Journal of Adolescence 70 (2019) 62–73
Contents lists available at ScienceDirect
Journal of Adolescence journal homepage: www.elsevier.com/locate/adolescence
School effectiveness and students' future orientation: A multilevel analysis of upper secondary schools in Stockholm, Sweden
T
Susanne Alma,∗, Sara Brolin Låftmanb, Julia Sandahlc, Bitte Modinb a b c
Swedish Institute for Social Research (SOFI), Stockholm University, SE-106 91, Stockholm, Sweden Department of Public Health Sciences, Stockholm University, SE-106 91, Stockholm, Sweden Department of Criminology, Stockholm University, SE-106 91, Stockholm, Sweden
ABSTRACT
Introduction: Future orientation (FO) refers to individuals' beliefs and feelings about their future. Earlier research has primarily investigated correlates of FO at the individual and family level, but it seems likely that FO is also shaped by other central agents or institutions, such as the school. Earlier studies have found positive associations between “school effectiveness” and student performance, and negative associations in relation to e.g., bullying, delinquency, and health risk behaviors. The current study investigated three teacher-reported features of school effectiveness - school leadership, teacher cooperation and consensus, and school ethos - and their links with student-reported FO. Methods: Survey data were collected in 2016 among 5131 students (aged 17–18 years) and 1061 teachers in 46 upper secondary schools in Stockholm, Sweden, and merged with school-level register data. Two-level binary logistic regression analyses were performed. Results: The analyses showed that higher teacher ratings of school leadership and school ethos were associated with a greater likelihood of reporting an optimistic FO among students. Teacher cooperation and consensus was however not associated with students' FO. Conclusion: The findings indicate that the school environment contributes to shaping students' beliefs about their future. Thus, enhancing features of school effectiveness may be a way of promoting a positive development and brighter objective future prospects for the young, via pathways such as good student-teacher relations and academic motivation and achievement.
Future orientation can be understood as individuals' thoughts, plans, motivations, hopes and feelings about their future (Stoddard, Zimmerman, & Bauermeister, 2011). Even though people orient themselves towards the future throughout their lives, thoughts and plans about the future are especially significant during adolescence and young adulthood, since this is when choices about education, occupation, and other central aspects of life are formed (Chua, Milfont & Jose, 2015; Nurmi, 1991). Earlier studies have demonstrated that adolescents' future orientation differs by socioeconomic background (Alm, 2011; Alm & Estrada, 2018; Alm & Låftman, 2016; Halleröd, 2011), but also that the way in which the individual perceives his or her future, in itself, has a bearing upon actual outcomes later in life (Alm, 2011; Halleröd, 2011; see also; Brezina, Tekin & Topali, 2009; Piquero, 2016; Alm & Estrada, 2018). This makes the study of future orientation and its determinants important from a policy perspective. From a theoretical point of view, it seems possible that an adolescent's future orientation is shaped not only by living conditions in the family of origin, but also by other central agents and institutions in his or her environment, such as the school. This has however been studied to a much lesser extent. Adolescents spend a considerable part of their time in school. School offers opportunities to establish social relations outside the family, primarily with peers, but also with adults who can act as role models and provide guidance and support in relation to future life choices. The possibilities on part of the school to act as a favourable social arena for the students can be expected to be linked to its capacities of providing “school effectiveness”. The literature on school effectiveness has asserted that some schools are more successful than others in creating a positive school environment that is beneficial for student outcomes, irrespective of the schools' Corresponding author. E-mail addresses:
[email protected] (S. Alm),
[email protected] (S.B. Låftman),
[email protected] (J. Sandahl),
[email protected] (B. Modin). ∗
https://doi.org/10.1016/j.adolescence.2018.11.007 Received 5 June 2018; Received in revised form 2 October 2018; Accepted 27 November 2018 Available online 10 December 2018 0140-1971/ © 2018 The Authors. Published by Elsevier Ltd on behalf of The Foundation for Professionals in Services for Adolescents. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
sociodemographic student composition (Mortimore, Sammons, Stoll, Lewis, & Russel, 1989; Rutter, Maughan, Mortimore, Ouston, & Smith, 1979; West, Sweeting, & Leyland, 2004). A strong and well targeted school leadership, a high degree of cooperation and consensus among teachers, and a positive school ethos are central aspects of effective schools, which are interrelated but presenting themselves at different levels of the school organization (Granvik Saminathen, Låftman, Almquist, & Modin, 2018; Ramberg, Låftman, Almquist, & Modin, 2018a). Earlier studies have shown that school-contextual features influence adolescents and their well-being in various ways (Bonell et al., 2013; Eriksson & Sellström, 2010; Låftman & Modin, 2012; Modin & Östberg, 2009; Sellström & Bremberg, 2006). To the best of our knowledge, however, no study has hitherto investigated the links between school-contextual characteristics and students' future orientation. The aim of the current study is to analyze three teacher-reported school-contextual features reflecting school effectiveness – school leadership, teacher cooperation and consensus, and school ethos – and their links with student-reported future orientation. 1. Future orientation There are many ways to approach individuals' thoughts and feelings about the future. A useful umbrella term is future orientation, originally introduced by Trommsdorff (1983) and by Nurmi (1991). Future orientation can be divided into a cognitive, a motivational and an affective component, where the affective one, which primarily concerns a person's feelings in terms of optimism/pessimism regarding the future, is perhaps the most intuitive and the most commonly studied aspect. The affective component is also the focus of the present study. Earlier studies have demonstrated that adolescents' future orientation differs by socioeconomic background, in that those from lower social strata tend to have a less positive view of their future compared with their peers with a more advantaged social background (Alm, 2011; Halleröd, 2011). One interpretation of this is that individuals already at a quite early age are aware of the unequal life chances in relation to socioeconomic status (Halleröd, 2011). In addition, previous research has indicated that the future orientation among adolescents may also in itself be independently related to their later life chances. Using longitudinal survey and register data, Alm (2011) found that when socioeconomic background was taken into consideration, an individual's future orientation in adolescence was related to actual outcomes in adult life, with respect to labor market attachment, and even to premature death. An optimistic future orientation may thus be one of the mechanisms explaining the inheritance of advantageous living conditions. Departing from this, it is important to study other possible determinants of future orientation among the young. 