RESEARCH ARTICLE

Chronic Work Stress and Depressive Symptoms:Assessing the Mediating Role of Teacher BurnoutMary A. Steinhardt1*†, Shanna E. Smith Jaggars2, Kathryn E. Faulk1 & Christian T. Gloria1

1Kinesiology and Health Education, The University of Texas at Austin, Austin, TX, USA2Teachers College, Columbia University, New York, NY, USA

Abstract

A conceptual model of the relationship between stress, the mediating role of burnout, and depressive symptomswas examined. Results indicated that teachers (n = 267) experiencing greater stress were more burned out. Thesubscale emotional exhaustion was moderately related to depressive symptoms, whereas depersonalization andreduced personal accomplishment had small positive relationships. After controlling for burnout anddemographics, the relationship between stress and depressive symptoms was small but significant. Emotionalexhaustion mediated the association between stress and depressive symptoms. The total effect of stress ondepressive symptoms, taking together the direct and indirect effects via burnout, accounted for 43% of the totalvariance. Copyright © 2011 John Wiley & Sons, Ltd.

Received 9 May 2010; Accepted 16 January 2011; Revised 11 January 2011

Keywords

stressors; chronic work stress; teacher burnout; depressive symptoms

*Correspondence

Mary A. Steinhardt, Department of Kinesiology and Health Education, Bellmont Hall 222, MC: D3700, The University of Texas at Austin,

Austin, TX 78710, USA.†Email: msteinhardt@austin.utexas.edu

Published online 11 February 2011 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/smi.1394

Introduction

Public school teachers experience a great deal of stressthroughout their careers with up to one‐third indicatingthat teaching is a ‘very or extremely stressful’ profession(Borg & Riding, 1991; Kyriacou, 2001). Stress, burnoutand teacher attrition have reached alarming levels,threatening quality education and subsequent studentachievement. Teacher attrition due to work stress isincreasing, with 40–50% of new teachers leaving theprofession after only 3 years (Ingersoll & Smith, 2003).Stressors that teachers regularly encounter include roleoverload, disruptive students, non‐supportive parents,lack of support from the administration, poor relation-ships with colleagues, being evaluated and high‐stakesstudent testing (Kyriacou, 2001; Manthei, Gilmore,Tuck, & Adair, 1996; Montgomery & Rupp, 2005).These stressors create and exacerbate stress, burnout andhigh turnover rates in the teaching profession.

Teacher burnout is recognized as a prolongedexposure to emotional and interpersonal stressors onthe job, often accompanied by insufficient recovery,resulting in previously committed teachers disengagingfrom their work. Burnout represents a response to anintolerable work situation (Cherniss, 1980; Evers, Tomic,

420

& Brouwers, 2004; Hakanen, Bakker, & Schaufeli, 2006;Maslach & Leiter, 2008) and is defined as a psychologicalsyndrome of three interrelated components: emotionalexhaustion, depersonalization and reduced personalaccomplishment (Maslach, 1993; Maslach, Jackson,& Leiter, 1996; Maslach, Schaufeli, & Leiter, 2001).Emotional exhaustion is the stress component andcentral quality of burnout. It occurs when a teacher isextremely fatigued and feels overextended by work anddrained of emotional and physical resources (Schwarzer& Hallum, 2008). Depersonalization represents theinterpersonal component and is described as feelingcynical, irritable and negative towards others. Teachersare more likely to develop an indifferent or cynicalattitude and detached response on the job when feelingemotionally exhausted and discouraged (Maslach et al.,2001). The self‐evaluation component of reducedpersonal accomplishment includes feelings of incompe-tence and a lack of productivity and achievement at work.It is difficult to feel a sense of accomplishment at workwhen feeling emotionally exhausted and indifferent orcynical towards others (Maslach & Leiter, 2008).Although this burnout syndrome can occur in anyprofession, it is most common in the helping professions,especially among teachers (Sakharov & Farber, 1983).

Stress and Health 27: 420–429 (2011) © 2011 John Wiley & Sons, Ltd.

