Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

School Belongingness and Mental Health Functioning across the Primary-Secondary Transition in a Mainstream Sample: Multi-Group Cross-Lagged Analyses

School Belongingness and Mental Health Functioning across the Primary-Secondary Transition in a... The relationship between school belongingness and mental health functioning before and after the primary-secondary school transition has not been previously investigated in students with and without disabilities. This study used a prospective longitudinal design to test the bi-directional relationships between these constructs, by surveying 266 students with and without disabilities and their parents, 6-months before and after the transition to secondary school. Cross-lagged multi-group analyses found student perception of belongingness in the final year of primary school to contribute to change in their mental health functioning a year later. The beneficial longitudinal effects of school belongingness on subsequent mental health functioning were evident in all student subgroups; even after accounting for prior mental health scores and the cross-time stability in mental health functioning and school belongingness scores. Findings of the current study substantiate the role of school contextual influences on early adolescent mental health functioning. They highlight the importance for primary and secondary schools to assess students’ school belongingness and mental health functioning and transfer these records as part of the transition process, so that appropriate scaffolds are in place to support those in need. Longer term longitudinal studies are needed to increase the understanding of the temporal sequencing between school belongingness and mental health functioning of all mainstream students. Citation: Vaz S, Falkmer M, Parsons R, Passmore AE, Parkin T, et al. (2014) School Belongingness and Mental Health Functioning across the Primary-Secondary Transition in a Mainstream Sample: Multi-Group Cross-Lagged Analyses. PLoS ONE 9(6): e99576. doi:10.1371/journal.pone.0099576 Editor: Kenji Hashimoto, Chiba University Center for Forensic Mental Health, Japan Received December 19, 2013; Accepted May 16, 2014; Published June 26, 2014 Copyright:  2014 Vaz et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This project was funded by a Doctoral scholarship provided by the Centre for Research into Disability and Society and the School of Occupational Therapy and Social Work, Curtin University, Perth, Australia. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * Email: [email protected] four times more often than their typically developing peers; with Mental Health Problems in Adolescence psychiatric disorders in young people with a disability often Worldwide estimates of mental health problems in children and undiagnosed and untreated [8]. Household-SES influences phys- youth range from 10–20% [1]. Australian figures report a 14% ical and mental health across the lifespan, with socially and prevalence in a national sample of 4–12 year olds, which rises to economically disadvantaged children and adults found to be an 19% in the 13–17 year old category [2] and 27% in the 18–24 increased risk for both physical and mental health problems [9– year old group [3]. These figures suggest that approximately one 12]. Thus, it is imperative that research studies account for within in four to five young Australians have a mental health problem [4]. group variability in mental health functioning of children and Mental health functioning of children and youth has been shown youth. to vary due to gender, presence of disability and household socio- Of concern is the growing evidence on the stability of mental economic standing (SES). For example, conduct disorder is the health problems in children and adolescents [13,14] and its most common psychiatric disorder in childhood, with three times longitudinal effects on mental health disorders, delinquency, as many boys as girls being affected [5]. During adolescence, girls crime, unemployment, homelessness and suicidal behaviour in have a higher prevalence of depression and eating disorders, and adulthood [14–22]. Mental health problems in children and engage more in suicidal ideation and suicide attempts than boys, adolescents could be antecedents of chronic, complex, disabling who are more prone to engage in high risk behaviours and commit and expensive complications in adult life. For these reasons, early suicide more frequently [6,7]. Young people with an intellectual detection of clinical and subclinical mental health issues is disability manifest behaviours and experiences which may be important. Most mental health disorders that are likely to persist indicative of mental health or psychological impairment three to PLOS ONE | www.plosone.org 1 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth into adult life emerge between ages 12 and 25 [23,24]. While early in primary school and a remixing of friendship networks and social intervention is more economical and cost-effective than later hierarchies. It is likely that students are forced to redefine their action [25], its effectiveness in some cases is modest [26,27] or fails sense of school belongingness after they transition to secondary to reach the majority of those most in need [26]. Australian data school. Whether poorer mental health functioning before the suggest that only one in four youth who need professional help transition is associated with poorer school belongingness thereafter actually get the help they need [28]. These facts, underscore the remains mainly unexplored. need to gain a deeper understanding of pathways in and out of childhood mental health problems [15,28–30]. Schools are an Aim and Objectives ideal setting for efficiently detecting children and adolescents with The current study extends the existing knowledge base on unidentified mental health problems because they offer the primary-secondary school transition by explicitly examining the opportunity to reach large numbers of students [2,30,31]. temporal relationships between school belongingness and overall mental health functioning after one year, by tracking a cohort with School belongingness and overall mental health and without disability, enrolled in the regular school system in functioning across primary-secondary school transition Western Australia (WA). We also tested the equivalence of these In recent years, school belongingness, referring to students’ relationships across gender, disability and household-SES. It was beliefs of being ‘‘personally accepted, respected, included, and hypothesised that: supported by others in the school social environment’’ [32], has emerged as an important factor associated with positive health (a) direct relationships would exist between concurrent percep- outcomes [33,34]. Cross-sectional studies document moderate tion of school belongingness and overall mental health associations between school belongingness and emotional distress functioning, before and after the transition; and and depression in typically developing adolescents [35–38], before (b) primary school belongingness would be related to overall and after accounting for personal and contextual factors, such as mental health functioning in early secondary school, even family-parent-belongingness, self-esteem and grade point average after accounting for prior mental health scores. [36,37]. Short term longitudinal studies present mixed findings on the directional relationships between these constructs. In some No hypothesis was made regarding the predictive role of mental studies [39,40] a unidirectional relationship has been documented; health in primary school on school belongingness a year later, due with school belongingness predicting selective prospective mental to the inconsistent empirical evidence on this issue. health components, depending on gender. For example, Shocket and colleagues [40] found that early adolescents’ perception of Method school belongingness predicted future depressive symptoms in boys and girls; anxiety in girls; and conduct problems in boys. Other A cohort study using a prospective, longitudinal design with two studies suggest bidirectional relationships between these con- data collection points was used [Primary school = Wave 1, and structs, which vary depending on the type of mental health domain Secondary school = Wave 2]. Students enrolled in the final year being measured [41]. Loukas and colleagues [41] presented of primary school in WA (class 6 or 7), in the academic years evidence of a bidirectional loop between school belongingness and commencing January 2006 or 2007, and due to transition to either conduct problems, but not depressive symptoms in 10–14 year old middle or secondary school in January 2007 or 2008, were typically developing youth [41]. Also, the positive effects of school considered for inclusion in the study. Inclusion was limited to belongingness have been found to extend to students’ home lives; regular schools in the educational districts of metropolitan Perth or concomitantly buffer the effects of family disadvantage on other major city centres of Western Australia (WA). Several functioning [42]; and prospectively protect them from involve- recruitment strategies were used to maximize reach and repre- ment in risk behaviours [43,44]. Consequently, a growing body of sentativeness. The current study is part of a larger study on the evidence with typically developing youth supports the interrela- factors associated with student academic, social-emotional and tionship between school belongingness and positive mental health participatory adjustment across the primary-secondary school outcomes. transition [49]. Details on the study design, recruitment and data Conspicuous in the above cited investigations on school collection have been published elsewhere [50]. For the ease of belongingness and mental health functioning, is the exclusion of readership, a brief overview is described below. students with disabilities in the study samples, despite their Wave 1 data collection occurred six months prior to the presence in the regular school system for several decades. transition to either middle or secondary school, with data collected Additional research is needed to authenticate the role of school from students (with and without disabilities) and a primary belongingness in the disability subgroup. Preliminary findings are caregiver (parent or guardian). Wave 2 data were collected 6 hopeful, showing school belongingness to be negatively associated months after the transition, using the same procedure and sample with emotional stress, suicide attempts, and violence amongst as Wave 1. students with learning disabilities [45]. Yet another gap is the Information was collected via survey questionnaires, primarily absence of evidence on the prospective benefits of fostering paper and pencil format. Informed written consent was obtained belongingness in primary school on overall mental health from school principals, parents, teachers, and written assent was functioning of all students after the transition to secondary school, obtained from students to participate in this study. In situations or whether there are student subgroups, based on gender, where the student declined to participate, even with parental disability status, or household-SES, that need additional support. consent, they were not included. All participants were made aware Students in western societies, including Australia, negotiate the that they were not obliged to participate, and were free to primary-secondary school transition at a time in development withdraw from the study at any time without justification or when they are striving to gain independence from their parents, prejudice. Ethics approval was obtained from Curtin University establish their unique identity [46,47], and gain approval and support from peers [48]. As a result of this school transition, Health Research Ethics Committee, in Western Australia (WA) (Reference number HR 194/2005). students experience a disruption of the secure peer network forged PLOS ONE | www.plosone.org 2 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth At Wave 1, data were collected from 395 students from 75 scores correlate positively with school success [32,59], lower levels primary schools across the Perth metropolitan area and major city of depression [40], and lower levels of anxiety [34]. Higher PSSM centres across WA. An attrition rate of 32.7% resulted in a Wave 2 scores indicate better perceived school belongingness. sample of 266 participants from 52 primary schools and Family demographics and school contextual 152 secondary schools. Chi-square and paired sample t-tests characteristics. Items were drawn from the Indicators of showed that the participants who continued to be involved in Social and Family Functioning Instrument Version-1 (ISAFF) the study at Wave 2 did not differ from those who discontinued [60] and Australian Bureau of Statistics (ABS, 2001) surveys, and involvement, on gender, disability, household SES-level, school used to provide family demographic information. Parents reported belongingness, and mental health functioning scores. The current details on the family demographic characteristics, residence post study uses data from the 266 students that answered both Wave 1 code, and their child’s disability. Information on the school sector, and 2 questionnaires. Access to the complete dataset can be post code number of students enrolled in each school, and obtained by contacting the first author. organisational structure at each school was obtained from The mean age of students sample at Wave 1 was 11.89 years Department of Education and Training, WA records. The sample (SD = 0.45 years, median = 12 years), and that at Wave 2 was 12.9 was categorised into three-income groups as per the median years (SD = 0.57 