Bullying victimisation and adolescents: implications for school-based intervention programs.

Lester, Leanne ; Cross, Donna ; Dooley, Julian 等

Abstract

Schools have become increasingly aware of the prevalence, seriousness and negative impacts of bullying. Investigation into the direction and strength of the relationship between social health factors and bullying victimisation during early adolescence and the determination of a critical time to focus school-based bullying intervention programs is a high priority. Data were collected using a self-completion questionnaire four times over 3 years from 3459 students aged 11-14 years during the transition from primary to the end of the second year of secondary school. Results show the path coefficients for bullying victimisation to social heath factors were stronger at the beginning of secondary school than the reverse paths, with bullying victimisation associated with greater loneliness, less peer support, less connectedness to school and feeling less safe at school. Reciprocal relationships between bullying victimisation and social health were found during the first 2 years of secondary school. Consequently, the time prior to the transition to secondary school and within the first 2 years of secondary school appears to be a critical time to implement a whole-school bullying intervention program to reduce victimisation.

Keywords

Connectedness, loneliness, peer support, safety, school bullying, victimisation

**********

Schools have become increasingly aware of the prevalence, seriousness and negative impacts of bullying--a type of aggressive behaviour involving the systematic abuse of power through unjustified and repeated acts intended to inflict harm (Smith, 2004). Approximately 10% of Australian school students in Grades 4-9 reported being bullied most days or more often, with 25% reporting being bullied every few weeks or more often in the previous term (10 weeks) at school (Cross et al., 2009). This bullying victimisation impacts on social health (the ability to get along with others, how others react to them, how they react to social institutions and societal mores), physical health (Tremblay et al., 2004) as well as mental health (Gini & Pozzoli, 2009; Kaltiala-Heino, Rimpela, Rantanen, & Rimpela, 2000). Cross-sectional studies found victimisation was associated with many social health factors including a loss of friendships (D. Espelage, Bosworth, & Simon, 2000), feelings of isolation and hopelessness (D. Espelage et al., 2000), unhappiness and lack of self-esteem (Glover, Gough, Johnson, & Cartwright, 2000) and disruptions to learning (Bosworth, 1999).

The application of social-ecological theory to the conceptualisation of bullying and victimisation suggests that bullying perpetration and victimisation are reciprocally influenced by the individual, family, school, peer group, community and society (Swearer et al., 2006). Social health factors that have been identified as protective against victimisation include the ability to make new friends (Akos & Galassi, 2004), the number and quality of friends (Pellegrini & Bartini, 2000), peer support (Pellegrini, 2002), feeling comfortable in new social situations (Cohen & Smerdon, 2009), positive evaluation of self by others (Storch, Brassard, & Masia-Warner, 2003) and having a positive self-image (Mizelle, 2005). Liking school (Barber & Olsen, 2004), school belonging (Benner & Graham, 2009), connectedness to school (O'Brennan & Furlong, 2010) and feeling safe at school (D. Espelage et al., 2000) are also protective against victimisation.

There is evidence that during a student's school life there are periods of time when the risk of being bullied is higher than at other times. For example, victimisation peaks twice for Australian school students first at age 10 and then again following their transition to secondary school, around age 12 (Cross et al., 2009). The transition from primary to secondary school provides both challenges and opportunities for adolescents as they experience environmental, physiological, cognitive and social changes (Barton & Rapkin, 1987). During this period, students need to learn to adapt to new organisational and social structures within their school environment, while having mixed feelings of fear and anticipation about the social relationships which dominate the school transition experience (Pereira & Pooley, 2007). Friendship and peer support have been identified as important contributors to a successful transition from primary school to secondary school (Crockett, Petersen, Graber, Schulenberg, & Ebata, 1989), with the social support of peers, parents and teachers helping to mediate the relationship between victimisation and school adjustment (Malecki, Demaray, & Davidson, 2008).

Factors that can contribute to bullying and victimisation during this time include social comparisons between peers (Pellegrini, 2002), the number and quality of friends, being disliked by peers and the establishment of hierarchy and new social roles in new social groups (Pellegrini & Bartini, 2000). The transition period can result in increased feelings of isolation as a major change in social structure occurs with adolescents often having to develop new friendships and lose friends at a time when great importance is placed on peer relationships (Pellegrini & Bartini, 2000). The dependence on peer relationships and reliance on peers for social support comes with increasing pressures to attain high social status (D Espelage & Holt, 2001). The social and emotional changes associated with moving into secondary school can manifest in frustration and anxiety and have been associated with negative and disruptive social behaviours (Cohen & Smerdon, 2009).

Of all the factors mentioned, two individual and two school-based variables related to social health, victimisation and transition were opportunistically examined in this study. These variables included loneliness at school, peer support, connectedness to school and feeling safe at school. Evidence suggests lonely children are more likely to be victimised by peers (Berguno, Leroux, McAinsh, & Shaikh, 2004) and, in turn, those victimised are more likely to be lonely as other peers avoid them for fear of being bullied themselves or losing social status among their peers (Nansel et al., 2001). Having many friends, having friends who are willing and able to support and protect and being accepted by the peer group are the main social factors identified as protective against victimisation (Hodges & Perry, 1996). Importantly, friendship moderates the relationship between victimisation and loneliness (Storch et al., 2003), which involves both the circumstance of aloneness and the feeling of sadness (Cassidy & Asher, 1992), while poor friendship quality is associated with high levels of loneliness (Woods, Done, & Hardeep, 2009). The perception of peer support--which refers to the quality of students' friendships, both the level of validation and social support they receive through their friends (Ladd, Kochenderfer, & Coleman, 1996)--was also found to be protective against victimisation (Pellegrini, 2002). Furthermore, victimised students perceive less peer support and place greater importance on peer support than those who bully or are uninvolved (Malecki & Demaray, 2004).