M. A. Steinhardt et al. Stress, Burnout and Depressive Symptoms

Chronic work stress and subsequent burnoutsyndrome stem from the teachers’ appraisal ofdemands and the inability to cope effectively withthese demands (Kyriacou & Sutcliffe, 1978). TheTransactional Model of Stress and Coping (Lazarus,1966; Lazarus & Folkman, 1984) provides a frameworkto understand the complex cognitive processes thatteachers undergo as they attempt to cope withdemands at work. The basic premise is that stressfulexperiences are construed as transactions between theenvironment and the individual. The experience ofstress results when teachers appraise the environmentaldemands as threatening and feel they do not have thecoping resources available to meet those demands. Ifstressful situations cannot be alleviated through activeand meaning‐based coping strategies, defensive copingstrategies provide a much needed psychological escape.For example, if empathic and caring teachers remaincommitted in work situations where failure anddisappointment are frequent, the prolonged levels ofstress may result in teachers who are emotionallyexhausted and who become apathetic, cynical or rigid;if students are ungrateful or even abusive and theteacher’s active coping strategies are ineffective, theteacher–student relationship may become mechanicaland distant.

Both chronic stress and burnout negatively influencehealth status and are significantly related to mentalillness, especially depression (Schaufeli & Greenglass,2001; WHO, 1992). Psychosomatic disorders areamong the predominant reasons teachers give forleaving the profession, with depression among themost commonly listed (Weber, Weltle, & Lederer,2002). When work stress reaches an unbearable level,which nearly 20% of teachers admit experiencing(Hammen & deMayo, 1982), exhaustion, poor perfor-mance (Ahola et al., 2005) and depression can result(Bakker & Schaufeli, 2000). Because depression inteachers is one of the primary causes for increasedteacher absenteeism and high attrition rates, itthreatens the quality of our educational system(Hammen & deMayo, 1982; Hastings & Bham, 2003).

Studies have found work stress to be an independentpredictor of all three components of teacher burnout(Chan, 2003; Friesen & Sarros, 1989) and of depressivesymptoms (Melchior et al., 2007). Although teacherstend to display high levels of emotional exhaustion, thecomponent of burnout most strongly related todepression (Maslach et al., 2001; Schaufeli & Enzmann,1998), Maslach et al. suggest that the phenomenon ofburnout must include all three components. Theypropose that an exclusive focus on the stresscomponent of emotional exhaustion fails to capturethe relationships that teachers have with their work.When teachers are overloaded, they distance them-selves emotionally and cognitively from their work inan effort to cope with feelings of exhaustion. Exhaustedteachers do not have the capacity to proactively

Stress and Health 27: 420–429 (2011) © 2011 John Wiley & Sons, Ltd.

respond to the needs of their students and, as a result,are often indifferent or cynical towards others and findit difficult to feel a sense of accomplishment and pridein their work. They feel victimized, have low self‐esteem and have no long‐term goals to strive for, oftencharacterized by extended sick leaves or positionaltransfer requests (Maslach, 1993).

The view of burnout as a psychological syndrome isalso supported utilizing principles derived from equitytheory. According to this theory, individuals have atendency to pursue reciprocity in interpersonalrelationships and feel distressed when relationshipsare inequitable. For teachers, their investments ofenthusiasm and effort are reciprocated when studentsrespond with attention, respect and gratitude (Blase,1982, 1986; Blau, 1964). A lack of reciprocity candeplete the teachers’ emotional resources and fuel thedevelopment of the burnout syndrome (Bakker et al.,2000; Farber, 1991). Theoretically, chronic work stressleads to teacher burnout and hence to depression(Bakker et al., 2000). We follow this conceptualframework in our study, allowing burnout to mediatethe relationship between stress and depression. We alsoindicate a direct effect from work stress to depressivesymptoms as work stress has been identified as a riskfactor for the development of major depressiveepisodes (Wang, 2005) and shown to precipitate thedevelopment of anxiety and depression in previouslyhealthy individuals (Melchior et al., 2007).

In addition to the observed relationships amongchronic work stress, burnout and depression, a varietyof demographic characteristics have been found to relateto burnout and subsequent depression. Although studiesare relatively few, and findings are generally inconsistent(see reviews by Chang, 2009; Schaufeli & Enzmann,1998), these characteristics include sociodemographic(age, gender, ethnicity and education level) andjob‐related characteristics (grade level taught and yearsof teaching experience). Age has been most consistentlyrelated to burnout, with younger employees reportinghigher levels of burnout than older employees (Maslachet al., 2001). With regard to gender, some studies havefound nodifference in the prevalence of burnout betweengenders (Ahola et al., 2005; Hastings & Bham, 2003),whereas others suggest that women experience higherlevels of emotional exhaustion and men experiencehigher levels of depersonalization (Lackritz, 2004;Schwarzer & Hallum, 2008). In one study, men whoreported serious burnout were three times more likelyto experience major depression than women were(Ahola et al., 2005). There have only been a few studiesexamining ethnicity and burnout, and they have resultedin mixed findings: no difference (Lackritz, 2004), greaterburnout among minorities (Dyrbye et al., 2007) andgreater burnout among non‐minorities (McCarty, Zhao,& Garland, 2007). Similarly, the education level ofteachers has an uncertain relationship with burnout.Burnout has been found to occur more frequently in