2. School effectiveness In their seminal work on “effective schools”, based on empirical studies of schools in London in the 1970s, Michael Rutter and his colleagues demonstrated that there were organizational and environmental factors in the school setting that had a positive influence on students' performance, even when the students' social background was taken into consideration (Rutter et al., 1979). Central features of effective schools are high expectations of students, transparent goals, constructive feedback, regular monitoring of students' progress, an orderly environment, and strong parent-school relations (Edmonds, 1979; Rutter et al., 1979; Wang, Walters, & Thum, 2012; West et al., 2004). More recent studies have broadened the focus of the implications of effective schools and negative associations have been found between aspects of school effectiveness and, for example, delinquency (Gottfredson, 2001; Sandahl, 2016), bullying (Ertesvåg & Roland, 2015; Låftman, Östberg, & Modin, 2017a; Modin, Låftman, & Östberg, 2017), truancy (Ramberg, Låftman, Fransson, & Modin, 2018b) and health risk behaviors (Bonell, Fletcher, & McCambridge, 2007; West et al., 2004). Possible mechanisms between features of school effectiveness and student outcomes may include students' experiences of respectful student-teacher relationships and different types of teacher support, but also academic motivation and achievement. Indeed, it has been shown that school leadership, teacher cooperation and consensus and school ethos are all positively related to students' perceived teacher caring (Ramberg et al., 2018a), as well as to students' academic achievement (Granvik Saminathen et al., 2018). It is reasonable to assume that features of school effectiveness also promote students' self-esteem and an optimistic future orientation, via their positive impact on, for instance, strong and respectful student-teacher relationships and academic motivation and achievement. 2.1. School leadership Research on school effectiveness has underlined the crucial role of good school leadership and pointed to the importance of its capacity to formulate a vision for the school, as well as being in charge of the measures taken to implement it in practice (Blair, 2002). Providing structures for effective collegial work and continuously offering clear directions for its further improvement are two ways of promoting a positive and efficient work environment (Ertesvåg & Roland, 2015).
63
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
Research has shown positive effects of school leadership on student learning and achievement (Granvik Saminathen et al., 2018; Leithwood, Seashore Louis, Anderson, & Wahlstrom, 2004), but there are also studies indicating that school leadership may have other implications for students in addition to the purely academic. For example, high teacher ratings of the school leadership have been shown to be related to a lower prevalence of bullying among students (Ertesvåg & Roland, 2015; Låftman et al., 2017a; Roland & Galloway, 2004). 2.2. Teacher cooperation and consensus As discussed above, a successful school leadership can be considered a prerequisite for obtaining the necessary conditions for efficient work on part of the teachers. And for work to indeed be efficient, good communication and cooperation between members of staff are central components. Not surprisingly, a majority of teachers consider teacher cooperation to be important, since it enables exchange of both skills and support and increases the teacher morale (Johnson, 2003). In addition, consensus regarding the school's goals and the means to reach these goals has shown to be of importance for effective schools (Sammons, Hillman, & Mortimore, 1995). Although, perhaps not surprisingly, teacher cooperation and consensus tend to first and foremost improve the quality of work for the teachers themselves, there are indications that it also affects students in terms of enhanced learning and achievement (Granvik Saminathen et al., 2018; Vangrieken, Dochy, Raes, & Kyndt, 2015). Furthermore, in a similar manner as for school leadership, earlier studies have demonstrated that schools where the teacher collaboration is rated as strong have a more positive social climate as reflected through the fact that bullying among students is less common (Ertesvåg & Roland, 2015; Roland & Galloway, 2004). 2.3. School ethos According to Rutter et al. (1979, p. 179), the concept of school ethos refers to norms, values, attitudes and behaviors characterizing the patterns of interaction of both teachers and students in a given school (see also Modin et al., 2017). The term school climate is related to that of school ethos in that both concepts refer to social aspects and the overall “atmosphere” of the school environment. However, in empirical studies, school climate is often conceptualized in terms of students' experiences of the school environment, and measured from student-reported information which has been aggregated to the class or school level (Gower, McMorris, & Eisenberg, 2015; Låftman, Östberg, & Modin, 2017b; Modin & Östberg, 2009). By contrast, school ethos, as one core feature of effective schools, reflects school-contextual features that are imposed from higher levels in the school structure through a well targeted leadership (Modin et al., 2017). Originally, a favourable school ethos was shown to interrelate with higher academic performance (Granvik Saminathen et al., 2018; Mortimore et al., 1989; Rutter et al., 1979) but more recent studies have also demonstrated associations with, for example, less student involvement in health risk behaviors (Aveyard, Markman & Cheng 2004; Bonell et al., 2007; West et al., 2004) as well as with less traditional bullying and cyberbullying behaviour among students (Modin et al., 2017). 3. The Swedish school In Sweden, compulsory school encompasses grades 1–9 (preceded by a mandatory preschool year). Children enrol in the first grade the calendar year when they turn seven, and graduate from the ninth grade the year they turn 16. Upper secondary school is not mandatory but the majority of adolescents enrol in secondary education. The upper secondary programmes are three years of length. There are 18 national programmes in upper secondary school of which six academic and 12 vocational. Schools are run by municipalities (public schools) or by other actors (independent schools), but irrespective of the mandatorship, through a school voucher system, all education is free of charge in Sweden. The market oriented reforms of the Swedish school system in recent decades has led to increased school segregation. This is particularly true in the large urban areas, where the less attractive schools have had to deal with a decreasing number of motivated students and fewer students whose parents hold a higher education. As a consequence of this, the principal's capacity to build a common value system for the school and to provide the necessary conditions for their teachers to work “collectively” in a studentcentered manner is likely to differ substantially between schools. Indeed, a recent study by Granvik Saminathen et al. (2018) shows that teachers' ratings of school effectiveness vary significantly with the sociodemographic composition of students, in that schools with a more socioeconomically privileged student body enjoy higher teacher ratings of school effectiveness compared with schools dominated by students with a more disadvantaged social background. In order to combat inequalities in life chances as a consequence of socioeconomic differences, schools are not only obliged to offer all children equal quality of schooling, but they also have a compensatory duty in relation to the students' social background (Granvik Saminathen et al., 2018; Swedish National Agency for Swedish National Agency of Education, 2012). To strengthen the prerequisites of schools with a socioeconomically less privileged student composition to fulfil this compensatory duty, extra financial resources are allocated by the municipality. Typically, these resources are used to increase the number of teachers, resulting in a lower studentteacher ratio in schools with a less socioeconomically privileged student composition (Granvik Saminathen et al., 2018).