School connectedness describes the quality of the social relationships within the school: the extent to which a student feels like he/she belongs at and feels cared for by the school (McNeely, Nonnemaker, & Blum, 2002). Students who are physically, verbally and relationally victimised are more likely to report feeling disconnected from school (O'Brennan & Furlong, 2010; You et al., 2008). In contrast, students who feel more connected are more considerate and accepting of others (Osterman, 2000) and report less peer harassment (Eisenberg, Neumark-Sztainer, & Perry, 2003). Students' perception of safety at school is negatively influenced by bullying (Bradshaw, O'Brennan, & Sawyer, 2008) and the behavioural reactions of their peer group (Gini, Pozzoli, Borghi, & Franzoni, 2008), if they hear others being mean (Beran & Tutty, 2002), and if they feel adults at school are not supportive (Beran & Tutty, 2002). Feeling unsafe at and disconnected to school may be the result of being bullied but, conversely, may also result in perpetration of bullying behaviours (Bradshaw et al., 2008).

As the majority of research in this area is cross-sectional in design, the relationship between bullying victimisation and social health factors as examined in this research during and 2 years following school transition is not well established. This paper examines the evolving nature of the association between the social health outcomes and victimisation over and beyond the transition period, to determine whether the associations are reciprocal or unidirectional and whether they increase or decrease following transition. Through examining the relationships between the social health variables and victimisation over time, we hope to determine the critical times to intervene to prevent victimisation.

Methods

Sample and procedure

Data were drawn from a longitudinal study, the Supportive Schools Project (SSP). These data included adolescents' knowledge, attitudes and experiences of bullying victimisation and perpetration during the transition from primary to secondary school. Data were collected in four waves from 2005 to 2007, from students in Catholic schools in Perth, Western Australia. Catholic schools represent 11% of all secondary schools within Perth. The student cohort completed a baseline self-completed questionnaire in Grade 7, the last year of primary school (11 years old). After the transition to secondary school, the cohort completed questionnaires at the beginning (Term 1) of Grade 8 (12 years old), the end (Term 3) of Grade 8 (13 years old) and Term 3 of Grade 9 (14 years of age).

Catholic secondary schools were chosen to participate in the SSP study to reduce the rate of transition attrition, as students within Australian Catholic primary schools are more likely to move to their local Catholic secondary school than is the case within government schools. Schools were stratified according to the number of students enrolled and each school's socio-economic status and then randomly assigned to the intervention or comparison group. Of the 29 eligible schools, 21 schools consented to participate with ten schools randomly allocated to the intervention group (n=1789) and eleven to the comparison group (n=1980). Eight schools declined because of other priorities and demanding staff workloads.

Active followed by passive consent (Ellickson & Hawes, 1989) was sought via mail from parents of Grade 7 students (in over 400 primary schools) enrolled for the following year in the 21 recruited secondary schools (five of the schools included Grade 7, with only about 4% of students transitioning within the same school setting). Students in this study were transitioning from primary schools with an average student population of 165 to secondary schools with an average student population of 790. Of the 3769 students eligible to participate, parental consent was obtained for 92% (n=3462). The numbers of students and response rates at each data collection wave are given in Table 1. Approximately 3100 (90%) students responded on at least three data collection points and 1771 responded on all four data collection points (51%). Over the study period, participants comprised approximately 50% males and 70% attended a co-educational (n=8) versus single sex (n=3) secondary school. Ethics approval was granted by the Edith Cowan University Ethics Committee.

Measures

Victimisation. To assess physical, relational and verbal bullying victimisation, a 7-item categorical index adapted from items/scales developed by Rigby and Slee (1998) and Olweus (1996) was used. Students were asked how often they were bullied in the following ways: hit, kicked and pushed around; had money or other things broken or taken away from them; made to feel afraid they would get hurt; made fun of and teased in a hurtful way; called mean and hurtful names; students ignored them, did not let them join in, or left them out on purpose; and students told lies about them and tried to make other students not like them, during the prior term at school. Each item sought responses on a 5-point scale equivalent to that recommended by Solberg and Olweus (2003) (1=never, 2=once or twice, 3=every few weeks, 4=about once a week, 5=most days). A factor analysis performed on the victimisation scale confirmed its unidimensionality (comparative fit index (CFI)>0.9, SMR<0.10 at all time points). A victimisation score was calculated at each time point for each student by averaging the 7 items with a higher score reflecting more experiences of victimisation (average alpha=0.86).

