = .22) Discussion Using a large-scale probability sample in Taiwan, this study examined how student psychological health is associated with exposure to school violence and with peer social support in an Asian junior-high school setting. This study proposed that student psychological health is directly associated with exposure to school violence (i.e., student victimization by students

Victimization by Students -.28* -.08

School Peer .50* .55*

-.02 -.02 .50* .46*

.08 .08

Psychological Health R2= .26 .22

Social Support R2= .06 .01 -.09* -.08

Maltreatment by Teachers

Fig. 3. Structural equation modeling of direct and mediating effects on Han Chinese and Indigenous junior-high students’ psychological health. The coefficients in regular print and those in bold italics print, represent, respectively, the results for the Han Chinese and the Indigenous samples.

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

259

and student maltreatment by teachers) as well as indirectly mediated through perception of peer social support. This study expected the model to fit male and female students as well as Han Chinese and Indigenous students, with certain paths in the model differing. The findings of this study provide empirical support for the proposed theoretical model. Overall model The results of this study show good indices of model fit. This suggested that the theoretical model of how the interrelationships of exposure to school violence perpetrated by students and teachers and perception of peer support associated with student psychological health is supported in an Asian context. Furthermore, the variables in this model, particularly the variable of school peer social support, account for a relatively large amount of the explained variance in student psychological health. This suggested that school peer social support in this model plays an important role in explaining psychological health among junior-high students in Taiwan. The overall results of this study indicate that the social support deterioration model is relevant to school violence in an Asian culture, especially in Taiwan. The results show that student psychological health is not directly associated with victimization by students and student maltreatment by teachers. The findings are consistent with large-scale empirical studies on school victimization conducted in Taiwan or other countries (e.g., Chen & Wei, 2011b; Malecki et al., 2008), but conflict with theories and most Western studies indicating strong direct associations between school victimization and student psychological health (e.g., Baldry, 2003, 2004; Chen & Wei, 2011a; Flannery et al., 2004; Gladstone et al., 2006; Graham & Bellmore, 2007). The findings may imply that the direct impact of exposure to school violence on student psychological health in Taiwan is weaker than in most Western countries (Chen & Wei, 2011b). It is perhaps that Taiwanese students are taught from childhood the philosophies of Confucianism, Taoism and Buddhism, which involves a series of teachings to increase students’ tolerance and forgiveness in the face of unexpected disaster, trauma or victimization, in order to enable students to cope with and stop further development of negative psychological outcomes (Nisbett, 2003; Tiwari, Wong, & Ip, 2001). Indeed, some empirical studies have shown that Chinese people have a greater ability than Western people to moderate and control their emotions when facing conflict or trauma (e.g., Tsai & Levenson, 1997), which may explain the direct association of exposure to school violence with psychological health being weaker or insignificant in Taiwan, in comparison to Western cultures. However, the results of this study show that student psychological health is indirectly associated with victimization by students, mediated through perception of peer social support. The findings support our proposition that social support can be a psychosocial mechanism in the relationships between school violence and its negative outcomes. Specifically, consistent with previous studies, the findings suggest that the perception of peer social support can play an important mediating role in the relationships between school victimization and psychological distress (Malecki et al., 2008; Seeds et al., 2010). That is, students in Taiwan who experience school violence perpetrated by their school peers are more likely to perceive less available peer social support; in turn, their risk of suffering from psychological distress is increased (Malecki et al., 2008; Seeds et al., 2010). The findings also imply that it is not the exposure to school violence itself that directly contributes to psychological distress among Taiwanese students; rather, it may be that perceptions of these student victims about how their available social support reacts to violence, which in turn influences level of psychological distress. The findings provide empirical evidence that school peer social support is a unique factor contributing directly and indirectly to outcomes of school violence in Taiwanese junior-high schools (Chen & Astor, 2010, 2011a, 2011b, 2012; Chen & Wei, 2011b). The results of this study show that student maltreatment by teachers is a weak predictor of perception of peer social support. These findings were unexpected, because previous studies and theories suggested that child maltreatment by adults should be a strong predictor of poor perception of social support (e.g., Pepin & Banyard, 2006). Possible explanations for the inconsistent findings may be that Taiwanese students tend to perceive maltreatment by teachers as a proper disciplinary method to correct student misbehavior, which leads these students to perceive that teachers’ maltreatment does not affect their peer relationships or resources (Chen & Wei, 2011a, 2011b). Another possible explanation is that maltreatment by teachers often takes place in a context of student-teacher interaction rather than a peer context (Chen & Wei, 2011a, 2011b). Thus, students may be more likely to associate maltreatment by teachers with perception of social support from teachers or adults than with support from peers. Future studies may consider further examining the role of teacher support in the model to confirm this proposition. Gender differences The results show that the overall theoretical model is applicable across both genders. These findings suggest that how males’ and females’ exposure to school violence and perception of peer social support affect mental health are quite similar. It is interesting to note that, among female students, student victimization by students had a stronger association with perception of peer social support. The link between perception of peer social support and psychological health was also stronger for females. These findings suggest that perception of peer social support had a stronger mediating effect between victimization by students and psychological health for female students. That is, compared to male students, female victims are more likely to perceive less peer social support; in turn, the risk of poorer psychological health is increased. These findings are consistent with previous studies suggesting that girls’ psychological health is more related to social support than is boys’, and boys’ perception of social support is less likely to be influenced by traumatic experiences than is girls’