64
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
Features of school effectiveness are associated with other school contextual characteristics, such as the student composition, and potentially also the student-teacher ratio. Furthermore, it has been shown that socio-economic characteristics of an individual's neighborhood have an impact on adolescents' future orientation (Alm & Brännström, 2011), a tendency which may also be found for schools. Therefore, in studies on school effectiveness and students' future orientation, we believe it is important to adjust for school contextual characteristics related to the socio-demographic composition, in order to try to isolate the effect of the variables of interest as far as possible. 4. The current study Using a new, large data material based on survey information from 1061 teachers and 5131 students in 46 upper secondary schools in Stockholm as well as register data on schools, the main aim of this study is to examine whether teachers' ratings of three aspects of school effectiveness – school leadership, teacher cooperation and consensus, and school ethos – are related to students' reports on their future orientation, even when adjusting for individual and school sociodemographic characteristics. Since the data material was not originally collected for the specific purpose of the current study, it is by definition a secondary data analysis. Building on knowledge from earlier studies in related areas, three hypotheses were formulated: H1. Students attending schools with higher teacher ratings of school leadership are more likely to report an optimistic future orientation compared with students in schools with lower teacher ratings of school leadership. H2. Students attending schools with higher teacher ratings of teacher cooperation and consensus are more likely to report an optimistic future orientation compared with students in schools with lower teacher ratings of teacher cooperation and consensus. H3. Students attending schools with higher teacher ratings of school ethos are more likely to report an optimistic future orientation compared with students in schools with lower teacher ratings of school ethos. 5. Study participants and procedure The data material is based on combined information from two separate surveys performed in Stockholm, Sweden, in 2016: the Stockholm School Survey (SSS) and the Stockholm Teacher Survey (STS). To the data, we also linked school-level official register information from the Swedish National Agency for Education, including information on school level on, for example, proportion of parents with tertiary education, proportion of parents with foreign background, and number of students per teacher. The SSS is a cross-sectional survey conducted biannually among students in the ninth grade of compulsory school (ages 15–16 years) and the second grade of upper secondary school (ages 17–18 years) in all public schools and a large number of independent schools in the Stockholm municipality. Students fill in the questionnaires in the classroom and return them to the teacher upon completion. The SSS covers questions on risk behaviours including alcohol consumption, drug use, smoking and delinquent behaviour, but also topics such as the social situation in school and psychological health (Låftman et al., 2017b; Låftman & Modin, 2012; Modin & Östberg, 2009; Sandahl, 2016). In the 2016 questionnaire, a question on students' future orientation was also included. The response rate has been estimated to approximately 78 per cent (Stockholm Municipality, 2016). The STS is a web survey performed among teachers. In 2014, the survey was carried out among teachers in the senior-level schools participating in the SSS, and in 2016 among teachers in senior-level and upper secondary schools participating in the SSS. The overall aim of the STS was to collect information about school-contextual features, in terms of teachers' assessments of aspects such as the school leadership, cooperation and consensus among colleagues, and school ethos, but also about other aspects of teachers' working conditions. The purpose was to analyze if these school-contextual characteristics were linked with student-reported outcomes such as psychological health, bullying, and academic achievement. In the 2016 survey, 1247 senior level teachers and 1414 upper secondary school teachers participated, corresponding to a response rate of 54% and 58%, respectively. School-level measures were constructed by aggregating teachers' ratings to the school level. These measures were subsequently merged to the SSS studentlevel data. In the current study, we use data from the SSS of 2016 collected among students in the second grade of upper secondary school (17–18 years) with linked school-level information collected from upper secondary school teachers in the STS of 2016 and official register information on schools from the Swedish National Agency for Education. Information from all three sources was available for 46 upper secondary schools, with responses from 5604 students and 1061 teachers. The relatively large number of teachers in comparison to students – almost one teacher in every five students – is a consequence of the fact that only students in the second grade of upper secondary school were surveyed, while the teacher survey was sent to all the teachers at each upper secondary school. For the current study, we excluded 473 students who had not replied to all of the questions used in the analyses, resulting in a study sample of 5131 students distributed across 46 schools (i.e., 91.6% of the original sample). For more information on the data material, please see the Technical Report (Kjellström, Holmin von Saenger, Löfgren Jarl, & Modin, 2018).
