Social Indicators Research

, Volume 95, Issue 3, pp 421–436

School Attachment Among Taiwanese Adolescents: The Roles of Individual Characteristics, Peer Relationships, and Teacher Well-Being

Authors

    • Department of Social WorkNational Taipei University
  • Ji-Kang Chen
    • Department of Social WorkChinese University of Hong Kong
Article

DOI: 10.1007/s11205-009-9529-3

Cite this article as:
Wei, H. & Chen, J. Soc Indic Res (2010) 95: 421. doi:10.1007/s11205-009-9529-3

Abstract

This study examines the effects of individual characteristics (school grade and gender), peer relationships (peer support and peer victimization), and the subjective well-being of teachers (depression and job satisfaction) on students’ attachment to school. Twenty-four classes in grades 7 through 9 at two middle schools in Taipei were selected as the sample, and survey data were obtained from students and homeroom teachers. Hierarchical linear modeling was used to conduct a two-level analysis on 720 students and 24 teachers with valid data on all research variables. A series of models were constructed and tested stepwise. The results indicated that students’ average school attachment scores varied significantly among classes. A higher school grade was associated with reduced attachment while no gender difference was found. Peer support had a positive influence and peer victimization had a negative effect on school attachment. Finally, job satisfaction of the homeroom teachers positively contributed to students’ attachment to school, but teachers’ depression had no significant effect. Implications for creating a positive classroom environment were discussed.

Keywords

School attachmentPeer victimizationHomeroom teacherWell-beingJob satisfaction

1 Introduction

School is a crucial environment that has a significant influence on student well-being and development (Roeser and Eccles 1998). Children and adolescents now spend a considerable amount of time in school each day, and are likely to have more interactions with classmates and teachers than with their own parents. School shapes youngsters’ beliefs and behaviors profoundly, which further affects their lifespan trajectory and career prospects. As a central part of life, school is regarded as a major area of students’ life satisfaction, while educational performance often serves as a core indicator of child well-being (Gilman et al. 2000; Ghysels and Van Vlasselaer 2008). Research has shown that negative school experiences impact students’ psychosocial adjustment and can result in a variety of consequences including low self-esteem, loneliness, and depression (Brand et al. 2008; Locker and Cropley 2004; McDougall et al. 2001). Such findings apply not only to Western countries but also in Eastern contexts as well. For example, a large-scale survey of 9,586 adolescents in Taiwan suggested that lower satisfaction with peer relations and less connectedness to school were associated with higher risk of depression (Lin et al. 2008).

Schools may play an even more important role for student well-being in certain Asian regions compared to Western countries. One study compared correlations among domains of life satisfaction for Korean and U.S. adolescents and found that satisfaction with school was highly associated with global life satisfaction for Korean, but not American, students (Park and Huebner 2005). As with many other Confucian-based cultures, the education system of Taiwan is highly academically-oriented and students are preoccupied by competitive exams and evaluations. Parents also have high expectations for their children’s educational achievement. As a result, Taiwanese adolescents spend a considerable part of every day in school-related activities. In a survey of 2,764 junior and senior high school students in one metropolitan area of Taiwan, 53.7% of the respondents were attending cram schools after regular school hours (Wei et al. 2006). In the same survey, 84.9% of the respondents reported school performance as a personal concern, which made it the major source of stress for local youth.

Positive attitudes and feelings toward school are beneficial for student well-being and learning. It is human nature to seek connections and bonding, and school is one major social institution that provides a sense of belonging for individuals (Baumeister and Leary 1995). Feelings of emotional connectedness, support, and inclusion in the school environment are often referred to as school attachment or school bonding (Jimerson et al. 2003), and have been shown to impact students’ psychological and behavioral adjustment. For example, attachment to school was found to contribute to lower delinquency and aggression as well as higher motivation for achievement (Eccles 2007; Frey et al. 2009). School attachment is also positively linked to better academic performance and self-esteem while negatively correlated with a variety of problem behaviors including substance use and gang involvement (Maddox and Prinz 2003; Catalano et al. 2004; Simons-Morton et al. 1999). For example, one study of sixth to eighth graders found differences in school attachment between bullies, victims, bully-victims, and non-involved peers, and such feelings were associated with their risk for bullying and investment in prosocial beliefs and behaviors (Cunningham 2007). So far, little research has been done in Asian regions to investigate the personal and social factors of school attachment despite the common emphasis on education in Confucian-based cultures. Therefore, the present study hopes to fill in this knowledge gap, examining the effects of individual, peer, and teacher contexts on Taiwanese adolescents’ attachment to school.