years, median = 13 years). Boys constituted income distribution based on the Australian Bureau of Statistics 46.6% (n = 124) of the sample; and 25.9% (n = 69) were reported [51] data. by a parent or primary caregiver to have a disability. The predominant disabilities included asthma (18.8%), auditory Data Management disability (15.9%), Attention Deficit Hyperactivity Disorder/ Data were managed and analysed using the SPSS Version 20.0 Attention Deficit Disorders (14.5%), learning disability (11.6%), and SAS Version 9.2 software packages. Only 1.8–2.5% of data Autism Spectrum Disorders (10.1%), and cerebral palsy (8.7%). were missing at scale levels. The estimation maximization The majority of the sample came from mid-range households, and algorithm and Little’s chi-square statistic revealed that the data reported a weekly income of $600–1,999 (58.3%, n = 154) [51]. were missing completely at random [61,62]. Missing data Under one-third of the sample (33%, n = 87) came from high-SES replacement was undertaken using guidelines recommended by households ($ 2000+/week) and 8.7%, n = 23 were from low-SES the SDQ tool developers (http://www.sdqinfo.org/c1.html). In groupings ($ 1–599 per week). the case of the school belongingness questionnaire, individual The sample represented 52 different primary schools and 77 mean score substitution was used [63]. The validity of the data different classes distributed across metropolitan Perth and other substitution techniques used was substantiated using sensitivity city centres of WA. Based on the Commonwealth Department of analyses. Education, Employment, and Workplace Relations measure of In the present study, the bidirectional associations between relative socio-economic advantage and disadvantage [67], 21.4% school belongingness and mental health functioning over one year (n = 57) of the sample came from schools located in the most were estimated by cross-lagged analyses, using the structural th affluent areas across Australia (10 decide), 44% (n = 117) came equation modelling program, AMOS 5.0. A critical preliminary th th from the 9 decile; 17.7% (n = 47) were from the 7–8 decide and step in the analysis was to investigate if data met the normality th 16.9% (n = 45) came from more disadvantaged areas (1–6 assumption. With regard to the normality assumptions of the Full decide). Forty-seven percent of the sample (n = 125) were enrolled Information Maximum Likelihood estimation procedure, the in the public schools, 29% (n = 77) in Catholic schools, and the normality of each variable was investigated in terms of its kurtosis remaining 24% (n = 64) in independent/private schools. There and skewness [64]. Box-cox transformations were undertaken to was a movement out of government schools into Catholic and normalise the PSSM and SDQ scores. In order to provide independent schools for secondary education; with 11.2% of clinically relevant information, standardized Beta values from students (n = 14) moving into Catholic schools and 28.8% (n = 36) multiple linear regression analyses have also been presented, using moving into independent schools for their secondary education. the original data. Data collection instruments Statistical Analyses Mental health functioning. The 25-item parent version of the Strengths and Difficulties Questionnaire (SDQ) was used to Characteristics of the sample measure student overall functioning across hyperactivity, emo- Descriptive statistics were used to summarise the profile of tional health, conduct problem and peer problem domains [52]. participants. This version has moderate to high internal consistency scores (a = .70–80) [53], and is reported to be more sensitive than the Testing for the effects of nesting of students on mental Child Behaviour Check List [54] in detecting inattention and health and school belongingness hyperactivity, and equally effective in detecting internalising and In order to test for the effect of clustering of students, i.e., externalising problems in children and adolescents [55]. Estab- nesting of students in classes within schools on their school lished reliability and validity of the SDQ makes it a useful brief belongingness (PSSM) and mental health functioning (SDQ) screening measure of adjustment and psychopathology in children scores, a Hierarchical Linear Model was fitted using the mixed and adolescents [53,55–57]. Higher SDQ scores indicate poorer procedure in SAS. The class-level Intra Class Correlation mental health functioning. Coefficients (ICC) for PSSM and SDQ were obtained, after School belongingness. The 18-item, Psychological Sense of adjustment for gender, disability, and household-SES. School Membership scale (PSSM) was used to assess students’ Interrelationship between school belongingness (PSSM) and perceptions of belongingness in school [32]. The PSSM has satisfactory internal consistency (a = .803) [32]. Test-retest reli- mental health functioning scores (SDQ). Pearson correlation ability indices of .78 (4-week interval) [58], and .56 and .60 for coefficients were used to identify associations between the SDQ boys and girls respectively (12-month interval) have been and PSSM scores at and between each wave. A two-factor analysis documented in early adolescent samples [40]. The total PSSM of variance with and without interaction terms was run to test the PLOS ONE | www.plosone.org 3 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth within-group variability in SDQ due to gender, disability, and data without comparison to a reference model whereas an household-SES. incremental fit index compares the target model to a more Testing the hypothesized model of the relationship between restricted baseline model [67]. Both these indexes take into school belongingness and mental health functioning. Autoregres- account model complexity, which is an important property for sive cross-lagged panel analysis was performed to study the comparing several alternative models with different degrees of reciprocal relationship between school belongingness and mental complexity. According to criteria outlined by Hu and Bentler [67], health functioning across the primary-secondary school transition. a good fitting model has NNFI values of .95 or greater, RMSEA The path-diagram of the autoregressive cross-lagged model used in values smaller than .06, and a CFI greater than or equal to .95. In this study is presented in Figure 1. reporting on evidence of invariance, two criteria were used. Firstly, Cross-lagged panel analysis allows examination of the cross- the multi-group model must exhibit an adequate fit to the data. lagged paths while controlling for cross-time stability of each of the Secondly, the determination of multi-group invariance is based on variables. In each of the models, the exogenous variables of Wave delta CFI; that is, when the differences in CFI values between 1, which included school belongingness (PSSM) and mental health models are less than .01 [68]. functioning (SDQ), were freely correlated. The residuals (error variances) of all Wave 2 variables were also correlated, due to Results auto-correlation effects. Stability paths from each of the Wave 1 constructs to their respective Wave 2 outcomes were included to Testing for the effects of nesting of students on mental partial out the effects of baseline adjustment problems. The health and school belongingness inclusion of stability paths provides a stringent test of the Wave 1 A total of 52 different schools, and 77 different classes were influences and results in the examination of change in the variable involved in Wave 1. In order to test for the effect of clustering of of interest. To test the contribution of prior school belongingness students, i.e., nesting of students in classes within schools, a (PSSM) to future mental health functioning (SDQ), paths from Hierarchical Linear Model was fitted using the mixed procedure Wave 1-PSSM to Wave 2-SDQ were included. The opposite in SAS. The class-level Intra Class Correlation Coefficients (ICC) direction of associations, path from Wave 1-SDQ to Wave 2- for school belongingness and mental health functioning scores PSSM was also simultaneously estimated, but not presented in were obtained (after adjustment for the demographic data: gender, Figure 1. disability, and household-SES). The ICC for each model was low, Multi-group invariance (equivalence) of the baseline ranging from 0–12%, showing that the contribution of the model (Model 1). To examine the equivalence of the hypoth- clustering to the overall variance was small, and therefore the esized model across subgroups, namely, gender, disability and clustering appeared to have minimal effect on the relationships household-SES, parameters were simultaneously estimated for between the student-level variables and school belongingness and each subgroup, respectively. The fit of this simultaneously mental health functioning scores. Hence, further analyses were estimated unconstrained model provides the baseline value for undertaken at the level of the individual student. each subgroup against which all subsequently specified models are compared. A fully constrained model, in which all parameters Characteristics of the sample: Within-group variability in (factor variances, factor covariances, and error covariances) were mental health functioning constrained or specified to be equivalent across subgroups, was The mental health functioning scores (SDQ) of the students then calculated. x difference tests were used to determine involved in the current study was better than those found in an significant differences between the unconstrained and constrained Australian community sample for this age range [53,69]. Within- models of each subgroup. group variability interactions were not statistically significant; Model Evaluation Criteria. To determine the fit of the hence only the main effects were included in the final models. In models, criteria were adopted from several sources. Because x is the case of Wave 2-SDQ scores, significant differences due to influenced by sample size, we examined the x /degrees of freedom 2 2 gender, F (1,256) = 4.30, p = 0.04, disability, F (1,256) = 49.95, (df) ratio (x /df) rather than the significance of the x alone [65]. p =,.001, and household-SES, F (2,254) = 3.77, p = 0.02 were Furthermore, we also used the Non-Normed Fit Index (NNFI) [66]. Additionally, fit was evaluated by one absolute fit index (the found. Boys (M= 8.88, SE = .45) had worse Wave2-SDQ scores than girls (M= 7.65, SE = .44); and students with disability had Root Mean Square Error of Approximation, RMSEA) and one incremental fit index (the Comparative Fit Index, CFI). An worse scores (M= 10.61, SE = .58) than those without disability absolute fit index assesses how well a model reproduces the sample (M= 5.91, SE = .35). Students from low-SES households Figure 1. Cross-lagged relationship between PSSM and SDQ across the primary-secondary school transition. doi:10.1371/journal.pone.0099576.g001 PLOS ONE | www.plosone.org 4 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth (M= 8.90, SE = 1.18) had significantly poorer Wave2-SDQ scores excellent fit to the data. x difference tests found no significant differences between the unconstrained and constrained models of than their peers from high-SES (M = 7.12, SE = .55, p = .05), but not mid-SES households (M= 7.52, SE = .420, p..05). each subgroup, suggesting invariance of the baseline model (Figure 2) across gender, disability and household-SES. Interrelationship between school belongingness (PSSM) Discussion and mental health functioning scores (SDQ) The means, standard deviations and correlation matrix for all The present study extends existing research by providing study variables without adjustment for gender, disability and evidence that students’ ratings of belongingness in the final year household-SES are presented in Table 1. School belongingness of primary school contributes to change in their mental health (PSSM) was concurrently and longitudinally associated with functioning a year later. The beneficial effect of primary school mental health functioning (SDQ) at both waves of the study. belongingness on subsequent mental health functioning was Early adolescents reporting higher levels of school belongingness evident for the entire population of mainstream students, even (higher PSSM) also reported better mental health functioning after accounting for their prior mental health scores and the cross- (lower SDQ). Examination of the cross-time stability of the time stability in mental health functioning and school belonging- variables indicated that the magnitude of the correlations was ness scores. moderate for students’ perceptions of school belongingness (PSSM, Findings of the current study corroborate a large body of r=.49), and larger for mental health functioning (SDQ, r=.77). evidence on the significance of boosting school belongingness as a mental health promotion strategy not only in typically developing Testing the hypothesized model of the relationship students [36,39,40,70,71], but also students with disabilities. These results are of significance given current estimates that psychiatric between school belongingness and mental health disorders in young people with disabilities are often undiagnosed functioning (Figure 1) and untreated, despite the fact that these students manifest Figure 2 presents the most parsimonious baseline model that 2 behaviours and experiences indicative of mental illness or best fitted the data [x (1, n = 266) = .716, n.s.; CFI = 1.00; psychological impairment three to four times more often than RMSEA = .000; AIC = 26.716]. As shown in Fig. 2, the stability their typically developing peers [8]. Several theoretical underpin- paths were positive and significant, and inter-correlations among nings may explain the results. Students who sense a bonding in the two exogenous variables were significant. The error variance