260

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

(Cumsille & Epstein, 1994; Guay et al., 2006; Kendler et al., 2005; Norris et al., 2002; Reevy & Maslach, 2001; Sharir et al., 2007). It is interesting to note that the association between student maltreatment by teachers and perception of peer support is significantly positive for females but insignificant for males. That means female students experiencing maltreatment by teachers are more likely to perceive more social support from peers, which promotes psychological health. However, those findings are not relevant to male students. It is possible that boys are more likely than girls to misbehave, and that teacher maltreatment is perceived in Taiwan as a common and acceptable teaching practice to correct male misbehavior (Chen & Wei, 2011b). When girls experience maltreatment by teachers, female victims and their peers may be more likely than males to perceive it as inappropriate, which attracts more peer concern and care to support female victims. Ethnicity The results of this study show that the overall theoretical model is applicable across both ethnicities. For Han Chinese and Indigenous students, the perception of peer social support had similar mediating effects on outcomes of school violence. These findings imply that the mediating mechanism of social support is similar between ethnicities. These findings suggest that ethnicity in Taiwan does not account for the different interrelationships of exposure to school violence, social support and psychological mental health (e.g., Chen & Wei, 2011b; Cheng & Jacob, 2008). However, it is interesting to note that the link between victimization by students and perception of peer social support is weaker for Indigenous students. These findings indicate that Han Chinese victims are more likely than Indigenous ones to perceive themselves as having less social support resources available. These findings do not support the proposition that inequality in social resources accounts for the minority’s weaker perception of social support when they encounter traumatic events or life stressors (Cheng & Jacob, 2008; Chuang & Huang, 2007; Davison and Demaray, 2007; Yeh & Chiang, 2005). It is possible that the Indigenous students’ positive attitudes about dealing with interpersonal conflict and relationships helped them reduce the negative impact of school violence on perception of peer support (Shih, 2011). Limitations It is important to acknowledge some limitations of this study. The analysis is based on cross-sectional and self-reported data, which restricts the establishment of causality and increases the possibility of students’ under-or-over reporting victimization (Benbenishty & Astor, 2005). Further longitudinal research is required to examine the cause-effect relationships of exposure to school violence, perception of social support and student mental health. Future researchers should consider collecting data from multiple sources, such as teacher or peer reports, in order to avoid under- or over-reporting (Benbenishty & Astor, 2005). The present study has examined peer social support. Other important sources of social support, such as parents and teachers, were not assessed in this study. Future researchers should consider examining if these sources of social support play a mediating role between school victimization and outcomes. Future research may also examine how these sources of social support differentiate the mediation mechanism (Seeds et al., 2010). The results of that examination may provide fruitful information to school practitioners in determining which source(s) of social support should be the focus of intervention. Finally, this study is based on a random sample of junior-high schools in one diverse county in Taiwan. Only with caution can the results of this study be generalized to younger or older students, those in other counties or areas, or those in other Chinese or Asian societies. Furthermore, the Taiwanese Indigenous peoples include fourteen tribes, each of which has its own ancestors, history, culture and language (Taiwan Council of Indigenous Peoples Executive Yuan, 2008). The sample of the current study was drawn from Yilan County, where the Ataya tribe is the most numerous minority. Future researchers may consider examining how the theoretical model applies to other tribes, age groups, areas and countries to confirm similarities or differences in school-violence findings. Implications Perhaps due to a lack of empirical evidence, social support has been recognized as potentially effective element to reduce violence in school, but it has not yet been comprehensively incorporated in current school-violence prevention or intervention programs (Malecki et al., 2008). Our findings provide strong evidence and information for school practitioners and policymakers to justify developing or incorporating social support into prevention and intervention strategies. Overall, our findings imply that social support plays an important role in influencing the level of negative outcomes of school violence. This indicates that school-violence interventions focusing on increasing social support for victims may be effective. In particular, increasing student perception of available school peer support may be a focus of cognitive and behavioral intervention for school victimization in Taiwan’s junior-high schools (Seeds et al., 2010). In addition, school practitioners may consider providing a series of training sessions to help students learn social supportive behavior (i.e., in emotional, informational, appraisal and instrumental dimensions) (Malecki et al., 2008; Tardy, 1985) and build up supportive school peer networks when school violence occurs, which may be able to help students overcome victimization and reduce the further risk of mental health problems (Malecki et al., 2008). School practitioners may also consider forming self-help groups for victims. In such groups, victims may share their victimization experiences and provide mutual support and