1.1 Individual Characteristics

On an individual level, gender is one variable that has been extensively examined in the school attachment literature. Girls are constantly found to have more positive feelings and attitudes toward school. One large-scale survey on 431,331 public school students in Arizona showed that girls had higher school satisfaction scores than boys (Okun et al. 1990). Another study surveyed 4,263 6th through 8th graders in seven U.S. middle schools and found female students to have higher school bonding and school adjustment while male students had more problem behaviors (Simons-Morton et al. 1999). Verkuyen and Thijs (2002) examined school satisfaction among 1,090 Dutch children aged 10–12 years and found girls were significantly more satisfied with their schools than boys were. Still another study analyzed data from the National Longitudinal Study of Adolescent Health and showed that girls had higher school attachment than boys in the middle school years, although such differences were only statistically significant in grade 8 (Johnson et al. 2006). Girls were also found to have higher grades and school engagement. These findings are interesting because research in general has shown that girls often have lower self-esteem, higher depression, increased loneliness, and more suicidal ideation during adolescence compared to boys (Polce-Lynch et al. 2001; Mahon et al. 2006; Galambos et al. 2004; Liu et al. 2005). It seems that school is one area where girls have higher enjoyment and adjust better than boys.

Age and school grade differences in school attachment are also shown to be salient in the literature. Student satisfaction and bonding with school have been shown to decrease during middle school years, with the lowest levels in 8th or 9th grade (Johnson et al. 2006; Hawkins et al. 2001; Okun et al. 1990). The common decline in school attachment and motivation during early adolescence following the transition to junior high school has been stated to result from a misfit between student psychological needs and the opportunities provided by the school environment (Eccles et al. 1996). One study of 641 American children from 3rd to 6th grade showed that students’ self-reported relatedness to teachers increased between 3rd and 5th grade but fell significantly after entering middle school, as did their classroom engagement (Furrer and Skinner 2003). Another large-scale study revealed a similar trend in which delinquency and problem behaviors went up while students’ school bonding and perceived climate continually dropped from 6th to 8th grade (Simons-Morton et al. 1999).

Little research has been done to investigate the potential effect of age and school grade on the well-being and school attachment of Taiwanese adolescents. However, the 9th grade is expected to be a difficult and stressful time for local students since the Basic Competence Test for senior high school entrance is usually held two times in May and July while the middle school students graduate in June. In accordance with this schedule, middle schools in Taiwan increasingly focus on academic performance as adolescents move along the school grade levels, which can result in great distress and negative feelings toward school for students in higher grades. Interestingly, some studies on students from other cultures found exams and evaluations not necessarily threatening or harmful. For example, one UK study of 520 secondary school students found no significant difference in depression between school year (year 9 and year 11) and time points (regular term time and the week immediately prior to the examinations; Locker and Cropley 2004). On the other hand, given the cultural value of academic achievement and high educational expectations from parents, performance pressure and test anxiety are common for students in Asian countries such as Japan, Korea, and Singapore (Ang et al. 2007; Juon et al. 1994; Hawkins and Tanaka 1992). One study compared public middle school pupils’ perceptions of test anxiety in the United States and China, and Chinese students reported significantly stronger concerns for the possibility of annoying their parents with unsatisfactory academic achievement than American students (Xing et al. 2005). The pressure from family and school is likely to increase as students approach the critical exam in 9th grade and may make students more likely to avoid attending schools. It is thus hypothesized in the present study that school grade is negatively associated with students’ school attachment.