421

Stress, Burnout and Depressive Symptoms M. A. Steinhardt et al.

teachers with lower levels of education (Demir, Ulusoy, &Ulusoy, 2003; Friedman, 1991), higher levels of education(Maslach et al., 2001) or to have no relationship at all witheducation (Embich, 2001). With regard to job‐relatedcharacteristics, high school teachers report greaterburnout than teachers of lower grades (Beer & Beer,1992; Russell, Altmaier, &VanVelzen, 1987). Conflictingresults have been found for level of teaching experience,with some studies finding that greater experiencedecreases burnout (Demir et al., 2003; Maslach et al.,2001) whereas others suggest teachers with greaterexperience are more burned out (Friedman, 1991).

In summary, teachers experiencing burnout feeldissatisfied and exhausted, alienate themselves fromfellow teachers and their students, are less productiveand experience greater health problems (Chan, 2003;Maslach et al., 2001). Although previous studies haveexamined the independent relationships between workstress, teacher burnout and depressive symptoms, weknow of no study that has examined the chronic natureof teacher stress on burnout and depressive symptomsin a single model while controlling for a variety ofdemographic characteristics. Therefore, the purpose ofthis study was to test a conceptual model examiningthe three burnout subscales (emotional exhaustion,depersonalization and reduced personal accomplish-ment) as mediators of the association between chronicwork stress and depressive symptoms while controllingfor a variety of demographic characteristics. Based onprevious research, we hypothesized that after controllingfor demographic characteristics, (1) chronic work stresswould have a significant direct effect on depressivesymptoms and each subscale of burnout, (2) eachsubscale of burnout would have a direct effect ondepressive symptoms and (3) each subscale of burnoutwould mediate the relationship between chronic workstress and depressive symptoms.

Method

Participants and procedure

Participants

The study participants (n= 267) were primarily female(75%) with a mean age of 45 years, ranging from 23 to68 years. Demographically, 86% were White, 8%Hispanic or Latino, 3% African‐American and less than1% each of Native American and Asian American. Mostteachers had a bachelor’s degree (55%), although 43%held a master’s degree, and 2% had attained a doctoratedegree. The majority were high school teachers (75%),with the remaining teaching either middle school (13%)or elementary school (12%). Overall, the sample washighly experienced, with an average of 18 years ofteaching experience. Only 12% of the sample had lessthan 5 years of teaching experience; 69% had over10 years of experience, and 43% had over 20 years ofexperience. Further, 22% of the teachers had received a

422

prestigious Teaching Excellence Award from the univer-sity conducting the study.

Procedures

A cross‐sectional self‐report survey methodology wasused. Researchers obtained a convenience sample ofpublic school teachers who had received teaching awardsfrom an alumni association at a large research institutionin Texas. Since 1987, the alumni association hasrecognized public school teachers representing everyschool size, from every Texas geographic area, in everysubject matter and with degrees from any university inthe United States on an annual basis. Each year, schoolprincipals can nominate one teacher to represent theirschool. To be eligible for the award, teachers must be fulltime, have a minimum of 10 years of classroom teachingexperience and return to teaching the following year. Tenhigh school (grades 9–12) and two elementary or middleschool (grades 1–8) teachers are awarded each year;selection of the award‐winning teachers is made by apanel composed of former recipients, education faculty, afuture teacher and alumni leaders.

