Journal of Adolescence 35 (2012) 929–939

Contents lists available at SciVerse ScienceDirect

Journal of Adolescence

journal homepage: www.elsevier .com/locate/ jado

Gendered pathways in school burnout among adolescents

Katariina Salmela-Aro a,*, Lotta Tynkkynen b

aUniversity of Helsinki, Helsinki Collegium for Advanced Studies, PO Box 4, 00014 Helsinki, FinlandbUniversity of Jyväskylä, Finland

Keywords:AdolescenceSchool burnoutExhaustionCynicismInadequacy trajectoriesLongitudinalEducational transitionGenderAcademic trackVocational track

* Corresponding author. Tel.: þ358 50 4155283.E-mail address: katariina.salmela-aro@helsinki.fi

0140-1971/$ – see front matter � 2012 The Foundadoi:10.1016/j.adolescence.2012.01.001

a b s t r a c t

The aim of this study is to examine differences in student burnout by gender, time statuswith two time points before and after an educational transition, and educational track(academic vs. vocational). The definition of burnout is based on three components:exhaustion due to school demands, a disengaged and cynical attitude toward school, andfeelings of inadequacy as a student (Salmela-Aro, Kiuru, Leskinen, & Nurmi, 2009). A totalof 770 Finnish adolescents (M age ¼ 16) were examined at the beginning of their last yearin comprehensive school, and three times annually during their secondary education bothon academic and vocational tracks. Among boys on the academic track, overall schoolburnout and its three components, exhaustion, cynicism and inadequacy, increased,whereas among boys on the vocational track, no changes in school burnout emerged.Among girls on the academic track, overall school burnout and inadequacy increased,whereas among girls on the vocational track, cynicism decreased. Finally, school burnoutwas highest among girls on the academic track, but increased most among boys on theacademic track.� 2012 The Foundation for Professionals in Services for Adolescents. Published by Elsevier

Ltd. All rights reserved.

Adolescents gendered pathways of multidimensional school burnout during academic and vocational tracks

Recent surveys show that there have been dramatic increases in school-related stress and stress-related health problemsamong adolescents in the Nordic countries (OECD, 2009; Schrami, Peski, Grossi & Simonsson-Sarnecki, 2011); over the lastdecade the number of 16-year-olds reporting stress has been rising steadily. These symptoms are especially prevalent amonggirls (Basow & Rubin, 1999; Hagquist, 2009). This alarming trend should be taken seriously; it has recently been suggestedthat educational transitions and educational tracks may play a destabilizing role in school stress and burnout (Salmela-Aro,Kiuru, & Nurmi, 2008; Tram & Cole, 2006). In the present four-wave longitudinal study we examined whether certainadolescents have difficulty and experience feelings of stress in navigating school transitions and educational tracks. Weanalyze differences in gender and educational tracks in school burnout, defined as exhaustion owing to school demands,a disengaged and cynical attitude toward school, and feelings of inadequacy as a student (Salmela-Aro, Kiuru, et al., 2009;Schaufeli, Salanova, Gonzales-Roma, & Bakker, 2002; see also Fredricks, Blumenfeld, & Paris, 2004).

A recent survey in Sweden revealed that the perceived stress-related demands of adolescents were most often connectedto school (ULF, 2009). Academic school-related burnout can be seen as a consequence of deteriorating energy resources andincreasing school demands. Although burnout was originally considered awork-related disorder (Maslach, Schaufeli, & Leiter,2001), it is also relevant in the context of school (see also Finn, 1989: Modin, Österbeg, Toivanen, & Sandell, 2011; Skinner,Furrer, Marchand, & Kindermann, 2008). School is a place where students work: they attend classes and carry out assign-ments in order to pass examinations and obtain a degree. According to the demands-resources model (Schaufeli & Bakker,

(K. Salmela-Aro).

tion for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939930

2004), the investment of a great deal of energy and performance without any return depletes one’s personal energy store. Ifthe imbalance between spending and regaining energy persists over a long period, burnout may occur. In accordancewith thedemands-resourcesmodel (Schaufeli & Bakker, 2004), two processes play a role in burnout: an effort-driven energetic processof overtaxing and wearing out in which the demands of studying exhaust a student’s energy and the motivational process inwhich there is a lack of resources for dealing effectively those study demands.

In many European educational systems, the transition from comprehensive school to either an academic or a vocationaltrack is the key educational change during adolescence, although the actual age at transition varies considerably (Mastenet al., 1999; Schulenberg, Maggs, & O’Malley, 2003; Shanahan, 2000). Moreover, as adolescents approach the end of upper-secondary school, they come under increasing pressure to succeed academically as they face another transition to tertiaryeducation (Lefkowitz, 2005). There is substantial evidence of a decline in academic motivation, attachment to school, andacademic achievement during school transition in early adolescence (e.g., Anderman & Maehr, 1994; Eccles & Midgley, 1989).Most of these studies, however, were conducted in the U.S. Less is known about what happens in middle and late adolescenceduring upper-secondary school (the academic track) or vocational school (the vocational track), which are the typical patternsin the Nordic countries. The track taken determines the quality and the kind of learning opportunities open to the student(Oakes, Gamoran, & Page,1992) as well as the peer exposure and thus, to a certain extent, the nature of the social relationshipsformed at school (Fuligni, Eccles, & Barber, 1995). On the academic track the demands at school increase, and adolescentsbecomemore aware of individual differences in their abilities and achievements (Wigfield, Eccles, Schiefele, Roesner, & Davis-Kean, 2006). In line with demands-resources (Schaufeli & Bakker, 2004) and stage-environment fit theory (Eccles & Midgley,1989), the academic track might create a misfit that provides less caring and supportive teacher–student relationships andincreasing social comparisons and competition compared to the vocational track. This might lead to an increase in schoolburnout, while the less demanding vocational track might lead to decrease in school burnout. Adolescents choosing theacademic track might experience the classroom as more competitive, feel an increased emphasis on grades, and perceivehigher teacher expectations. This could lead to feelings of exhaustion and inadequacy. The vocational track offers fewerdemands and a safer, more supportive environment and thereby supports feelings of competence and relatedness (Deci &Ryan, 2002). It can thus lead to a better fit and less cynical attitudes toward school.