1.2 Peer Relationships

As adolescents’ major social environment, school is a multilayer context with a variety of subsystems, among which peers constitute a critical one. Peers become increasingly significant during adolescence and have a considerable influence on children’s well-being and development. Peer support, friendships, and positive classroom climate have been found to be associated with higher school liking and school satisfaction (Erath et al. 2008; Verkuyten and Thijs 2002). Positive relationships with peers is also associated with healthy and adaptive outcomes including higher self-esteem, better school performance, and stronger achievement motivation (de Bruyn and van den Boom 2005; Harter et al. 1998; Eshel et al. 2003; Nelson and DeBacker 2008). A large-scale study of 18,735 students in Dutch junior high schools indicated that students with peer acceptance had lower risk of being held back from advancing to the next grade or to go downward in the track system (Lubbers et al. 2006). On the other hand, peer rejection can affect students’ learning and result in maladjustment and problem behaviors (Wentzel 2003; Ladd 2006; Pedersen et al. 2007), and can also lead to reduced class engagement and detachment from school (Buhs et al. 2006).

Getting along with peers is not always an easy task for adolescents, and many are actually abused or maltreated by classmates and fellow students. Recent research has clearly shown that peer victimization is prevalent in Western and Eastern countries (Chen and Astor 2009a; Due and Holstein 2008; Koo et al. 2008; Wong et al. 2008; Wei and Chen 2009). Wei and others surveyed a sample of Taiwanese 7th graders and showed that approximately 31% of the students reported being bullied at least once by classmates during their first three months in middle school while almost 24% were bullied by students outside their classes (Wei et al. 2007). Such experiences have a considerable impact on the victims’ psychosocial well-being (Hawker and Boulton 2000; Karin Natvig et al. 2001). Students’ academic achievement, attachment to school, and even life satisfaction are all likely to be affected (You et al. 2008; Beran et al. 2008). Wei and Williams (2004) analyzed data on 1,022 US 6th grade students and found that peer victimization was associated with lower school attachment, which in turn led to inattentive school behaviors and poor academic performance. Another study in the UK showed that victimization and the fear of future victimization predicted disrupted concentration on class work (Boulton et al. 2008).

The influence of peers is probably more salient in Taiwanese schools than in certain Western counterparts. Middle school students in Taiwan stay with a fixed group of peers every day throughout the school years, which makes peer reputation and status in class extremely important for quality of life in school. Under the current school schedule, Taiwanese students attend school approximately two hundred days a year, and their time with classmates can be as long as nine hours a day, which is much longer compared to many other countries. In addition, Taiwan maintains much of its collectivism characteristics, such as group-orientation and emphasis on interpersonal relations, despite the profound transformation toward Westernization (Ali et al. 2005; Chiou 2001). It is therefore hypothesized in the present study that peer relationship factors, including peer support and peer victimization, are associated with Taiwanese adolescents’ attachment to school.

1.3 Teacher Factors in School Attachment

Besides peers, teachers are another major factor affecting students’ sense of connectedness and attachment to school. Past research has largely focused on the influence of the teacher–student relationship on school-related feelings. Perceived teacher support has been shown to be associated with positive perceptions of school climate, school satisfaction, and academic motivation (Danielsen et al. 2009; Frey et al. 2009). A study of 805 American students in 4th to 7th grade found that teacher support led to increased emotional engagement and decreased disaffection (Skinner et al. 2008). Having a warm and trustful relationship with a teacher predicted positive school outcomes for at-risk students (Baker et al. 2008). Davis and Lease (2007) employed a peer-rating procedure with middle school students to map their perceived status of teacher likeability in classrooms. The position in this “teacher-liking space” was found to be associated with students’ achievement and teacher relationship quality. Furthermore, Hughes and Kwok (2006) suggested that teacher–student relationship quality contributes to classroom engagement, which in turn impacted students’ peer acceptance.