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

261

help (Demaray & Malecki, 2006; Newman, Horne, & Bartolomucci, 2000). The professional-expert model is recommended for these self-help groups because previous theories and studies have suggested that school victims are more likely to be aggressive through poor quality of peers, without appropriate monitoring (Astor, Meyer, Benbenishty, Marachi, & Rosemond, 2005; Chen & Astor, 2010, 2011a, 2011b, 2012). Finally, the findings show that the overall theoretical model is applicable across both genders and ethnicities. The findings imply that national-level interventions or policies promoting social support could be effective across genders and ethnicities in Taiwan. In addition, our findings suggest that peer social support has a stronger mediating effect for female students, implying that interventions aimed at increasing victims’ social support from peers would reduce negative outcomes of school violence and would be more likely to help female students.

References Alhija, F. N., & Wisenbaker, J. (2006). A Monte Carlo study investigating the impact of item parceling strategies on parameter estimates and their standard errors in CFA. Structural Equation Modeling: A Multidisciplinary Journal, 13(2), 204–228. Astor, R. A., Meyer, H. A., Benbenishty, R., Marachi, R., & Rosemond, M. (2005). School safety interventions: Best practices and programs. Children and Schools, 24(1), 17–32. Baldry, A. C. (2003). Bullying in schools and exposure to domestic violence. Child Abuse & Neglect, 27(7), 713–732. Baldry, A. C. (2004). The impact of direct and indirect bullying on the mental and physical health of Italian youngsters. Aggressive Behavior, 30(5), 343–355. Benbenishty, R., & Astor, R. A. (2005). School violence in context: Culture, neighborhood, family school, and gender. New York: Oxford University Press. Bentler, P. M. (1988). Casual modeling via structural equation systems. In J. R. Nesselroad, & R. B. Cattell (Eds.), Handbook of multivariate experimental psychology (2nd ed., pp. 317–335). New York: Plenum. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246. Bentler, P. M. (1992). On the fit of models to covariances and methodology to the Bulletin. Psychological Bulletin, 112, 400–404. Bentler, P. M., & Bonnett, D. G. (1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin, 88, 588–606. Berrera, M., Jr. (1986). Distinction between social support concepts, measures, and models. American Journal of Community Psychology, 14, 413–445. Bollen, K. A. (1989). A new incremental fit index for general structural models. Sociological Methods & Research, 17, 303–316. Chan, D. W. (1993). The Chinese General Health Questionnaire in a psychiatric setting: The development of Chinese scaled version. Social Psychiatry and Psychiatric Epidemiology, 28(3), 123–129. Chen, X. (2000). Growing up in a collectivistic culture: Socialization and socioemotional development in Chinese children. In A. L. Comunian, & U. P. Gielen (Eds.), Human development in cross-cultural perspective (pp. 331–353). Lengerich, Germany: Pabst Science. Chen, M.J. (2005). Xin tai wan zhi zi de she hui zhi chi, zi zun yu sheng huo man yi du zhi yan jiu. [The relationships among social support, self-esteem and life satisfaction in New Taiwanese Children]. Unpublished master’s thesis, Da-Yeh University, Changhua, Taiwan. Chen, J. K., & Astor, R. A. (2009a). The perpetration of school violence in Taiwan: An analysis of gender, grade level, school type. School Psychology International, 30(6), 568–584. Chen, J. K., & Astor, R. A. (2009b). Students’ reports of violence against teachers in Taiwanese schools. Journal of School Violence, 8(1), 2–17. Chen, J. K., & Astor, R. A. (2010). School violence perpetration in Taiwan: Examining how western risk factors predict school violence in an Asian culture. Journal of Interpersonal Violence, 25(8), 