school are more likely to forge supportive relationships with between the endogenous variables was significant and in the teachers [32,72], associate with pro-social peer groups [73] and expected direction. Regarding the cross-lagged paths, Wave 1- are more likely to have better mental health functioning [74]. PSSM was associated with lower levels of Wave 1-SDQ, even after Students with social attachment to the school could be expected to controlling for baseline levels of all variables and for their cross- feel committed to its goals, norms, and morals [75–77]. Hence, time stability. Clinically, this means that even after accounting for they are more likely to be involved in activities that enhance school past mental health functioning (SDQ), a unit increase of Wave 1 belongingness. For this reason, they show fewer mental health school belongingness (PSSM) is associated with a corresponding problems than their counterparts who are not participating to the 0.11 standard unit deviation (Beta) reduction in Wave 2-SDQ same extent. The positive effect of school belongingness on mental (based on multiple regression analyses). These results suggest that health may also represent the degree to which schools are meeting promoting school belongingness before the transition to secondary the developmental needs of their students [78,79]. The associa- school has a beneficial effect on post-transition mental health tions between school belongingness and subsequent mental health functioning. The pathway from Wave 1-mental health (SDQ) to functioning found in the current study suggest that both primary Wave 2-belongingness (PSSM) was not significant, as expected. and secondary schools have a responsibility to foster school belongingness of all students from an early age, to safeguard future Step 2: Multi-group invariance (equivalence) of the mental health. baseline model (Model 1) Our results are consistent with the work of Shochet and Several additional models were examined to determine the colleagues [40] who reported significant relationships between equivalence of Figure 2 across gender (Table 2), disability (Table 3), prior school belongingness and future mental health symptoms in and household-SES (Table 4). The fit of the unconstrained model a large community sample (N = 2,200) of 12–14 year old in each analysis was compared to the fit of a fully constrained Australian high school students. Shocket et al., [40] however used model. Imposing the equality constraints did not significantly hierarchical linear modeling to test the relationship between the deteriorate the fit of the model. Both models represented an study variables, independently for boys and girls. The current Table 1. Means, Standard Deviations, and correlations between the Strength and Difficulties Questionnaire (SDQ) and Psychological Scale of School Membership (PSSM) at Wave 1 and Wave 2. M SD Wave 1 -SDQ Wave 1 -PSSM Wave 2-SDQ Wave 2-PSSM Wave1-SDQ 6.90 5.56 1 2.42** .77** 2.28** Wave1-PSSM 3.90 .70 1 2.40** .49** Wave2-SDQ 7.11 5.24 1 2.33** Wave2-PSSM 3.84 .64 1 **Correlation is significant at the 0.01 level (2-tailed). Note that higher SDQ indicate worse mental health functioning; higher PSSM indicate better school belongingness. doi:10.1371/journal.pone.0099576.t001 PLOS ONE | www.plosone.org 5 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth Figure 2. Cross-lagged relationship between PSSM and SDQ across the primary-secondary school transition, using data from the entire sample. doi:10.1371/journal.pone.0099576.g002 study extends Shocket and colleagues work [40] in two ways. [40]. This could be attributed to the school sector change noted in Firstly, it explicitly tested the role of gender, disability and the current study, i.e., shift from government to independent household-SES as a moderator of the associations between school schooling for secondary education. Nonetheless, unlike previous belongingness and early adolescent mental health functioning. studies that report significant declines in mean school belonging- Secondly, it applied multi-group cross-lagged panel analysis and ness scores as students’ progress through the secondary years of took into account the commonly reported co-variation between school [43,84,85], there was no significant between-group change, school belongingness and mental health, at all points in time, along or within-subject change in school belongingness scores across the with the cross-lagged and cross-time stability of the variables. In transition in the current study (Please contact first author for doing so, confidence that the obtained associations reflect the detailed results). Taken together, these findings highlight the need unique contributions of the relationship between the study for schools to assess all students’ perceptions of school belonging- variables increases. ness and mental health functioning before the transition, and The current study’s findings are however contrary to those ensure that student records are transferred as part of their reported by Loukas et al., [41], who in a US sample of 9–14 year Individualised Education Plan, so that appropriate scaffolds are in old youth, reported bidirectional relationships between school place to support those in need. belongingness and conduct problems. One possible explanation for the absence of the significant cross-lagged association between Limitations prior mental health functioning and future school belongingness in Detailed limitations of the current study have been discussed in the current study could be the highly stable level of mental health an earlier publication [50]. Key points are hereafter discussed. For functioning across time. Alternative possible explanation for the example, the current study’s population was drawn from null finding may be that other variables not examined in this study, metropolitan and other major city centres across WA, and did such as grade point average, motivational variables, teachers’ not involve other rural and regional populations, or major classroom management strategy, etc. are better predictors of metropolitan cities in Australia; thus limiting generalizability. change in early adolescents’ school belongingness [80,81]. Albeit, Despite several recruitment efforts, 70% of the schools declined to the high co-variations between Wave 1 measures could suggest participate in the study, which may have introduced a possible that baseline reports of school belongingness contribute to initial bias. The study’s cohort was different to the profile of all schools in levels of mental health functioning, which then remain stable WA. The number of students in the lower SES subgroup was across time. This could mean that failure to connect to the school relatively small for sub-group differences to be identified. during the early school years year may contribute to concurrent Furthermore, the criterion for inclusion into the disability category levels of mental health functioning, which are maintained across (i.e., limiting inclusion to those with a medical diagnosis who time. However, a parallel possibility is that initial of mental health attended regular school for 80% of school hours) could have functioning may influence initial perceptions of school belonging- resulted in the exclusion of students with more disability related ness, which are then maintained across time and re-enforce future physical, cognitive, social, and emotional restrictions [86]. mental health, cannot be ignored. For example, earlier studies have shown students with externalizing mental health problems to Furthermore, parents were asked to report on their child’s disability and overall mental functioning. Additional studies that be more likely to experience peer rejection [82], higher levels of student–teacher conflict, and decreased levels of closeness [83]. involves multisource data from students, parents, teachers, and The two-point study design precluded studying the longitudinal possibly validation using clinical interviews, medical and school records, are warranted to validate the findings [87]. The two-point relationship between these constructs. Longer term time series analyses are desirable to parcel out these contributions and identify longitudinal study design did not permit the study of the longer- term effect of transition on school belongingness and mental health time-snaps that are ideal to intervene. functioning. Future research into the relationship between The high stability of students’ mental health functioning over time, noticed in our study and past research, highlights the need covariates (gender, disability and household-SES) in the cross- lagged model is desirable. The small size of the sub-group samples, for primary and secondary schools to transfer students’ mental health functioning details as part of the transition transfer process, together with the absence of any significant interactions between and factor them into their Individual Education Plans. The 12- the covariates, precluded the need for those analyses in the present month school belongingness stability correlation was slightly lower study. Longer term longitudinal studies that track students along than previously documented in Australian community samples the educational continuum are desirable to increase our under- PLOS ONE | www.plosone.org 6 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth PLOS ONE | www.plosone.org 7 June 2014 | Volume 9 | Issue 6 | e99576 Table 2. Multi-group analyses of the unconstrained and constrained longitudinal relationship between school belongingness and mental health functioning across the primary- secondary school transition (Gender as the grouping variable). Model no Name of model x df p NFI CFI RMSEA AIC 1 Unconstrained boys/girls 1.709 3 .635 .997 1.000 .000 55.709 2 Constrained 10.787 21 .967 .983 1.000 .000 28.787 3 Difference between 1 and 2 9.078 18 .957 0.000 doi:10.1371/journal.pone.0099576.t002 Table 3. Multi-group analyses of the unconstrained and constrained longitudinal relationship between school belongingness and mental health functioning across the primary- secondary school transition (Disability as the grouping variable). Model no Name of model x df p NFI CFI RMSEA AIC 1 Unconstrained disability/no disability 1.877 3 .598 .997 1.000 .000 55.877 2 Constrained 16.35 21 .750 .974 1.000 .000 34.350 3 Difference between 1 and 2 14.477 18 .697 0.000 doi:10.1371/journal.pone.0099576.t003 School Belongingness and Mental Health in Youth standing of how and when risk is expressed as disorder; to determine the ideal time to intervene; and the relevance of intervention on student outcomes. Conclusions The current study adds to the growing body of research examining the role of school contextual influences in early adolescent mental health functioning. Adolescent experiences of belonging to, and closeness with, others at the school may buffer or offset the subsequent negative mental health functioning, above and beyond prior mental health functioning. The current study’s findings highlight the importance for both primary and secondary schools to assess the belongingness and mental health needs of all their students. Such assessments could allow schools to pay special attention to those with poorer mental health functioning and school belongingness scores, as these are more likely to continue to be disadvantaged over time. Author Contributions Conceived and designed the experiments: SV AEP RP. Performed the experiments: SV RP. Analyzed the data: TP SV RP. Contributed reagents/materials/analysis tools: SV RP AEP. Wrote the paper: SV TF MF RP AEP TP. Critical review: TF MF RP. References 1. Kieling C, Baker-Henningham H, Belfer M, Conti G, Ertem I, et al. (2011) Child and adolescent mental health worldwide: evidence for action. The Lancet 378: 1515–1525. 2. Sawyer MG, Arney FM, Baghurst PA, Clark JJ, Graetz BW, et al. (2000) The mental health of young people in Australia: Child and adolescent component of the National Survey of Mental Health and Wellbeing. Canberra, Australia: Department of Health and Aged Care, Mental Health and Special Programs Branch. 3. McLennan W (1997) Mental Health and Wellbeing: Profile of Adults. Canberra, Australia: Australian Bureau of Statistics. 4. Australian Institute of Health and Welfare (AIHW) (2003) Australia’s young people: Their health and wellbeing. Canberra: Australian Institute of Health and Welfare. 5. Scott S (1998) Aggressive behaviour in childhood. BMJ 316: 202–206. 6. Hawton K, Rodham K, Evans E, Weatherall R (2002) Deliberate self harm in adolescents: self report survey in schools in England. BMJ 23: 1207–1211. 7. Parker G, Roy K (2001) Adolescent depression: A review. Aust N Z J Psychiatry 35: 572–580. 8. Sturgeon L (2007) Mental health and intellectual disability in children and adolescents: working with this challenging population, (PPT), Available: www. psychdd.com.au/IDandPsycillnessforDADHC.ppt. 9. Berkman LF, Kawachi I (2000) Social Epidemiology. New York: Oxford University Press. 10. Wickrama KAS, Conger RD, Lorenz FO, Jung T (2008) Family Antecedents and Consequences of Trajectories of Depressive Symptoms from Adolescence to Young Adulthood: A Life Course Investigation. Journal of Health and Social Behavior 49: 468–483. 11. Herd P, Goesling B, House JS (2007) Socioeconomic Position and Health: The Differential Effects of Education versus Income on the Onset versus Progression of Health Problems. Journal of Health and Social Behavior 48: 223–238. 12. Oakes JM, Rossi PH (2003) The measurement of SES in health research: current practice and steps toward a new approach. Social science & medicine (1982) 56: 769–784. 13. Moffitt TE, Caspi A, Dickson N, Silva P, Stanton W (1996) Childhood-onset versus adolescent-onset antisocial conduct problems in males: Natural history from ages 3 to 18 years. Development and Psychopathology 8: 399–424. 