Under the current educational system in Taiwan, one homeroom teacher is assigned to each class in middle school, and this teacher interacts with these students in homeroom on a daily basis for as long as 3 years (Chen and Astor 2009b; Taiwan Ministry of Education 2006). The homeroom teachers play a crucial role in the classroom climate and students’ learning due to the great power endorsed to them by the institution and cultural traditions (Fwu and Wang 2002). The Confucian heritage culture of Taiwan emphasizes respect for teachers and an obedient role for students. Teachers are regarded as a parent-like figure and pupils are supposed to follow their instructions without questioning (Wu 2002). As a symbol of authority, they are in control of classroom dynamics and their conduct sets an example for students to imitate. However, teaching is a stressful occupation in Eastern as well as Western countries. A study based on a sample of senior high school teachers in China found that their somatization and depression scores were higher than the Chinese norm (Huang et al. 2006). A large-scale telephone survey in Hong Kong also showed that the prevalence rates of generalized anxiety disorder and major depressive episodes among teachers were higher than the general population, and these symptoms were associated with job-specific stressors (Lee et al. 2007).

Teachers’ subjective well-being, such as psychological adjustment and job satisfaction, is closely related to performance and teaching quality. Lower well-being was found to be associated with teacher burnout (Milfont et al. 2008). Betoret (2009) suggested that emotional exhaustion was a key element of burnout, which was largely accounted for by teachers’ job stressors. Teachers who perceived their job as highly demanding and low in control were at higher risk of emotional exhaustion and burnout (Santavirta et al. 2007), and teachers who were experiencing burnout were shown to have attention difficulties (van der Linden et al. 2005). A study of Spanish secondary school teachers found emotional exhaustion and job dissatisfaction to be associated with psychiatric sick leave (Moriana and Herruzo 2006). The relationship between job satisfaction and job performance has been well documented (see Judge et al. 2001 for a review), and teachers’ distress can interfere with their motivation and ability to teach and care for students, but so far there is little research directly investigating the effect of teacher well-being on student attachment to school. The present study addresses this issue and evaluates the hypothesis that homeroom teachers’ depression and job satisfaction predict students’ general school attachment in that class.

1.4 The Present Study

A school is a complex environment with multiple layers of contexts (Furlong et al. 2003). From an ecological perspective, the present study aims to integrate individual, peer, and teacher factors into a multilevel model and examine their contributions to middle school students’ school attachment. One limitation of previous studies is that they often rely on a single source of data, especially students’ self-reports, to assess variables on different levels including class and school contexts. Such practices increase the concern of shared-method variance (Donaldson and Grant-Vallone 2002). This study employs data from both teachers and students in order to reduce the risk of self-report bias. It also delivers a more direct analysis for the link between teachers’ well-being and students’ feelings toward school, which is a core index of well-being among adolescents. Specifically speaking, this study will test the following research hypotheses:
  1. 1.

    On the individual level, girls have higher school attachment than boys and students in higher grades are less attached to school than students in lower grades.

     
  2. 2.

    School attachment is positively affected by peer support and negatively affected by peer victimization.

     
  3. 3.

    Homeroom teacher’s depression has a negative effect and his/her job satisfaction has a positive effect on students’ attachment to school.

     

Finally, this study focuses on Taiwanese youth and accounts for the specific educational characteristics under a Confucian heritage culture. The results can provide valuable insights regarding potential cultural differences in school dynamics between Western and Eastern countries.

2 Methods

2.1 Participants

Two middle schools in Taipei were invited to participate in the present study, in which 834 students and 24 homeroom teachers from 24 classes in grades 7–9 were recruited as the research respondents. Written consent was obtained and self-report questionnaires were administered. Only those with complete data on all research variables were included in the following analysis, which results in a final sample of 720 students and 24 teachers. About half (49.7%) of the students (n = 358) are male and 50.3% (n = 362) are female. Their ages range from 10 to 15 years old with a mean of 13.34 and a standard deviation of .97. The majority of teacher respondents are female (83.3%, n = 20) and in their forties (79.2%, n = 19).

