The Association Between School Stress, Life ...

Soc Indic Res DOI 10.1007/s11205-014-0842-0

The Association Between School Stress, Life Satisfaction and Depressive Symptoms in Adolescents: Life Satisfaction as a Potential Mediator Unni K. Moksnes • Audhild Løhre • Monica Lillefjell Don G. Byrne • Gørill Haugan

•

Accepted: 29 November 2014 Ó Springer Science+Business Media Dordrecht 2014

Abstract The aim of the present study was to investigate the interrelationships between school-related stress (school performance, teacher interaction), life satisfaction, and depressive symptoms, as well as the potential mediating role of life satisfaction on the association between school-related stress and depressive symptoms. A total of 1,239 adolescents (13–18 years of age) from public elementary and secondary schools in midNorway participated in the school-based survey. The data were analysed using structural equations modelling. The present study showed that stress of school performance was significantly and positively related to depressive symptoms and significantly and inversely related to life satisfaction. At the bivariate levels, stress of teacher interaction was associated with more depressive symptoms and reduced life satisfaction. However, these associations were non-significant in the multivariate analyses, controlled for stress of school performance. A significant inverse association was found between life satisfaction and depressive symptoms. Further, life satisfaction partly mediated the association between stress of school performance and depressive symptoms. The results reflect the complexity

U. K. Moksnes (&) A. Løhre M. Lillefjell D. G. Byrne G. Haugan Center for Health Promotion Research, Mauritz Hansens gt 2, 7030 Trondheim, Norway e-mail: [email protected] U. K. Moksnes G. Haugan Faculty of Nursing, Sør-Trøndelag University College, Trondheim, Norway A. Løhre Department of Teacher and Interpreter Education, Sør-Trøndelag University College, Trondheim, Norway M. Lillefjell Department of Occupational Therapy, Faculty of Health Education and Social Work, Sør-Trøndelag University College, Trondheim, Norway D. G. Byrne Research School of Psychology, Australian National University, Canberra, Australia

123

U. K. Moksnes et al.

of the interaction between adolescents’ experience of school performance stress and mental health, and the role of life satisfaction as a potentially relevant mediator of this association. Keywords symptoms

Youth Subjective well-being Life satisfaction Stress Emotional

1 Introduction Adolescence is a crucial developmental life phase, marked by a multitude of significant physical, psychological, and social changes (Byrne et al. 2007; Grant et al. 2006; Moksnes et al. 2010a). While the transition through adolescence is inevitable, the speed and magnitude of these changes may overtax the capacity of many young people to cope, and the resulting phenomenon of adolescent stress is now well recognized (Compas and Reeslund 2009; Grant et al. 2006). The present study focuses on perceived stress and describes the individual’s subjective experience of pressure, implying an evaluation of the outcome of a process. This is in line with the transactional view, by which stress is the condition that results when person-environment transactions lead the individual to perceive a discrepancy—whether real or not—between the demands of a situation and the resources of the person’s biological, psychological, or social systems (Lazarus and Folkman 1984). The situations and pressures that cause stress are known as stressors (Grant et al. 2004). Education is fundamental in the lives of adolescents and school is an important developmental context for adolescents’ psychological functioning (Eccles and Roeser 2011; Lo¨hre et al. 2013). Some studies have drawn a parallel between the work context in adult life and the school context, emphasizing the importance of increasing social and academic demands in school as normative stressors in adolescents’ lives (Eccles and Roeser 2011; Lo¨hre et al. 2010; Suldo et al. 2009; Undheim and Sund 2005). Potential stressors in the school environment include troubled interaction with peers and teachers, demands of academic performance and school rules, as well as school/leisure conflict (Byrne et al. 2007; Hjern et al. 2008; Moksnes et al. 2010b; Moksnes and Espnes 2011). Although exposure to some stressful negative events is considered a normal part of development, stressors remain central as a potential threat to the well-being and healthy development of adolescents (Grant et al. 2004). Stressors related to academic performance and perceiving schoolwork as highly demanding are found to be associated with psychological complaints including depressive symptoms (Byrne et al. 2007; Fro¨jd et al. 2008; Moksnes et al. 2010b; Undheim and Sund 2005). The same association has been found with regard to perceiving stressful interaction with teachers (Byrne et al. 2007; Undheim and Sund 2005). Moreover, in line with the broader base of evidence on adolescent stress, girls seem to be more psychologically vulnerable when faced with adversity than boys (Charbonneau et al. 2009; Shih et al. 2006; Thapar et al. 2012). Nevertheless, in order to promote positive functioning during the adolescent years, it is essential to focus on factors that may act as resources in relation to potential stressful experiences and adversity. The life satisfaction construct (LS) represents the cognitive component of subjective well-being (Pavot and Diener 2008; Proctor et al. 2009a). Pavot and Diener (1993) define LS as ‘‘a judgemental process, in which individuals assess the quality of their lives on the basis of their own unique criteria (p. 164).’’ Evaluation of LS is thus a cognitive appraisal of the overall quality of a person’s life based on self-selected

123

Life Satisfaction as a Potential Mediator

standards. The individual’s perception of LS is regarded as a key indicator sensitive to the entire spectrum of functioning and mental health, and it might extend our understanding of adolescents’ ability to cope with developmental tasks and challenges (Goldbeck et al. 2007; Proctor et al. 2009a). In the face of stressful experiences, adolescents’ perception of LS might be a relevant mediator intervening on the relation between stress and depressive symptoms (Wu and Zumbo 2008). However, based on evaluation of articles published in data bases of PsychINFO, Psycharticles, Pubmed and Google Scholar, using search terms such as ‘‘satisfaction with life,’’ ‘‘subjective well-being,’’ ‘‘emotional health,’’ and ‘‘stress’’ much of the research conducted on LS has primarily been focusing on adult populations. Comparatively, limited work has examined LS in children and adolescents (Gilman and Huebner 2006; Proctor et al. 2009a, 2010). Previous cross-sectional and longitudinal studies have shown that high LS relates to a range of positive personal, behavioural, psychological, and social outcomes, just as low LS is associated with increased stress, psychological disorders, and behavioural problems (Gilman and Huebner 2006; Proctor et al. 2009a, 2010). Research findings from adult populations have shown a strong inverse association between LS and depression (Koivumaa-Honkanen et al. 2004; Pavot and Diener 2008; Schimmack et al. 2004). The same findings have been substantiated in research with children and adolescents (Gilman and Huebner 2006; Park 2004; Proctor et al. 2009a, 2010). School is a crucial context for adolescents’ well-being (Garcia-Moya et al. 2013). Accordingly, previous studies suggest that LS is linked with various indicators of positive functioning in school (Danielsen et al. 2009; Hueber and Gilman 2006; Suldo et al. 2006). High LS is related to cognitive engagement, academic achievement, and support from classmates and teachers, as well as perception of school connectedness (Lewis et al. 2011; Danielsen et al. 2009; Oberle et al. 2011; Proctor et al. 2010; Suldo et al. 2006). The study by Lo¨hre et al. (2010) showed that perception of well-being in school was positively associated with enjoying school work and receiving necessary help from the teacher. Moreover, increased feelings of school alienation and negative attitudes toward teachers and the school have been found to be related to lower LS (Gilman and Huebner 2006; Natvig et al. 2003; Salmela-Aro and Tuominen-Soini 2010). Although few studies have used the term ‘‘stressors’’ in association with LS, young people’s negative evaluations of academic and interpersonal variables in school could be perceived as potential stressors in adolescents’ lives (Byrne et al. 2007). The LS literature provides clear evidence to suggest that LS is more than just an outcome of various psychological states (e.g., positive affect), it is also an indicator of psychological states (e.g., depression) (Gilman and Huebner 2006; Park 2004; Proctor et al. 2009a, 2010). The mediating role of LS in relation to adolescents’ experiences and their adaptive and maladaptive outcomes has been examined (Park 2004; Proctor et al. 2009a), providing support for conceptualizations of LS as more than just an epiphenomenon of deeper underlying psychological constructs. LS is found to mediate the relation between parental social support and both internalizing (withdrawal and anxiety/depression) and externalizing (delinquent and aggressive behavior) problems in adolescents (Suldo and Huebner 2004b). A study of McKnight et al. (2002) found that LS mediated on the relation between major stressful life events and internalizing symptoms (depression and anxiety) and externalizing symptoms (aggressive and delinquent behavior). Based on recent reviews and empirical findings (Gilman and Huebner 2006; McKnight et al. 2002; Park 2004; Proctor et al. 2009a, 2010; Suldo and Huebner 2004b; Suldo et al. 2006, 2009), no studies investigating LS as a potential mediator between school-related stress and depressive symptoms were found, indicating that further research in this area is needed. An increased