14. Orvaschel H, Lewinsohn PM, Seeley JR (1995) Continuity of Psychopathology in a Community Sample of Adolescents. Journal of the American Academy of Child and Adolescent Psychiatry 34: 1525–1535. 15. Zubrick S, Silburn SR, Gurrin L, Teoh H, Shepherd C, et al. (1997) The Western Australian Child Health Survey: Education, health and competence. Perth, WA: Australian Bureau of Statistics and the Institute for Child Health Research. 16. Zubrick S, Silburn SR, Burton P, Blair EM (2000) Mental health disorders in children and young people: Scope, cause and prevention. Australian and New Zealand Journal of Psychiatry 34: 570–578. 17. Rutter M (1995) Relationships between mental disorders in childhood and adulthood. Acta Psychiatrica Scandinavica 91: 73–85. PLOS ONE | www.plosone.org 8 June 2014 | Volume 9 | Issue 6 | e99576 Table 4. Multi-group analyses of the unconstrained and constrained longitudinal relationship between school belongingness and mental health functioning across the primary- secondary school transition (Household-SES as the grouping variable). Model no Name of model x df p NFI CFI RMSEA AIC 1 Unconstrained for each SES group 10.877 27 .621 .983 1.000 .000 64.877 2 Constrained 31.266 9 .453 .952 1.000 .004 49.266 3 Difference between 1 and 2 20.389 18 .311 0.000 doi:10.1371/journal.pone.0099576.t004 School Belongingness and Mental Health in Youth the National Longitudinal Study of Adolescent Health. Journal of Adolescent 18. Petticrew M, Chisholm D, Thomson H, Jane-LLlopis E (2005) Generating Health 27: 340–348. evidence on determinants, effectiveness and cost effectiveness. In: Herrman I, Saxena H, Moodie R, editors. Promoting Mental Health: Concepts, Emerging 46. Erikson EH (1968) Identity: Youth and crisis. New York: Norton. Evidence, Practice. Geneva: World Health Organisation. 47. Weiss RS (1973) Loneliness: The experience of emotional and social isolation. 19. Geller B, Zimerman B, Williams M, Bolhofner K, Craney JL (2001) Bipolar Cambridge, MA: MIT Press. disorder at prospective follow-up of adults who had prepubertal major depressive 48. Larson RW, Verma S (1999) How children and adolescents spend time across disorder. American Journal of Psychiatry 158: 125–127. the world: Work, play, and developmental opportunities. Psychological Bulletin 20. Geller B, Zimerman B, Williams M, Bolhofner K, Craney JL (2001) Adult 125: 701–736. Psychosocial Outcome of Prepubertal Major Depressive Disorder. Journal of the 49. Vaz S (2010) Factors affecting student adjustment as they transition from American Academy of Child and Adolescent Psychiatry 40: 673–677. primary to secondary school: A longitudinal investigation. Perth: Curtin University. 21. Cicchetti D, Rogosch FA (1999) Psychopathology as Risk for Adolescent 50. Vaz S, Parsons R, Falkmer T, Passmore AE, Falkmer M (2014) The impact of Substance Use Disorders: A Developmental Psychopathology Perspective. personal background and school contextual factors on academic competence Journal of Clinical Child Psychology 28: 355–365. and mental health functioning across the primary-secondary school transition. 22. Fordwood SR, Asarnow JR, Huizar DP, Reise SP (2007) Suicide attempts PLOS ONE. among depressed adolescents in primary care. Journal of Clinical Child and 51. Australian Bureau of Statistics [ABS] (2001) ABS 2001 Census Dictionary (Cat. Adolescent Psychololgy 36: 392–404. No. 2901.0). 23. Patel V, Flisher AJ, Hetrick S, McGorry P (2007) Mental health of young people: 52. Goodman R (1997) The strengths and difficulties questionnaire: A research note. a global public health challenge. The Lancet 369: 1302–1313. Journal of Child Psychology and Psychiatry 38: 581–586. 24. Hickie IB, Groom G, Davenport T (2004) Investing in Australia’s future: The 53. Mellor D (2005) Normative data for the strengths and difficulties questionnaire personal, social and economic benefits of good mental health. Canberra: Mental in Australia. Australian Psychologist 40: 215–222. Health Council of Australia. 54. Achenbach TM (1991) Manual for the youth self-report and 1991 profile. 25. Davis C, Martin G, Kosky R, O’Hanlon A (2000) Early Intervention in the Burlington, VT: University of Vermont Department of Psychiatry. Mental Health of Young People: A Literature Review. Monograph published by 55. Goodman R, Scott S (1999) Comparing the Strengths and Difficulties The Australian Early Intervention Network for Mental Health in Young People. Questionnaire and the Child Behaviour Checklist: Is small beautiful? Journal 26. Weisz JR, McCarty CA, Valeri SM (2006) Effects of psychotherapy for of Abnormal Child Psychology: 1. depression in children and adolescents: A meta-analysis. Psychological Bulletin 56. Goodman R (1999) The extended version of the strengths and difficulties 132: 132–149. questionnaire as a guide to child psychiatric caseness and consequent burden. 27. Zalsman G, Brent DA, Weersing VR (2006) Depressive Disorders in Childhood Journal of Child Psychology and Psychiatry 40: 791–801. and Adolescence: An Overview: Epidemiology, Clinical Manifestation and Risk 57. Goodman R, Ford T, Simmons H, Gatward R, Meltzer H (2003) Using the Factors. Child and Adolescent Psychiatric Clinics of North America 15: 827– Strengths and Difficulties Questionnaire (SDQ) to screen for child psychiatric disorders in a community sample. International Review of Psychiatry 15: 166– 28. Sawyer MG, Arney FM, Baghurst PA, Clark JJ, Graetz RJ, et al. (2001) The mental health of young people in Australia: Key findings from the child and 58. Hagborg WJ (1994) An exploration of school membership among middle- and adolescent component of the national survey of mental health and well-being. high-school students. Journal of Psychoeducational Assessment 12: 312–323. Australian and New Zealand Journal of Psychiatry 35: 806–814. 59. McMahon SD, Parnes AL, Keys CB,Viola JJ (2008) School belonging among 29. Verhulst FC, Koot HM (1995) The Epidemiology of Child and Adolescent low-income urban youth with disabilities. Psychology in the Schools 45: 26–47. Psychopathology. London: Oxford University Press. 60. Zubrick S, Williams A, Silburn SR, Vimpani G (2000) Indicators of social and 30. Sawyer MG, Pfeiffer S, Spence SH (2009) Life events, coping and depressive family functioning. Canberra: Commonwealth of Australia. symptoms among young adolescents: A one-year prospective study. Journal of 61. Meyers LS, Gamst G, Guarino AJ (2006) Applied multivariate research: Design affective disorders 117: 48–54. and implication. CA: Sage Publications, Inc. 31. World Health Organisation [WHO] (1996) Promoting health through schools - 62. Tabachnick B, Fidell L (2007) Using multivariate statistics. Boston, MA: Pearson The World Health Organisation’s global school health initiative. Geneva: World Education Inc. and Allyn & Bacon. Health Organisation. 63. McKnight P, McKnight K, Sidani S, Figueredo A (2007) Missing data: A gentle 32. Goodenow C (1993) The psychological sense of school membership among introduction New York: Guilford. adolescents: Scale development and educational correlates. Psychology in the 64. West SG, Finch JF, Curran PJ (1995) Structural equation models with non- Schools 30: 79–90. normal variables: Problems and Remedies. In R.H. Hoyle (ed.) Structural 33. Blum RW (2005) A case for school connectedness. Educational Leadership 62: Equation Modeling: Concepts, Issues and Applications. Sage Publications; 16–20. Thousand Oaks CA. 34. Hargreaves A, Earl L, Ryan J (1996) Schooling for change: Reinventing 65. Kline RB (2005) Principles and practice of structural equation modelling. New education for early adolescents. London: The Falmer Press. York: Guilford Press. 35. Jacobson KC, Rowe DC (1999) Genetic and environmental influences on the 66. Bentler PM, Bonett DG (1980) Significance tests and goodness-of-fit in the relationship between family connectedness, school connectedness and adolescent analysis of covariance structures. Psychological Bulletin, 88, 588–600. depressed mood: Sex differences. Developmental Psychology 35: 926–939. 67. Hu L, Bentler PM (1999) Cutoff criteria for fit indexes in covariance structure 36. Resnick M, Bearman P, Blum RW, Bauman KE, Harris K, et al. (1997) analysis: Conventional criteria versus new alternatives. Structural Equation Protecting adolescents from harm: Findings from the National Longitudinal Modeling, 6, 1–55. Study on Adolescent Health. JAMA 278: 823–832. 68. Cheung GW, Rensvold RB (2002) Evaluating goodness-of-fit indexes for testing 37. Johnson MK, Crosnoe R, Thaden LL (2006) Gendered Patterns in Adolescents’ measurement invariance. Structural equation modeling 9: 233–255. School Attachment. Social Psychology Quarterly 69: 284–295. 69. SDQ (2007) Information for researchers and professionals about the Strengths & 38. Ma X (2003) Sense of Belonging to School: Can Schools Make a Difference? Difficulties Questionnaires. The Journal of Educational Research 96: 340–349. 70. Anderman EM (2002) School effects on psychological outcomes during 39. Kuperminic GP, Leadbeater BJ, Blatt SJ (2001) School Social Climate and adolescence. Journal of Educational Psychology 94: 795–808. Individual Differences in Vulnerability to Psychopathology among Middle 71. Loukas A, Ripperher-Suhler KG, Horton KD (2009) Examining temporal School Students. Journal of School Psychology 39: 141–159. associations between school connectedness and early adolescent adjustment 40. Shochet IM, Dadds MR, Ham D, Montague R (2006) School Connectedness Is Journal of Youth and Adolescence 38: 808–812. an Underemphasized Parameter in Adolescent Mental Health: Results of a 72. Whitlock JL (2006) Youth perceptions of life at school: Contextual correlates of Community Prediction Study. Journal of Clinical Child & Adolescent school connectedness in adolescence. Applied Developmental Science 10: 13–29. Psychology 35: 170–179. 73. Battistich V, Solomon D, Watson M, Schaps E (1997) Caring school 41. Loukas A, Ripperger-Suhler KG, Horton KD (2009) Examining Temporal communities. Educational Psychologist 32: 137–151. Associations Between School Connectedness and Early Adolescent Adjustment. 74. Burton ES, Stice E, Seeley JR (2004) A prospective test of the stress-buffering Journal of Youth and Adolescence 38: 804–812. model of depression in adolescent girls: No support once again. Journal of 42. Loukas A, Roalson LA, Herrera DE (2010) School connectedness buffers the Consulting and Clinical Psychology 72: 689–697. effects of negative family relations and poor effortful control on early adolescent 75. Cernkovich SA, Giordano PC (1992) School bonding, race, and delinquency. conduct problems. Journal of Research on Adolescence 20: 13–22. Criminology 30: 261–291. 43. Catalano RF, Oesterle S, Fleming CB, Hawkins JD (2004) The Importance of 76. Wade TJ, Brannigan A (1998) The genesis of adolescent risktaking: Pathways Bonding to School for Healthy Development: Findings from the Social through family, school, and peers. Canadian Journal of Sociology 23: 1–19. Development Research Group. Journal of School Health 74: 252–261. 77. Hirschi T (1969) Causes of delinquency. Berkeley, CA University of California 44. Bond L, Butler H, Thomas L, Carlin J, Glover S, et al. (2007) Social and School Press. Connectedness in Early Secondary School as Predictors of Late Teenage 78. Eccles JS, Lord S, Midgley C (1991) What are we doing to early adolescents? Substance Use, Mental Health, and Academic Outcomes. The Journal of The impact of educational contexts on early adolescence. American Journal of adolescent health: official publication of the Society for Adolescent Medicine 40: Education 99: 521–542. 357. e359–357. e318. 79. Eccles JS, Midgley C (1989) Stage-environment fit: Developmentally approriate 45. Svetaz MV, Ireland M, Blum RW (2000) Adolescents with learning disabilities: classrrooms for young adolescents. In: Ames R, Ames C, editors. Research on Risk and protective factors associated with emotional well-being: Findings from motivation in education. New York: Academic Press. 139–181. PLOS ONE | www.plosone.org 9 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth 80. Anderman LH (2003) Academic and social perceptions as predictors of change 84. Anderman LH, Freeman TM (2004) Students’ sense of belonging in school. In: Pintrich PR, Maehr ML, editors. Motivating students, improving schools: The in middle school students’ sense of school belonging. Journal of Experimental legacy of Carol Midgley Oxford, UK: Elsvier, Ltd. 27–63. Education 72: 5–22. 85. Monahan K, Oesterle S, Hawkins JD (2010) Predictors and consequences of 81. Morrison GM, Cosden MA, O’Farrell SL, Campos E (2003) Changes in Latino school connectedness: The case for prevention. The Prevention Researcher 17: students’ perceptions of school belonging over time: Impact of language 3–6. proficiency, self-perceptions and teacher evaluations. The California School 86. Bell M, Dempsey I (2001) Enrolment and placement practices for students with Psychologist 8. special needs in NSW government schools: A critical analysis of current policy. 82. Dishion TJ, Patterson GR, Stoolmiller M, Skinner ML (1991) Family, school, Special Education Perspectives 10: 3–14. and behavioral antecedents to early adolescent involvement with antisocial 87. Stone LL, Otten R, Engels RC, Vermulst AA, Janssens JM (2010) Psychometric peers. Developmental Psychology 27: 172–180. properties of the parent and teacher versions of the strengths and difficulties 83. Birch SH, Ladd GW (1998) Children’s interpersonal behaviors and the teacher– questionnaire for 4- to 12-year-olds: a review. Clinical child and family child relationship. Developmental Psychology 34: 934–946. psychology review 13: 254–274. PLOS ONE | www.plosone.org 10 June 2014 | Volume 9 | Issue 6 | e99576 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PLoS ONE Unpaywall

School Belongingness and Mental Health Functioning across the Primary-Secondary Transition in a Mainstream Sample: Multi-Group Cross-Lagged Analyses

Download PDF
10 pages

Loading next page...