65
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
6. Ethics Data from the Stockholm School Survey are filled in anonymously and there is no information on personal identification. According to a decision by the Regional Ethical Review Board of Stockholm (2010/241-31/5), this means that the data are not subject to consideration for ethical approval. The Regional Ethical Review Board of Stockholm has provided ethical permission for the Stockholm Teacher Survey (2015/1827-31/5). 7. Student-level variables 7.1. Future orientation Future orientation was assessed through the SSS and measured by the question “If you compared your future prospects to those of most others of your age, do you believe that yours are worse, just as good, or better?” The reponse categories were (1) “Much worse”, (2) “Slightly worse”, (3) “Just as good”, (4) “Slightly better”, and (5) “Much better”. We dichotomized the measure with the first two response categories classified as a pessimistic future orientation, and the three latter response categories classified as an optimistic future orientation. The question has been used in previous studies of adolescents, both as a predictor (Alm & Estrada, 2018; Halleröd, 2011) and as an outcome (Låftman, Alm, Sandahl, & Modin, 2018). 7.2. Controls Gender. Gender was measured by the question “Are you a boy or a girl?” and the response categories “Boy” and “Girl”. Family Structure. Family structure was captured by the question “Which people do you live with?” with a list of boxes to be ticked. Those who ticked “Mother” and “Father” were classified as living with two parents in one household and were contrasted against all others. Parental University Education. Parental university education was assessed through the question “What is the highest education your parents have?” The response categories, to be ticked for mother and father separately, were: “Old elementary school (folkskola) or compulsory school (max 9 years schooling)”, “Upper secondary school”, “University and university college”, and “Don't know”. Due to large proportions of students reporting “Don't know” or skipping the question (17.2% for the mother and 21.6% for the father in the study sample), a dichotomous measure was created, where students who ticked “University and university college” for one or both parents were classified as having at least one parent with university education, and categorized against all others. Parental Unemployment. Parental unemployment was measured through the question “What do your parents do?” with a number of response categories, to be ticked for mother and father separately. Those who ticked “Unemployed” for one or both parents were categorized as having at least one parent who is unemployed, and were contrasted to all others. Migration Background. Migration background was measured by the question “How long have you lived in Sweden?” The response categories were: “All my life”, “10 years or more”, “5–9 years” and “Less than 5 years”. The last two categories were merged due to small numbers. 8. School-level variables 8.1. School leadership School leadership was measured by an index constructed from ten items. The items were: a) “The management has an interest in pedagogical questions”; b) “The management shows an understanding of my work problems”; c) “The school leaders have high expectations of me as a teacher”; d) “When the management makes decisions on important issues they first discuss it with the teaching staff”; e) “The majority of teachers' understanding of school goals and policies align with the management's”; f) “The management allows room for teachers' pedagogical freedom”; g) “I regularly receive feedback from the management about my performance as a teacher”; h) “The management is a good support for teachers experiencing difficulties with a class”; i) “The distribution of responsibility between teachers is clear at this school”; and j) “This school is led in a good way”. The response categories were: (5) “Strongly agree”; (4) “Agree”; (3) “Neither agree nor disagree”; (2) “Disagree”; and (1) “Strongly disagree”. Values from all items were summed, thus forming a scale with the possible range 10–50, with higher values indicating higher teacher ratings of the school leadership. The measure was based on exploratory and confirmatory factor analysis and demonstrated good psychometric properties (RMSEA = 0.056; CFI = 0.988; TLI = 0.985) and high internal consistency (Cronbach's alpha = 0.88).1 To capture teachers' ratings of the school leadership at the school-level, we used the mean value on the school leadership index for each school. 1 For the indices of school leadership, teacher cooperation and consensus, and school ethos, the psychometric properties are reported for the full data material, i.e. before selections. More detailed information on the indices is found in the Technical Report (Kjellström et al., 2018).
66
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
8.2. Teacher cooperation and consensus Teacher cooperation and consensus was measured by seven items: a) “Teachers meet regularly to discuss and plan lessons”; b) “There is consistency in the approach to school goals among teachers”; c) “Teachers at this school usually use the same methods to deal with students who break school rules”; d) “Teachers at this school usually use the same methods to deal with and prevent bullying among students”; e) “There is agreement among teachers in the teaching philosophy of this school”; f) “Teachers are in agreement that it is important to work well with parents”; and g) “I can discuss work problems with my colleagues”. The response categories were: (5) “Strongly agree”; (4) “Agree”; (3) “Neither agree nor disagree”; (2) “Disagree”; and (1) “Strongly disagree”. Values from the seven items were summed to a scale with the possible range of 7–35, with higher values indicating higher ratings of the teacher cooperation and consensus. The measure was based on exploratory and confirmatory factor analysis and showed good psychometric properties (RMSEA = 0.109; CFI = 0.966; TLI = 0.949). The index also demonstrates high internal consistency (Cronbach's alpha = 0.83). To measure teacher cooperation and consensus at the school-level, the mean value on the teacher cooperation and consensus index for each school was used. 8.3. School ethos School ethos was captured by an index based on twelve items: a) “At this school we have a value system (värdegrund) which is clear to students”; b) “At this school the teachers make an effort to provide positive feedback about students' performance”; c) “Teachers have high expectations of student performance”; d) “Teachers at this school take their time with students even if they want to discuss something other than school work”; e) “At this school we actively work on issues such as violence, bullying and harassment among students”; f) “This school provides a stimulating learning environment”; g) “The teachers at this school have a strong work ethic”; h) “The teachers work with strong enthusiasm”; i) “At this school the students are treated with respect”; j) “The teachers at this school feel confident as classroom leaders”; k) “At this school students' motivation is a stimulating part of work” and l) “There is high staff turnover amongst teachers at this school”. The response categories were (5) “Strongly agree”; (4) “Agree”; (3) “Neither agree nor disagree”; (2) “Disagree”; and (1) “Strongly disagree”, except for the last item i) which had a four point scale with the reponse categories (1) “Agree completely”, (2) “Agree somewhat”, (3) “Disagree somewhat”, and (4) “Disagree completely”. Values from all items were summed to form an index with the possible range 12–59, with higher values indicating stronger school ethos. The measure was based on exploratory and confirmatory factor analysis and showed good psychometric properties (RMSEA = 0.079; CFI = 0.931; TLI = 0.916) as well as high internal consistency (Cronbach's alpha = 0.88). The mean value of the index for each school was used as a school-level measure of school ethos. 8.4. Controls The controls on school level are all based on official register information provided by the Swedish National Agency for Education. School Type. School type includes two categories that indicate whether the school was public or independent. Proportion of Students with Parents with Tertiary Education. Information on the schools' proportion of students with parents with tertiary education is reported in per cent. Proportion of Students with Foreign-Born Parents. Information on the schools' proportion of students with foreign-born parents is reported in per cent. Number of Students per Teacher. The measure indicates the student-teacher ratio and is continuous. 9. Statistical method Since our data included students who were nested in schools, and since our main independent variables of interest operate at the school level and our dependent variable at the student level, the statistical method used was multilevel modelling. The software used was Stata, version 15. Two-level binary logistic regression models were estimated using the “xtmelogit” command. Since the three measures of school effectiveness were highly correlated (r = 0.59–0.78), we performed analyses including one aspect of school effectiveness at the time rather than including them in the same model. To detect potential non-linear associations between teacher-rated school effectiveness and student-reported future orientation, each of the three indices measuring school leadership, teacher cooperation and consensus, and school ethos were divided into three categories of about equal size, in order to distinguish schools with weak, intermediate, and strong teacher ratings of each aspect. However, as an additional check, we also ran models with the continuous measures. These results are not presented in tables but are reported in the text, after the discussion of the main results. 10. Results Descriptive statistics of our study sample are presented in Table 1. According to our operationalization, 18.2% of the students report a pessimistic future orientation, and 81.8% an optimistic one. Table 2 demonstrates proportions of students reporting an optimistic future orientation, by teachers' ratings of the school leadership, teacher cooperation and consensus, and school ethos. The distribution of an optimistic future orientation differed by school leadership and school ethos in a gradient manner. In schools where the teachers' ratings of the school leadership were comparatively 67
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
Table 1 Descriptives of variables included in the analyses. n = 5131 distributed across 46 upper secondary schools. Student level Future orientation Pessimistic Optimistic Gender Boy Girl Family structure Two parents in the same household Other Parental university education No parent At least one parent Parental unemployment No parent At least one parent Migration background ≥10 years in Sweden < 10 years in Sweden School level School type Public Independent School leadership Weak (n = 1900) Intermediate (n = 1651) Strong (n = 1580) Teacher cooperation and consensus Weak (n = 2005) Intermediate (n = 1657) Strong (n = 1469) School ethos Weak (n = 1736) Intermediate (n = 1743) Strong (n = 1652) % students with parents with tertiary education % students with foreign-born parents Number of students per teacher
n
%
4195 936
18.2 81.8
2383 2748
46.4 53.6
3279 1852
63.9 36.1
1709 3422
33.3 66.7
4814 317
93.8 6.2
4656 475 n
90.7 9.3 %
3101 2030 Mean
60.4 39.6 S.d.