Besides the educational track, gender might play a role in determining the different trajectories of school burnout duringadolescence (Ge, Conger, & Elder, 2001; Nolen-Hoeksema & Girgus, 1994). Previous research has identified gender differencesin school adjustment with 60 percent of girls in high school reporting the feeling of stress as a result of the demands they putupon themselves, whereas the corresponding figure for boys was 38 percent (ULF, 2009; see also Jose & Ratcliffe, 2004;Matud, 2004). Girls and boys may experience school stressors differently: girls have been found to experience more inter-nalized symptoms (e.g., Hoffmann, Powlishta, & White, 2004; Nolen-Hoeksema & Girgus, 1994; Pomerantz, Altermatt, &Saxon, 2002), such as depression (Moksnes, Moljord, Espnes, & Byrne, 2010), exhaustion, and inadequacy (Salmela-Aro et al.,2008), whereas boys typically show more problem behaviors and externalized symptoms (Masten et al., 1999), such ascynicism (Salmela-Aro et al., 2008). There is some evidence to suggest that girls respond more negatively to competitivelearning conditions, are more exposed to stressful events, and more vulnerable to the negative effects (Ge et al., 2001; Kessler& McLeod, 1984). As a consequence, girls may suffer more from school burnout. However, girls also tend to perform better atschool than boys (e.g., Pomerantz et al., 2002), attribute greater importance to academic achievement (Murberg & Bru, 2004),and be more engaged at school (Martin, 2004). For boys, in turn, courses may be taught in a manner that they find eitherboring or irrelevant to their interests (Eccles & Midgley, 1989), which could lead to the development of a cynical attitudetoward education on the academic track. The problem of disengagement is particularly acute among boys during the highschool years (Wigfield et al., 2006). According to the dual impact model of gender and career (Abele, 2003), gendered societalexpectations and gendered self-concept can direct these gendered paths in school burnout. Gender differences might alsoemerge if traditional gender role behaviors and values are emphasized (Eccles, 2011), and girls and boys might engage indifferent activities and value different subjects (Chow& Salmela-Aro, 2011). However, the research thus far has not focused ongendered changes in overall school burnout or on its three components, exhaustion, cynicism, and inadequacy, on theacademic and vocational tracks in late adolescence.

Research aim and hypotheses

The aim of the study is to examine the differences in student burnout (overall, exhaustion, cynicism, and inadequacy) bygender, time (twice before and after the school transition), and educational track (academic vs. vocational).

(H1) Track: School burnout, in particular inadequacy, increases on the academic track because of increasing academicdemands.

(H2) Track: School burnout, in particular cynicism, decreases on the vocational track because of decreasing academicdemands, better demands-resources fit, and better stage-environment fit.

(H3) Gender: Girls feel the effects of school burnout, in particular, inadequacymore than boys because girls internalize stress.(H4) Gender: Boys feel more cynicism (external symptoms) than girls.(H5) Track and gender: On the academic track, cynicism increase among boys and feelings of inadequacy increase among

girls.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939 931

Method

Participants

The present study is part of the ongoing FinEdu project, the focus of which is on life-planning and well-being duringeducational transitions in middle and late adolescence. At the beginning of the present study the participants were in theirninth school year (median age ¼ 15), facing the transition to secondary education. All the ninth-year students in a medium-sized town (population 88,000) in central Finland were contacted (N ¼ 954) and asked to participate in all four stages. InFinland school children uniformly receive a similar basic education up until the age of 16; the education is provided by thestate and tuition is free. After comprehensive school, the educational trajectories begin to diverge: about 55 percent of alladolescents enter upper-secondary schools, while 37 percent choose vocational schools, each trajectory takes three to fouryears. Two percent stay on for a voluntary tenth year, and 6 percent exit formal education (School statistics, 2003). Theaverage academic achievement in the ninth year of comprehensive school is the minimum requirement for admission toupper-secondary school: matriculation, in turn, is a bridge to further education, most likely, higher education. Vocationaleducation offers a wide variety of options and serves as a route to working life in academically less demanding occupationsand also to tertiary-level education, most often in vocational institutions. Girls are more likely to matriculate and enteruniversities than boys (School statistics, 2003).

The first two measurements were carried out before the transition to upper-secondary education (the academic track) orvocational education (the vocational track) education. The first was taken at the beginning of the ninth grade, which is thefinal year of comprehensive school (Time 1), and the second was taken at the end of that school year (Time 2). The final twomeasurements were carried out after the transition to secondary schooling: the first was taken six months after the transition(Time 3), and the second was made a year later (Time 4). The target sample was all the ninth grade students (N ¼ 954) of thetown. At Time 1, 687 (327 girls, 360 boys) of the 954 students participated in the study (response rate 72%). The number ofparticipants at Time 2was 642 (317 girls, 325 boys; response rate 67%); at Time 3, it was 818 (396 girls, 422 boys, response rate86%), and at Time 4, the number was 749 (368 girls, 381 boys; response rate 79%). In the analyses we included only thoseparticipants about whomwe had information on their educational tracks after Time 2 (N¼ 770). Of these participants, 99 girls(12.9%) were on the vocational track, 274 girls (35.6%) were on the academic track, 154 boys (20%) were on the vocationaltrack, and 243 boys (31.6%) were on the academic track.

The majority of the participants (99%) were Finnish-speaking. This ratio agrees well with the figures for ethnic minoritiesin the area. The questionnaires were group-administered to the students in their classrooms during regular school hours orsent to their postal address.

Attrition analyses

In order to assess the extent of attrition between the measurements, we compared the students who participated at leastonce (N¼ 463) with thosewho participated at each time point (N¼ 307) with regard to school burnout and educational track.The analysis revealed no selection effect with respect to burnout or its three components (Time 1–Time 4) or school track. Weused the missing-data procedure (for details, see the description below of our analytical strategy) and were thus able to applydata on all the participants for the analyses.