2.2 Procedures

Data collection for the present study was conducted during the fall semester of 2007. Student respondents were asked to fill out the research questionnaire in class. The self-report questionnaire takes approximately 30 min to complete. The students were advised to read the instructions carefully and not to discuss answers with classmates during the survey. It was clearly stated that respondents could withdraw from the survey at any time with no penalty. The teacher version of the questionnaires was distributed to homeroom teachers and returned after completion.

2.3 Measures

2.3.1 Individual Characteristics

Information on student characteristics, including gender and school grade, were collected through corresponding items on the self-report questionnaire. Male and female were dichotomously coded while the three school grades were coded as 7, 8, and 9, respectively.

2.3.2 Peer Victimization

Students’ experiences with peer victimization during the last semester were assessed via seven items on a five-point scale (never–five times or more). These seven items in this scale were selected and adapted from the scale of victimization based on the scale of school violence in Taiwan (Chen and Astor 2009a) as well as adaptive version of California School Climate and Safety Survey (CSCSS; e.g., Benbenishty and Astor 2005; Furlong et al. 1998). The original items in adaptive version of California School Climate and Safety Survey were translated from English to Chinese in Standardized Version. A standard back-translation procedure was employed in order to ensure translation accuracy. Example items include “My classmates beat, kick, punch me or use other instruments to hurt me”, “My classmates curse or insult me”, and “My classmates boycott me and do not want to play or talk with me”. The coefficient alpha was calculated as .91. Regarding validity, California School Climate and Safety Survey was shown to have good construct validity (Furlong et al. 2005).

2.3.3 School Attachment

School attachment was measured by six items on the questionnaire adopted from previous studies (Chen and Astor 2010) and CSCSS (Benbenishty and Astor 2005; Furlong et al. 1998, 2005; Murray and Greenberg 2001) with modifications to assess students’ general interests and feelings toward school. Example items include “I feel unsafe in school”, “I feel that my school doesn’t care about students”, and “I like school life”. A five-point scale was used (strongly disagree–strongly agree). Negative items were reverse scored and an aggregate score was calculated to represent school attachment. The coefficient alpha of the 6 items in the research sample equals .69. CSCSS demonstrated adequate construct validity (Furlong et al. 2005).

2.3.4 Peer Support

Peer support was measured by two items on the questionnaire: “My classmates encourage and support me in my time of difficulty and frustration” and “My classmates would like to share their thoughts with me”. A five-point scale was used (strongly disagree–strongly agree). The coefficient alpha equals .78.

2.3.5 Teacher Depression

Teacher depression was assessed using seven items adopted from the Brief Symptom Rating Scale (BSRS), a Chinese questionnaire used to screen for mental health problems (Lee et al. 1990). The whole battery consists of 50 items, which are divided into ten dimensions such as phobia, depression, and obsession. Only the depression subscale of BSRS was employed for the present study. Respondents are asked to report their status over the past week on a five-point scale (never–very severe). Example items include “I have suicidal thoughts”, “I feel lonely”, and “I lose interest in daily activities”. The coefficient alpha was calculated as .94. BSRS was suggested by the authors to be a valid screening tool and its scores were found highly correlated with the parental form of the Symptom Check List-90-R (SCL-90-R) among medical populations for each symptom dimension (Lee et al. 1990).

2.3.6 Teacher Job Satisfaction

The Job Satisfaction Scale (JSS; Liu 2004) was adopted to assess teachers’ feelings about their current job, including administration and leadership, salary and promotions, relationships with colleagues, psychological rewards, and job content. Example items include “opportunities of job promotion”, “campus climate for faculty”, and “efficiency of school administration”. JSS has a total of 35 items measured on a five-point Likert scale (very unsatisfied–very satisfied). The coefficient alpha was found to be .96. The five factors of JSS accounted for approximately 60% of the variance and good construct validity was suggested by the author (Liu 2004).