Range
29.97 34.37 37.99
2.59 0.88 2.09
24.67-32.65 32.71-35.40 35.52-44.60
23.46 25.04 26.70
1.25 0.24 1.21
19.67-24.62 24.64-25.42 25.57-32.00
41.30 45.97 51.23 51.63 41.21 16.47
2.71 1.11 1.57 25.01 21.58 3.29
34.83-44.28 44.35–48.00 48.10–54.25 7.00–86.27 6.00–95.68 5.10–26.70
Table 2 Proportions of students reporting an optimistic future orientation, by teacher-rated school leadership, teacher cooperation and consensus, and school ethos. n = 5131 distributed across 46 upper secondary schools. For each of the three features of school effectiveness, students in schools with “weak” teacher ratings constitute the reference category. Optimistic future orientation % School leadership Weak (ref.) Intermediate Strong Teacher cooperation and consensus Weak (ref.) Intermediate Strong School ethos Weak (ref.) Intermediate Strong
***p < 0.001 **p < 0.01 *p < 0.05.
68
79.8 80.9 84.9
– n.s. *
80.1 84.1 81.3
– n.s. n.s.
76.4 82.2 87.0
– ** ***
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
Table 3 Student-reported future orientation regressed on teacher-rated school leadership. Odds ratios from two-level binary logistic regression models demonstrating the likelihood of reporting an optimistic future orientation (omitted reference category = pessimistic future orientation). n = 5131 students distributed across 46 upper secondary schools. Model 1
Model 2
OR Student level Gender Boy (ref.) Girl Family structure Two parents in the same household (ref.) Other Parental university education No parent (ref.) At least one parent Parental unemployment No parent (ref.) At least one parent Migration background ≥10 years in Sweden (ref.) < 10 years in Sweden School level School leadership Weak (ref.) Intermediate Strong School type Public (ref.) Independent % students with parents with tertiary education % students with foreign-born parents Number of students per teacher School-level variance (s.e.)
95% CI
1.00 1.03 1.36
– 0.74-1.44 0.98-1.87
0.12 (0.04)
OR
95% CI
1.00 1.13
– 0.97-1.31
1.00 0.72***
– 0.62-0.84
1.00 1.48***
– 1.27-1.73
1.00 0.85
– 0.64-1.13
1.00 1.07
– 0.83-1.38
1.00 1.01 1.31*
– 0.78-1.31 1.03-1.68
1.00 0.80 1.00 1.00 1.02 0.04 (0.02)
– 0.61-1.06 1.00–1.01 0.99-1.01 0.99-1.05
***p < 0.001 **p < 0.01 *p < 0.05.
strong, larger proportions of students had an optimistic future orientation, compared with schools where the leadership was comparatively weak (84.9 per cent vs. 79.8 per cent). The gradient pattern was even more pronounced in relation to school ethos: 87.0 per cent of the students in schools with a strong ethos had an optimistic future orientation, compared with 76.4 per cent of students in schools with a weak ethos. For teacher cooperation and consensus, there were no statistically significant differences in the distribution of students' future orientation. Subsequently, we assessed the associations between the three aspects of school effectiveness and students' future orientation in a multilevel framework, adjusting for student and school-level sociodemographic characteristics as well as school type (public/independent). Results from analyses of school leadership and students' future orientation are presented in Table 3. The odds ratios demonstrate the likelihood of reporting an optimistic future orientation, compared to a pessimistic one. Model 1 included the three levels of school leadership only, and displays no statistically significant association with future orientation. In Model 2, which included student- and school-level control variables, the estimates show that attending a school in which the teachers rated the leadership as strong, compared to weak, was associated with a significantly increased likelihood of reporting an optimistic future orientation (OR 1.31, 95% CI 1.03–1.68). In regards to the student-level control variables, there were no statistically significant differences in future orientation by gender, parental unemployment or migration background. However, students who did not live with two parents in the same household had a lower likelihood of reporting an optimistic future orientation compared with those living with two parents, and students with at least one university-educated parent displayed an increased likelihood of reporting an optimistic future orientation. Given the lack of response categories for those not identifying themselves as boys or girls, and for those living with two mothers or two fathers, unfortunately we cannot say anything about the future orientation among these students. None of the school-level control variables showed a statistically significant association with future orientation. Table 4 presents results from analyses of teachers' ratings of teacher cooperation and consensus and student-reported future
69
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
Table 4 Student-reported future orientation regressed on teacher-rated teacher cooperation and consensus. Odds ratios from two-level binary logistic regression models demonstrating the likelihood of reporting an optimistic future orientation (omitted reference category = pessimistic future orientation). n = 5131 students distributed across 46 upper secondary schools. Model 1
Model 2
OR Student level Gender Boy (ref.) Girl Family structure Two parents in the same household (ref.) Other Parental university education No parent (ref.) At least one parent Parental unemployment No parent (ref.) At least one parent Migration background ≥10 years in Sweden (ref.) < 10 years in Sweden School level Teacher cooperation and consensus Weak (ref.) Intermediate Strong School type Public (ref.) Independent % students with parents with tertiary education % students with foreign-born parents Number of students per teacher School-level variance (s.e.)