 
/lp/unpaywall/school-belongingness-and-mental-health-functioning-across-the-primary-8BlFqGnlZ6

References (92)

Publisher
Unpaywall
ISSN
1932-6203
DOI
10.1371/journal.pone.0099576
Publisher site
See Article on Publisher Site

Abstract

The relationship between school belongingness and mental health functioning before and after the primary-secondary school transition has not been previously investigated in students with and without disabilities. This study used a prospective longitudinal design to test the bi-directional relationships between these constructs, by surveying 266 students with and without disabilities and their parents, 6-months before and after the transition to secondary school. Cross-lagged multi-group analyses found student perception of belongingness in the final year of primary school to contribute to change in their mental health functioning a year later. The beneficial longitudinal effects of school belongingness on subsequent mental health functioning were evident in all student subgroups; even after accounting for prior mental health scores and the cross-time stability in mental health functioning and school belongingness scores. Findings of the current study substantiate the role of school contextual influences on early adolescent mental health functioning. They highlight the importance for primary and secondary schools to assess students’ school belongingness and mental health functioning and transfer these records as part of the transition process, so that appropriate scaffolds are in place to support those in need. Longer term longitudinal studies are needed to increase the understanding of the temporal sequencing between school belongingness and mental health functioning of all mainstream students. Citation: Vaz S, Falkmer M, Parsons R, Passmore AE, Parkin T, et al. (2014) School Belongingness and Mental Health Functioning across the Primary-Secondary Transition in a Mainstream Sample: Multi-Group Cross-Lagged Analyses. PLoS ONE 9(6): e99576. doi:10.1371/journal.pone.0099576 Editor: Kenji Hashimoto, Chiba University Center for Forensic Mental Health, Japan Received December 19, 2013; Accepted May 16, 2014; Published June 26, 2014 Copyright:  2014 Vaz et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This project was funded by a Doctoral scholarship provided by the Centre for Research into Disability and Society and the School of Occupational Therapy and Social Work, Curtin University, Perth, Australia. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * Email: [email protected] four times more often than their typically developing peers; with Mental Health Problems in Adolescence psychiatric disorders in young people with a disability often Worldwide estimates of mental health problems in children and undiagnosed and untreated [8]. Household-SES influences phys- youth range from 10–20% [1]. Australian figures report a 14% ical and mental health across the lifespan, with socially and prevalence in a national sample of 4–12 year olds, which rises to economically disadvantaged children and adults found to be an 19% in the 13–17 year old category [2] and 27% in the 18–24 increased risk for both physical and mental health problems [9– year old group [3]. These figures suggest that approximately one 12]. Thus, it is imperative that research studies account for within in four to five young Australians have a mental health problem [4]. group variability in mental health functioning of children and Mental health functioning of children and youth has been shown youth. to vary due to gender, presence of disability and household socio- Of concern is the growing evidence on the stability of mental economic standing (SES). For example, conduct disorder is the health problems in children and adolescents [13,14] and its most common psychiatric disorder in childhood, with three times longitudinal effects on mental health disorders, delinquency, as many boys as girls being affected [5]. During adolescence, girls crime, unemployment, homelessness and suicidal behaviour in have a higher prevalence of depression and eating disorders, and adulthood [14–22]. Mental health problems in children and engage more in suicidal ideation and suicide attempts than boys, adolescents could be antecedents of chronic, complex, disabling who are more prone to engage in high risk behaviours and commit and expensive complications in adult life. For these reasons, early suicide more frequently [6,7]. Young people with an intellectual detection of clinical and subclinical mental health issues is disability manifest behaviours and experiences which may be important. Most mental health disorders that are likely to persist indicative of mental health or psychological impairment three to PLOS ONE | www.plosone.org 1 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth into adult life emerge between ages 12 and 25 [23,24]. While early in primary school and a remixing of friendship networks and social intervention is more economical and cost-effective than later hierarchies. It is likely that students are forced to redefine their action [25], its effectiveness in some cases is modest [26,27] or fails sense of school belongingness after they transition to secondary to reach the majority of those most in need [26]. Australian data school. Whether poorer mental health functioning before the suggest that only one in four youth who need professional help transition is associated with poorer school belongingness thereafter actually get the help they need [28]. These facts, underscore the remains mainly unexplored. need to gain a deeper understanding of pathways in and out of childhood mental health problems [15,28–30]. Schools are an Aim and Objectives ideal setting for efficiently detecting children and adolescents with The current study extends the existing knowledge base on unidentified mental health problems because they offer the primary-secondary school transition by explicitly examining the opportunity to reach large numbers of students [2,30,31]. temporal relationships between school belongingness and overall mental health functioning after one year, by tracking a cohort with School belongingness and overall mental health and without disability, enrolled in the regular school system in functioning across primary-secondary school transition Western Australia (WA). We also tested the equivalence of these In recent years, school belongingness, referring to students’ relationships across gender, disability and household-SES. It was beliefs of being ‘‘personally accepted, respected, included, and hypothesised that: supported by others in the school social environment’’ [32], has emerged as an important factor associated with positive health (a) direct relationships would exist between concurrent percep- outcomes [33,34]. Cross-sectional studies document moderate tion of school belongingness and overall mental health associations between school belongingness and emotional distress functioning, before and after the transition; and and depression in typically developing adolescents [35–38], before (b) primary school belongingness would be related to overall and after accounting for personal and contextual factors, such as mental health functioning in early secondary school, even family-parent-belongingness, self-esteem and grade point average after accounting for prior mental health scores. [36,37]. Short term longitudinal studies present mixed findings on the directional relationships between these constructs. In some No hypothesis was made regarding the predictive role of mental studies [39,40] a unidirectional relationship has been documented; health in primary school on school belongingness a year later, due with school belongingness predicting selective prospective mental to the inconsistent empirical evidence on this issue. health components, depending on gender. For example, Shocket and colleagues [40] found that early adolescents’ perception of Method school belongingness predicted future depressive symptoms in boys and girls; anxiety in girls; and conduct problems in boys. Other A cohort study using a prospective, longitudinal design with two studies suggest bidirectional relationships between these con- data collection points was used [Primary school = Wave 1, and structs, which vary depending on the type of mental health domain Secondary school = Wave 2]. Students enrolled in the final year being measured [41]. Loukas and colleagues [41] presented of primary school in WA (class 6 or 7), in the academic years evidence of a bidirectional loop between school belongingness and commencing January 2006 or 2007, and due to transition to either conduct problems, but not depressive symptoms in 10–14 year old middle or secondary school in January 2007 or 2008, were typically developing youth [41]. Also, the positive effects of school considered for inclusion in the study. Inclusion was limited to belongingness have been found to extend to students’ home lives; regular schools in the educational districts of metropolitan Perth or concomitantly buffer the effects of family disadvantage on other major city centres of Western Australia (WA). Several functioning [42]; and prospectively protect them from involve- recruitment strategies were used to maximize reach and repre- ment in risk behaviours [43,44]. Consequently, a growing body of sentativeness. The current study is part of a larger study on the evidence with typically developing youth supports the interrela- factors associated with student academic, social-emotional and tionship between school belongingness and positive mental health participatory adjustment across the primary-secondary school outcomes. transition [49]. Details on the study design, recruitment and data Conspicuous in the above cited investigations on school collection have been published elsewhere [50]. For the ease of belongingness and mental health functioning, is the exclusion of readership, a brief overview is described below. students with disabilities in the study samples, despite their Wave 1 data collection occurred six months prior to the presence in the regular school system for several decades. transition to either middle or secondary school, with data collected Additional research is needed to authenticate the role of school from students (with and without disabilities) and a primary belongingness in the disability subgroup. Preliminary findings are caregiver (parent or guardian). Wave 2 data were collected 6 hopeful, showing school belongingness to be negatively associated months after the transition, using the same procedure and sample with emotional stress, suicide attempts, and violence amongst as Wave 1. students with learning disabilities [45]. Yet another gap is the Information was collected via survey questionnaires, primarily absence of evidence on the prospective benefits of fostering paper and pencil format. Informed written consent was obtained belongingness in primary school on overall mental health from school principals, parents, teachers, and written assent was functioning of all students after the transition to secondary school, obtained from students to participate in this study. In situations or whether there are student subgroups, based on gender, where the student declined to participate, even with parental disability status, or household-SES, that need additional support. consent, they were not included. All participants were made aware Students in western societies, including Australia, negotiate the that they were not obliged to participate, and were free to primary-secondary school transition at a time in development withdraw from the study at any time without justification or when they are striving to gain independence from their parents, prejudice. Ethics approval was obtained from Curtin University establish their unique identity [46,47], and gain approval and support from peers [48]. As a result of this school transition, Health Research Ethics Committee, in Western Australia (WA) (Reference number HR 194/2005). students experience a disruption of the secure peer network forged PLOS ONE | www.plosone.org 2 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth At Wave 1, data were collected from 395 students from 75 scores correlate positively with school success [32,59], lower levels primary schools across the Perth metropolitan area and major city of depression [40], and lower levels of anxiety [34]. Higher PSSM centres across WA. An attrition rate of 32.7% resulted in a Wave 2 scores indicate better perceived school belongingness. sample of 266 participants from 52 primary schools and Family demographics and school contextual 152 secondary schools. Chi-square and paired sample t-tests characteristics. Items were drawn from the Indicators of showed that the participants who continued to be involved in Social and Family Functioning Instrument Version-1 (ISAFF) the study at Wave 2 did not differ from those who discontinued [60] and Australian Bureau of Statistics (ABS, 2001) surveys, and involvement, on gender, disability, household SES-level, school used to provide family demographic information. Parents reported belongingness, and mental health functioning scores. The current details on the family demographic characteristics, residence post study uses data from the 266 students that answered both Wave 1 code, and their child’s disability. Information on the school sector, and 2 questionnaires. Access to the complete dataset can be post code number of students enrolled in each school, and obtained by contacting the first author. organisational structure at each school was obtained from The mean age of students sample at Wave 1 was 11.89 years Department of Education and Training, WA records. The sample (SD = 0.45 years, median = 12 years), and that at Wave 2 was 12.9 was categorised into three-income groups as per the median years (SD = 0.57 years, median = 13 