1388–1410. Chen, J. K., & Astor, R. A. (2011a). The influence of students’ personal traits, victimization, family factors, and school dynamics on the perpetration of school violence in Taiwanese elementary schools. Health Education Research, 26(1), 150–166. Chen, J. K., & Astor, R. A. (2011b). School engagement, risky peer, and student–teacher relationships as mediators of school violence in Taiwanese vocational vs. academically oriented high schools. Journal of Community Psychology, 39(1), 10–30. Chen, J. K., & Astor, R. A. (2012). School variables as mediators of the effect of personal and family factors on school violence in Taiwanese junior high schools. Youth & Society, 44(2), 175–200. Chen, J. K., & Wei, H. S. (2011a). Student victimization by teachers in Taiwan: Prevalence and associations. Child Abuse & Neglect, 35, 382–390. Chen, J. K., & Wei, H. S. (2011b). The impact of school violence on self-esteem and depression among Taiwanese junior high school students. Social Indicator Research, 100(3), 479–498. Cheng, S. Y., & Jacob, W. J. (2008). American Indian and Taiwan aboriginal education: Indigenous identity and career aspiration. Asia Pacific Education Review, 9(3), 233–247. Chuang, H. L., & Huang, W. C. (2007). A re-examination of the suicide rates in Taiwan. Social Indicators Research, 83(3), 465–485. Cobb, S. (1976). Social support as a mediator of life stress. Psychosomatic Medicine, 38, 300–314. Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310–357. Cowie, H., & Olafsson, R. (2000). The role of peer support in helping the victims of bullying in a school with high levels of aggression. School Psychology International, 21(1), 79–95. Cumsille, P. E., & Epstein, N. (1994). Family cohesion, family adaptability, social support, and adolescent depression symptoms in outpatient clinic families. Journal of Family Psychology, 8(2), 202–214. Davison, L. M., & Demaray, M. K. (2007). Social support as moderator between victimization and internalizing–externalizing distress from bullying. School Psychology Review, 36(3), 383–405. Demaray, M. K., & Malecki, C. K. (2003). Perceptions of the frequency and importance of social support by students classified as victims, bullies, and bully/victims in an urban middle school. School Psychology Review, 32, 471–489. Demaray, M. K., & Malecki, C. K. (2006). A review of the use of social support in anti-bullying programs. Journal of School Violence, 5(3), 51–70. Flannery, D. J., Singer, M. I., & Wester, K. L. (2004). Impact of exposure to violence in school on child and adolescent mental health and behavior. Journal of Community Psychology, 32(5), 559–573. Furlong, M., Whipple, A., Jean, G., Simental, J., Soliz, A., & Punthuna, S. (2003). Multiple contexts of school engagement: Moving toward a unifying framework for educational research and practice. California School Psychologist, 8, 99–113. Furlong, M. J., Chung, A., Bates, M. P., Whipple, A. D., Jimenez, T. C., & Morrison, R. L. (2005). Development of California school climate and safety survey-short form. Psychology in the Schools, 42(2), 137–149. Gladstone, G. L., Parker, G. B., & Malhi, G. S. (2006). Do bullied children become anxious and depressed adults? A cross-sectional investigation of the correlates of bullying and anxious depression. Journal of Nervous & Mental Disease, 194(3), 201–208. Graham, S., & Bellmore, A. D. (2007). Peer victimization and mental health during early adolescence. Theory into Practice, 46(2), 138–146. Guay, S., Billete, V., & Marchand, A. (2006). Exploring the links between posttraumatic stress disorder and social support: Processes and potential research avenues. Journal of Traumatic Stress, 19(3), 327–338. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1–55. Kaniasty, K., & Norris, F. H. (1993). A test of the social support deterioration model in the context of nature disaster. Journal of Personality and Social Psychology, 64(3), 395–408.