2.4 Data Analysis

This study employed hierarchical linear modeling (HLM) as the major analytical approach. HLM is suitable for analyzing data that has nested structures (Raudenbush and Bryk 2002). In this study, the students were nested in classes, and each class had only one homeroom teacher. The student and teacher variables thus constitute two levels of data. The fact that students belonged to classes violates the statistical assumption that all observations are independent of one another. Unlike many traditional regression approaches, HLM takes this non-independence into account and is able to analyze the variance in the outcome variable at multiple levels. In the present study, a stepwise strategy was used in which student-level models were first computed to assess the effects of individual characteristics and peer relationships on school attachment. One regression equation was calculated for each class with intercepts and slopes, which were then used in a class-level analysis to further examine the influence of teacher well-being on students’ attachment to school.

3 Results

The descriptive statistics of the research variables are shown in Table 1. As a first step of HLM, a basic two-level model of school attachment without any predictors was specified and tested. The results show a significant difference in average school attachment scores between classes (tau = .98, Chi-square = 65.14, p < .01), which provides a rationale to conduct further multi-level analysis. In this unconditional model, the within-group (student-level) variance was found to be 16.02 while between-group variance (teacher-level) was .98. The Intraclass Correlation Coefficient (ICC) was calculated as .06, suggesting that class-level factors account for approximately 6% of the variance in students’ school attachment.
Table 1

Descriptive statistics of research variables

 

Mean

SD

Minimum

Maximum

Outcome variable

 School attachment

20.31

4.12

6.00

30.00

Student-level variables

 Gendera

N/A

   

 School grade

8.18

0.80

7.00

9.00

 Peer support

7.07

1.97

2.00

10.00

 Peer victimization

10.97

3.99

7.00

28.00

Teacher-level variables

 Depression

31.71

4.30

16.00

35.00

 Job satisfaction

122.13

18.77

79.00

165.00

aGender is a variable with only two categories (1 = male and 2 = female)

Secondly, a model with student-level predictors only was specified, in which gender, school grade, peer support, and peer victimization were included. The results of the model testing are summarized in Table 2. School grade and peer victimization were found to be negatively associated with school attachment while peer support positively contributed to school attachment. An inspection of the mean attachment scores of the three school grades revealed a downward tendency with 7th grade being the highest and 9th grade the lowest. No significant gender difference was found. A supplementary ANOVA was conducted, in which girls were found to have slightly higher school attachment scores than boys although the difference was non-significant [F(1,718) = .05, p = .83]. Gender was therefore dropped from the consequent analysis.
Table 2

HLM results of the model with student-level predictors only

Student-level variables

Coefficient

Standard error

T-ratio

p value

Gender

−0.371

0.369

−1.005

0.316

School grade

−0.592

0.267

−2.216

0.027

Peer support

0.704

0.081

8.610

0.00

Peer victimization

−0.095

0.0417

−2.275

0.023

Finally, teacher depression and job satisfaction were added as teacher-level predictors, and the complete model with variables from the two levels was specified as follows:

Student-level model
$$ {\text{School}}\;{\text{attachment}} = {\text{B}}0 + {\text{B}}1 * \left( {{\text{school}}\;{\text{grade}}} \right) + {\text{B}}2 * \left( {{\text{peer}}\;{\text{support}}} \right) + {\text{B}}3 * \left( {{\text{peer}}\;{\text{victimization}}} \right) + {\text{R}} $$
Teacher-level model
$$ {\text{B}}0 = {\text{G}}00 + {\text{G}}01 * \left( {{\text{teacher}}\;{\text{depression}}} \right) + {\text{G}}02 * \left( {{\text{teacher}}\;{\text{job}}\;{\text{satisfaction}}} \right) + {\text{U}}0 $$
$$ {\text{B}}1 = {\text{G}}10 + {\text{U}}1 $$
$$ {\text{B}}2 = {\text{G}}20 + {\text{U}}2 $$
$$ {\text{B}}3 = {\text{G}}30 + {\text{U}}3 $$

B0 represents the intercept of the student-level equation while B1, B2 and B3 refer to the slopes of school grade, peer support and peer victimization, respectively. The model specifies that teacher-level variables affect the intercept of the student-level equation. In other words, teacher’s depression and job satisfaction were used to predict after controlling for individual-level factors including school grade, peer support and peer victimization. All the predictors were centered on the grand mean to obtain more interpretable estimates.