Measures

School burnoutBurnout was assessed in accordance with the School Burnout Inventory (Salmela-Aro, Kiuru, et al., 2009) and the

adolescents’ self-reports for the last month. The burnout scale comprises nine uni dimensional items measuring the threedimensions: (1) exhaustion at school (e.g., I feel overwhelmed by my school work; I brood over matters related to my schoolwork a lot during my free time); (2) disengagement or cynicism with regard to the meaning of school (e.g., I feel a lack ofmotivation in my school work and often think of giving up; I feel that I am losing interest in my school work), and (3) feelingsof inadequacy at school (e.g., I often have feelings of inadequacy with regard to my school work). There were three items foreach of the three dimensions, all of which were rated on a six-point scale (1 ¼ strongly disagree; 6 ¼ strongly agree). Wecalculated separate sum scores at each of the four time points for the three dimensions and the overall burnout. The Cron-bach’s alpha reliability for the overall school burnout scale was between .78 and .86; for exhaustion, between .75 and .78; forcynicism, between .83 and .87; and for inadequacy, between .74 and .84.

The sum score of overall burnout (criteria> 29) showed that at Time 1, 6.9% of boys and 10.3% of girls suffered from severeburnout. At Time 4,16.9% of girls on the academic track and 12.4% of boys on the academic track suffered from severe burnout,while 10.8% of girls on the vocational track and 9.9% of boys on the vocational track suffered from severe burnout.

GenderOn the questionnaire the students were asked to circle their gender (1 ¼ girl, 2 ¼ boy). This variable was dummy-coded

further so that 1 indicated a girl and 0 indicated a boy.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939932

Educational tracksThe following question concerned the participants’ educational track after comprehensive school (Time 3 and Time 4): (1)

“Are you in school at the moment?” (1¼ yes, 0¼ no); (2) “If you are in school, what is the name of your school?” The next stepwasto circle an educational-track variable by comparing students on the academic track to those on the vocational track.Vocational-track students included those who were studying in a vocational school full-time and also those who were invocational school plus were taking courses in upper-secondary school, thus they were aiming for a double degree. Adoles-cents in upper-secondary schools (academic track; n ¼ 517) were coded 1 and those in vocational schools (vocational track;n ¼ 253) were coded 0. Twenty-one (3%) of the participants had dropped out of the educational system one year aftercomprehensive school. They were not included in the data.

Analysis strategy

The aim of the present study was to examine the differences and changes girls and boys show in their overall burnout andits three dimensions after transition from comprehensive school to academic vs. vocational track by using latent growth curvemodeling (LGCM; Muthén & Muthén, 1998–2007). LGCM is a statistical technique used in the structural equation frameworkto estimate growth trajectory over a period of time. First, we fitted the linear growth curve models in the data for overallburnout and its three dimensions: exhaustion, cynicism, and inadequacy. Thereafter, we added a quadratic term in all themodels and applied a chi-square difference test to see whether the less parsimonious model would provide a better fit. Themodels were run separately for four groups: girls and boys on the vocational track and girls and boys on the academic track.Finally, we ran multigroup (pairwise) analyses to examine whether there were significant differences in the slopes and initiallevels of burnout and its three dimensions among the four groups. The differences between groups were tested pairwise bya chi-square difference test: the slopes and intercepts of two groups were constrained to be equal, and the fit of the non-constrained model was compared to the fit of the constrained model. Each group (gender*trajectory) was contrasted withall the other groups one by one, unless the parameters to be tested were non-significant in both groups.

We used the Mplus statistical package (Version 5; Muthén & Muthén, 1998–2007). The full information maximum like-lihood (FIML) estimationmethod was applied tomissing data. This missing-data method does not assign values for those thatare missing, but uses all the data that are available to estimate the model. Because the variables were skewed toward thelower end (g1 ¼ .32–.97), the parameters of the models were estimated using maximum likelihood estimation with non-normality robust standard errors (MLR estimator; Muthén & Muthén, 1998–2007).

The goodness-of-fit of the estimated latent growth models was evaluated according to the following four indicators: (a)the chi-square test, (b) the Comparative Fit Index (CFI), (c) the Root Mean Square Error of Approximation (RMSEA), and (d) theStandardized Root Mean Square Residual (SRMR).

Results

Correlations, means, and standard deviations for burnout variables are shown in Table 1. For all four groups (girls and boyson academic vs. vocational tracks), the linear model showed a better overall fit than the quadratic model, and thus the linearmodel was chosen as a basis for further analyses. Linear growth curve parameters for each of the four groups are shown inTable 2.

Table 1Sample correlation matrix and means and standard deviations for the observed variables.

Variable 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

1. Exhaustion T1 12. Cynicism T1 .39 13. Inadequacy T1 .54 .75 14. Overall Burnout T1 .76 .86 .91 15. Exhaustion T2 .48 .28 .33 .45 16. Cynicism T2 .17 .54 .46 .48 .54 17. Inadequacy T2 .26 .48 .50 .50 .67 .82 18. Overall Burnout T2 .34 .50 .50 .54 .83 .89 .94 19. Exhaustion T3 .47 .20 .30 .39 .47 .21 .30 .37 110. Cynicism T3 .17 .42 .32 .36 .21 .43 .41 .40 .45 111. Inadequacy T3 .23 .35 .37 .38 .33 .38 .44 .44 .58 .78 112. Overall Burnout T3 .34 .37 .38 .44 .41 .40 .45 .48 .78 .87 .91 113. Exhaustion T4 .45 .14 .27 .37 .44 .15 .24 .31 .59 .21 .36 .46 114. Cynicism T4 .15 .27 .27 .29 .14 .24 .24 .24 .29 .52 .45 .49 .35 115. Inadequacy T4 .24 .26 .36 .35 .23 .20 .30 .28 .34 .46 .52 .52 .51 .77 116. Overall Burnout T4 .34 .27 .36 .40 .24 .24 .32 .33 .49 .48 .53 .59 .74 .85 .91 1M 2.72 2.27 2.45 2.47 2.70 2.34 2.42 2.47 2.68 2.25 2.46 2.46 2.83 2.32 2.61 2.59SD 1.00 1.11 1.05 .88 1.23 1.27 1.23 1.10 1.05 1.12 1.03 .91 1.09 1.12 1.08 .92

Note. All correlations are statistically significant p < .001.

Table 2Linear growth curve parameters.