Subject recruitment for the current study includedsending a survey packet in the mail to all award‐winningteachers who were still teaching and had currentmailing addresses (n = 170). The packet contained acover letter explaining the purpose of the study, sixcopies of the survey and stamped envelopes for theconvenient return of the surveys to the researchers.The award‐winning teachers were directed to fill outone survey and distribute the other five packets tofellow teachers that they felt represented a variety ofteaching experience and ability. Included in each surveypacket was a small deck of inspirational quote cards(valued at $5) that provided a little incentive forcompleting the survey. To protect the anonymity ofeach participant, an additional stamped and addressedpostcard was included in each survey packet to allowthe potential participant to be entered into a draw forthe chance to receive an autographed university foot-ball (valued at $100), a university T‐shirt (valued at $19)or a gift card (valued at $15) for completing the survey.One incentive was awarded for every 10 surveysreceived. All responses were anonymous and completedat a time and location of the teacher’s convenience.The final sample consisted of n= 267 teachers,representing a moderate response rate of 26% (Alreck& Settle, 2004).

Instruments

The survey included demographic characteristics andself‐report measures of perceived chronic work stress,teacher burnout and depressive symptoms. Each ofthese measures is described below.

Demographics

Teachers reported their age, gender, ethnicity, educa-tion level, grade level taught, years of teaching experience

Stress and Health 27: 420–429 (2011) © 2011 John Wiley & Sons, Ltd.

M. A. Steinhardt et al. Stress, Burnout and Depressive Symptoms

and whether they had received a teaching award from theuniversity conducting the study. Because these demo-graphic characteristics may be related to burnout, thecurrent study statistically controlled for them in theestimation of the impacts of chronic work stress andteacher burnout on depressive symptoms. Because of thehigh correlation between age and years of teachingexperience (r=0.81), age was omitted as a controlvariable.

Perceived chronic work stress

The unrelenting nature of chronic stressors createsmore detrimental effects on physical and mental healththan acute stressors (Lepore, 1998; McEwen, 2002).Although previous studies have examined the sourcesof stress in teachers (Kyriacou, 2001; Maslach et al.,2001; Montgomery & Rupp, 2005), none have used acomprehensive chronic stress measure that assessesexposure, appraisal and duration of each stressor. Ourperceived chronic work stress measure is a modifiedversion of the Teacher Stress Inventory (Fimian, 1984),adapted to include additional potential stressfulsituations as well as an assessment of exposure,appraisal and duration of each stressful situation. Themodified measure assessed 41 commonly occurringstressful situations and asked whether the teacher hadexperienced each situation within the past year (codedas 1) or had not experienced the situation (0). If astressful situation was experienced, the teacher alsoindicated how negative, undesirable or upsetting thesituation was on a Likert scale from 0 (not at all) to 4(extremely). Finally, teachers reported the number ofmonths that each stressful situation had been experi-enced during the current academic year, ranging fromthe beginning of school in September (1 month) towhen the survey was completed in March (7 months).Four sample items are included in Table I. To scoreeach item, the exposure rating was multiplied by boththe appraisal and duration ratings. The chronic workstress score was then calculated by averaging all theitems. The internal consistency of the modified

Table I. Sample items for the perceived chronic work stress measure

Experienced? If Yes, ho

No Yes Not at all A li

1. Not enough time to

get things done

No Yes 0 1

2. Too much paperwork No Yes 0 1

3. Administration not

supportive enough

No Yes 0 1

4. Attending to students who

are poorly motivated

No Yes 0 1

Stress and Health 27: 420–429 (2011) © 2011 John Wiley & Sons, Ltd.

measure was strong (Cronbach’s alpha = 0.93). A copyof the survey may be obtained by contacting the firstauthor. To examine the concurrent validity of ourchronic work stress measure, a four‐item version of thePerceived Stress Scale (PSS; Cohen, Kamarck, &Mermelstein, 1983) was administered; our measurecorrelated well (r = 0.54, p< 0.001) with the PSS.

Teacher burnout

The Maslach Burnout Inventory‐Educators Survey(MBI‐ES) assessed burnout among the teachers(Maslach et al., 1996). The MBI‐ES measures all threesubscales of burnout: emotional exhaustion, deperson-alization and reduced personal accomplishment. Theteachers were asked how often they experiencedemotional exhaustion (nine items), depersonalization(five items) and reduced personal accomplishment(eight items) on a Likert scale ranging from 0 (never)to 6 (every day). Sample subscale items included ‘I feelemotionally drained from my work’, ‘I don’t really carewhat happens to some students’ and ‘I deal veryeffectively with the problems of my students.’ Thescore for each subscale was the mean across the itemsfor that particular subscale, with higher scoresindicating greater burnout. Reliability was strong foremotional exhaustion (α= 0.90) and acceptable fordepersonalization (α= 0.75) and reduced personalaccomplishment (α= 0.76).