Table 3 summarizes the results of hierarchical linear modeling. The three student-level variables all significantly contributed to school attachment. It was expected that students in lower grades were more attached to school, and the results supported this hypothesis. Another hypothesis was also supported, in which higher peer support and less peer victimization were associated with higher attachment to school. However, as mentioned earlier, gender was found having no significant effect on student’s school attachment, which was inconsistent with one of the student-level hypotheses. At the same time, teacher-level hypotheses were partly supported by the results, in which job satisfaction of the homeroom teacher was found significantly associated with school attachment while no significant relationship was found between depression of the teacher and students’ school attachment.
Table 3

HLM results of the final model

 

Coefficient

Standard error

T-ratio

p value

Student-level variables

 School grade

−0.541

0.227

−2.378

0.018

 Peer support

0.686

0.079

8.64

0.000

 Peer victimization

−0.088

0.039

−2.219

0.027

Teacher-level variables

 Depression

−0.0185

0.037

−0.498

0.62

 Job satisfaction

0.028

0.009

2.93

0.008

4 Discussion

School is a major institution that plays a critical role in youngsters’ education and socialization, and students’ feelings toward school are closely related to their learning and well-being. The present study utilizes a multi-level approach and examines the effect of individual characteristics, peer experiences, and teacher well-being on students’ school attachment. These factors have not only been studied extensively in past research but also have cultural relevance in the context of Taiwan. Predictors are categorized into two levels and hierarchical linear modeling is the major analytic strategy. The results largely support the research hypotheses. First, higher school grade was found to be significantly associated with lower school attachment, which confirmed one of our student-level hypotheses. As mentioned earlier, middle school students in Taiwan are required to take the Basic Competence Test in 9th grade, and this major exam determines their future in high school and thus becomes a major source of stress for many students. School classes are more and more exam-oriented as the exam approaches, and teachers also put stronger emphasis on students’ academic performance and test scores. Based on the results of this study, it seems that such pressures actually reduced students’ positive attitudes toward school, which may be damaging for their learning and well-being. The study also highlights the need for mental health services among students in higher school grades. However, this study did not assess students’ academic stress nor examine the class contents and schedules between different school grades, which constitutes a limitation and makes it difficult to draw any definitive conclusion. In order to better understand the decrease in school attachment with increasing grades, future investigation is needed to reveal the underlying mechanisms through which the school grade impacts those youngsters’ bonding with school.

Another student-level hypothesis concerning gender difference was rejected by the results. Although girls had slightly higher scores on school attachment, no significant gender difference was found in this study, which is inconsistent with past research in Western countries. Interestingly, a recent cross-national survey on adolescents showed that American and Irish male respondents reported significantly lower school satisfaction scores than their female peers while no such differences were found for Chinese and South Korean adolescents (Gilman et al. 2008). A closer examination of the present data revealed that girls in 7th grade had higher school attachment than boys (although the difference was not significant), but the situation was reversed for 8th graders, in which girls had lower scores than boys. Both genders reached an identical lowest level in 9th grade. The middle school transition and school environment seem to be somewhat gender-unfriendly for female students in Taiwan. One contributing factor may be the patriarchic culture and gender-insensitive practice in many Taiwanese schools. Lee (2007) examined the organization of a Taiwanese high school and suggested the existence of gendered discourses in which a men-strong/women-weak differentiation was reinforced and a sexist school culture was reproduced. Further investigation is called for to identify the exact factors causing girls’ rapid decline in school attachment during this period, and more efforts are needed in gender education and gender mainstreaming in order to promote a friendly environment for both boys and girls.