Variable Intercept Var Slope Var Intercept M Slope M Intercept*slope

Girls, Vocational TrackExhaustion .67*** .03* 3.46*** �.15 �.36*Cynicism .92*** .03* 2.64*** �.43* �.66***Inadequacy .46*** .01 4.01*** �.96 �.17Overall Burnout .52*** .02 3.73*** �.43 �.59***

Girls, Academic TrackExhaustion .50*** .02*** 4.21*** .19 �.32*Cynicism .74*** .04*** 2.59*** .14 �.46***Inadequacy .53*** .02** 3.37*** .41** �.34*Overall Burnout .39*** .02** 4.07*** .30** �.30

Boys, Vocational TrackExhaustion .42*** .02 3.70*** �.17 �.13Cynicism .66*** .04* 2.87*** �.19 �.60***Inadequacy .54*** .02* 3.24*** �.15 �.44*Overall Burnout .47*** .02* 3.40*** �.15 �.42**

Boys, Academic TrackExhaustion .62*** .01 3.14*** .46*** �.32Cynicism .82*** .05*** 2.17*** .45*** �.51***Inadequacy .70*** .03** 2.57*** .62*** �.44***Overall Burnout .55*** .02*** 2.97*** .57*** �.42***

Note. ***p < .001, **p < .01, *p < .05.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939 933

Overall burnout

First, we examined the models for overall burnout individually for each group. For girls on both the vocational track(Chi ¼ 2.16, p > .05, CFI¼ 1.00, RMSEA ¼ .00, SRMR¼ .03) and the academic track (Chi¼ 10.58, p > .05, CFI¼ .96, RMSEA ¼ .10,SRMR ¼ .09) and for boys on both the vocational track (Chi ¼ 4.87, p > .05, CFI ¼ .98, RMSEA ¼ .04, SRMR ¼ .08) and theacademic track (Chi ¼ 7.21, p > .05, CFI ¼ .98, RMSEA ¼ .06, SRMR ¼ .05), the linear growth curve model fit the data well. Theoverall level of burnout remained stable among girls and boys on the vocational track, but it increased among both girls andboys on the academic track (Table 2, Fig. 1). The c2 difference tests showed that the slopes of girls and boys on the academictrack significantly differed statistically from the slopes in the other groups (Table 3). Burnout increased most among boys onthe academic track. Moreover, the initial levels of burnout were significantly different between girls and boys (Table 3): girlson the academic track scored higher on overall burnout than boys on the academic track. There was statistically significantvariance in the level and change of burnout in all groups except among girls on the vocational track (Table 2). The correlationbetween the intercept and the linear slope was negative and statistically significant in all groups except among girls on theacademic track: the higher the level of overall burnout, the less it increased (Table 2). Overall burnout was highest in girls ongirls, but increased most among boys on the academic track.

Exhaustion

The linear growth curve model for exhaustion fit the data well for girls on both the vocational track (Chi ¼ 6.75, p > .05,CFI ¼ .97, RMSEA ¼ .06, SRMR ¼ .09) and the academic track (Chi ¼ 15.48, p > .01, CFI ¼ .94, RMSEA ¼ .09, SRMR ¼ .07) and forboys on both the vocational track (Chi¼ 6.07, p> .05, CFI¼ .98, RMSEA¼ .04, SRMR¼ .07) and the academic track (Chi¼ 13.22,p > .05, CFI ¼ .96, RMSEA ¼ .08, SRMR ¼ .06). The results showed that among boys on the academic track, there wasa statistically significant increase in exhaustion, which differed from the other groups (Table 4, Fig. 2), in which exhaustionwas stable. The initial levels of exhaustionwere significantly different between girls and boys (Table 4): girls on the academictrack scored higher on exhaustion than boys on the academic track. The results further showed that there was significantvariance in the level of exhaustion in all groups and in the slope among girls on both tracks (Table 2). The correlation betweenthe intercept and the linear slope in exhaustion was negative and statistically significant among girls: the higher the level ofexhaustion, the less it increased (Table 2).

Cynicism

The linear growth curve model for cynicism fit the data well for girls, both on the vocational track (Chi ¼ 6.90, p > .05,CFI ¼ .96, RMSEA ¼ .06, SRMR ¼ .05) and the academic track (Chi ¼ 7.23, p > .05, CFI ¼ .98, RMSEA ¼ .04, SRMR ¼ .06), and forboys both on the vocational track (Chi¼ 5.84, p> .05, CFI¼ .94, RMSEA¼ .05, SRMR¼ .08) and the academic track (Chi¼ 12.98,p> .05, CFI¼ .95, RMSEA¼ .08, SRMR¼ .05). The results showed that cynicism decreased among girls on the vocational track,which was significantly different statistically from girls and boys on the academic track; cynicism increased among boys onthe academic track, which was significantly different statistically from all the other groups (Table 5, Fig. 3). Among girls on theacademic track and boys on the vocational track, cynicism remained stable. The initial levels of cynicism were significantly

2,9

2,5

2,7

2,3Girls-VocGirls-AcadBoys-Voc

1,9

2,1 Boys-Acad

1,7Time 1 Time 2 Time 3 Time 4

Fig. 1. Overall school burnout estimates for the four groups. Note. Time 1 ¼ the beginning of the 9th year; Time 2 ¼ the end of the 9th year; Time 3 ¼ the first yearin secondary education (Acad ¼ Academic track, upper-secondary education; Voc ¼ Vocational track); Time 4 ¼ the second year in secondary education.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939934

different among all groups, except between girls and boys on the vocational track, girls on the academic track, and boys on thevocational track. At the last measurement cynicism was highest among boys on the academic track. The results showedfurther that there was statistical variance in both the level and slope of cynicism in all of the groups (Table 2): there wassignificant heterogeneity in cynicism. Finally, the correlation between the intercept and the linear slope in cynicism wasnegative: the higher the level of cynicism, the less it increased (Table 2).