123


understanding of how stressors in the school context are related to adolescents’ LS, as well as to depressive symptoms, would underscore the importance of academic and interpersonal school factors for adolescents‘ emotional health and overall perception of life as a whole. Further, investigating LS as a potential mediator would enhance the understanding of positive factors that indirectly explain the stress-health relationship. By using structural equation modelling (SEM), this study investigated the interrelationships between two possible school stressors (school performance and teacher interaction), LS and depressive symptoms, as well as the potential mediating role of LS between school stress and depressive symptoms. Previous empirical findings support the positive association between academic and interpersonal school stressors and depressive symptoms (Byrne et al. 2007; Fro¨jd et al. 2008; Moksnes et al. 2010b). Therefore, we hypothesized that positive associations would be found between school stressors and depressive symptoms. Furthermore, based on empirical findings showing that LS is related to positive school functioning, including teacher interaction and academic achievement (Lewis et al. 2011; Danielsen et al. 2009; Oberle et al. 2011; Proctor et al. 2010; Suldo et al. 2006), we expected negative associations would be found between LS and school stressors. As the previous literature have supported a negative association between LS and depressive symptoms (Gilman and Huebner 2006; Proctor et al. 2009a, 2010), we expected the same association for the present study. Given that previous support for LS as a relevant mediator in relation to stressful events and internalizing problems has been found (McKnight et al. 2002), we expected that LS would mediate the relation between school stress and depressive symptoms. This is in line with the view of LS as not only an outcome of various psychological states but also an indicator of psychological states. Hence, as portrayed in Fig. 1, the following hypotheses were proposed: H1 and H2: Stress of school performance and stress of teacher interaction is positively associated with depressive symptoms. H3 and H4: Stress of school performance and stress of teacher interaction is inversely associated with LS. H5:

LS is inversely associated with depressive symptoms.

H6 and H7: LS mediates the association between each of the school-related stressors and depressive symptoms.

Fig. 1 Hypotheses to be tested in the present study

123


2 Methods 2.1 Participants Every fifth year since 1996 a school based survey has been conducted by the Norwegian University of Science and Technology in Trondheim based on a convenience sampling of adolescents living in rural areas in the county of Sør-Trøndelag, mid-Norway. This crosssectional study uses data collected in 2011 by convenience sampling of schools from inland to coastal areas in five of the county’s 25 municipalities. A total of 1924 students from 12 public lower and upper secondary schools were asked to participate in the study and N = 1,289 completed the questionnaire (response rate 67 %). Nonresponses were mainly due to students being absent from schools when the questionnaire was administered, or students who declined to answer the questionnaire. No detailed information is available on students who did not fill in the questionnaire. Students younger than 13 or older than 18 (n = 50) were excluded, leaving n = 1,239 (64 %) to participate in the study with an age range 13–18 years. In the sample, 634 (51.2 %) were girls and 603 (48.7 %) were boys (gender was not identified for two participants). The number and percentage of participants in each age group were as following: 13 years: n = 293 (23.6 %); 14 years: n = 247 (19.9 %); 15 years: n = 250 (20.2 %); 16 years: n = 180 (14.5 %); 17 years: n = 149 (12 %); 18 years: n = 120 (9.7 %). The mean age for the total sample was 15.00 (SD 1.62); for boys 14.99 (SD 1.63) and for girls 15.02 (SD 1.63). 2.2 Procedure Data collection was approved by the Regional Committee for Medical Research Ethics and the Norwegian Social Science Data Services. The headmaster at each school approved the content of the questionnaire prior to agreeing to participate in the survey. All students and parents to students younger than 16 years received an information letter that briefly explained the purpose of the study. It was emphasized that participation was voluntary and that collected information was confidential. According to research ethical guidelines, written consent was claimed from all participants and also from parents when students were younger than 16. Questionnaire administration was completed in whole class groups during one regular school period of 45 min. The data were collected during October and November 2011. This time of year is appropriate for collecting data because it is a lesshectic period for the students in the academic year. 2.3 Measures 2.3.1 School Stress School stress was assessed using two sub dimensions included in the Norwegian version of the Adolescent Stress Questionnaire (ASQ-N), focusing on stressors related to school performance (five items) and teacher interaction (six items). The original Australian version of the Adolescent Stress Questionnaire (ASQ) is a 56-item inventory made up of items originally designed to measure common stressors that adolescents may experience in their daily life (Byrne et al. 2007). It also allows adolescents to report the extent to which any recent stressor experience has constituted a psychological challenge for them. All items are rated on a five-point Likert scale, ranging from (1) not at all stressful or is irrelevant to me