Peer support. The peer support at school scale (adapted from the 24-item Perceptions of Peer Social Support Scale; Ladd et al., 1996) comprised eleven items: How often would students: choose you on their team; tell you you're good at things; explain something if you did not understand; invite you to do things with them; help you if you are hurt; miss you if you were not at school; help you if something is bothering you; ask to work with you; help you if other students treat you badly; ask you to join in when alone; and share things with you? Items sought responses on a 3-point scale (1=never, 2=sometimes, 3=lots of times). A factor analysis performed on the adapted peer support scale confirmed its unidimensionality (CFI > 0.9, SMR <0.10 at all time points). A peer support score at each time point was calculated for each student by averaging all items, higher scores reflecting greater feelings of peer support (average alpha=0.88).

Loneliness. Loneliness was measured using 7 items adapted from Cassidy and Asher's (1992) 15-item loneliness at school scale. The 7 items (I feel alone at school; I have lots of friends to talk to at school; It is hard for me to make friends at school; I have nobody to talk to in my classes; I do not have anyone to spend time with at school; I am lonely at school; I feel left out of things at school) invited responses on a 5-point scale ranging from 1=strongly disagree to 5 = strongly agree. A factor analysis performed on the adapted loneliness scale confirmed its unidimensionality (CFI>0.9, SMR<0.10 at all time points). A mean loneliness score was calculated at each time point for each student, higher scores reflected greater feelings of loneliness (average alpha=0.72).

Connectedness. The connectedness to school scale comprised 4 items adapted from the Resnick and McNeely (1997) 6-item School Connectedness Scale (I feel close to people at school; I feel like I am part of this school; I am happy to be at school; the teachers treat students fairly) with a 5-point response scale (1=never, 2=unsure, 3=sometimes, 4=usually, 5=always). The unidimensionality of the adapted scale was confirmed in a factor analysis (CFI>0.9, SMR<0.10 at all time points). For each student at each time point an average school connectedness score was calculated, with a higher score reflecting greater feelings of connectedness to their school (average alpha=0.80).

Safety. Safety at school was a single item adapted from the Peer Relations Questionnaire (Rigby & Slee, 1998) and measured on a 3-point scale (1=No, I never feel safe at school, 2=Yes, some of the time, 3=Yes, all or most of the time) for each time point with a higher value reflecting greater feelings of safety at school.

Data collection. All Grade 7 students enrolled in the 21 recruited secondary schools were invited to participate in the SSP via a package mailed to their parent. Each package contained a parental information statement and consent form, a student questionnaire, instructions on how to complete the questionnaire, a reply paid envelope for them to return their questionnaire once completed and a contact phone number of a researcher. In Grades 8 and 9, students with consent completed hard copy questionnaires in the classroom during a normal school period administered by trained research personnel according to a strict procedural and verbal protocol. Students without consent were given alternate learning activities.

Statistical analysis

Analyses were conducted using MPlus v6 and SPSS v17. Cross-lagged models within the Structural Equation Modelling (SEM) framework were used to model causal paths with longitudinal data between the social health factors and victimisation. These models combine autoregressive and cross-lagged effects. The autoregressive effects describe the stability in the outcome (e.g., victimisation) over time. The cross-lag effects describe the associations between the constructs between one time point and the next controlling for prior levels of the outcome (e.g., from victimisation to loneliness), controlling for previous levels of victimisation, and vice versa with loneliness as the outcome. In addition, correlations between the residuals at the second and subsequent time points describe the concurrent associations between the non-stable components of the constructs. Due to the skewed nature of the victimisation and social health variables, the MLR estimator (robust maximum likelihood parameter estimator) was used within the cross-lagged models as it implements non-normality robust standard error calculations. Differences between study condition and gender were examined within the cross-lagged models to ensure models fit equally well and the associations were the same in the different groups. Social health factors were modelled separately to determine the individual relationships of the different social health variables and bullying victimisation. All 4 time-points were represented in all cross-lagged models tested to determine the direction of the associations between social heath factors and victimisation as observed at a later time point (see Figures 1-4). Additionally, the results from the cross-lagged models were examined to determine the point at which the relationships between victimisation and social health were the strongest, to determine whether there is a critical time point in which to intervene with targeted bullying intervention programs. Missing data ranged from 0.5% to 3% for each variable used in the analysis at each time point. Missing data were handled using the Expectation-Maximisation (EM) procedure in SPSS and Full Information Maximum Likelihood (FIML) estimation in MPlus, enabling the use of all students with at least one valid score in the analyses (data coverage ranged from 51% to 95% for each of the variance-covariance estimates).

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

Results

Table 2 lists the means and standard deviations for the social health factors and victimisation at the 4 time points. On average, students were not lonely (range of means 1.94 to 2.00, maximum score of 5), believed they were supported by their peers (range of means 2.44 to 2.58, maximum of 3), felt connected to their school (range of means 4.00 to 4.40, maximum of 5) and felt safe at school (range of means 2.70 to 2.82, maximum of 3) over the 4 time points. Most students did not report being frequently victimised (range of mean 1.28 to 1.47, maximum of 5), with 8-12% of students reporting that they were bullied frequently (every few weeks or more often) at each time point. A significant drop in bullying victimisation was found between the end of primary school and the beginning of secondary school (t=4.44, p<0.001), with victimisation significantly increasing between the start of secondary school and the end of Grade 8 (t=6.28, p<0.001).

As shown in Table 3, at all time points significant correlations were found between the 4 social health factors (all p<.01). Bivariate correlation coefficients describing the concurrent relationships between the social health factors and bullying victimisation were also significant (all p<.01), and showed a temporary significant decrease in association over the transition from primary to secondary school, as the beginning of Grade 8 values were lower than for the other time points for each of the social health variables' associations with victimisation.