years). Boys constituted income distribution based on the Australian Bureau of Statistics 46.6% (n = 124) of the sample; and 25.9% (n = 69) were reported [51] data. by a parent or primary caregiver to have a disability. The predominant disabilities included asthma (18.8%), auditory Data Management disability (15.9%), Attention Deficit Hyperactivity Disorder/ Data were managed and analysed using the SPSS Version 20.0 Attention Deficit Disorders (14.5%), learning disability (11.6%), and SAS Version 9.2 software packages. Only 1.8–2.5% of data Autism Spectrum Disorders (10.1%), and cerebral palsy (8.7%). were missing at scale levels. The estimation maximization The majority of the sample came from mid-range households, and algorithm and Little’s chi-square statistic revealed that the data reported a weekly income of $600–1,999 (58.3%, n = 154) [51]. were missing completely at random [61,62]. Missing data Under one-third of the sample (33%, n = 87) came from high-SES replacement was undertaken using guidelines recommended by households ($ 2000+/week) and 8.7%, n = 23 were from low-SES the SDQ tool developers (http://www.sdqinfo.org/c1.html). In groupings ($ 1–599 per week). the case of the school belongingness questionnaire, individual The sample represented 52 different primary schools and 77 mean score substitution was used [63]. The validity of the data different classes distributed across metropolitan Perth and other substitution techniques used was substantiated using sensitivity city centres of WA. Based on the Commonwealth Department of analyses. Education, Employment, and Workplace Relations measure of In the present study, the bidirectional associations between relative socio-economic advantage and disadvantage [67], 21.4% school belongingness and mental health functioning over one year (n = 57) of the sample came from schools located in the most were estimated by cross-lagged analyses, using the structural th affluent areas across Australia (10 decide), 44% (n = 117) came equation modelling program, AMOS 5.0. A critical preliminary th th from the 9 decile; 17.7% (n = 47) were from the 7–8 decide and step in the analysis was to investigate if data met the normality th 16.9% (n = 45) came from more disadvantaged areas (1–6 assumption. With regard to the normality assumptions of the Full decide). Forty-seven percent of the sample (n = 125) were enrolled Information Maximum Likelihood estimation procedure, the in the public schools, 29% (n = 77) in Catholic schools, and the normality of each variable was investigated in terms of its kurtosis remaining 24% (n = 64) in independent/private schools. There and skewness [64]. Box-cox transformations were undertaken to was a movement out of government schools into Catholic and normalise the PSSM and SDQ scores. In order to provide independent schools for secondary education; with 11.2% of clinically relevant information, standardized Beta values from students (n = 14) moving into Catholic schools and 28.8% (n = 36) multiple linear regression analyses have also been presented, using moving into independent schools for their secondary education. the original data. Data collection instruments Statistical Analyses Mental health functioning. The 25-item parent version of the Strengths and Difficulties Questionnaire (SDQ) was used to Characteristics of the sample measure student overall functioning across hyperactivity, emo- Descriptive statistics were used to summarise the profile of tional health, conduct problem and peer problem domains [52]. participants. This version has moderate to high internal consistency scores (a = .70–80) [53], and is reported to be more sensitive than the Testing for the effects of nesting of students on mental Child Behaviour Check List [54] in detecting inattention and health and school belongingness hyperactivity, and equally effective in detecting internalising and In order to test for the effect of clustering of students, i.e., externalising problems in children and adolescents [55]. Estab- nesting of students in classes within schools on their school lished reliability and validity of the SDQ makes it a useful brief belongingness (PSSM) and mental health functioning (SDQ) screening measure of adjustment and psychopathology in children scores, a Hierarchical Linear Model was fitted using the mixed and adolescents [53,55–57]. Higher SDQ scores indicate poorer procedure in SAS. The class-level Intra Class Correlation mental health functioning. Coefficients (ICC) for PSSM and SDQ were obtained, after School belongingness. The 18-item, Psychological Sense of adjustment for gender, disability, and household-SES. School Membership scale (PSSM) was used to assess students’ Interrelationship between school belongingness (PSSM) and perceptions of belongingness in school [32]. The PSSM has satisfactory internal consistency (a = .803) [32]. Test-retest reli- mental health functioning scores (SDQ). Pearson correlation ability indices of .78 (4-week interval) [58], and .56 and .60 for coefficients were used to identify associations between the SDQ boys and girls respectively (12-month interval) have been and PSSM scores at and between each wave. A two-factor analysis documented in early adolescent samples [40]. The total PSSM of variance with and without interaction terms was run to test the PLOS ONE | www.plosone.org 3 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth within-group variability in SDQ due to gender, disability, and data without comparison to a reference model whereas an household-SES. incremental fit index compares the target model to a more Testing the hypothesized model of the relationship between restricted baseline model [67]. Both these indexes take into school belongingness and mental health functioning. Autoregres- account model complexity, which is an important property for sive cross-lagged panel analysis was performed to study the comparing several alternative models with different degrees of reciprocal relationship between school belongingness and mental complexity. According to criteria outlined by Hu and Bentler [67], health functioning across the primary-secondary school transition. a good fitting model has NNFI values of .95 or greater, RMSEA The path-diagram of the autoregressive cross-lagged model used in values smaller than .06, and a CFI greater than or equal to .95. In this study is presented in Figure 1. reporting on evidence of invariance, two criteria were used. Firstly, Cross-lagged panel analysis allows examination of the cross- the multi-group model must exhibit an adequate fit to the data. lagged paths while controlling for cross-time stability of each of the Secondly, the determination of multi-group invariance is based on variables. In each of the models, the exogenous variables of Wave delta CFI; that is, when the differences in CFI values between 1, which included school belongingness (PSSM) and mental health models are less than .01 [68]. functioning (SDQ), were freely correlated. The residuals (error variances) of all Wave 2 variables were also correlated, due to Results auto-correlation effects. Stability paths from each of the Wave 1 constructs to their respective Wave 2 outcomes were included to Testing for the effects of nesting of students on mental partial out the effects of baseline adjustment problems. The health and school belongingness inclusion of stability paths provides a stringent test of the Wave 1 A total of 52 different schools, and 77 different classes were influences and results in the examination of change in the variable involved in Wave 1. In order to test for the effect of clustering of of interest. To test the contribution of prior school belongingness students, i.e., nesting of students in classes within schools, a (PSSM) to future mental health functioning (SDQ), paths from Hierarchical Linear Model was fitted using the mixed procedure Wave 1-PSSM to Wave 2-SDQ were included. The opposite in SAS. The class-level Intra Class Correlation Coefficients (ICC) direction of associations, path from Wave 1-SDQ to Wave 2- for school belongingness and mental health functioning scores PSSM was also simultaneously estimated, but not presented in were obtained (after adjustment for the demographic data: gender, Figure 1. disability, and household-SES). The ICC for each model was low, Multi-group invariance (equivalence) of the baseline ranging from 0–12%, showing that the contribution of the model (Model 1). To examine the equivalence of the hypoth- clustering to the overall variance was small, and therefore the esized model across subgroups, namely, gender, disability and clustering appeared to have minimal effect on the relationships household-SES, parameters were simultaneously estimated for between the student-level variables and school belongingness and each subgroup, respectively. The fit of this simultaneously mental health functioning scores. Hence, further analyses were estimated unconstrained model provides the baseline value for undertaken at the level of the individual student. each subgroup against which all subsequently specified models are compared. A fully constrained model, in which all parameters Characteristics of the sample: Within-group variability in (factor variances, factor covariances, and error covariances) were mental health functioning constrained or specified to be equivalent across subgroups, was The mental health functioning scores (SDQ) of the students then calculated. x difference tests were used to determine involved in the current study was better than those found in an significant differences between the unconstrained and constrained Australian community sample for this age range [53,69]. Within- models of each subgroup. group variability interactions were not statistically significant; Model Evaluation Criteria. To determine the fit of the hence only the main effects were included in the final models. In models, criteria were adopted from several sources. Because x is the case of Wave 2-SDQ scores, significant differences due to influenced by sample size, we examined the x /degrees of freedom 2 2 gender, F (1,256) = 4.30, p = 0.04, disability, F (1,256) = 49.95, (df) ratio (x /df) rather than the significance of the x alone [65]. p =,.001, and household-SES, F (2,254) = 3.77, p = 0.02 were Furthermore, we also used the Non-Normed Fit Index (NNFI) [66]. Additionally, fit was evaluated by one absolute fit index (the found. Boys (M= 8.88, SE = .45) had worse Wave2-SDQ scores than girls (M= 7.65, SE = .44); and students with disability had Root Mean Square Error of Approximation, RMSEA) and one incremental fit index (the Comparative Fit Index, CFI). An worse scores (M= 10.61, SE = .58) than those without disability absolute fit index assesses how well a model reproduces the sample (M= 5.91, SE = .35). Students from low-SES households Figure 1. Cross-lagged relationship between PSSM and SDQ across the primary-secondary school transition. doi:10.1371/journal.pone.0099576.g001 PLOS ONE | www.plosone.org 4 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth (M= 8.90, SE = 1.18) had significantly poorer Wave2-SDQ scores excellent fit to the data. x difference tests found no significant differences between the unconstrained and constrained models of than their peers from high-SES (M = 7.12, SE = .55, p = .05), but not mid-SES households (M= 7.52, SE = .420, p..05). each subgroup, suggesting invariance of the baseline model (Figure 2) across gender, disability and household-SES. Interrelationship between school belongingness (PSSM) Discussion and mental health functioning scores (SDQ) The means, standard deviations and correlation matrix for all The present study extends existing research by providing study variables without adjustment for gender, disability and evidence that students’ ratings of belongingness in the final year household-SES are presented in Table 1. School belongingness of primary school contributes to change in their mental health (PSSM) was concurrently and longitudinally associated with functioning a year later. The beneficial effect of primary school mental health functioning (SDQ) at both waves of the study. belongingness on subsequent mental health functioning was Early adolescents reporting higher levels of school belongingness evident for the entire population of mainstream students, even (higher PSSM) also reported better mental health functioning after accounting for their prior mental health scores and the cross- (lower SDQ). Examination of the cross-time stability of the time stability in mental health functioning and school belonging- variables indicated that the magnitude of the correlations was ness scores. moderate for students’ perceptions of school belongingness (PSSM, Findings of the current study corroborate a large body of r=.49), and larger for mental health functioning (SDQ, r=.77). evidence on the significance of boosting school belongingness as a mental health promotion strategy not only in typically developing Testing the hypothesized model of the relationship students [36,39,40,70,71], but also students with disabilities. These results are of significance given current estimates that psychiatric between school belongingness and mental health disorders in young people with disabilities are often undiagnosed functioning (Figure 1) and untreated, despite the fact that these students manifest Figure 2 presents the most parsimonious baseline model that 2 behaviours and experiences indicative of mental illness or best fitted the data [x (1, n = 266) = .716, n.s.; CFI = 1.00; psychological impairment three to four times more often than RMSEA = .000; AIC = 26.716]. As shown in Fig. 2, the stability their typically developing peers [8]. Several theoretical underpin- paths were positive and significant, and inter-correlations among nings may explain the results. Students who sense a bonding in the two exogenous variables