123


to (5) very stressful, where a higher score indicates higher stress level. The ASQ has been continually developed and validated since the mid-1990s (Byrne et al. 2007), and the Norwegian version of the instrument (ASQ-N) has been tested for use in a Norwegian adolescent sample (Moksnes et al. 2010a). Further validations of the ASQ-N have reduced the scale to 30 items and the instrument has been appropriately tested with reference to internal consistency and construct validity. The 30-item instrument reflects seven stress dimensions: Teacher interaction, peer pressure, home life, romantic relationships, school attendance, school/leisure conflict, and school performance (Moksnes and Espnes 2011). The inventory was first translated from English to Norwegian by three native bilingual Norwegian translators who completed this procedure independently, and then it was backtranslated from Norwegian to English by two other translators who had not seen the original version. In all steps of the translation, the different versions were compared to ensure that the translations were as precise and complete as possible with reference to semantic and conceptual equivalence (Moksnes et al. 2010a). In the present study, the focus was on school-related stress and therefore two of the original sub scales in the instrument were included reflecting stress of teacher interaction and stress of school performance. However, the sub scales were modified, and some items were excluded after the measurement models had been tested in CFA (see Table 1; Sect. 2.4). The items included in the modified scale to assess stress of school performance were as follows: Having to study things you don’t understand; Having difficulties with some subjects; and Teachers expecting too much from you. The items included in the scale to assess stress of teacher interaction were as follows: Teachers hassling you; Not being listened to by teachers; Lack of respect from teachers; and Not getting along with your teachers. In the present study, the internal consistency of the scales for stress of school performance and stress of teacher interaction was a = .83 and a = .86 respectively (see Table 2). 2.3.2 Life Satisfaction Life satisfaction (LS) was assessed using the Satisfaction with Life Scale (SWLS) (Diener et al. 1985). The scale consists of five items, rated on a seven-point Likert scale, ranging from (1) strongly disagree to (7) strongly agree, where a higher score indicates higher LS. The scale has been extensively used in adult samples (Pavot and Diener 2008) as well as among adolescents (Proctor et al. 2009b). The internal consistency of the SWLS has been found to be high, generally exceeding Cronbach’s alpha values of .80 (Pavot and Diener 2008; Proctor et al. 2009b). The Norwegian version of the scale is from Ed Diener’s official webpage (http://internal.psychology.illinois.edu/*ediener/SWLS.html). Norwegian validations of the scale have supported a single-factor structure, and the scale has also been observed to be appropriate for use across a broad age range, including adolescence (Clench-Aas et al. 2011; Moksnes et al. 2013). For the present study, the original scale was tested in CFA and one item was excluded (see Table 1; Sect. 2.4). Items included in the scale for the present study were: In most ways my life is close to my ideal; The conditions of my life are excellent; I am satisfied with my life and If I could live my life over, I would change almost nothing. The internal consistency of the SWLS in the present study was a = .83 (Table 2). 2.3.3 Depressive Symptoms A nonclinical depression scale was used to assess depressive symptoms. While specific instruments are available to assess adolescent depression (the Reynolds Adolescent

123


Depression Scale (Reynolds and Mazza 1998) is probably the archetype), these are typically far lengthier than is convenient for a screening assessment of depression in a large sample, and oriented too far toward the diagnosis of clinical depression to be useful as a simple index of (state) nonclinical depressive mood. Clinically focussed instruments also have the potential to alienate otherwise healthy participants in an essentially normal sample. A depression scale was therefore constructed specifically for the validation of the ASQ by Byrne et al. (2007). It consists of a short, 15-item questionnaire measuring respondents’ current level of depressive mood. Item choice was informed by reference to commonly experienced depressive features outlined in the Diagnostic and Statistical Manual: Fourth Edition TR (American Psychiatric Association 2000). Reference was also made to the Zung Self-Rating Depression Scale (Zung 1965). The items constructed describe a number of commonly experienced but essentially nonclinical depressive attributes, and respondents were asked to indicate the extent to which they had experienced these symptoms in the past week using a five-point Likert scale ranging from (1) never to (5) always, where a higher score indicates more symptoms of depression. The scale is found to correlate positively and significantly with anxiety (r = 0.67) and negatively and significantly with self-esteem (r = -0.63) (Byrne et al. 2007). The scale was translated from English to Norwegian and then back-translated by two native bilingual Norwegian translators who completed this procedure independently. The scale was modified after the measurement model had been tested in CFA, retaining six items: (1) I feel like crying; (2) I have lost interest in things that I use to; (3) I have lost confidence in myself or put myself down; (4) I have had difficulty making decisions; (5) I have felt like I have failed; and (6) I have felt like things always go wrong, no matter how hard I try (see Table 1; Sect. 2.4). The internal consistency of the scale in the present study was a = .89 (Table 2). 2.4 Statistics Data were screened and analysed using the SPSS, version 20.0 and LISREL 8.8. Descriptive statistics of means and standard deviations were calculated for all scales and Pearson product-moment correlation was used to test bivariate associations between the scales in the study. Internal consistency for each scale was examined with Cronbach’s alpha (a) and further investigated in CFA by means of composite reliability; values C0.60 are acceptable whereas values C0.70 are considered to be good (Hair et al. 2010). Structural equation modelling (SEM) was used to test the hypothesized relations between the two school stressors, LS and depressive symptoms. In SEM, there is no universal agreement about the sample size and no easy way to determine the sample size needed for confirmatory factor analysis (CFA). When using SEM, random measurement error is accounted for and psychometric properties of the scales in the model are more accurately derived. At the same time, all the direct, indirect and total effects throughout the model are estimated. SEM models combine measurement models (factor models) with structural models (regression), where a major issue is evaluation of model fit. The statistical significance of each estimated parameter is assessed by its t value (Sharma 1996). In accordance with criteria provided by Tabachnik and Fidell (2007), factor loadings below .32 were considered to be poor, .45 moderate, .55 good, .63 very good and .71 excellent. The conventional overall test of model-fit was evaluated by Chi-square (v2); a small and preferably a non-significant Chi-square correspond to good fit. Although SEM is a large sample technique ideal for the data in the present study (Tabachnik and Fidell 2007), Chisquare statistics are sensitive to large sample sizes, and it is therefore recommended to use this measure along with other measures of overall model fit. As normality of data is a

123

U. K. Moksnes et al. Table 1 Items included in the original scales Stress of school performance Item 1a Having to study things you don’t understand Item 2a Teachers expecting too much from you Item 3a Having difficulties with some subjects Item 4 Having to study things you are not interested in Item 5 Difficulty with some subjects Stress of teacher interaction Item 6a Teachers hassling you Item 7a Not being listened to by teachers Item 8a Lack of respect from teachers Item 9a Not getting along with your teachers Item 10 Lack of trust from adults Item 11 Parents hassling you about the way you look Depression Item 1: I have felt sad or unhappy Item 2a I feel like crying Item 3: I feel guilty without knowing why Item 4a I have lost interest in things that I use to Item 5: I have not enjoyed activities that I used to Item 6: I have felt uneasy, restless, or irritable Item 7a I have lost confidence in myself or put myself down Item 8: I have had difficulty concentrating Item 9a I have had difficulty making decisions Item 10a I have felt like I have failed Item 11a I have felt like things always go wrong, no matter how hard I try Item 12: My sleep has been disturbed—sleeping more or less, or broken sleep Item 13: My appetite has been disturbed—eating more or less Item 14: I have felt like it takes me greater effort to do things Item 15: I have felt tired or have had very little energy Satisfaction with life Item 1a In most ways my life is close to my ideal Item 2a The conditions of my life are excellent Item 3a I am satisfied with my life Item 4: So far I have gotten the important things I want in life Item 5a If I could live my life over, I would change almost nothing a

Items included in the scales in the study

premise in SEM, the non-normality (skewness and kurtosis of the data were significant) was corrected by applying the robust maximum likelihood (RML) in the estimate procedure, and report the Satorra–Bentler scaled Chi-square (v2). In addition, the following fit indices were used; the RMSEA and the SRMR with acceptable/good fit set to .08/.05 (Brown 2006; Byrne 2001), the CFI and the NNFI with acceptable/good fit at .97/.95 respectively, as well as the GFI and the NFI with acceptable fit/good fit indicated by .90/.95 (Brown 2006; Byrne 2001). For the AGFI acceptable fit was set to .85 and good fit at .90.