Cross-lagged models

As a first step, the cross-lagged models were tested for invariance across gender groups and study condition, to determine whether the associations between the social health variables and victimisation were equivalent across these groups and therefore, whether one model could be fitted for each social health variable using the combined gender and study condition data. Tests using the Satorra Bentler Scaled Chi-square were non-significant (all p>0.05, see Table 4), indicating it was not necessary to fit separate models for the gender groups or the study conditions. The subsequent cross-lagged models for each of the social health variables and victimisation are given in Figures 1-4. Fit indices indicate good model fit for all variables and victimisation (all CFI>0.9, all root mean square error of approximation (RMSEA)<0.08) (see Table 4).

Considering the associations over the transition period, the model results reveal that victimisation in the last year of primary school is associated with loneliness ([beta]=0.08, p=0.035, Figure 1), feeling less connected to secondary school ([beta]=-0.06, p<0.001, Figure 3) and feeling less safe at the start of secondary school ([beta]=-0.15, p<0.001, Figure 4). Peer support was the only variable to show an association in the opposite direction with greater peer support ([beta]=-0.09, p=0.003) in the last year of primary school associated with less victimisation at the start of secondary school (Table 3). The magnitude of these associations over the transition from primary to secondary school were significant, though small, with the magnitude of associations increasing over the first 2 years of secondary school.

Reciprocal relationships exist between social health factors and bullying victimisation in the first 2 years of secondary school. The coefficients between the beginning and end of the first year of secondary school for victimisation to the social heath factors (loneliness: [beta]=0.17, p<0.001; peer support: [beta]=-0.15, p<0.001; connectedness to school: [beta]=-0.10, p<0.001; feeling safe at school: [beta]=-0.20, p<0.001) were stronger than those from the social health factors to victimisation (loneliness: [beta]=0.09, p<0.001; peer support: [beta]=-0.11, p<0.001; connectedness to school: [beta]=-0.07, p<0.001; feeling safe at school: [beta]=-0.05, p=0.023).

Between the end of the first year of secondary school and the end of the second year of secondary school, the coefficient for bullying victimisation to loneliness ([beta]=0.14, p<0.001) was higher than that from loneliness ([beta]=0.06, p=0.023) to bullying victimisation. The reciprocal coefficients for peer support were similar in magnitude ([beta]=-0.10, p<0.001 and [beta]=-0.09, p<0.001, respectively), and the same for feeling safe at school ([beta]=-0.10, p<0.001). However, the coefficient from connectedness ([beta]=-0.11, p<0.001) to bullying victimisation was stronger than from bullying victimisation to connectedness ([beta]=-0.08, p<0.001. This suggests that after 1 to 2 years of secondary school, victimisation is more likely to occur when a student feels disconnected than feeling disconnected contributing to victimisation.

The increases in the correlations of the residuals between victimisation and social health in the cross-lag models indicate the strength of the associations increase over time i.e. victimisation is associated with increasingly higher levels of loneliness and increasingly lower levels of peer support, connected to school and feelings of safety at school.

Discussion

The significant reciprocal associations found in the cross-lag models in the first 2 years of secondary school indicate the need to intervene, not only during the transition period but also within the first 2 years of secondary school, to improve students' social health outcomes (and thereby reduce their victimisation in Grades 8 and 9) as well as reduce their victimisation experiences to avoid flow on effects on their social health. These associations were present even after controlling for the stability over time in victimisation levels and the social health outcomes, and therefore are not simply a consequence of associations in primary school between victimisation and the social health outcomes. Based on the magnitude of the coefficients (small, though statistically significant), the strongest associations in the direction from victimisation to the social health variables occurred from the beginning to the end of Grade 8. Reducing students' victimisation in Grade 8 may, therefore, protect against poorer outcomes on the social health variables, such as feeling safe at school, for students in the first year of secondary school. The fact that the auto-regressive coefficients for victimisation and social health are much larger than the cross-lag coefficients and the cross-lag coefficients are small in terms of effect sizes suggests that both victimisation and social health are fairly well-established by late primary school. The persistent and strengthening associations between bullying victimisation and social health in the period following transition indicate that intervention is needed both during and following transition.

Individual level social health variables and victimisation

The results of this study support the hypothesis that students who had less peer support at the end of primary school would experience higher levels of victimisation at the beginning of secondary school. However, this study found loneliness at the end of primary school was not associated with greater victimisation at the start of secondary school rather that victimisation at the end of primary school is associated with greater loneliness at the beginning of secondary school. Reciprocal relationships between victimisation and loneliness and peer support existed in the first 2 years of secondary school with the path from victimisation to loneliness stronger than the alternate path. These results support previous cross-sectional studies that show students' who are bullied report greater loneliness as they may experience avoidance by others who fear being bullied themselves or through losing social status amongst their peers (Nansel et al., 2001). However, similar to the findings reported by Nansel and colleagues (2001), associations were also found in the opposite direction, with students in this study who are socially isolated and lonely more likely to be victimised.