262

J.-K. Chen, H.-S. Wei / Child Abuse & Neglect 37 (2013) 252–262

Kendler, K. S., Myers, J., & Prescott, C. A. (2005). Sex differences in the relationship between social support and risk for major depression: A longitudinal study of opposite-sex twin pairs. American Journal of Psychiatry, 162, 250–256. Kim, H. S., Sherman, D. K., & Taylor, S. E. (2008). Culture and social support. American Psychologist, 63(6), 518–526. Kochenderfer-Ladd, B., & Skinner, K. (2002). Children’s coping strategies: Moderators of the effects of peer victimization? Developmental Psychology, 38, 267–278. Little, T., Lindenberger, U., & Nesselroade, J. R. (1999). On selecting indicators for multivariate measurement and modeling with latent variables. Psychological Methods, 4, 192–211. Malecki, C. K., & Demaray, M. K. (2002). Measuring perceived social support: Development of the Child and Adolescent Social Support Scale. Psychology in the Schools, 39, 1–18. Malecki, C. K., & Demaray, M. K. (2003). What type of support do they need? Investigating student adjustment as related to emotional, informational, appraisal, and instrumental support. School Psychology Quarterly, 18, 231–252. Malecki, C. K., Demaray, M. K., & Elliott, S. N. (2000). The Child and Adolescent Social Support Scale. DeKalb, IL: Northern Illinois University. Malecki, C. K., Demaray, M. K., & Davison, L. M. (2008). The relationship among social support, victimization, and student adjustment in a predominantly Latino sample. Journal of School Violence, 7(4), 48–71. Marsh, H. W., Hau, K. T., Balla, J. R., & Grayson, D. (1998). Is more ever too much? The number of indicators per factor in confirmatory factor analysis. Multivariate Behavioral Research, 33, 181–220. Naylor, P., & Cowie, H. (1999). The effectiveness of peer support systems in challenging school bullying: The perspectives and experiences of teachers and pupils. Journal of Adolescence, 22(4), 467–479. Newman, D. A., Horne, A. M., & Bartolomucci, C. L. (2000). Bully busters: A teacher’s manual for helping bullies, victims, and bystanders grade 6–8. Champaign, IL: Research Press. Nisbett, R. E. (2003). The geography of thought: How Asian and Westerners think differently . . . and why. New York: Free Press. Norris, F. H., & Kaniasty, K. (1996). Received and perceived social support in times of stress: A test of the social support deterioration deterrence model. Journal of Personality and Social Psychology, 71(3), 498–511. Norris, F. H., Foster, J. D., & Weisshaar, D. L. (2002). The epidemiology of sex differences in PTSD across developmental, societal and research contexts. In P. O. Kimerling, & J. Wolfe (Eds.), Gender and PTSD. New York: Guilford Press. Papalia, D. E., Olds, S. W., & Feldman, R. D. (2009). Human development (11th ed.). New York: McGraw Hill. Pepin, E. N., & Banyard, V. L. (2006). Social support: A mediator between child maltreatment and developmental outcomes. Journal of Youth and Adolescence, 35, 617–630. Prati, G., & Pietrantoni, L. (2010). An application of the social support deterioration deterrence model to rescue workers. Journal of Community Psychology, 38(7), 901–917. Prelow, H. M., Mosher, C. E., & Bowman, M. A. (2006). Perceived racial discrimination, social support, and psychological