123


Before examining the hypothesized relationships in the SEM analysis, the measurement models (scales) were tested by CFA as this is an important prerequisite for SEM in order to determine theoretical and empirical valid constructs and to derive a model that is reliable and meaningful to the data (Brown 2006). The measurement models were established based on theoretical considerations and validity reliability concerns. The SEM-literature supposes that 3–5 indicators per latent variable are sufficient (Kline 2011; Marsh et al. 1998). However, when reducing the number of indicator variables it is important to retain the theoretical substance and nuances of each latent variable (Brown 2006; Hair et al. 2010; Kline 2011). In the present study the statistical evaluation of the measurement models and exclusion of items were based on high Chi-square values, bad fit indices, low factor loadings, high residual variances and thus high cross-loadings of items. The CFA provided a good fit to the observed data for the two-factor model of stress of school performance (three items) and stress of teacher interaction (four items) (v2 = 71.83, df = 13, p \ .01, RMSEA = .063, p \ .063, SRMR = .039, NFI = .99, NNFI = .98, CFI = .99, GFI = .97, AGFI = .94). CFA found good model fit for the measurement model of depression comprising six items (v2 = 14.56, df = 9, p \ .10, v2/df = 1.62, RMSEA = .024, p \ .98, SRMR = .016, NFI = 1.00, NNFI = 1.00, CFI = 1.00, GFI = .99, AGFI = .98) and the LS measurement model comprising four items (v2 = 8.56, df = 2, p \ .014, v2/df = 4.28, RMSEA = .056, p \ .32, SRMR = .014, NFI = 1.00, NNFI = .99, CFI = 1.00, GFI = .99, AGFI = .97) (see items presented in Table 1). After modification, all parameter estimates in the measurement models were significant (p \ .05) and loaded positively and clearly on their intended latent variable with standardized factor loadings between .59 and .88.

3 Results 3.1 Descriptive Analysis Table 2 display the means (M), standard deviations (SD), Cronbach’s a and Pearson’s correlation matrix for the variables of stress of school performance, stress of teacher interaction, depressive symptoms, and LS. The correlations between all the variables were moderate to strong, significant and in the expected direction; stress of school performance and teacher interaction were positively correlated with depressive symptoms and inversely correlated with LS. An inverse correlation was found between LS and depressive symptoms. The Cronbach’s a for the various variables indicated a good inter-item consistency in the measures with coefficients varying between .83 and .89 (Table 2). Table 2 Means (M), standard deviations (SD), Cronbach’s alpha, and correlation coefficients for the study variables Construct

M

SD

Range

Cronbach’s a (no. of items)

Correlation coefficients Pearson’s r2 SP

School performance (SP)

7.79

3.27

3–15

.83 (3)

Teacher interaction (TI)

7.61

4.12

4–20

.86 (4)

Depression

12.15

5.40

5–30

.89 (6)

Life satisfaction (LS)

18.02

5.08

4–28

.83 (4)

–

TI .55** –

Depression

LS

.46**

-.28**

.36**

-.16**

–

-.51** –

** p \ .01

123


3.2 SEM Analyses of the Association Between School Stress, LS and Depressive Symptoms Model 1 comprising stress of school performance, stress of teacher interaction and depressive symptoms was evaluated initially in order to test the two first hypotheses (H1 and H2) (Fig. 1). Model 2 including stress of school performance, stress of teacher interaction, LS and depressive symptoms was evaluated in order to test hypotheses H3–H7 (Fig. 1). Model 1 (see Fig. 2; Table 3) offered a good fit to the observed data (v2 = 149.99, p \ .001, df = 62, RMSEA = .036, p value = 1.00, SRMR = .036, NFI = .99, NNFI = .99, CFI = .99, GFI = .97 and AGFI = .96). The model displayed a strong positive association between stress of school performance and depressive symptoms controlled for stress of teacher interaction, supporting the first hypothesis (H1: c1,1 = .49*). The relation between stress of teacher interaction and depressive symptoms (H2: c1,2 = .07), was non-significant, controlled for stress of school performance. Model 2 investigating the association between the two school stressors, LS and depressive symptoms (Fig. 3; Table 3) revealed a good model fit to the present data (v2 = 210.22, p \ .01, df = 113, v2/df = 1.86, RMSEA = .030, p value = 1.00, SRMR = .035, NFI = .99, NNFI = .99, CFI = .99, GFI = .97 and AGFI = .96). All factor loadings were strong, ranging between .59 and .88, with high explained variance indicated by R2-values ranging from .35 to .78. Accordingly, the composite reliability was excellent, showing values of .75 or higher (Table 4). The association between stress of school performance and depressive symptoms was strong, positive and significant, but considerably reduced when controlling for LS and stress of teacher interaction in the model (H1: c1,1 from 0.49* in Model 1 to 0.34* in Model 2, see Table 3). Stress of teacher interaction did not display any significant relation with depressive symptoms controlled for LS and stress of school performance (H2: c2,2 = .07). A significant inverse and moderate strong relation was found between stress of school performance and LS, supporting the

Fig. 2 Model 1 evaluating the relationship between school stress and depression. Note *p \ .05. Completely standardized factor loadings and gammas. v2 Chi square, df degrees of freedom, RMSEA root mean square error of approximation, SRMR standardized root mean residual, NFI Normed Fit Index, NNFI Non-normed Fit Index, CFI Comparative Fit Index, GFI Goodness of Fit Index, AGFI Adjusted Goodness of Fit Index

123

Life Satisfaction as a Potential Mediator Table 3 Direct relationships between school stress, depression, and life satisfaction for Model 1 and Model 2 Construct