These results are consistent with previous studies where the majority of students in late primary and early secondary school report supportive attitudes towards those being victimised, but these attitudes become less supportive with age (Gini et al., 2008), with an increasing number of students over time disliking those who are victimised (Rigby & Slee, 1991). Importantly, it has been found that the number and quality of friends and being liked by peers may protect against victimisation (Pellegrini & Bartini, 2000), suggesting the development of healthy and multiple friendship groups would reduce the negative impact of victimisation and reduce the probability of further victimisation and loneliness. Victimised students report lower peer acceptance and family support (Perren & Hornung, 2005) as well as lower teacher support (Rigby, 2000). Peers can reduce bullying by intervening and helping the person being victimised (Salmivalli, 1999), while student, parent and teacher support can buffer victimised students from internalising distress (Rigby, 2000). Given the impact of student victimisation and over the transition period, the development of stress coping skills will promote healthy coping in victimised students. Thus, intervention approaches that provide opportunities for students to practice and develop coping behaviours will greatly assist with dealing with victimisation.

School-level social health variables and victimisation

The strongest paths in this study were found between feeling connected and safe at school at the end of the first year of secondary school associated with decreased victimisation at the end of the second year of secondary school. Conversely, the paths leading from feeling less connected and safe at school to victimisation were stronger from the end of the first year of secondary school to the end of the second year of secondary school. Similarly, in a cross-sectional study of secondary school students, Bradshaw and colleagues (2008) found that victimised students were more likely to report feeling disconnected and unsafe at school and that victimisation at primary school was associated with lower feelings of school connectedness and safety across the transition period to secondary school. During transition, students typically experience a new social environment moving from small, personal primary school environments to secondary schools which are generally larger (Pereira & Pooley, 2007) and more impersonal (Mizelle, 2005), with teachers, classrooms and often classmates constantly changing (Simmons, Burgeson, Carlton-Ford, & Blyth, 1987). Consistent with this study, during this time students report a reduced sense of school belonging (Pereira & Pooley, 2007) and connectedness (O'Brennan & Furlong, 2010).

School connectedness and feeling safe at school need to be actively fostered in primary school and during the first few years of secondary school through a strong school ethos of care, clear social support systems where relationships promote health and well-being and positive classroom management (Cowie, Naylor, Talamelli, Chauhan, & Smith, 2002). Increasing adult supervision, enhancing and encouraging the ability of adults to prevent, detect and intervene in bullying incidents, enabling students to support victimised students, easily report bullying and effectively communicating to students and the other members of the school community the school's bullying prevention policy and actions will also help to reduce victimisation and increase the students' perceived sense of safety at school (Beran & Tutty, 2002; Bradshaw et al., 2008). Adult supervision needs to occur in an appropriate manner, which allows students to develop autonomy. Consistent and clear, but not intrusive, supervision is important in hotspots where bullying is known to occur. The perception that school staff are pro-active in their efforts to intervene and reduce bullying (Beran & Tutty, 2002) can influence the students' perceived sense of safety at school. In this study, school connectedness and feeling safe at school declined steadily after the transition to secondary school. This highlights the need for further longitudinal research to better understand the relationship between connectedness to school and feeling safe at school and bullying victimisation over time and how actions taken by the school increases feelings of connectedness and safety and their influences on victimisation.

The transition period

There was an unexpected significant decrease in overall reported victimisation between the end of Term 4 primary school and the start of secondary school, but then victimisation significantly increased again over the first year of secondary school. This increase suggests that the first year of secondary school represents a critical time to focus on strategies to prevent bullying and support students who are victimised and that the final years of primary school need to focus on building students' skills to prevent, discourage and manage bullying behaviour. During this time, reciprocal relationships between bullying victimisation and all measured social health factors commence.

Peer groups are likely to change during the transition period. This study suggests that this socially challenging time for adolescents is also when the associations between social health variables and victimisation may already be well-established from primary school. Due to the increased importance placed on peer relationships during early adolescence and the risk of sustained victimisation (Rueger, Malecki, & Demaray, 2011), provision of social support to reduce the negative impact of transition effects and minimise harm to higher risk students is essential at this time (D. Espelage et al., 2000).

Strengths and limitations

There are several strengths of this study. Most importantly, the longitudinal nature of the research design over the transition from primary into secondary school enabled the investigation of victimisation and social health at a time that can be challenging for many students. Moreover, these findings are robust due to the large sample of students (90%) who completed questionnaires in at least three data collection points. The limitations of this study include relying on student self-report of bullying victimisation and associated factors over the adolescent years. The research was conducted with a sample of Catholic secondary schools within the Perth metropolitan area which affects the generalisability of results given that they represent 11% of all secondary schools (n= 186) in Perth, Western Australia. Future research needs to include students from metropolitan and non-metropolitan areas and Government and non-Government schools to clarify and validate the nature of the relationships found in this research, as Catholic schools may support the transition process for students differently compared to other schools. The students who remained in Catholic schools during the transition to secondary school may also have different characteristics from those students who left. It is recommended that future examinations of adolescence and transition consider developmental issues that may impact on social relationships at this time.