Inadequacy

The linear growth curve model fit the data well for inadequacy for girls both on the vocational track (Chi ¼ 6.67, p > .05,CFI ¼ .95, RMSEA ¼ .06, SRMR ¼ .07) and the academic track (Chi ¼ 7.58, p > .05, CFI ¼ .97, RMSEA ¼ .06, SRMR ¼ .05), and forboys both on the vocational track (Chi¼ 7.33, p> .05, CFI¼ .92, RMSEA¼ .08, SRMR¼ .09) and the academic track (Chi¼ 11.48,p > .05, CFI ¼ .96, RMSEA ¼ .07, SRMR ¼ .06). The results showed that the level of inadequacy remained stable on thevocational track, whereas it increased on the academic track (Table 6, Fig. 4): an increase in inadequacy in girls on academictrack was not significantly different from the increase in boys, but both differed from boys and girls on the vocational track.The results further showed that there was a statistically significant variance among all groups in the level and in the linearslope of inadequacy except among girls on the vocational track (Table 2). The correlation between the intercept and the linearslope in inadequacy was negative and statistically significant among girls on the academic track and among boys on bothtracks: the higher the level of inadequacy, the less it increased (Table 2).

Exhaustion, cynicism, and inadequacy

All the slopes (overall, exhaustion, cynicism, inadequacy) in the means on the vocational track were negative, while on theacademic track the slopes were positive.

Finally, the correlations between the slopes of the three components analyzed only between those slopes that werestatistically significant, showed that among boys on the academic track, the correlation between the slopes of exhaustion andcynicism (r ¼ .90, p < .01) and between exhaustion and inadequacy (r ¼ .80, p < .05) were statistically significant: the moreexhaustion increased, the more cynicism and inadequacy also increased.

Table 3Overall burnout comparisons among the four groups.

Level c2difference (H0–H1) Slope c2difference (H0–H1)

Girls, Vocationalvs. Girls, Academic c2(1) ¼ 1.82 ns. c2(1) ¼ 11.52 p < .001vs. Boys, Vocational c2(1) ¼ 10.36 p < .01vs. Boys Academic c2(1) ¼ 24.25 p < .001 c2(1) ¼ 20.37 p < .001

Girls, Academicvs. Boys, Vocational c2(1) ¼ 6.79 p < .001 c2(1) ¼ 5.94 p < .05vs. Boys, Academic c2(1) ¼ 19.28 p < .001 c2(1) ¼ 4.27 p < .05

Boys, Vocationalvs. Boys, Academic c2(1) ¼ 1.97 ns. c2(1) ¼ 13.46 p < .001

Note. The chi-square difference tests were not conducted between two non-significant parameters.

Table 4Exhaustion comparison among the four groups.

Level c2difference (H0–H1) Slope c2difference (H0–H1)

Girls, Vocationalvs. Girls, Academic c2(1) ¼ 1.06 ns.vs. Boys, Vocational c2(1) ¼ 12.05 p < .001vs. Boys, Academic c2(1) ¼ 8.08 p < .001 c2(1) ¼ 4.82 p < .05

Girls, Academicvs. Boys, Vocational c2(1) ¼ 1.42 p < .001vs. Boys, Academic c2(1) ¼ 1.18 p < .001 c2(1) ¼ .64 ns.

Boys, Vocationalvs. Boys, Academic c2(1) ¼ .96 ns. c2(1) ¼ 5.06 p < .05

Note. The chi-square difference tests were not conducted between two non-significant parameters.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939 935

Discussion

The alarming trend of increasing stress among adolescents in a school context is a serious one. Recent studies show thatschool burnout can lead to depression later, and thus in life, school burnout should be taken seriously (Salmela-Aro,Savolainen, & Holopainen, 2009). The present study investigated the differences in academic school-related burnout bygender, points in time, and in academic vs. vocational educational tracks. We defined burnout on the basis of threecomponents: exhaustion owing to school demands, a cynical and disengaged attitude toward school, and feelings of inad-equacy as a student (Salmela-Aro, Kiuru, et al., 2009). In line with our hypotheses, our study demonstrates that educationaltrack, gender, and their interactions all play a role in the differences seen in multidimensional school burnout.

First, the academic track appears to play a critical role in school burnout. Overall school burnout and inadequacy increasedamong those in the academic track thereby supporting Hypothesis 1. Second, girls, in particular those on the academic track,suffered most from school burnout, supporting Hypothesis 3, whereas boys on the academic track suffered most fromcynicism, supporting Hypothesis 4. Third, both educational track and gender played a role in cynicism: cynicism increased inboys on the academic track, while it decreased among girls on the vocational track, thus supporting Hypothesis 5. Fourth,school burnout decreased, particularly cynicism among girls on the vocational track, supporting Hypothesis 2. Interestingly,among boys on the academic track overall school burnout and each of its dimensions increased, and burnout increased moreamong boys on the academic track than among girls. The results of the present study reveal that overall school burnout and alldimensions increased among boys on the academic track.

In accordance with the demands-resources theory (Schaufeli & Bakker, 2004), we assumed that the academic track wouldlead to stronger burnout and more feelings of inadequacy than the vocational track. Our results show that the academic trackis clearly a risk period in terms of developing burnout, whereas the vocational track seems to be protective, particularly forgirls. The vocational track might allow students to feel competent and relatedness and thus support their needs (Deci & Ryan,2002), whereas the academic track with its focus on competition, comparison, and relative ability might be detrimental forsome students. An effort-driven energetic process of overtaxing and wearing out inwhich study demands exhaust a student’senergy might play a role in increasing the burnout on the academic track, whereas the motivational process in which theresources dealing effectively with decreasing demands and thus decreasing burnout might play a role in the lower levels ofburnout on the vocational track. Finnish upper-secondary schools may be challenging and stressful for young people, who arefacing unfamiliar academic expectations, changes in sources of social support, and demanding social norms. The big-fishlittle-pond effect, a phenomenon in which attending academically selective high schools negatively affects the academicself-concept, might also play a role in these feelings of inadequacy (Marsch, Trautwein, Ludtge, Baumert, & Koller, 2007).

3,2

3

2,6

2,8Girls-VocGirls-Acad

2,4Boys-VocBoys-Acad

2

2,2

Time 3Time 1 Time 2 Time 4

Fig. 2. Exhaustion estimates for the four groups. Note. Time 1 ¼ the beginning of the 9th year; Time 2 ¼ the end of the 9th year; Time 3 ¼ the first year insecondary education (Acad ¼ Academic track, upper-secondary education; Voc ¼ Vocational track); Time 4 ¼ the second year in secondary education.