were significant. The error variance school are more likely to forge supportive relationships with between the endogenous variables was significant and in the teachers [32,72], associate with pro-social peer groups [73] and expected direction. Regarding the cross-lagged paths, Wave 1- are more likely to have better mental health functioning [74]. PSSM was associated with lower levels of Wave 1-SDQ, even after Students with social attachment to the school could be expected to controlling for baseline levels of all variables and for their cross- feel committed to its goals, norms, and morals [75–77]. Hence, time stability. Clinically, this means that even after accounting for they are more likely to be involved in activities that enhance school past mental health functioning (SDQ), a unit increase of Wave 1 belongingness. For this reason, they show fewer mental health school belongingness (PSSM) is associated with a corresponding problems than their counterparts who are not participating to the 0.11 standard unit deviation (Beta) reduction in Wave 2-SDQ same extent. The positive effect of school belongingness on mental (based on multiple regression analyses). These results suggest that health may also represent the degree to which schools are meeting promoting school belongingness before the transition to secondary the developmental needs of their students [78,79]. The associa- school has a beneficial effect on post-transition mental health tions between school belongingness and subsequent mental health functioning. The pathway from Wave 1-mental health (SDQ) to functioning found in the current study suggest that both primary Wave 2-belongingness (PSSM) was not significant, as expected. and secondary schools have a responsibility to foster school belongingness of all students from an early age, to safeguard future Step 2: Multi-group invariance (equivalence) of the mental health. baseline model (Model 1) Our results are consistent with the work of Shochet and Several additional models were examined to determine the colleagues [40] who reported significant relationships between equivalence of Figure 2 across gender (Table 2), disability (Table 3), prior school belongingness and future mental health symptoms in and household-SES (Table 4). The fit of the unconstrained model a large community sample (N = 2,200) of 12–14 year old in each analysis was compared to the fit of a fully constrained Australian high school students. Shocket et al., [40] however used model. Imposing the equality constraints did not significantly hierarchical linear modeling to test the relationship between the deteriorate the fit of the model. Both models represented an study variables, independently for boys and girls. The current Table 1. Means, Standard Deviations, and correlations between the Strength and Difficulties Questionnaire (SDQ) and Psychological Scale of School Membership (PSSM) at Wave 1 and Wave 2. M SD Wave 1 -SDQ Wave 1 -PSSM Wave 2-SDQ Wave 2-PSSM Wave1-SDQ 6.90 5.56 1 2.42** .77** 2.28** Wave1-PSSM 3.90 .70 1 2.40** .49** Wave2-SDQ 7.11 5.24 1 2.33** Wave2-PSSM 3.84 .64 1 **Correlation is significant at the 0.01 level (2-tailed). Note that higher SDQ indicate worse mental health functioning; higher PSSM indicate better school belongingness. doi:10.1371/journal.pone.0099576.t001 PLOS ONE | www.plosone.org 5 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth Figure 2. Cross-lagged relationship between PSSM and SDQ across the primary-secondary school transition, using data from the entire sample. doi:10.1371/journal.pone.0099576.g002 study extends Shocket and colleagues work [40] in two ways. [40]. This could be attributed to the school sector change noted in Firstly, it explicitly tested the role of gender, disability and the current study, i.e., shift from government to independent household-SES as a moderator of the associations between school schooling for secondary education. Nonetheless, unlike previous belongingness and early adolescent mental health functioning. studies that report significant declines in mean school belonging- Secondly, it applied multi-group cross-lagged panel analysis and ness scores as students’ progress through the secondary years of took into account the commonly reported co-variation between school [43,84,85], there was no significant between-group change, school belongingness and mental health, at all points in time, along or within-subject change in school belongingness scores across the with the cross-lagged and cross-time stability of the variables. In transition in the current study (Please contact first author for doing so, confidence that the obtained associations reflect the detailed results). Taken together, these findings highlight the need unique contributions of the relationship between the study for schools to assess all students’ perceptions of school belonging- variables increases. ness and mental health functioning before the transition, and The current study’s findings are however contrary to those ensure that student records are transferred as part of their reported by Loukas et al., [41], who in a US sample of 9–14 year Individualised Education Plan, so that appropriate scaffolds are in old youth, reported bidirectional relationships between school place to support those in need. belongingness and conduct problems. One possible explanation for the absence of the significant cross-lagged association between Limitations prior mental health functioning and future school belongingness in Detailed limitations of the current study have been discussed in the current study could be the highly stable level of mental health an earlier publication [50]. Key points are hereafter discussed. For functioning across time. Alternative possible explanation for the example, the current study’s population was drawn from null finding may be that other variables not examined in this study, metropolitan and other major city centres across WA, and did such as grade point average, motivational variables, teachers’ not involve other rural and regional populations, or major classroom management strategy, etc. are better predictors of metropolitan cities in Australia; thus limiting generalizability. change in early adolescents’ school belongingness [80,81]. Albeit, Despite several recruitment efforts, 70% of the schools declined to the high co-variations between Wave 1 measures could suggest participate in the study, which may have introduced a possible that baseline reports of school belongingness contribute to initial bias. The study’s cohort was different to the profile of all schools in levels of mental health functioning, which then remain stable WA. The number of students in the lower SES subgroup was across time. This could mean that failure to connect to the school relatively small for sub-group differences to be identified. during the early school years year may contribute to concurrent Furthermore, the criterion for inclusion into the disability category levels of mental health functioning, which are maintained across (i.e., limiting inclusion to those with a medical diagnosis who time. However, a parallel possibility is that initial of mental health attended regular school for 80% of school hours) could have functioning may influence initial perceptions of school belonging- resulted in the exclusion of students with more disability related ness, which are then maintained across time and re-enforce future physical, cognitive, social, and emotional restrictions [86]. mental health, cannot be ignored. For example, earlier studies have shown students with externalizing mental health problems to Furthermore, parents were asked to report on their child’s disability and overall mental functioning. Additional studies that be more likely to experience peer rejection [82], higher levels of student–teacher conflict, and decreased levels of closeness [83]. involves multisource data from students, parents, teachers, and The two-point study design precluded studying the longitudinal possibly validation using clinical interviews, medical and school records, are warranted to validate the findings [87]. The two-point relationship between these constructs. Longer term time series analyses are desirable to parcel out these contributions and identify longitudinal study design did not permit the study of the longer- term effect of transition on school belongingness and mental health time-snaps that are ideal to intervene. functioning. Future research into the relationship between The high stability of students’ mental health functioning over time, noticed in our study and past research, highlights the need covariates (gender, disability and household-SES) in the cross- lagged model is desirable. The small size of the sub-group samples, for primary and secondary schools to transfer students’ mental health functioning details as part of the transition transfer process, together with the absence of any significant interactions between and factor them into their Individual Education Plans. The 12- the covariates, precluded the need for those analyses in the present month school belongingness stability correlation was slightly lower study. Longer term longitudinal studies that track students along than previously documented in Australian community samples the educational continuum are desirable to increase our under- PLOS ONE | www.plosone.org 6 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth PLOS ONE | www.plosone.org 7 June 2014 | Volume 9 | Issue 6 | e99576 Table 2. Multi-group analyses of the unconstrained and constrained longitudinal relationship between school belongingness and mental health functioning across the primary- secondary school transition (Gender as the grouping variable). Model no Name of model x df p NFI CFI RMSEA AIC 1 Unconstrained boys/girls 1.709 3 .635 .997 1.000 .000 55.709 2 Constrained 10.787 21 .967 .983 1.000 .000 28.787 3 Difference between 1 and 2 9.078 18 .957 0.000 doi:10.1371/journal.pone.0099576.t002 Table 3. Multi-group analyses of the unconstrained and constrained longitudinal relationship between school belongingness and mental health functioning across the primary- secondary school transition (Disability as the grouping variable). Model no Name of model x df p NFI CFI RMSEA AIC 1 Unconstrained disability/no disability 1.877 3 .598 .997 1.000 .000 55.877 2 Constrained 16.35 21 .750 .974 1.000 .000 34.350 3 Difference between 1 and 2 14.477 18 .697 0.000 doi:10.1371/journal.pone.0099576.t003 School Belongingness and Mental Health in Youth standing of how and when risk is expressed as disorder; to determine the ideal time to intervene; and the relevance of intervention on student outcomes. Conclusions The current study adds to the growing body of research examining the role of school contextual influences in early adolescent mental health functioning. Adolescent experiences of belonging to, and closeness with, others at the school may buffer or offset the subsequent negative mental health functioning, above and beyond prior mental health functioning. The current study’s findings highlight the importance for both primary and secondary schools to assess the belongingness and mental health needs of all their students. Such assessments could allow schools to pay special attention to those with poorer mental health functioning and school belongingness scores, as these are more likely to continue to be disadvantaged over time. Author Contributions Conceived and designed the experiments: SV AEP RP. Performed the experiments: SV RP. Analyzed the data: TP SV RP. Contributed reagents/materials/analysis tools: SV RP AEP. Wrote the paper: SV TF MF RP AEP TP. Critical review: TF MF RP. References 1. Kieling C, Baker-Henningham H, Belfer M, Conti G, Ertem I, et al. (2011) Child and adolescent mental health worldwide: evidence for action. The Lancet 378: 1515–1525. 2. Sawyer MG, Arney FM, Baghurst PA, Clark JJ, Graetz BW, et al. (2000) The mental health of young people in Australia: Child and adolescent component of the National Survey of Mental Health and Wellbeing. Canberra, Australia: Department of Health and Aged Care, Mental Health and Special Programs Branch. 3. McLennan W (1997) Mental Health and Wellbeing: Profile of Adults. Canberra, Australia: Australian Bureau of Statistics. 4. Australian Institute of Health and Welfare (AIHW) (2003) Australia’s young people: Their health and wellbeing. Canberra: Australian Institute of Health and Welfare. 5. Scott S (1998) Aggressive behaviour in childhood. BMJ 316: 202–206. 6. Hawton K, Rodham K, Evans E, Weatherall R (2002) Deliberate self harm in adolescents: self report survey in schools in England. BMJ 23: 1207–1211. 7. Parker G, Roy K (2001) Adolescent depression: A review. Aust N Z J Psychiatry 35: 572–580. 8. Sturgeon L (2007) Mental health and intellectual disability in children and adolescents: working with this challenging population, (PPT), Available: www. psychdd.com.au/IDandPsycillnessforDADHC.ppt. 9. Berkman LF, Kawachi I (2000) Social Epidemiology. New York: Oxford University Press. 10. Wickrama KAS, Conger RD, Lorenz FO, Jung T (2008) Family Antecedents and Consequences of Trajectories of Depressive Symptoms from Adolescence to Young Adulthood: A Life Course Investigation. Journal of Health and Social Behavior 49: 468–483. 11. Herd P, Goesling B, House JS (2007) Socioeconomic Position and Health: The Differential Effects of Education versus Income on the Onset versus Progression of Health Problems. Journal of Health and Social Behavior 48: 223–238. 12. Oakes JM, Rossi PH (2003) The measurement of SES in health research: current practice and steps toward a new approach. Social science & medicine (1982) 56: 769–784. 13. Moffitt TE, Caspi A, Dickson N, Silva P, Stanton W (1996) Childhood-onset versus adolescent-onset antisocial conduct problems in males: Natural history from ages 3 to 18 years. Development and Psychopathology 8: 399–424. 