95% CI
1.00 1.38 1.13
– 0.98-1.96 0.83-1.54
0.13 (0.04)
OR
95% CI
1.00 1.12
– 0.96-1.30
1.00 0.72***
– 0.62-0.83
1.00 1.49***
– 1.27-1.74
1.00 0.86
– 0.65-1.14
1.00 1.08
– 0.84-1.40
1.00 1.26 1.14
– 0.97-1.63 0.88-1.47
1.00 0.81 1.00 1.00 1.02 0.04 (0.02)
– 0.62-1.07 0.99-1.01 0.99-1.01 0.99-1.05
***p < 0.001 **p < 0.01 *p < 0.05.
orientation. Neither of the models demonstrated any statistically significant association between teacher cooperation and consensus and future orientation. Analyses of the association between school ethos and future orientation are presented in Table 5. Model 1 showed that the likelihood of reporting an optimistic future orientation was associated with school ethos in a gradient manner and with statistically significant estimates. When adjusting for student- and school-level control variables in Model 2, the results remained robust. Students in schools with an intermediate school ethos had a higher likelihood of reporting an optimistic future orientation (OR 1.39, 95% CI 1.14–1.70) compared with those in schools with a relatively weak ethos, and the estimate was even stronger for students in schools with a strong ethos (OR 1.66, 95% CI 1.34–2.07). As an additional check, we also performed analyses with the continuous, z-standardized measures of school leadership, teacher cooperation and consensus, and school ethos, adjusted for all student- and school-level control variables (not presented in table). For school leadership, the association with future orientation was statistically significant only at the 10%-level (OR 1.10, 95% CI 0.99–1.22, p = 0.065). The rather weak estimate was a result of the association not being linear, i.e. the change in students' future orientation was not consistent across all levels of teacher reported school leadership. This means that analyzing categories of the measure was a more appropriate strategy than using a continuous measure. The continuous measure of teacher cooperation and consensus did not show any statistically significant association with future orientation (OR 1.08, 95% CI 0.98–1.19, p = 0.134). The continuous measure of school ethos was however shown to be associated with students' future orientation and statistically significant at the 1%-level (OR 1.22, 95% CI 1.11–1.35, p < 0.001).
70
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
Table 5 Student-reported future orientation regressed on teacher-rated school ethos. Odds ratios from two-level binary logistic regression models demonstrating the likelihood of reporting an optimistic future orientation (omitted reference category = pessimistic future orientation). n = 5131 students distributed across 46 upper secondary schools. Model 1
Model 2
OR Student level Gender Boy (ref.) Girl Family structure Two parents in the same household (ref.) Other Parental university education No parent (ref.) At least one parent Parental unemployment No parent (ref.) At least one parent Migration background ≥10 years in Sweden (ref.) < 10 years in Sweden School level School ethos Weak (ref.) Intermediate Strong School type Public (ref.) Independent % students with parents with tertiary education % students with foreign-born parents Number of students per teacher School-level variance (s.e.)
95% CI
1.00 1.43** 1.90***
– 1.11-1.84 1.44-2.51
0.06 (0.03)
OR
95% CI
1.00 1.10
– 0.94-1.27
1.00 0.72***
– 0.62-0.84
1.00 1.47***
– 1.26-1.72
1.00 0.85
– 0.64-1.12
1.00 1.06
– 0.83-1.37
1.00 1.39** 1.66***
– 1.14-1.70 1.34-2.07
1.00 0.90 1.00 1.00 1.01 0.01 (0.02)
– 0.71-1.13 1.00–1.01 1.00–1.01 0.99-1.04
***p < 0.001 **p < 0.01 *p < 0.05.
11. Discussion The current study investigated how three teacher-rated aspects of school effectiveness – school leadership, teacher cooperation and consensus, and school ethos – were associated with students' future orientation. We hypothesized that higher teacher ratings of these school-contextual features were associated with a greater likelihood of reporting an optimistic future orientation among students. The results showed that school leadership and in particular school ethos were positively linked to students' future orientation. The results remained even when a set of student- and school-level sociodemographic characteristics were adjusted for, indicating a robust relationship. Teacher cooperation and consensus was however not related with students' future orientation. A central question is how our findings regarding school leadership and school ethos can be understood. One potential pathway may be students' academic motivation and school performance. Indeed, the initial aim of research on school effectiveness was to study the association with students' academic performance, and school leadership, teacher cooperation and consensus, and school ethos have been shown to be positively related with student learning and achievement (Granvik Saminathen et al., 2018). It is likely that high performance in upper secondary school is linked to high educational aspirations for tertiary education and for a good position in the labor market, and high aspirations are in turn probably closely related to an optimistic future orientation. Another potential mechanism may revolve around the positive social relationships that tend to prevail in schools with a high degree of effectiveness. While supportive social relations are beneficial for well-being and health (Cohen, Underwood & Gottlieb, 2000), this may be the case also for individuals' optimism for the future. As measured in the present study, high teacher ratings of the school leadership to a great extent capture a positive relationship between the school leaders and the teachers, expressed through teachers' experiences of being listened to and respected by the school leaders. Our measure of school ethos taps into relationships between teachers and students in terms of, for example, teacher support including feedback to students, respectful treatment of students, and active work against bullying, indicating that schools with a strong ethos tend to have good student-teacher relationships. It seems likely that positive student-teacher relationships may promote students' future orientation, and indeed, school ethos showed a stronger association with future orientation than did school leadership. However, it also seems possible that positive relations between the school leaders and the teachers imply an overall supportive and gratifying school atmosphere, which may in turn be positively related to students' future orientation. A recent study reported that teachers' ratings of the school leadership, teacher cooperation and consensus, and school ethos were all positively associated with students' perceived teacher caring (Ramberg et al., 2018a), implying that positive social relationships at different levels in the organization permeate the school as a whole. 71
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