Table 5Cynicism comparison among the four groups.

Level c2difference (H0–H1) Slope c2difference (H0–H1)

Girls, Vocationalvs. Girls, Academic c2(1) ¼ 7.54 p < .01 c2(1) ¼ 5.13 p < .05vs. Boys, Vocational c2(1) ¼ 2.13 ns. c2(1) ¼ .59 ns.vs. Boys, Academic c2(1) ¼ 28.42 p < .001 c2(1) ¼ 13.30 p < .001

Girls, Academicvs. Boys, Vocational c2(1) ¼ 1.12 ns.vs. Boys, Academic c2(1) ¼ 7.64 p < .001 c2(1) ¼ 7 p < .01

Boys, Vocationalvs. Boys, Academic c2(1) ¼ 20.68 p < .001 c2(1) ¼ 27.06 p < .001

Note. The chi-square difference tests were not conducted between two non-significant parameters.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939936

The results showed that girls experienced more overall school burnout than boys. An earlier cross-sectional study usinga sample of Norwegian adolescents also reported that girls experienced more school pressure than boys (Murberg & Bru,2004). The dual impact of gender and career (Abele, 2003) also suggests more pressure on girls on the academic trackcompared to those on the vocational track. One possibility is that girls may turn the stress inward and feel inadequacy on theacademic track, while boys may direct it outward toward school and other institutions and feel cynicism. Interestingly, theresults of the present study reveal that overall school burnout and all dimensions increased among boys on the academictrack. This is in line with recent results, which found boys at greater risk for school disengagement during high school(Wigfield et al., 2006). As boys approach graduation from upper-secondary school, they may experience increasing pressureto be successful academically (Lefkowitz, 2005). Boys on the academic track seem to catch up with girls in terms of schoolburnout, thus indicating a reduction in the gender difference over time during late adolescence. However, more research isneeded on this topic.

Finally, it might be that not all students follow the same trajectory andwe examined possible heterogeneity in the changesin academic burnout. The results of the present study revealed significant heterogeneity in the changes in multidimensionalacademic burnout. The greatest heterogeneity was found in the slopes in cynicism among those on the academic track, whilethe heterogeneity was the least in the slopes in exhaustion. These results show that there might be different trajectories incynicism. It may be that cynicism increases only for some of those on the academic track, and thus there may be a need to takea person-oriented approach in future studies.

Limitations

The results showed that school burnout among young peoplemay either intensify or decline following amajor educationaltransition (see also Ash &Huebner, 2001).However, it should be borne inmind that this major educational transition is not theonly change adolescents experience during this period: they also face neurobiological, social and autonomy transitions, forexample (Lefkowitz, 2005; Masten et al., 1999; Schulenberg et al., 2003; Shanahan, 2000), any of which could havecontributed to the changes in school burnout identified in this study. It has been shown that, aside from the genetic andbiological factors and the environmental risk factors (e.g., peer rejection, harsh parenting, stressful life-events; Bagwell,Newcomb, & Bukowski, 1998), cognitive vulnerabilities (e.g., dysfunctional attitudes, a negative attributional style, anda ruminative response style, for a review, see Hankin et al., 1998) are connected to maladjustment during adolescence.Previous research attributes the higher vulnerability of girls to factors ranging from poor well-being to changes in hormonallevels, lower body satisfaction, and entering puberty early (Angold, Costello, Erkanli, &Worthman,1999), being more reactive

2,9

2,5

2,7

2,3Girls-VocGirls-Acad

1,9

2,1Boys-VocBoys-Acad

1,7

,

Time 1 Time 2 Time 3 Time 4

Fig. 3. Cynicism toward school estimates for the four groups. Note. Time 1 ¼ the beginning of the 9th year; Time 2 ¼ the end of the 9th year; Time 3 ¼ the firstyear in secondary education (Acad ¼ Academic track, upper-secondary education; Voc ¼ Vocational track); Time 4 ¼ the second year in secondary education.

Table 6Inadequacy comparison among the groups.

Level c2difference (H0–H1) Slope c2difference (H0–H1)

Girls, Vocationalvs. Girls, Academic c2(1) ¼ 5.1 p < .05 c2(1) ¼ 14.96 p < .001vs. Boys, Vocational c2(1) ¼ 6.66 p < .001vs. Boys Academic c2(1) ¼ 20.37 p < .001 c2(1) ¼ 21.51 p < .001

Girls, Academicvs. Boys, Vocational c2(1) ¼ .72 ns. c2(1) ¼ 8.72 p < .01vs. Boys, Academic c2(1) ¼ 11.26 p < .001 c2(1) ¼ 1.99 ns.

Boys, Vocationalvs. Boys, Academic c2(1) ¼ 4.75 p < .01 c2(1) ¼ 15.01 p < .001

Note. The chi-square difference tests were not conducted between two non-significant parameters.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939 937

to stress and negative life events (Crick & Zahn-Waxler, 2003), having a higher tendency toward ruminative coping(Broderick, 1998; Burwell & Shirk, 2007), and differing in how adjustment of self-perceptions according to external feedbackon academic ability (Crosnoe, Riegle-Crumb, & Muller, 2007). However, more research is needed on this topic.

At least the following limitations should be taken into account in any effort to generalize the results. First, all of themeasures were based on self-reporting, which is not always themost valid or most reliable method of data collection (Shaffer,2002) and also exposes the data to common-method variance. Accordingly, it would have been useful to have had registerdata on achievement, as well as on the neurobiological and hormonal changes taking place during this age period. Second, wedid not have information about possible natural oscillations in school burnout. Consequently, the changes that took placeconcurrently with this particular educational transition may be attributable in part to such oscillations. Moreover, we had noinformation about other transitions, such as puberty, that might have occurred during the age period of interest. Fourth, theeducational trajectories under investigation spanned only the transition from comprehensive school to upper-secondary orvocational education. Future studies are thus needed to examine the role of school burnout during other educational tran-sitions. It would also be useful to study the relationship between school and work burnout. Fifth, the present study wascarried out in Finland, and thus caution should be exercised in generalizing the results to other school contexts. However,many European countries have a similar educational system, in which students attend comprehensive school and then go onto an academic or a vocational track. Nevertheless, it is very likely that those who dropped out of the educational system arethe same adolescents who dropped out of the longitudinal data collection for this study, and this bias should be taken intoaccount as a possible limitation. Moreover, variables such as family problems may affect both academic progress and schoolburnout, and this too should be taken into account in future studies. Finally, our results do not provide any causal expla-nations, and more research is needed to identify possible reverse paths from school trajectories and school burnout.