14. Orvaschel H, Lewinsohn PM, Seeley JR (1995) Continuity of Psychopathology in a Community Sample of Adolescents. Journal of the American Academy of Child and Adolescent Psychiatry 34: 1525–1535. 15. Zubrick S, Silburn SR, Gurrin L, Teoh H, Shepherd C, et al. (1997) The Western Australian Child Health Survey: Education, health and competence. Perth, WA: Australian Bureau of Statistics and the Institute for Child Health Research. 16. Zubrick S, Silburn SR, Burton P, Blair EM (2000) Mental health disorders in children and young people: Scope, cause and prevention. Australian and New Zealand Journal of Psychiatry 34: 570–578. 17. Rutter M (1995) Relationships between mental disorders in childhood and adulthood. Acta Psychiatrica Scandinavica 91: 73–85. PLOS ONE | www.plosone.org 8 June 2014 | Volume 9 | Issue 6 | e99576 Table 4. Multi-group analyses of the unconstrained and constrained longitudinal relationship between school belongingness and mental health functioning across the primary- secondary school transition (Household-SES as the grouping variable). Model no Name of model x df p NFI CFI RMSEA AIC 1 Unconstrained for each SES group 10.877 27 .621 .983 1.000 .000 64.877 2 Constrained 31.266 9 .453 .952 1.000 .004 49.266 3 Difference between 1 and 2 20.389 18 .311 0.000 doi:10.1371/journal.pone.0099576.t004 School Belongingness and Mental Health in Youth the National Longitudinal Study of Adolescent Health. Journal of Adolescent 18. Petticrew M, Chisholm D, Thomson H, Jane-LLlopis E (2005) Generating Health 27: 340–348. evidence on determinants, effectiveness and cost effectiveness. In: Herrman I, Saxena H, Moodie R, editors. Promoting Mental Health: Concepts, Emerging 46. Erikson EH (1968) Identity: Youth and crisis. New York: Norton. Evidence, Practice. Geneva: World Health Organisation. 47. Weiss RS (1973) Loneliness: The experience of emotional and social isolation. 19. Geller B, Zimerman B, Williams M, Bolhofner K, Craney JL (2001) Bipolar Cambridge, MA: MIT Press. disorder at prospective follow-up of adults who had prepubertal major depressive 48. Larson RW, Verma S (1999) How children and adolescents spend time across disorder. American Journal of Psychiatry 158: 125–127. the world: Work, play, and developmental opportunities. Psychological Bulletin 20. Geller B, Zimerman B, Williams M, Bolhofner K, Craney JL (2001) Adult 125: 701–736. Psychosocial Outcome of Prepubertal Major Depressive Disorder. Journal of the 49. Vaz S (2010) Factors affecting student adjustment as they transition from American Academy of Child and Adolescent Psychiatry 40: 673–677. primary to secondary school: A longitudinal investigation. Perth: Curtin University. 21. Cicchetti D, Rogosch FA (1999) Psychopathology as Risk for Adolescent 50. Vaz S, Parsons R, Falkmer T, Passmore AE, Falkmer M (2014) The impact of Substance Use Disorders: A Developmental Psychopathology Perspective. personal background and school contextual factors on academic competence Journal of Clinical Child Psychology 28: 355–365. and mental health functioning across the primary-secondary school transition. 22. Fordwood SR, Asarnow JR, Huizar DP, Reise SP (2007) Suicide attempts PLOS ONE. among depressed adolescents in primary care. Journal of Clinical Child and 51. Australian Bureau of Statistics [ABS] (2001) ABS 2001 Census Dictionary (Cat. Adolescent Psychololgy 36: 392–404. No. 2901.0). 23. Patel V, Flisher AJ, Hetrick S, McGorry P (2007) Mental health of young people: 52. Goodman R (1997) The strengths and difficulties questionnaire: A research note. a global public health challenge. The Lancet 369: 1302–1313. Journal of Child Psychology and Psychiatry 38: 581–586. 24. Hickie IB, Groom G, Davenport T (2004) Investing in Australia’s future: The 53. Mellor D (2005) Normative data for the strengths and difficulties questionnaire personal, social and economic benefits of good mental health. Canberra: Mental in Australia. Australian Psychologist 40: 215–222. Health Council of Australia. 54. Achenbach TM (1991) Manual for the youth self-report and 1991 profile. 25. Davis C, Martin G, Kosky R, O’Hanlon A (2000) Early Intervention in the Burlington, VT: University of Vermont Department of Psychiatry. Mental Health of Young People: A Literature Review. Monograph published by 55. Goodman R, Scott S (1999) Comparing the Strengths and Difficulties The Australian Early Intervention Network for Mental Health in Young People. Questionnaire and the Child Behaviour Checklist: Is small beautiful? Journal 26. Weisz JR, McCarty CA, Valeri SM (2006) Effects of psychotherapy for of Abnormal Child Psychology: 1. depression in children and adolescents: A meta-analysis. Psychological Bulletin 56. Goodman R (1999) The extended version of the strengths and difficulties 132: 132–149. questionnaire as a guide to child psychiatric caseness and consequent burden. 27. Zalsman G, Brent DA, Weersing VR (2006) Depressive Disorders in Childhood Journal of Child Psychology and Psychiatry 40: 791–801. and Adolescence: An Overview: Epidemiology, Clinical Manifestation and Risk 57. Goodman R, Ford T, Simmons H, Gatward R, Meltzer H (2003) Using the Factors. Child and Adolescent Psychiatric Clinics of North America 15: 827– Strengths and Difficulties Questionnaire (SDQ) to screen for child psychiatric disorders in a community sample. International Review of Psychiatry 15: 166– 28. Sawyer MG, Arney FM, Baghurst PA, Clark JJ, Graetz RJ, et al. (2001) The mental health of young people in Australia: Key findings from the child and 58. Hagborg WJ (1994) An exploration of school membership among middle- and adolescent component of the national survey of mental health and well-being. high-school students. Journal of Psychoeducational Assessment 12: 312–323. Australian and New Zealand Journal of Psychiatry 35: 806–814. 59. McMahon SD, Parnes AL, Keys CB,Viola JJ (2008) School belonging among 29. Verhulst FC, Koot HM (1995) The Epidemiology of Child and Adolescent low-income urban youth with disabilities. Psychology in the Schools 45: 26–47. Psychopathology. London: Oxford University Press. 60. Zubrick S, Williams A, Silburn SR, Vimpani G (2000) Indicators of social and 30. Sawyer MG, Pfeiffer S, Spence SH (2009) Life events, coping and depressive family functioning. Canberra: Commonwealth of Australia. symptoms among young adolescents: A one-year prospective study. Journal of 61. Meyers LS, Gamst G, Guarino AJ (2006) Applied multivariate research: Design affective disorders 117: 48–54. and implication. CA: Sage Publications, Inc. 31. World Health Organisation [WHO] (1996) Promoting health through schools - 62. Tabachnick B, Fidell L (2007) Using multivariate statistics. Boston, MA: Pearson The World Health Organisation’s global school health initiative. Geneva: World Education Inc. and Allyn & Bacon. Health Organisation. 63. McKnight P, McKnight K, Sidani S, Figueredo A (2007) Missing data: A gentle 32. Goodenow C (1993) The psychological sense of school membership among introduction New York: Guilford. adolescents: Scale development and educational correlates. Psychology in the 64. West SG, Finch JF, Curran PJ (1995) Structural equation models with non- Schools 30: 79–90. normal variables: Problems and Remedies. In R.H. Hoyle (ed.) Structural 33. Blum RW (2005) A case for school connectedness. Educational Leadership 62: Equation Modeling: Concepts, Issues and Applications. Sage Publications; 16–20. Thousand Oaks CA. 34. Hargreaves A, Earl L, Ryan J (1996) Schooling for change: Reinventing 65. Kline RB (2005) Principles and practice of structural equation modelling. New education for early adolescents. London: The Falmer Press. York: Guilford Press. 35. Jacobson KC, Rowe DC (1999) Genetic and environmental influences on the 66. Bentler PM, Bonett DG (1980) Significance tests and goodness-of-fit in the relationship between family connectedness, school connectedness and adolescent analysis of covariance structures. Psychological Bulletin, 88, 588–600. depressed mood: Sex differences. Developmental Psychology 35: 926–939. 67. Hu L, Bentler PM (1999) Cutoff criteria for fit indexes in covariance structure 36. Resnick M, Bearman P, Blum RW, Bauman KE, Harris K, et al. (1997) analysis: Conventional criteria versus new alternatives. Structural Equation Protecting adolescents from harm: Findings from the National Longitudinal Modeling, 6, 1–55. Study on Adolescent Health. JAMA 278: 823–832. 68. Cheung GW, Rensvold RB (2002) Evaluating goodness-of-fit indexes for testing 37. Johnson MK, Crosnoe R, Thaden LL (2006) Gendered Patterns in Adolescents’ measurement invariance. Structural equation modeling 9: 233–255. School Attachment. Social Psychology Quarterly 69: 284–295. 69. SDQ (2007) Information for researchers and professionals about the Strengths & 38. Ma X (2003) Sense of Belonging to School: Can Schools Make a Difference? Difficulties Questionnaires. The Journal of Educational Research 96: 340–349. 70. Anderman EM (2002) School effects on psychological outcomes during 39. Kuperminic GP, Leadbeater BJ, Blatt SJ (2001) School Social Climate and adolescence. Journal of Educational Psychology 94: 795–808. Individual Differences in Vulnerability to Psychopathology among Middle 71. Loukas A, Ripperher-Suhler KG, Horton KD (2009) Examining temporal School Students. Journal of School Psychology 39: 141–159. associations between school connectedness and early adolescent adjustment 40. Shochet IM, Dadds MR, Ham D, Montague R (2006) School Connectedness Is Journal of Youth and Adolescence 38: 808–812. an Underemphasized Parameter in Adolescent Mental Health: Results of a 72. Whitlock JL (2006) Youth perceptions of life at school: Contextual correlates of Community Prediction Study. Journal of Clinical Child & Adolescent school connectedness in adolescence. Applied Developmental Science 10: 13–29. Psychology 35: 170–179. 73. Battistich V, Solomon D, Watson M, Schaps E (1997) Caring school 41. Loukas A, Ripperger-Suhler KG, Horton KD (2009) Examining Temporal communities. Educational Psychologist 32: 137–151. Associations Between School Connectedness and Early Adolescent Adjustment. 74. Burton ES, Stice E, Seeley JR (2004) A prospective test of the stress-buffering Journal of Youth and Adolescence 38: 804–812. model of depression in adolescent girls: No support once again. Journal of 42. Loukas A, Roalson LA, Herrera DE (2010) School connectedness buffers the Consulting and Clinical Psychology 72: 689–697. effects of negative family relations and poor effortful control on early adolescent 75. Cernkovich SA, Giordano PC (1992) School bonding, race, and delinquency. conduct problems. Journal of Research on Adolescence 20: 13–22. Criminology 30: 261–291. 43. Catalano RF, Oesterle S, Fleming CB, Hawkins JD (2004) The Importance of 76. Wade TJ, Brannigan A (1998) The genesis of adolescent risktaking: Pathways Bonding to School for Healthy Development: Findings from the Social through family, school, and peers. Canadian Journal of Sociology 23: 1–19. Development Research Group. Journal of School Health 74: 252–261. 77. Hirschi T (1969) Causes of delinquency. Berkeley, CA University of California 44. Bond L, Butler H, Thomas L, Carlin J, Glover S, et al. (2007) Social and School Press. Connectedness in Early Secondary School as Predictors of Late Teenage 78. Eccles JS, Lord S, Midgley C (1991) What are we doing to early adolescents? Substance Use, Mental Health, and Academic Outcomes. The Journal of The impact of educational contexts on early adolescence. American Journal of adolescent health: official publication of the Society for Adolescent Medicine 40: Education 99: 521–542. 357. e359–357. e318. 79. Eccles JS, Midgley C (1989) Stage-environment fit: Developmentally approriate 45. Svetaz MV, Ireland M, Blum RW (2000) Adolescents with learning disabilities: classrrooms for young adolescents. In: Ames R, Ames C, editors. Research on Risk and protective factors associated with emotional well-being: Findings from motivation in education. New York: Academic Press. 139–181. PLOS ONE | www.plosone.org 9 June 2014 | Volume 9 | Issue 6 | e99576 School Belongingness and Mental Health in Youth 80. Anderman LH (2003) Academic and social perceptions as predictors of change 84. Anderman LH, Freeman TM (2004) Students’ sense of belonging in school. In: Pintrich PR, Maehr ML, editors. Motivating students, improving schools: The in middle school students’ sense of school belonging. Journal of Experimental legacy of Carol Midgley Oxford, UK: Elsvier, Ltd. 27–63. Education 72: 5–22. 85. Monahan K, Oesterle S, Hawkins JD (2010) Predictors and consequences of 81. Morrison GM, Cosden MA, O’Farrell SL, Campos E (2003) Changes in Latino school connectedness: The case for prevention. The Prevention Researcher 17: students’ perceptions of school belonging over time: Impact of language 3–6. proficiency, self-perceptions and teacher evaluations. The California School 86. Bell M, Dempsey I (2001) Enrolment and placement practices for students with Psychologist 8. special needs in NSW government schools: A critical analysis of current policy. 82. Dishion TJ, Patterson GR, Stoolmiller M, Skinner ML (1991) Family, school, Special Education Perspectives 10: 3–14. and behavioral antecedents to early adolescent involvement with antisocial 87. Stone LL, Otten R, Engels RC, Vermulst AA, Janssens JM (2010) Psychometric peers. Developmental Psychology 27: 172–180. properties of the parent and teacher versions of the strengths and difficulties 83. Birch SH, Ladd GW (1998) Children’s interpersonal behaviors and the teacher– questionnaire for 4- to 12-year-olds: a review. Clinical child and family child relationship. Developmental Psychology 34: 934–946. psychology review 13: 254–274. PLOS ONE | www.plosone.org 10 June 2014 | Volume 9 | Issue 6 | e99576

Journal

PLoS ONE – Unpaywall

Published: Jun 26, 2014

There are no references for this article.