Furthermore, both strong school leadership and strong school ethos have been related to lower rates of bullying (Ertesvåg & Roland, 2015; Låftman et al., 2017a; Modin et al., 2017; Roland & Galloway, 2004). The occurrence of bullying in a school class has been shown to be associated with poorer well-being not only among those who are exposed but among all students (Låftman & Modin, 2017; Meilstrup et al., 2015; Modin et al., 2018; Modin, Låftman, & Östberg, 2015). In a similar way, it is possible that a poor social climate may affect also students' future orientation in a negative manner. While both school leadership and school ethos were positively related to students' future orientation, no association was found for teacher cooperation and consensus. One interpretation of this null finding is that organizational aspects operating at higher levels in the school structure, reflected in both school leadership and school ethos, have a more direct influence on students' well-being than do the teachers' ratings of the collegial work. Scrutinizing the mechanisms at work in the association between school-contextual characteristics and students' future orientation is a promising task for future research. To empirically examine the role of potential mediators, structural equation modelling would be a useful tool. Another possible way of investigating how features of the school may influence students' future orientation is to apply qualitative methods, for instance personal interviews or focus groups with students. A strength of the study is that it was based on newly collected data derived from different sources, including survey responses from teachers and students as well as register information on schools, thus decreasing the risk of common methods variance. There are however also limitations. One drawback with the data is that we lack information on the students' track in upper secondary school. In Sweden, there are 18 national programmes in upper secondary school and it would have been useful to be able to distinguish between academic and vocational programmes. There are also some limitations concerning some of the control measures. First, as regards gender, a third, gender neutral option would have been desirable. From previous studies we know that the gender identity is of crucial importance for young persons' well-being, and thereby, perhaps also for their future orientation. In a similar manner, as regards the question of family type, an option for those living with two mothers or two fathers would have been desirable. With regard to parental education, adolescents do not always have correct information on this (Wardle, Robb, & Johnson, 2002), which in the present study is reflected in the large shares responding “Don't know”. Furthermore, the measure of parental education is not ideal, since it only contrasts between students who have reported that one or both of their parents have a university education, versus all others (including those who responded “Don't know”). Finally, the measure of migration background, distinguishing those who have lived their whole lives in Sweden from those having lived in the country up to, or more than ten years respectively, is a bit crude, and it would have been desirable also with information on the adolescents' and their parents' country of birth. It should also be mentioned that since the data were collected in Stockholm, the findings cannot be generalized to a wider population. Future studies should assess the links between school organizational characteristics and students' future orientation also in other geographical settings and school systems. To conclude, the present study finds support for the assumption that features of school environment, in terms of school leadership and school ethos, are related to students' perceptions of their future prospects. Accordingly, enhancing schools' leadership and ethos may benefit not only learning and achievement, as shown by earlier studies, but also other student outcomes that are not purely academic. From earlier studies it is known that the future orientation in adolescence is related to outcomes in adult life with respect to, for example, level of education, labor market attachment, criminal offending, and even mortality (Alm, 2011; Alm & Estrada, 2018; Halleröd, 2011). Thus, enhancing features of school effectiveness may be a way of promoting a positive development and brighter objective future prospects for the young, via pathways such as good student-teacher relations and academic motivation and achievement. Furthermore, the fact that features of school effectiveness are generally weaker in schools with a less privileged student composition (Granvik Saminathen et al., 2018) points to the importance of focusing on particularly on these schools. Such a strategy could help to buffer against processes of cumulative disadvantage and offer children from different backgrounds equal opportunities, both in relation to education specifically, as well as to life chances in general. Acknowledgments The study was financed by the Swedish Research Council, Forte, Formas and Vinnova (2014-10107). References Alm, S. (2011). The worried, the competitive and the indifferent. Approaches to the future in youth, their structural roots and outcomes in adult life. Futures, 43, 552–562. Alm, S., & Brännström, L. (2011). Future orientation – does it matter where one lives? [Framtidstro – spelar det roll var man bor?]. In S. Alm, O. Bäckman, A. Gavanas, & A. Nilsson (Eds.). Utanförskap [exclusion] (pp. 219–249). (Stockholm: Institute for Futures Studies/Dialogos). Alm, S., & Estrada, F. (2018). Future prospects, deprivation, and criminality—a longitudinal birth cohort study. Deviant Behavior. https://doi.org/10.1080/01639625. 2017.1410609. Alm, S., & Låftman, S.,B. (2016). Future orientation climate in the school class: Relations to adolescent delinquency, heavy alcohol use, and internalizing problems. Children and Youth Services Review, 70, 324–331. Aveyard, P., Markham, W. A., & Cheng, K. K. (2004). A methodological and substantive review of the evidence that schools cause pupils to smoke. Social Science & Medicine, 58, 2253–2265. Blair, M. (2002). Effective school leadership: The multi-ethnic context. British Journal of Sociology of Education, 23(2), 180–191. Bonell, C., Fletcher, A., & McCambridge, J. (2007). Improving school ethos may reduce substance misuse and teenage pregnancy. BMJ, 334, 614–616. Bonell, C., Parry, W., Wells, H., Jamal, F., Fletcher, A., Harden, A., et al. (2013). The effects of the school environment on student health: A systematic review of multilevel studies. Health & Place, 180–191. Brezina, T., Tekin, T., & Topalli, V. (2009). “Might not be a tomorrow”: A multimethods approach to anticipated early death and youth crime. Criminology, 47(4), 1091–1129.