Implications for future research and practice

Given that most longitudinal research conducted among adolescents focuses on the negative aspects of adolescentdevelopment, there is an evident need for longitudinal studies on strengths and well-being, such as school engagement.There is also a need to consider the different changes adolescents experience, such as in hormonal and cortical levels, andtheir impact on burnout. Moreover, studies should be conducted during stable time periods in order to investigate whetherburnout levels change in stable as well as in transitional periods. For the future, more intensive data collection, such as the useof diaries, would facilitate a more detailed tracking of school burnout among adolescents during educational transitions(Huebner, Funk, & Gilman, 2000).

2,7

2,9

2,5

,

Girls-VocGirls-Acad

2,3Boys-VocBoys-Acad

1,9

2,1

Time 3 Time 4Time 1 Time 2

Fig. 4. Feelings of inadequacy as a student estimates for the four groups. Note. Time 1 ¼ the beginning of the 9th year; Time 2 ¼ the end of the 9th year; Time3 ¼ the first year in secondary education (Acad ¼ Academic track, upper-secondary education; Voc ¼ Vocational track); Time 4 ¼ the second year in secondaryeducation.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939938

The study of burnout as a multidimensional construct and as an interaction between the individual and the schoolenvironment helps teachers to understand better the complexity of students’ experiences in school. It enables educators tocraft nuanced practices and environments to prevent academic burnout. As our study implies, adolescence seems to bea period of potential school burnout, the level of which increases for those on the academic track. Early interventions, inparticular, should target adolescents who exhibit increasing levels of burnout in order to prevent an accumulation ofproblems such as depression. Moreover, students who feel cynicism about school are likely to engage in problem behaviorssuch as dropping out of school (Finn & Rock, 1997), especially during the high school years (Wigfield et al., 2006) which canlead later on to marginalization.

Funding

Support from the Academy of Finland (134931, 139168) and the Jacobs Foundation.

References

Abele, A. (2003). The dynamics of masculine-agentic and feminine-communal traits: findings from a prospective study. Journal of Personality and SocialPsychology, 85, 768–776.

Anderman, E., & Maehr, M. (1994). Motivation and schooling in the middle grades. Review of Educational Research, 64, 287–309.Angold, A., Costello, E. J., Erkanli, A., & Worthman, C. M. (1999). Pubertal changes in hormone levels and depression in girls. Psychological Medicine, 29,

1043–1053.Ash, C., & Huebner, E. S. (2001). Environmental events and life satisfaction reports of adolescents. School Psychology International, 22, 20–36.Bagwell, C., Newcomb, A., & Bukowski, W. (1998). Preadolescent friendship and peer rejection as predictors of adult adjustment. Child Development, 69,

140–153.Basow, S., & Rubin, L. (1999). Gender influences on adolescent development. In N. Johnson, M. Roberts, & J. Worell (Eds.), Beyond appearance: A new look at

adolescent girls (pp. 25–52). Washington DC: American Psychological Association.Broderick, P. C. (1998). Early adolescent gender differences in the use of ruminative and distracting coping strategies. Journal of Early Adolescence, 18,

173–191.Burwell, R. A., & Shirk, S. R. (2007). Subtypes of rumination in adolescence: associations between brooding, reflection, depressive symptoms, and coping.

Journal of Clinical Child and Adolescent Psychology, 36, 56–65.Chow, A., & Salmela-Aro, K. (2011). Task-values across subject domains: a gender comparison using a person-centered approach. International Journal of

Behavioral Development, 35, 202–209.Crick, N. R., & Zahn-Waxler, C. (2003). The development of psychopathology in females and males: current progress and future challenges. Development and

Psychopathology, 15, 719–742.Crosnoe, R., Riegle-Crumb, C., & Muller, C. (2007). Gender, self-perception, and academic problems in high school. Social Problems, 54, 118–138.Deci, E., & Ryan, R. (2002). Handbook of self-determination research. Rochester, NY: University of Rochester Press.Eccles, J. (2011). Gendered educational and occupational choices: applying the Eccles et al. model of achievement-related choices. International Journal of

Behavioral Development, 35, 195–201.Eccles, J. S., & Midgley, C. (1989). Stage/environment fit: developmentally appropriate classrooms for early adolescents. In R. Ames, & C. Ames (Eds.),

Research on motivation in education, Vol. 3 (pp. 139–181). New York: Academic Press.Finn, J. (1989). Withdrawing from school. Review of Educational Research, 59, 117–142.Finn, J., & Rock, D. (1997). Academic success among students at risk for school failure. Journal of Applied Psychology, 82, 221–234.Fredricks, J., Blumenfeld, P., & Paris, A. (2004). School engagement: potential of the concept, state of the evidence. Review of Educational Research, 74, 59–

109.Fuligni, A., Eccles, J., & Barber, B. (1995). The long-term effects of seventh-grade ability groupings in mathematics. Journal of Early Adolescence, 15, 58–89.Ge, X., Conger, R. D., & Elder, G. H., Jr. (2001). Pubertal transition, stressful life events, and the emergence of gender differences in adolescent depressive

symptoms. Developmental Psychology, 37, 404–417.Hagquist, C. (2009). Psychosomatic health problems among adolescents in Sweden – are the time trends gender related? European Journal of Public Health,

19, 331–336.Hankin, B. L., Abramson, L. Y., Moffitt, T. E., Silva, P. A., McGee, R., & Angell, K. E. (1998). Development of depression from preadolescence to young adulthood:

478 Emerging gender differences in a 10-year longitudinal study. Journal of Abnormal Psychology, 107, 128–140.Hoffmann, M. L., Powlishta, K. K., & White, K. (2004). An examination of gender differences in adolescent adjustment: the effect of competence on gender

role differences in symptoms of psychopathology. Sex Roles: A Journal of Research, 50, 795–810.Huebner, E. S., Funk, B. A., & Gilman, R. (2000). Cross-sectional and longitudinal psychosocial correlates of adolescent life satisfaction reports. Canadian