adjustment among African American college students. Journal of Black Psychology, 32(4), 442–454. Punamaki, R. L., Komproe, I., Qouta, S., El-Masri, M., & de Jong, J. T. V. M. (2005). The deterioration and mobilization effects of trauma on social support: Childhood maltreatment and adulthood military violence in Palestinian community sample. Child Abuse & Neglect, 29, 351–373. Reevy, G. M., & Maslach, C. (2001). Use of social support: Gender and personality differences. Sex Roles, 44, 437–459. Seeds, P. M., Harkness, K. L., & Quilty, L. C. (2010). Parental maltreatment, bullying, and adolescent depression: Evidence for mediating role of perceived social support. Journal of Clinical Child & Adolescent Psychology, 39(5), 681–692. Seeman, T. E. (1996). Social ties and health: The benefits of social integration. Annals of Epidemiology, 6, 442–451. Sharir, D., Tanasescu, M., Turbow, D., & Maman, Y. (2007). Social support and quality of life among psychiatric patients in residential homes. International Journal of Psychosocial Rehabilitation, 11(1), 85–90. Shih, C.Y. (2011). Yuan zhu min yu fei yuan zhu min qing shao nian qing xu zhi neng yu xin li xing fu gan zhi bi jiao yan jiu. [A comparative study on emotional intelligence and psychological well-being between aboriginal and non-aboriginal adolescents]. Unpublished master’s thesis, National Taiwan Normal University, Taipei, Taiwan. Taiwan. (2008). Council of indigenous peoples executive Yuan (2008). Statistical year report: Indigenous peoples and health. Taipei Taiwan: Author. Tardy, C. H. (1985). Social support measurement. American Journal of Community Psychology, 13(2), 187–203. Tiwari, A., Wong, M., & Ip, H. (2001). Ren and yuan: A cultural interpretation of Chinese women’s response to battering. Canadian Journal of Nursing Research, 33, 63–79. Triandis, H. C. (1995). Individualism and collectivism. San Francisco, CA: Westview Press. Tsai, J. L., & Levenson, R. W. (1997). Cultural influences on emotional responding: Chinese American and European American dating couples during interpersonal conflict. Journal of Cross-cultural Psychology, 28, 600–625. Wei, H. S., & Chen, J. K. (2012). Exploring factors associated with peer sexual harassment victimization among Taiwanese adolescents. Sex Roles, 66, 66–78. Wei, H. S., Chen, G. R., & Chen, J. K. (2009). Yi-Lan County youth life conditions survey. Yi-Lan, Taiwan: Yi-Lan County Government. Wei, H. S., Williams, J. H., Chen, J. K., & Chang, H. Y. (2010). The effects of individual characteristics, teacher practice, and school organizational factors on students’ bullying: A multilevel analysis of public middle schools in Taiwan. Children and Youth Services Review, 32(1), 137–143. Williams, T., Connolly, J., Pepler, D., & Craig, W. (2005). Peer victimization, social support, and psychosocial adjustment of sexual minority adolescents. Journal of Youth and Adolescence, 34(5), 471–482. Ye, S. Q. (2009). Factor structure of the General Health Questionnaire (GHQ-12): The role of wording effects. Personality and Individual Differences, 46(2), 197–201. Yeh, M. Y., & Chiang, I. C. (2005). Comparison of the predictors of alcohol use and misuse among Han and aboriginal students in Taiwan. Addictive Behaviors, 30(5), 989–1000.