Parameter

Model 1

Model 2

School performance to depression (H1)

c1,1

0.49*

0.34*

t value

5.11

3.11

c1,2

0.07

0.07

Teacher interaction to depression (H2) School performance to life satisfaction (H3)

t value

0.93

0.27

c1,3

–

-0.29*

t value Teacher interaction to life satisfaction (H4)

c2,1

3.35 –

-0.04

–

-0.45*

t value Life satisfaction to depression (H6)

b1,2 t value

-0.27 -4.45

Listwise deletion n = 973 * p \ .05. Completely standardized Gamma1 and Beta2. 1Gamma (c); standardized regression coefficients representing directional relationships between school stress (school performance and teacher interaction), depression and life satisfaction. 2Beta (b); standardized regression coefficients representing direct relationships between the dependent variables of life satisfaction and depression. Model 1: relationships between school stress (school performance and teacher interaction), life satisfaction and depression. Model 2: equal to Model 1 but excluding the dependent variable life satisfaction

Fig. 3 Model 2 evaluating the association between school stress, LS and depression. Note *p \ .05. Completely standardized factor loadings, gammas and beta. v2 Chi square, df degrees of freedom, RMSEA root mean square error of approximation, SRMR standardized root mean residual, NFI Normed Fit Index, NNFI Non-normed Fit Index, CFI Comparative Fit Index, GFI Goodness of Fit Index, AGFI Adjusted Goodness of Fit Index

third hypothesis (H3: c1,2 = -.29*), however, the relation between stress of teacher interaction and LS was not significant, contradicting hypothesis four (H4: c2,2 = -.04). The fifth hypothesis was supported (see Fig. 3; Table 3), showing a significant inverse association between LS and depressive symptoms (H5: b1,2 = -.45*). The results

123

U. K. Moksnes et al. Table 4 Measurement models for school stress, depression and life satisfaction (Model 2) Items

Parameter

Lisrel estimate1

t value2

R2

Stress of school performance Item 1

kx1,1

0.79

29.69*

0.63

Item 2

kx2,1

0.83

36.01*

0.68

Item 3

kx3,1

0.78

30.96*

0.61

Stress of teacher interaction Item 6

kx4,2

0.62

11.23*

0.38

Item 7

kx5,2

0.84

33.33*

0.68

Item 8

kx6,2

0.86

31.81*

0.74

Item 9

kx7,2

0.84

27.71*

0.70

Life satisfaction Item 1

ky1,1

0.71

Item 2

ky2,1

0.82

21.39*

0.67

Item 3

ky3,1

0.88

22.13*

0.78

Item 5

ky4,1

0.64

17.56*

0.41

–

0.51

Depression Item 2

ky5,2

0.69

Item 4

ky6,2

0.59

17.38*

0.35

Item 7

ky7,2

0.82

21.18*

0.68

–

0.47

Item 9

ky8,2

0.76

20.36*

0.58

Item 10

ky9,2

0.88

23.50*

0.78

23.08*

0.72

Item 11

ky10,2

0.85

qc School performance

q3c

0.84

qc Teacher interaction

qc

0.88

qc Depression

qc

0.93

qc Life satisfaction

qc

0.79

Listwise deletion, n = 973. * p \ .05 1

Completely standardized factor loadings (c),

2

Squared multiple correlations (R2), P 2 ð kÞ Composite reliability qC ¼ P 2 P ð kÞ þ ðhÞ

3

presented in Table 3, show that the association between stress of school performance and depressive symptoms (H6) was partly mediated by LS, indicated by a reduction in the b coefficient from 0.49* to 0.34*, supporting the sixth hypothesis (H6). However, in contradiction to the seventh hypothesis, LS did not act as a mediator between stress of teacher interaction and depressive symptoms (H7). Looking at the total and indirect effects between the latent variables in Model 2 (Table 5), a significant indirect effect of school performance stress (.13*) in relation to depression was found, indicating that this association is affected by a significant intervening variable, in this case LS. The total effect of stress of school performance in relation to depression was .49* in Model 1 (Fig. 2). In model 2, significant total effects appeared of school performance both in relation to depression (.47*) and LS (-.29*) (Table 5), indicating that school performance stress explains variance both in depression and LS. Interestingly, the total effect of school performance stress on depression reduced from .49

123

Life Satisfaction as a Potential Mediator Table 5 Total and indirect effects on the study variables for Model 2 School performance

Teacher interaction

Total effects of school stress on depression and life satisfactiona Depression

0.47* 9.03

-0.04 -0.62

Life satisfaction

-0.29* -4.91

0.08 1.61

Indirect effects of school stress on depression and life satisfactionb Depression

0.13* 4.54

0.02 0.62

Life satisfaction

–

–

* p \ .05 a Total Effects represent the total influence of the explanatory variable school stress (direct ? indirect effects) b

Indirect effects represent the influence of School stress mediated by intervening variables (mediators). Standardized Lisrel estimates and t values

in Model 1 to .47 in Model 2 after LS was included. The small reduction in total effect indicate that stress of school performance explains most of the variance in depressive symptoms controlled for LS, and that the additive value of LS in explaining the school performance stress to depression association seems to be weak although a significant indirect effect was found.

4 Discussion The present study was designed to extend our knowledge of the association between school-related stressors, LS and depressive symptoms in adolescents, and the role of LS as a potential mediator between school stress and depressive symptoms. Seven hypotheses were formulated and tested, and four of them were supported by the present data. In line with our first hypothesis (H1), the results showed a positive and significant strong relation between stress of school performance and depressive symptoms. The second hypothesis (H2) referring to a positive association between stress of teacher interaction and depressive symptoms was supported in the bivariate results from the correlation analysis, but rejected in the multivariate results, once controlling for LS and stress of school performance. While exposure to some stressful negative events is considered a normal part of life during the adolescent years, the overall experience of cumulative and simultaneous stressors is strongly related to depression during adolescence (Byrne et al. 2007; Charbonneau et al. 2009; Moksnes et al. 2010b; Shih et al. 2006). More specifically, the association found between stress of school performance and depressive symptoms in both models tested in the present study is closely in line with previous findings (Byrne et al. 2007; Fro¨jd et al. 2008; Undheim and Sund 2005; Moksnes et al. 2010b). As adolescents make the transition to a higher school level, they might also perceive the academic demands as more competitively stressful and put greater emphasis on academic achievement (Byrne et al. 2007; Hjern et al. 2008). Failure in academic performance might be perceived to be stressful because it represents a threat to the adolescents’ goal of learning and to succeed in performance. High expectations from adolescents themselves and others