Journal of School Psychology, 16, 53–64.Jose, P. E., & Ratcliffe, V. (2004). Stressor frequency and perceived intensity as predictors of internalizing symptoms: gender and age differences in

adolescence. New Zealand Journal of Psychology, 33, 145–154.Kessler, R. C., & McLeod, J. D. (1984). Sex differences in vulnerability to undesirable life events. American Sociological Review, 49, 620–631.Lefkowitz, E. S. (2005). ‘Things have gotten better’ developmental changes among emerging adults after the transition to university. Journal of Adolescent

Research, 20, 40–63.Marsch, H., Trautwein, U., Ludtge, O., Baumert, J., & Koller, O. (2007). The big-fish-little-pond effect: persistent negative effects of selective high schools on

self-concept after graduation. American Psychologist, 44, 631–634.Martin, A. (2004). School motivation of boys and girls: differences of degree, differences of kind, or both? Australian Journal of Psychology, 56, 133–146.Maslach, C., Schaufeli, W., & Leiter, M. (2001). Job burnout. Annual Review of Psychology, 52, 397–422.Masten, A. S., Hubbard, J. J., Gest, S. D., Tellegen, A., Garmezy, N., & Ramirez, M. (1999). Competence in the context of adversity: pathways to resilience and

maladaptation from childhood to late adolescence. Development and Psychopathology, 11, 143–169.Matud, M. P. (2004). Gender differences in stress and coping styles. Personality and Individual Differences, 37, 1401–1415.Modin, B., Östberg, V., Toivanen, S., & Sandell, K. (2011). Psychosocial working conditions, school sense of coherence and subjective health. A multilevel

analysis of ninth grade pupils in the Stockholm area. Journal of Adolescence, 34, 129–139.Moksnes, U., Moljord, I., Espnes, G., & Byrne, D. (2010). The association between stress and emotional states in adolescents: the role of gender and self-

esteem. Personality and Individual Differences, 49, 430–435.Murberg, T. A., & Bru, E. (2004). School-related stress and psychosomatic symptoms among Norwegian adolescents. School Psychology International, 25,

317–332.Muthén, L., & Muthén, B. O. (1998–2007). Mplus user’s guide & Mplus Version 5. Retrieved from. http://www.statmodel.com/index2.html.Nolen-Hoeksema, S., & Girgus, J. S. (1994). The emergence of gender differences in depression during adolescence. Psychological Bulletin, 115, 424–443.

K. Salmela-Aro, L. Tynkkynen / Journal of Adolescence 35 (2012) 929–939 939

Oakes, J., Gamoran, A., & Page, R. N. (1992). Curriculum differentiation: opportunities, outcomes, and meanings. In P. W. Jackson (Ed.), Handbook of researchon curriculum (pp. 570–608). New York: Macmillan.

OECD. (2009). Doing better for children. OECD.Pomerantz, E. M., Altermatt, E. R., & Saxon, J. L. (2002). Making the grade but feeling distressed: gender differences in academic performance and internal

distress. Journal of Educational Psychology, 94, 396–404.Salmela-Aro, K., Kiuru, N., Leskinen, E., & Nurmi, J.-E. (2009). School burnout Inventory: reliability and validity. European Journal of Psychological Assessment,

25, 48–57.Salmela-Aro, K., Kiuru, N., & Nurmi, J.-E. (2008). The role of educational track in adolescents’ school burnout. British Journal of Educational Psychology, 78,

663–689.Salmela-Aro, K., Savolainen, H., & Holopainen, L. (2009). Depressive symptoms and school burnout during adolescence: evidence from two cross-lagged

longitudinal studies. Journal of Youth and Adolescence, 38, 1316–1327.Schaufeli, W., & Bakker, A. (2004). Job demands, job resources and their relationship with burnout and engagement; a multi-sample study. Journal of

Organizational Behavior, 25, 293–299.Schaufeli, W., Salanova, M., Gonzales-Roma, V., & Bakker, A. (2002). The Measurement of engagement and burnout: a two sample confirmatory factor

analytic approach. Journal of Happiness Studies, 3, 71–92.School statistics (2003). Helsinki: Central Statistical Office of Finland. Retrieved fromhttp://www.stat.fi/til/khak/2005/khak_2005_2006-12-21_tie_001.html.Schrami, K., Peski, A., Grossi, G., & Simonsson-Sarnecki, M. (2011). Stress symptoms among adolescents: the role of subjective psychosocial conditions, life

style and self-esteem. Journal of Adolescence, 34(5), 987–996.Schulenberg, J. E., Maggs, J. L., & O’Malley, P. M. (2003). How and why the understanding of developmental continuity and discontinuity is important: the

sample case of long-term consequences of adolescent substance abuse. In J. T. Mortimer, & M. J. Shanahan (Eds.), Handbook of the life course. New York:Kluewer Academic/Plenum Publishers.

Shaffer, D. R. (2002). Developmental psychology: Childhood and adolescence (6th ed.). Belmont: Wadsworth/Thomson Learning.Shanahan, M. J. (2000). Pathways to adulthood in changing societies: variability and mechanisms in life course perspective. Annual Review of Sociology, 26,

667–692.Skinner, E., Furrer, C., Marchand, G., & Kindermann, T. (2008). Engagement and disaffection in the classroom: part of a larger motivational dynamic? Journal

of Educational Psychology, 100, 765–781.Tram, J. M., & Cole, D. A. (2006). A multimethod examination of the stability of depressive symptoms in childhood and adolescence. Journal of Abnormal

Psychology, 115, 674–686.ULF. (2009). Undersökning av levnadsförhållanden bland barn 2009. www.sch.se.Wigfield, A., Eccles, J., Schiefele, U., Roesner, R., & Davis-Kean, P. (2006). Development of achievement motivation. In W. Damon, & N. Eisenberg (Eds.),

Handbook of child psychology (6th ed).. Social, emotional, and personality development, Vol. 3 (pp. 933–1002). New York: John Wiley.