School activities over the transition period will likely influence students" experiences. This current study collected information from participating secondary schools on transition activities offered to incoming students (e.g. school information, orientation day, student strategies, student support, family strategies) but did not collect information on transition activities offered by primary schools. This study also lacks information on the association between social health variables and victimisation before Grade 7 and therefore is limited in its ability to draw causal inference. All students involved in the study were included in this analysis regardless of study condition. The inclusion of the intervention students did not statistically impact on the results (factor invariance in the models demonstrated equivalence of correlations between the groups) and hence have not impacted on the findings. The confounding of aggressive victims on the victimisation results in this study is also thought to be minimal due to the small proportion (between 0 and 4%) of students who reported being victimised at any time point and also reported victimising others 'every few weeks' or more often. Finally, the questionnaires used in this study were not specifically designed for the secondary analyses reported in this paper.

Conclusion

School-based bullying prevention and intervention programs are a high priority to minimise student harm from bullying during adolescence. The findings from this study suggest a critical time to intensify whole-school bullying intervention programs, focused on decreasing bullying and loneliness, while increasing peer support, school connectedness and school safety, is during the transition to and particularly within the first 2 years of secondary school for higher risk students. The results indicate that increasing peer support and feelings of safety at school during the first year of secondary school may help to reduce victimisation.

DOI: 10.1177/0004944113485835

Acknowledgements

We thank Melanie Epstein and Stacey Waters for their contributions to the Supportive Schools Project (SSP), and the SSP study schools and their staff, parents, and students.

Declaration of conflicting interests

None declared.

Funding

The SSP Project and this study were funded by the Western Australian Health Promotion Foundation (Healthway) and the research supported by the Child Health Promotion Research Centre (CHPRC) at Edith Cowan University, Western Australia. This research was funded through a Western Australian Health Promotion Foundation Scholarship and supported by the Department of Industry, Innovation, Science, Research and Tertiary Education. Special thanks are given to all CRN partners for the contribution to the development of this work.

References

Akos, P., & Galassi, J. (2004). Middle and high school transitions as viewed by students, parents and teachers. Professional School Counselling, 7(4), 213-221.

Barber, B., & Olsen, J. (2004). Assessing the transitions to middle and high school. Journal of Adolescent Research, 19(3), 3-30.

Barton, J., & Rapkin, B. (1987). The transition to junior high school: A longitudinal study of self-esteem, psychological symptomatology, school life, and social support. Child Development, 58(5), 1235-1243.

Benner, A., & Graham, S. (2009). The transition to high school as a developmental process among multiethnic urban youth. Child Development, 80(2), 356-376.

Beran, T., & Tutty, L. (2002). Children's reports of bullying and safety at school. Canadian Journal of School Psychology, 17(2), 1-14.

Berguno, G., Leroux, P., McAinsh, K., & Shaikh, S. (2004). Children's experience of loneliness at school and its relation to bullying and the quality of teacher interventions. The Qualitative Report, 9, 483-499.

Bosworth, K. (1999). Factors associated with bullying behavior in middle school students. Journal of Early Adolescence, 19(3), 341-362.

Bradshaw, C., O'Brennan, L., & Sawyer, A. (2008). Examining variation in attitudes toward aggressive retaliation and perceptions of safety among bullies, victims, and bully/victims. Professional School Counseling, 12, 1-21.

Cassidy, J., & Asher, S. (1992). Loneliness and peer relations in young children. Child Development, 63(2), 350-365.

Cohen, J., & Smerdon, B. (2009). Tightening the dropout tourniquet: Easing the transition from middle to high school. Preventing School Failure, 53(3), 177-184.

Cowie, H., Naylor, P., Talamelli, L., Chauhan, P., & Smith, P. (2002). Knowledge, use of and attitudes towards peer support. Journal of Adolescence, 25, 453-467.

Crockett, L., Petersen, A., Graber, J., Schulenberg, J., & Ebata, A. (1989). School transitions and adjustment during early adolescence. The Journal of Early Adolescence, 9(3), 181-210.

Cross, D., Shaw, T., Hearn, L., Epstein, M., Monks, H., Lester, L. & Thomas, L. (2009). Australian covert bullying prevalence study. Perth, Australia: Child Health Promotion Research Centre, Edith Cowan University.

Eisenberg, M., Neumark-Sztainer, D., & Perry, C. (2003). Peer harassment, school connectedness, and academic achievement. Journal of School Health, 73(8), 311-316.

Ellickson, P., & Hawes, J. (1989). An assessment of active versus passive methods for obtaining parental consent. Evaluation Review, 13(1), 45-55.

Espelage, D., Bosworth, K., & Simon, T. (2000). Examining the social context of bullying behaviors in early adolescence. Journal of Counseling and Development, 78(3), 326-333.

Espelage, D., & Holt, M. (2001). Bullying and victimization during early adolescence: Peer influences and psychosocial correlates. In R. A. Geffner, M. Loring, & C. Young (Eds.), Bullying behavior: Current issues, research, and interventions (pp. 123-142). New York, NY: Haworth Press.

Gini, G., & Pozzoli, T. (2009). Association between bullying and psychosomatic problems: A meta-analysis. Pediatrics, 123(3), 1059-1065.

Gini, G., Pozzoli, T., Borghi, F., & Franzoni, L. (2008). The role of bystanders in students' perception of bullying and sense of safety. Journal of School Psychology, 46(6), 617-638.

Glover, D., Gough, G., Johnson, M., & Cartwright, N. (2000). Bullying in 25 secondary schools: Incidence, impact and intervention. Educational Research, 42(2), 141-156.

Hodges, E., & Perry, D. (1996). Victims of peer abuse: An overview. Journal of Emotional and Behavioural Problems, 5, 23-28.