123


(e.g., parents, teachers, or peers) might, in the long run, give youngsters a chronic feeling of inadequacy in the face of their academic performance and contribute to the development of depressive symptoms. However, due to the cross-sectional design of the present study, the possibility of bi-directionality of the present association cannot be discounted. Experience of psychological problems can influence both reporting of stressor experience and assessment of stressor impact, in that adolescents who score higher on emotional problems may perceive the situation as more stressful. Indeed, it is likely that both these types of factors contribute to explanations of the relationship between perceived academic stress and depressive symptoms in the present study. The non-significant association found between stress of teacher interaction and depressive symptoms in the multivariate results in Model two shows that adolescents’ perception of interaction with teachers may not be evaluated as intensively stressful with significance for their emotional health controlled for school performance and LS. The items in this stress dimension reflect difficult interactions with teachers represented by perceived lack of respect, not being listened to as well as being hassled by the teacher. Thus, the present findings does not support previous studies showing that perceived stressful interaction with teachers is related to more emotional problems in adolescents, controlled for the influence of other stressors (Eccles and Roeser 2011; Hjern et al. 2008; Moksnes et al. 2014; Undheim and Sund 2005). Meanwhile, the results correspond with a study of Moksnes et al. (2014), showing that stress of teacher interaction did not predict depressive symptoms when controlling for other school-related and interpersonal stressors in adolescents’ lives (Moksnes et al. 2014). Explanations for the present findings are not straightforward and may be related to the complexity of internal and contextual factors in adolescents’ lives, such as social adjustment, pubertal changes, and interaction with peers and adults (Oberle et al. 2011; Undheim and Sund 2005). The individual’s evaluation of the importance of the potential stressor, as well as the resources available to cope, is fundamental for the impact of the stressor. These evaluations will affect the way the response takes place and this further influences the individual’s health and well-being (Charbonneau et al. 2009; Shih et al. 2006). It should be noted that modifications of the stress sub scales may have accounted for the non-significant finding in the present study. Further, the overlap in conceptual and item wording as well as shared unexplained variance between the stressors could partly explain the small unique variance of stress of teacher interaction in the multivariate analyses. The bivariate correlation between the two stressors is likely to represent a series of reciprocal relations where students experiencing stressful interaction with teachers also have academic problems and vice versa (Byrne et al. 2007; Moksnes and Espnes 2011). In line with the third hypothesis (H3), the findings from the multivariate analyses in Model two showed that higher level of stress of school performance was associated with lower LS, represented by the second strongest b-value in the SEM model. Stress of teacher interaction was inversely related to LS in the bivariate results, however, the unique effect of relational stress with teachers was not significant when controlling for school performance and depressive symptoms, contradicting the fourth hypothesis (H4). Support for the role of LS in relation to stress of school performance is sustained by related research findings showing that high LS is positively associated with cognitive engagement, academic achievement, and perceived academic competence (Lewis et al. 2011; Danielsen et al. 2009; Proctor et al. 2010; Suldo et al. 2006). The significant bivariate association between stress of teacher interaction and LS is in line with previous studies showing that high LS is related to positive interaction with teachers, as well as perception of support and help from teachers (Danielsen et al. 2009; Eccles and Roser 2011; Suldo et al. 2006).

123


However, the non-significant association found in the multivariate results suggests that compared to perceptions of academic stress, adolescents’ perception of interaction with teachers may not be evaluated as intensively stressful with significance for their LS in the present study population. As mentioned previously, this could party be explained by overlap in conceptual meaning and wording of the items, as well as shared unexplained variance between the two school stressors. In line with the fifth hypothesis (H5), the results showed an inverse strong association between LS and depressive symptoms in Model two, represented by the strongest b-value in the SEM model. The results are highly supported by previous findings (Gilman and Huebner 2006; Proctor et al. 2009a, 2010), which show that LS is an important indicator in relation to adolescents’ emotional health. The results also support reviews that have highlighted the importance of LS for adolescents’ positive adjustment, health, and wellbeing (Proctor et al. 2009a, 2010; Suldo and Huebner 2004a). The findings of Proctor et al. (2010) and Gilman and Huebner (2006) showed that in comparison to students with average LS, those with high LS also scored higher on all indicators of adaptive, positive psychosocial functioning and reported lower levels of emotional symptoms, including depression. The results from the present study gave support (although weak) to the sixth hypothesis (H6) in showing that LS partly mediated the relation between stress of school performance and depressive symptoms. The small reduction in total effect of stress of school performance to depression after including LS indicate that LS may not have a strong additive value as a mediator on the particular relationship. The last hypothesis was rejected by showing that LS did not mediate the association between stress of teacher interaction and depressive symptoms (H7). The non-significant result of LS as a mediator between stress of teacher interaction and depressive symptoms may not be surprising since stress of teacher interaction was not significantly related to either LS or depressive symptoms. Meanwhile, the significant indirect effect found indicates that stress of school performance is indirectly related to depressive symptoms through adolescents’ experience of LS. Previous findings have showed inconsistent results of the mediating role of LS during adolescence (McKnight et al. 2002; Park 2004; Proctor et al. 2009a; Suldo and Huebner 2004b). The present findings are supported by a related study showing that LS mediated the association between stressful life events and internalizing symptoms (McKnight et al. 2002). Moreover, it is postulated that individuals with high LS will have a general confidence that resources are available to meet the demands posed by stressful situations. Consequently, they will potentially consider a stressor more as a challenge than as a threat. Further, individuals perceiving high LS effectively promote good and effective coping mechanisms by focusing on finding solutions, which increases the possibility of resolving tension associated with stress (Park 2004; Proctor et al. 2009a, 2010). The study findings contribute to the literature by providing insight into the importance of successfully managing stressors related to academic performance based on the significant associations found between school performance stress, LS and depressive symptoms, controlled for stress of teacher interaction. The results also extend previous findings by investigating the unique role of stress of teacher interaction in relation to depressive symptoms and LS after controlling for stressors pertinent to academic performance. This study provides additional evidence that adolescent LS is relevant not only in and of itself but also in relation to stress of school performance and emotional health (Danielsen et al. 2009; Proctor et al. 2010; Suldo et al. 2006, 2009). Knowing that stress of school performance significantly and indirectly relates to depressive symptoms through adolescents’ perception of LS may indicate that LS is a relevant factor to promote in reference to coping