Results of this study supported our hypothesis that school attachment is positively affected by peer support and negatively affected by peer victimization. The findings are consistent with the previous studies and confirmed the impact of peer relationships on students’ attitude toward school. Given the nature of class structure in Taiwanese schools, where students spend all of their time with the same group of peers, these classmates are expected to play an important role in adolescents’ learning and well-being. At the same time, the collectivist cultural tradition of Taiwan puts great emphasis on maintaining relational harmony and interdependence (Lu 2008), which may make peer maltreatment more problematic for victims. Akiba (2004) suggested that acceptance by the group was crucial in a collective society like Japan, and ijime (the Japanese word for bullying) was mostly a collective behavior in which a group of perpetrators abuse one single student. Sometimes the whole class may join in as bullies against the victim. More in-depth studies should be conducted to analyze group dynamics in the classroom and the cultural implications of peer relationships for Taiwanese adolescents.

In a Confucian heritage tradition, teachers are regarded as a symbol of authority while the students assume a passive and obedient role to be hard-working, humble, and silent (Ho et al. 2001). Teachers’ conduct and manners serve as a model example for youngsters to look up to. Given their central role in education, the well-being of teachers is expected to be tied to students’ feelings toward school. This study tested such hypotheses through a direct assessment of teachers’ subjective well-being, which provides some advantages over previous research. In partial support of our hypothesis, homeroom teachers’ job satisfaction, but not their depression, was associated with students’ school attachment. The considerable correlation between depression and job satisfaction (r = .50) implies their close relationship, but the latter appears to be a stronger predictor of students’ connectedness to school. In a sense, students’ attitudes correspond to teacher’s attitudes since those with unsatisfied homeroom teachers are less attached to school. However, the small ICC mentioned earlier suggests that only a small portion of the variance in students’ school attachment was accounted for by teacher-level variables. This is possibly due to the small sample size (720 students and 24 teachers) of the present study, which constitutes a limitation and calls for further investigation to better evaluate the magnitude of teacher influence in students’ bonding to school.

The findings of this study point to a number of implications for research and practice. First, the specific mechanisms and processes connecting teachers’ and students’ well-being need to be identified. Teachers experiencing job stress are likely to have negative perceptions toward school, which may impact their teaching competence and reduce their ability to care about the students. This underscores the necessity of mental health screening and care for school teachers. One study in Hong Kong showed that only 8.5% of local teachers with mental disorders sought professional help due to concerns of stigmatization (Lee et al. 2007). Schools should provide adequate and confidential mental health services not only for students but also for teachers in order to ensure their psychosocial adjustment and well-being. Since much of the teachers’ dissatisfaction and emotional exhaustion is associated with job stress, the above discussion highlights another important issue, which is the improvement of the school organization and culture. Lu (2007) found that middle school teachers’ perceptions of organizational politics affected their burnout through job satisfaction. School support is another important factor for teachers’ emotional well-being and professional disengagement (Galand et al. 2007).

Students need a warm and friendly environment in order to fulfill their potential, as do teachers. Therefore, effective strategies to promote school attachment must be inclusive, and a whole school approach is recommended in order to build a caring community for all individuals involved (Rowe et al. 2007; Battistich et al. 2004). Adolescents’ developmental demands and life changes during this stage should also be taken into consideration. Social support for the needs of both autonomy and relatedness is likely to enhance students’ quality of life over their transition to secondary school (Gillison et al. 2008). Teachers can be active participants in creating a positive school environment. Increased teacher support and respect for student perspectives have been found to be associated with less drug use and depression as well as perception of a respectful climate among students (LaRusso et al. 2008). Children and adolescents are valuable assets in our society and efforts have been made to enhance their positive development beyond mere survival. This study provides a comprehensive framework to explore adolescents’ school attachment in an Asian context, which can help in the development of effective and culture-sensitive programs to promote the well-being and growth of youth in Taiwan.

Copyright information

© Springer Science+Business Media B.V. 2009