Kaltiala-Heino, R., Rimpela, M., Rantanen, P., & Rimpela, A. (2000). Bullying at school--An indicator of adolescents at risk for mental disorders. Journal of Adolescence, 23(6), 661-674.

Ladd, G., Kochenderfer, B., & Coleman, C. (1996). Friendship quality as a predictor of young children's early school adjustment. Child Development, 67(3), 1103-1118.

Malecki, C., & Demaray, M. (Eds.). (2004). The role of social support in the lives of bullies, victims, and bully-victims. Mahwah, N J: Lawrence Erlbaum Associates.

Malecki, C., Demaray, M., & Davidson, L. (2008). The relationship among social support, victimization, and student adjustment in a predominantly latino sample. Journal of School Violence, 7(4), 48-71.

McNeely, C., Nonnemaker, J., & Blum, J. (2002). Promoting school connectedness: Evidence from the National Longitudinal Study of Adolescent Health. Journal of School Health, 72(4), 138-147.

Mizelle, N. B. (2005). Moving out of middle school. Educational Leadership, 62(7), 56-60.

Nansel, T., Overpeck, M., Pilla, R., Ruan, J., Simons-Morton, B., & Scheidt, P. (2001). Bullying behaviors among US youth: Prevalence and association with psychosocial adjustment. Journal of the American Medical Association, 285(16), 2094-2100.

O'Brennan, L., & Furlong, M. (2010). Relations between students' perceptions of school connectedness and peer victimization. Journal of School Violence, 9(4), 375-391.

Olweus, D. (1996). The revised Olweus Bully/Victim Questionnaire. Bergen, Norway: Mimeo, Research Center for Health Promotion (HEMIL), University of Bergen.

Osterman, K. (2000). Students' need for belonging in the school community. Review of Educational Research, 70, 323-367.

Pellegrini, A. (2002). Bullying, victimisation, and sexual harassment during the transition to middle school. Educational Psychologist, 37(3), 151-163.

Pellegrini, A., & Bartini, M. (2000). A longitudinal study of bullying, victimization, and peer affiliation during the transition from primary school to middle school. American Educational Research Journal, 37(3), 699-725.

Pereira, A., & Pooley, J. (2007). A qualitative exploration of the transition experience of students from a high school to a senior high school in rural Western Australia. Australian Journal of Education, 51(2), 162-177.

Perren, S., & Hornung, R. (2005). Bullying and delinquency in adolescence: Victims' and perpetrators' family and peer relations. Swiss Journal of Psychology/Schweizerische Zeitschrift fur Psychologie/ Revue Suisse de Psychologie, 64(1), 51-64.

Resnick, M., Bearman, P., Blum, R., Bauman, K., Harris, K., Jones, J., ... & Udry, J. R. (1997). Protecting adolescents from harm findings from the National Longitudinal Study on Adolescent Health. The Journal of the American Medical Association, 278(10): 823-832.

Rigby, K. (2000). Effects of peer victimization in schools and perceived social support on adolescent well-being. Journal of Adolescence, 23(1), 57-68.

Rigby, K., & Slee, P. (1991). Bullying among Australian school children: Reported behavior and attitudes toward victims. Journal of Social Psychology, 131(5), 615-627.

Rigby, K., & Slee, P. (1998). The Peer Relations Questionnaire. Point Lonsdale, Australia: The Professional Reading Guide for Educational Administrators.

Rueger, S., Malecki, C., & Demaray, M. (2011). Stability of peer victimization in early adolescence: Effects of timing and duration. Journal of School Psychology, 49(4), 443-464.

Salmivalli, C. (1999). Participant role approach to school bullying: Implications for interventions. Journal of Adolescence, 22, 453-459.

Simmons, R., Burgeson, R., Carlton-Ford, S., & Blyth, D. (1987). The impact of cumulative change in early adolescence. Child Development, 58(5), 1220-1234.

Smith, P. K. (2004). Bullying: Recent developments. Child and Adolescent Mental Health, 9(3), 98-103.

Solberg, M., & Olweus, D. (2003). Prevalence estimation of school bullying with the Olweus Bully/ Victim Questionnaire. Aggressive Behavior, 29(3), 239-268.

Storch, E., Brassard, M., & Masia-Warner, C. (2003). The relationship of peer victimization to social anxiety and loneliness in adolescence. Child Study Journal, 33(1), 1-18.

Swearer, S., Peugh, J., Espelage, D., Siebecker, A., Kingsbury, W., & Bevins, K. (2006). A social-ecological model for bullying prevention and intervention in early adolescence: An exploratory examination. In S. Jimerson, & M. Furlong (Eds.), Handbook of school violence and school safety: From research to practice (pp. 333-356). Mahwah, N J: Lawrence Erlbaum Associates.

Tremblay, R., Nagin, D., Seguin, J., Zoccolillo, M., Zelazo, P., Boivin, M., ... Japel, C. (2004). Physical aggression during early childhood: trajectories and predictors. Paediatries, 114(1), 43-50.

Woods, S., Done, J., & Hardeep, K. (2009). Peer victimisation and internalising difficulties: The moderating role of friendship quality. Journal of Adolescence, 32(2), 293-308.