123


with academic challenges. However, on the basis of the weak additive value of LS, more research is needed regarding the role of LS as a potential mediator and preferably by longitudinal studies. The range of LS correlates is so broad (e.g., self-efficacy, self-esteem, and social support) (Moksnes and Espnes 2013; Proctor et al. 2009a) that the question arises as to which are the primary correlates and which are the secondary; the need for investigation of the essential predictors of LS is needed. The main goal of health-promotion strategies among adolescents is generally to build cognitive, psychological, and social assets that prepare adolescents to navigate life’s pathways and to handle the challenges experienced during adolescence (Proctor et al. 2009a). On the basis of the present findings, an important focus in reference to the school context is to work on how to promote coping among adolescents through a supportive and positive school climate and to provide motivating and good learning conditions for students in school (Eccles and Roeser 2011). 4.1 Strengths and Limitations A major strength of the present study is the large sample size and high response rate. The study builds on data analysed by advanced statistical methods of SEM to investigate the hypotheses proposed for the present study. However, the results should be considered with some limitations in mind. The present results have tested model data consistency by comparing measurement models and their assumptions to the present data. Although the study used validated instruments for use on the adolescent group, the scales were modified in the present study, something that reduces the ability to link this study with prior relevant research using the same measures. However, evaluation of the measurement models is an important prerequisite for determining theoretically and empirically valid constructs and to derive a model that is meaningful for the sample data. A well-fitting model will ultimately increase the ability of making valid study conclusions on the basis of the hypothesis made. It should also be noted that the theoretical substance and nuances of the latent variables were retained although, and the SEM-literature states that three to five indicators for each latent variable are considered to be sufficient (Brown 2006; Hair et al. 2010). However, on the basis of the modified scales (especially the depression scale), the psychometric properties should be further evaluated in order to assess the applicability of the instruments on Norwegian adolescents. More information is needed about the nature and directionality of the relationship between LS, school-related stress, and depressive symptoms beyond the cross-sectional design used and the preliminary nature of results in the present study. It is therefore likely to assume that the relationships between stress of school performance, stress of teacher interaction, LS, and depressive symptoms are reciprocal, and in a cross-sectional study, this markedly reduces the capacity to comment on possible directions of causality. The related issues of a cross-sectional design and of indicators of causality should therefore be contemplated in future research. Another limitation of the study is the exclusive reliance on self-reports from adolescents of 13–18 years, which may lead to potential self-reporting bias (Rothman 2002). First, selfreports require that adolescents are at a level of cognitive maturity where they are able to reflect and understand concepts of health and illness. Second, there is a challenge regarding the adolescents’ ability to evaluate and report reliably on feelings and complaints through self-report (e.g., social desirability). This is especially so in the youngest adolescents, as they might have difficulty comprehending the abstract concepts and therefore are subject to over- or under reporting. The measurement of depressive symptoms was conducted using

123


an instrument assessing non-clinical depressive attributes within the last week. This instrument was regarded appropriate in reference to the study population of general adolescents with no known indications of mental health problems. Therefore the results do not allow for firm conclusions with regard to clinical emotional affect which could be a potential bias. Nevertheless, we believe that the results are relevant for levels of emotional affect that represent a significant impairment in the individual’s psychological well-being. A study by Haugland and Wold (2001) concluded that adolescents between 14 and 16 years of age are able to evaluate and give reliable information about their subjective health using questionnaires, and this methodology is well used in large sample studies. Meanwhile, it should be acknowledged that all such data might be prone to bias due to the possible influence of social desirability factors (Derdikman-Eiron et al. 2011).

5 Conclusion The present study found support for stress of school performance as significantly and positively related to depressive symptoms and significantly and inversely related to LS. At the bivariate levels, stress of teacher interaction was associated with more depressive symptoms and reduced LS. However, in the multivariate analyses stress of teacher interaction did not have significant unique associations with these outcomes after the influence of stress of school performance was taken into account. A significant inverse association was found between LS and depressive symptoms. Further, LS partly mediated the association between stress of school performance and depressive symptoms, whereas the mediating role of LS in relation to stress of teacher interaction and depressive symptoms was not significant. The study provides evidence of a significant linkage between LS, school performance stress and depressive symptoms, and LS as a mediator of this particular relationship, although the additive value of LS was weak. Further research is needed based on lack of studies in this area and the preliminary nature of the present findings. Longitudinal research investigating the reciprocal and dynamic relations between school stress, LS, and emotional outcomes is needed to investigate causality and the generalizability of the results. More studies investigating potential intervening variables (including LS) on the relation between school stressors and depressive symptoms is also needed. Further studies may lead to a more comprehensive understanding of how academic and interpersonal school stressors affect adolescents’ emotional health and functioning which can be used to develop appropriate interventions during adolescence.

References American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders DSM-IVTR (4th ed.). Washington, DC: American Psychiatric Association. Brown, T. (2006). Confirmatory factor analysis for applied research. New York: The Guilford Press. Byrne, M. B. (2001). Structural equation modeling with AMOS. Basic concepts, applications, and programming. USA: Taylor AND Francis Group. Byrne, D. G., Davenport, S. C., & Mazanov, J. (2007). Profiles of adolescent stress: The development of the adolescent stress questionnaire (ASQ). Journal of Adolescence, 30, 393–416. Charbonneau, A. M., Mezulis, A. H., & Hyde, J. S. (2009). Stress and emotional reactivity as explanations for gender differences in adolescents’ depressive symptoms. Journal of Youth and Adolescence, 38, 1050–1058.

123

U. K. Moksnes et al. Clench-Aas, J., Nes, R., Dalgard, O. S., & Aarø, L. E. (2011). Dimensionality and measurement invariance in the satisfaction with life scale in Norway. Quality of Life Research, 20, 1307–1317. Compas, B. E., & Reeslund, K. L. (2009). Processes of risk and resilience during adolescence. In R. M. Lerner & L. Steinberg (Eds.), Handbook of adolescent psychology (3rd ed., pp. 561–588). New Jersey: Wiley. Danielsen, A. G., Samdal, O., Hetland, J., & Wold, B. (2009). School-related social support and student’s perceived life satisfaction. The Journal of Education Research, 102, 303–318. Derdikman-Eiron, R., Indredavik, M. S., Bratberg, G. H., Taraldsen, G., Bakken, I. J., & Colton, M. (2011). Gender differences in subjective well-being, self-esteem and psychosocial functioning in adolescents symptoms of anxiety and depression: Findings from the Nord-Trøndelag health study. Scandinavian Journal of Psychology, 52, 261–267. Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality and Assessment, 49, 71–75. Eccles, J., & Roser, R. W. (2011). Schools as developmental contexts during adolescence. Journal of Research on Adolescence, 21, 225–241. Fro¨jd, S. A., Nissinen, E. S., Pelkonen, M. U. I., Marttunen, M. J., Koivisto, A. M., & Kaltiala-Heino, R. (2008). Depression and school performance in middle adolescent boys and girls. Journal of Adolescence, 31, 485–498. Garcia-Moya, I., Rivera, F., & Moreno, C. (2013). School context and health in adolescence: The role of sense of coherence. Scandinavian Journal of Psychology, 54, 243–249. Gilman, R., & Huebner, S. (2006). Characteristics of adolescents who report very high life satisfaction. Journal of Youth and Adolescence, 35, 311–319. Goldbeck, L., Schmitz, T. G., Besier, T., Herschbach, P., & Henrich, G. (2007). Life satisfaction decreases during adolescence. Quality of Life Research, 16, 696–979. Grant, K. E., Compas, B. E., Thurm, A. E., McMahon, S. D., & Gipson, P. Y. (2004). Stressors and child and adolescent psychopathology: Measurement issues and prospective effects. Journal of Clinical Child and Adolescent Psychology, 33, 412–425. Grant, K. E., Compas, B. E., Thurm, A. E., McMahon, S. D., Gipson, P. Y., Campbell, A. J., et al. (2006). Stressors and child and adolescent psychopathology: Evidence of moderating and mediating effects. Clinical Psychology Review, 26, 257–283. Hair, J., Black, W., Babin, B., & Anderson, R. (2010). Multivariate data analysis. Upper Saddle River: Prentice Hall. Haugland, S., & Wold, B. (2001). Subjective health complaints in adolescence—reliability and validity of survey methods. Journal of Adolescence, 24, 611–624. ¨ stberg, V. (2008). School stressors, psychological complaints and psychosomatic Hjern, A., Alfven, G., & O pain. Acta Paediatrica, 97, 112–117. Hueber, E. S., & Gilman, R. (2006). Students who like and dislike school. Applied Research in Quality of Life, 1, 139–150. Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). New York, London: The Guilford Press. Koivumaa-Honkanen, H., Kaprio, J., Honkanen, R., Viinama¨ki, H., & Koskenvuo, M. (2004). Life satisfaction and depression in a 15-year follow up of healthy adults. Social Psychiatry and Psychiatric Epidemiology, 39, 994–999. Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal and coping. New York: Springer Publication. Lewis, A. D., Huebner, E. S., Malone, P. S., & Valois, R. F. (2011). Life satisfaction and student engagement in adolescents. Journal of Youth and Adolescence, 40, 249–262. Lo¨hre, A., Lydersen, S., & Vatten, L. J. (2010). School wellbeing among children in grades 1–10. BMC Public Health, 10, 526. Lo¨hre, A., Moksnes, U. K., & Lillefjell, M. (2013). Gender differences in predictors of school wellbeing? Health Education Journal,. doi:10.1177/0017896912470822. Marsh, H. W., Kit-Tai, H., Balla, J. R., & Grayson, D. (1998). Is more ever too much? The number of indicators per factor in confirmatory factor analysis. Multivariate Behavioral Research, 33, 181. McKnight, C. G., Huebner, E. S., & Suldo, S. (2002). Relationships among stressful life events, temperament, problem behavior, and global life satisfaction in adolescents. Psychology in the Schools, 39, 677–687. Moksnes, U. K., Byrne, D. G., Mazanov, J., & Espnes, G. A. (2010a). Adolescent stress: Evaluation of the factor structure of the Adolescent Stress Questionnaire. Scandinavian Journal of Psychology, 51, 203–209. Moksnes, U. K., Moljord, I. E., Espnes, G. A., & Byrne, D. G. (2010b). The association between stress and emotional states in adolescents. The role of gender and self-esteem. Personality and Individual Differences, 49, 430–435.