72
Journal of Adolescence 70 (2019) 62–73
S. Alm et al.
Chua, L. W., Milfont, T. L., Milfont, & Jose, P. E. (2015). Coping skills help explain how future-oriented adolescents accrue greater well-being over time. Journal of Youth and Adolescence, 44, 2028–2041. Cohen, S., Underwood, L. G., & Gottlieb, B. H. (Eds.). (2000). Social support measurement and intervention: A guide for health and social scientists. Oxford: Oxford University Press. Edmonds, R. (1979). Effective schools for the urban poor. Educational Leadership, 37, 15–27. Eriksson, U., & Sellström, E. (2010). School demands and subjective health complaints among Swedish schoolchildren: A multilevel study. Scandinavian Journal of Public Health, 38(4), 344–350. Ertesvåg, S. K., & Roland, E. (2015). Professional cultures and rates of bullying. School Effectiveness and School Improvement, 26(2), 195–214. Gottfredson, D. C. (2001). Schools and delinquency. New York: Cambridge University. Gower, A. L., McMorris, B. J., & Eisenberg, M. E. (2015). School-level contextual predictors of bullying and harassment experiences among adolescents. Social Science & Medicine, 147, 47–53. Granvik Saminathen, M., Låftman, S. B., Almquist, Y. B., & Modin, B. (2018). Effective schools in a segregated landscape: The link with academic achievement among ninth grade students in Stockholm. School Effectiveness and School Improvementhttps://doi.org/10.1080/09243453.2018.1470988. Halleröd, B. (2011). What do children know about their future? Do children's expectations predict outcomes in middle age? Social Forces, 90(1), 65–84. Johnson, B. (2003). Teacher collaboration: Good for some, not so good for others. Educational Studies, 29(4), 337–350. Kjellström, J., Holmin von Saenger, I., Löfgren Jarl, E., & Modin, B. (2018). Technical report for the teacher survey with linkage to the Stockholm school survey. Report RRPHS 2018:2 https://www.su.se/publichealth/publikationer/research-reports-in-public-health-sciences. Låftman, S. B., Alm, S., Sandahl, J., & Modin, B. (2018). Future orientation among students exposed to school bullying and cyberbullying victimization. International Journal of Environmental Research and Public Health, 15, 605. Låftman, S. B., & Modin, B. (2012). School-performance indicators and subjective health complaints: Are there gender differences? Sociology of Health & Illness, 34, 608–625. Låftman, S. B., & Modin, B. (2017). Peer victimization among classmates: Associations with students' internalizing problems, self‐esteem, and life satisfaction. International Journal of Environmental Research and Public Health, 14, 1218. Låftman, S. B., Östberg, V., & Modin, B. (2017a). School leadership and cyberbullying–A multilevel analysis. International Journal of Environmental Research and Public Health, 14, 1226. Låftman, S. B., Östberg, V., & Modin, B. (2017b). School climate and exposure to bullying: A multilevel study. School Effectiveness and School Improvement, 28, 153–164. Leithwood, K., Seashore Louis, K., Anderson, S., & Wahlstrom, K. (2004). How leadership influences student learning – Review of research. Learning from Leadership Project. Center for Applied Research and Educational Improvement, University of Minnesota. Meilstrup, C., Ersbøll, A. K., Nielsen, L., Koushede, V., Bendtsen, P., Due, P., et al. (2015). Emotional symptoms among adolescents: Epidemiological analysis of individual-, classroom- and school-level factors. European Journal of Public Health, 25, 644–649. Modin, B., Låftman, S. B., & Östberg, V. (2015). Bullying in context: An analysis of psychosomatic complaints among adolescents in Stockholm. Journal of School Violence, 14, 382–404. Modin, B., Låftman, S. B., & Östberg, V. (2017). Teacher rated school ethos and student reported bullying – a multilevel study of upper secondary schools in Stockholm, Sweden. International Journal of Environmental Research and Public Health, 14, 1565. Modin, B., & Östberg, V. (2009). School climate and psychosomatic health: a multilevel analysis. School Effectiveness and School Improvement, 20, 433–455. Modin, B., Plenty, S., Låftman, S. B., Gustafsson, P. A., Hjern, A., & Bergström, M. (2018). School contextual features of social disorder and mental health complaints – a multilevel analysis of sixth grade students. International Journal of Environmental Research and Public Health, 15, 156. Mortimore, P., Sammons, P., Stoll, L., Lewis, D., & Russel, E. (1989). A study of effective junior schools. International Journal of Educational Research, 13, 753–768. Nurmi, J.-E. (1991). How do adolescents see their future? A review of the development of future orientation and planning. Developmental Review, 11, 1–59. Piquero, A. (2016). “Take my license n’ all that jive, I can't see…35!” Little hope for the future encourages offending over time. Justice Quarterly, 33(1), 73–99. Ramberg, J., Låftman, S. B., Almquist, Y. B., & Modin, B. (2018a). School effectiveness and students' perceptions of teacher caring: A multilevel study. Improving Schools. https://doi.org/10.1177/1365480218764693. Ramberg, J., Låftman, S. B., Fransson, E., & Modin, B. (2018b). School effectiveness and truancy: A multilevel study of upper secondary schools in Stockholm. International Journal of Adolescence and Youth. https://doi.org/10.1080/02673843.2018.1503085. Roland, E., & Galloway, D. (2004). Professional cultures in schools with high and low rates of bullying. School Effectiveness and School Improvement, 15(3–4), 241–260. Rutter, M., Maughan, B., Mortimore, P., Ouston, J., & Smith, A. (1979). Fifteen thousand hours. Cambridge: Harvard University Press. Sammons, P., Hillman, L., & Mortimore, P. (1995). Key characteristics of effective schools: A Review of school effectiveness research. London: Institute of Education. Sandahl, J. (2016). School climate and delinquency – on the significance of the perceived social and learning climate in school for refraining from offending. Journal of Scandinavian Studies in Criminology and Crime Prevention, 2, 110–130. Sellström, E., & Bremberg, S. (2006). Is there a “school effect” on pupil outcomes? A review of multilevel studies. Journal of Epidemiology & Community Health, 60(2), 149–155. Stockholm Municipality (2016). Stockholmsenkäten Årsrapport 2016 Avdelningen för stadsövergripande sociala frågor. http://www.stockholm.se/PageFiles/738426/ Stockholmsenk%C3%A4ten%20%C3%85rsrapport%202016.pdf, Accessed date: 16 September 2017. Stoddard, S. A., Zimmerman, M. A., & Bauermeister, J. A. (2011). Thinking about the future as a way to succeed in the present. A longitudinal study of future orientation and violent behaviors among African American youth. American Journal of Community Psychology, 48(3–4), 238–246. Swedish National Agency of Education (2012). Equal education in Swedish comprehensive school? A qualitative analysis of equity over time [Likvärdig utbildning i svensk grundskola? En kvantitativ analys av likvärdighet över tid]. (Report No. 374). Stockholm: Fritzes. Trommsdorff, G. (1983). Future orientation and socialization. International Journal of Psychology, 18, 381–406. Vangrieken, R., Dochy, F., Raes, E., & Kyndt, E. (2015). Teacher collaboration: A systematic review. Educational Research Review, 15, 17–40. Wang, A. H., Walters, A. M., & Thum, Y. M. (2012). Identifying highly effective urban schools: Comparing two measures of school success. International Journal of Educational Management, 27(5), 517–540. Wardle, J., Robb, K., & Johnson, F. (2002). Assessing socioeconomic status in adolescents: The validity of a home affluence scale. Journal of Epidemiology & Community Health, 56, 595–599. West, P., Sweeting, H., & Leyland, A. (2004). School effects on pupils' health behaviors: Evidence in support of the health promoting school. Research Papers in Education, 19(3), 261–291.
73