You, S., Furlong, M., Felix, E., Sharkey, J., Tanigawa, D., & Green, J. (2008). Relations among school connectedness, hope, life satisfaction, and bully victimization. Psychology in the Schools, 45(5), 446-460.

Leanne Lester

Post Doctoral Research Fellow and Biostatistician, Child Health Promotion Research Centre, Edith Cowan University, Australia

Donna Cross

Professor of Child and Adolescent Health, Child Health Promotion Research Centre, Edith Cowan University, Australia

Julian Dooley

Associate Director, Sellenger Centre for Research in Law, Justice and Social Change, Edith Cowan University, Australia

Therese Shaw

Senior Research Fellow and Biostatistician, Child Health Promotion Research Centre, Edith Cowan University, Australia

Corresponding author:

Leanne Lester, Child Health Promotion Research Centre, Edith Cowan University, Mount Lawley, WA 6050, Australia.

Email: I.lester@ecu.edu.au

Table 1. Consent and questionnaire completion rates.

                                           Beg. of
                          End of Grade 7   Grade 8
                          (primary)        (secondary)
                          n (%)            n (%)

Total eligible            3769             3462
Total with                3462 (92)        3462 (100)
  parental consent
Left school               0                83
Total sample              3462             3379
Total completed           2078 (60)        3317 (98)
Total completed male      1003 (48)        1675 (50)
Total completed female    1074 (52)        1639 (50)

                          End of        End of
                          Grade 8       Grade 9
                          n (%)         n (%)

Total eligible            3379          3332
Total with                3379 (100)    3332 (100)
  parental consent
Left school               47            163
Total sample              3332          3169
Total completed           3263 (98)     3025 (95)
Total completed male      1651 (51)     1542 (51)
Total completed female    1608 (49)     1478 (49)

Table 2. Descriptive statistics of social health factors
and victimisation.

                      End of Grade 7    Beg. of Grade 8
                      (primary)         (secondary)
                      n = 2037-2070     n = 3213-3301
                      Mean (SD)         Mean (SD)

Loneliness (1-5)      1.94 (0.47)       1.97 (0.52)
Peer support (1-3)    2.58 (0.34)       2.52 (0.35)
Connectedness (1-5)   4.40 (0.60)       4.20 (0.75)
Safety (1-3)          2.82 (0.41)       2.78 (0.43)
Victimisation (1-5)   1.42 (0.59)       1.33 (0.52)

Frequency of victimisation in the past term (10 weeks).

                      n (%)             n (%)

Not victimised        1305 (64)         2310 (70)
Once or twice         509(25)           744(22)
Every few weeks       127(6)            95 (3)
Once a week           46 (2)            74 (3)
Most days             54 (3)            61 (2)

                      End of Grade 8   End of Grade 9
                      n = 3162-3254    n = 2935-3015
                      Mean (SD)        Mean (SD)

Loneliness (1-5)      1.96 (0.55)      2.00 (0.62)
Peer support (1-3)    2.44 (0.39)      2.57 (0.43)
Connectedness (1-5)   4.20 (0.77)      4.00 (0.84)
Safety (1-3)          2.72 (0.50)      2.70 (0.52)
Victimisation (1-5)   1.45 (0.62)      1.53 (0.79)

Frequency of victimisation in the past term (10 weeks).

                      n (%)            n (%)

Not victimised        2051 (64)        1910 (64)
Once or twice         818 (25)         719 (24)
Every few weeks       167(5)           129 (4)
Once a week           100 (3)          123 (4)
Most days             93 (3)           120 (4)

Table 3. Bivariate correlations between social health factors
and victimisation.

                    Peer
                  support     Loneliness   Connectedness    Safety

End of Grade 7
Peer support     I
Loneliness       -0.604 **    I
Connectedness    0.535 **     -0.490 **    I
Safety           0.349 **     -0.411 **    0.430 **        I
Victimisation    -0.497 **    0.558 **     -0.495 **       -0.485 **
Beginning of
  Grade 8
Peer support     I
Loneliness       -0.545 **    I
Connectedness    0.500 **     -0.520 **    I
Safety           0.363 **     -0.437 **    0.465**         I
Victimisation    -0.412 **    0.435 **     -0.369**        -0.417 **
End of Grade 8
Peer support     I
Loneliness       -0.560 **    I
Connectedness    0.457 **     -0.480 **    I
Safety           0.373 **     -0.426 **    0.487 **        I
Victimisation    -0.472 **    0.546 **     -0.397 **       -0.428 **
End of Grade 9
Peer support     I
Loneliness       -0.585 **    I
Connectedness    0.487 **     -0.449 **    I
Safety           0.420 **     -0.419 **    0.509 **        I
Victimisation    -0.539 **    0.595 **     -0.401 **       -0.461 **

* p <.05.

** p <.01.

Table 4. Satorra-Bentler scaled chi-square model fit test of gender
and study condition invariance for the cross-lagged models.

                                    Gender

                              T      df      p

Loneliness                  16.349    12    0.176
Peer support                 9.147    12    0.690
Connectedness to school     10.301    12    0.590
Safety at school            19.905    12    0.069

                                Study condition

                              T      df      p

Loneliness                   9.150    12    0.690
Peer support                 7.394    12    0.831
Connectedness to school      3.793    12    0.987
Safety at school             9.569    12    0.654

* p <.05.

** p <.01.