123

Life Satisfaction as a Potential Mediator Moksnes, U. K., & Espnes, G. A. (2011). Evaluation of the Norwegian version of the Adolescent Stress Questionnaire (ASQ-N): Factorial validity across samples. Scandinavian Journal of Psychology, 52, 601–608. Moksnes, U. K., & Espnes, G. A. (2013). Self-esteem and life satisfaction in adolescents-gender and age as potential moderators. Quality of Life Research,. doi:10.1007/s11136-013-0427-4. Moksnes, U. K., Espnes, G. A., & Haugan, G. (2014). Stress, sense of coherence and emotional symptoms in adolescents. Psychology & Health, 29, 32–49. Moksnes, U. K., Lo¨hre, A., Byrne, D. G., & Haugan, G. (2013). Satisfaction with life in adolescents: Evaluation of the factor structure in a Norwegian sample. Social Indicators Research,. doi:10.1007/ s11205-013-0451-3. Natvig, G. K., Albrektsen, G., Anderssen, N., & Quarnstrøm, U. (2003). Associations between psychosocial factors and happiness among school adolescents. International Journal of Nursing Practive, 9, 166–175. Oberle, E., Schonert-Reichl, A., & Zumbo, B. D. (2011). Life satisfaction in early adolescence: Personal, neighbourhood, school, and peer influences. Journal of Youth and Adolescence, 40, 889–901. Park, N. (2004). The role of subjective well-being in positive youth development. The Annals of the Americal Academy, 591, 25–39. Pavot, W., & Diener, E. (1993). Review of the Satisfaction With Life Scale. Psychologial Assessment, 5, 164–172. Pavot, W., & Diener, E. (2008). The satisfaction with life scale and the emerging construct of life satisfaction. The Journal of Positive Psychology, 3, 137–152. Proctor, C. L., Linley, P. A., & Maltby, J. (2009a). Youth life satisfaction: A review of the literature. Journal of Happiness Studies, 10, 583–630. Proctor, C. L., Linley, P. A., & Maltby, J. (2009b). Youth life satisfaction measures: A review. The Journal of Positive Psychology, 4, 128–144. Proctor, C., Linley, P., & Maltby, J. (2010). Very happy youths: Benefits of very high life satisfaction among adolescents. Social Indicators Research, 98, 519–532. Reynolds, W. M., & Mazza, J. J. (1998). Reliability and validity of the Reynolds Adolescent Depression Scale with young adolescents. Journal of School Psychology, 36, 295–312. Rothman, K. J. (2002). Epidemiology. An introduction. Oxford: Oxford University Press. Salmela-Aro, K., & Tuominen-Soini, H. (2010). Adolescents’ life satisfaction during the transition to postcomprehensive education: Antecedents and consequences. Journal of Happiness Studies, 11, 683–701. Schimmack, U., Oishi, S., Furr, R. M., & Funder, D. C. (2004). Personality and life satisfaction: A facetlevel analysis. Personality and Social Psychology Bulletin, 30, 1062–1075. Sharma, S. (1996). Applied multivariate techniques. New York: Wiley. Shih, J. H., Eberhart, N. K., Hammen, C. L., & Brennan, P. A. (2006). Differential exposure and reactivity to interpersonal stress predict sex differences in adolescent depression. Journal of Clinical Child and Adolescent Psychology, 35, 103–115. Suldo, S. M., & Huebner, E. S. (2004a). Does life satisfaction moderate the effects of stressful life events on psychopathological behaviour during adolescence? School Psychology Quarterly, 19, 93–105. Suldo, S. M., & Huebner, E. S. (2004b). The role of life satisfaction in the relationship between authoritative parenting dimensions and adolescent problem behaviour. Social Indicators Research, 66, 165–195. Suldo, S. M., Riley, K. N., & Shaffer, E. J. (2006). Academic correlates of children and adolescents’ life satisfaction. School Psychology International, 27, 567–582. Suldo, S. M., Shaunessy, E., Thalji, A., Michalowski, J., & Shaffer, E. (2009). Sources of stress for students in high school college preparatory and general education programs: Group differences and associations with adjustment. Adolescence, 44, 925–948. Tabachnik, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). USA: Pearson Education Inc. Thapar, A., Collishaw, S., Pine, D. S., & Thapar, A. K. (2012). Depression in adolescence. Lancet, 379, 1056–1067. Undheim, A. M., & Sund, A. M. (2005). School factors and the emergence of depressive symptoms among young Norwegian adolescents. European Child and Adolescent Psychiatry, 14, 446–453. Wu, A. D., & Zumbo, B. D. (2008). Understanding and using mediators and moderators. Social Indicators Research, 87, 367–392. Zung, W. W. K. (1965). A self-rating depression scale. Archives of General Psychiatry, 12, 63–70.

123