Depressive symptoms

The Center for Epidemiological Studies DepressionScale (CES‐D) measured depressive symptoms such asa depressed mood, feelings of guilt, worthlessness,helplessness, loss of appetite and restless sleep (Radloff,1977). The teachers responded to 20 items that askedhow often they felt or behaved according to each itemduring the past week. Sample items included ‘I wasbothered by things that usually don’t bother me’ and‘I felt depressed.’ Responses ranged from 0 (rarely ornone of the time, less than 1 day) to 3 (most or all ofthe time, 5–7 days) and were averaged across items for

w negative, undesirable or upsetting? If Yes, how long?

ttle Moderately Very Extremely How many months

experienced?

<1 to 7 months

2 3 4 ____months

2 3 4 ____months

2 3 4 ____months

2 3 4 ____months

423

Stress, Burnout and Depressive Symptoms M. A. Steinhardt et al.

a total score. A CES‐D mean score of 0.8 or greater isconsidered a moderately severe level of depressivesymptoms (Radloff, 1977). The reliability of the CES‐Dfor the current study was strong (α= 0.92).

Analysis

We performed a path analysis to test the relationshipsdepicted in Figure 1 while controlling for demographicvariables (gender, minority status, years of teachingexperience, whether the teacher was an awardwinner andwhether the teacher was employed at a high school versusan elementary or middle school). In Figure 1, therelationship between chronic work stress and depressivesymptoms is mediated by three different subscales ofburnout. Thus, themodel tests several types of effects: thedirect effects of chronic work stress on depressivesymptoms and each of the three subscales of burnout,and of the three subscales of burnout on depressivesymptoms; the specific indirect effect of chronic workstress on depressive symptoms through each subscale ofburnout (or the unique mediating effect of eachcomponent of burnout controlling for the other twocomponents of burnout as well as the demographiccontrols); the total indirect effect of chronic work stress ondepressive symptoms (or the sum of each of the threespecific indirect effects); and the total effect of work stresson depressive symptoms (or the sum of the direct andtotal indirect effect).

Social science researchers have traditionally esti-mated a given indirect effect through a series ofregression models (Baron & Kenny, 1986) and testedits significance using the well‐known Sobel (1982,1986) test or other tests relying on the standard normal

Note: *p < 0.05, ** p < 0.01, and *** p < 0.001 † Controlling for Years taught, Award, High school, Female, and Minority

Chronic WorkStress

Emotionexhaustio

Depersonali

Reduced peaccomplish

0.61***

0.38***

0.28***

Figure 1 Path analysis of the relationship between chronic work stress a

424

distribution (e.g. MacKinnon, Lockwood, Hoffman,West, & Sheets, 2002). However, the assumption of anormal sampling distribution for the indirect effect istypically violated with relatively small sample sizes suchas the one in this study. Accordingly, many methodol-ogists now recommend a bootstrapping approach,which does not rely on a standard normal distribution,to calculate the standard errors of indirect paths (e.g.Briggs, 2006; Williams & MacKinnon, 2008).

In simple multiple‐mediator models such as the oneunder study, it is sensible to assume a saturated model,which then allows the researcher to test the uniqueeffects of each mediating variable (Preacher & Hayes,2008). Saturated mediation models including onlyobserved variables may be estimated through either thetraditional regression‐based approach or a structuralequation modelling package; provided that standarderrors are appropriately calculated via bootstrapping,the two approaches yield the same results. Accord-ingly, we conducted our analysis via the traditionalregression‐based approach with bootstrapped standarderrors for the indirect effects, using an SPSS macro(SPSS Inc., Chicago, IL, USA) explicitly designed toestimate multiple‐mediation models (Preacher &Hayes, 2008). For indirect paths, this analysis producespoint estimates and three varieties (percentile, bias‐corrected, and bias‐corrected and accelerated) ofbootstrapped 95% confidence intervals; the analysisalso allows the researcher to conduct pairwisecomparisons of the indirect effects to determinewhether a particular mediator has a significantlystronger unique indirect effect in comparison toanother mediator. All variables were standardized

Model R-squared = 0.43Indirect effect of stress ondepressive symptoms = 0.32***Total effect of stress ondepressive symptoms = 0.44***

aln

zation

Depressive† Symptoms

rsonalment

0.38***

0.13*

0.13*

0.11*

nd depressive symptoms and the mediating role of teacher burnout

Stress and Health 27: 420–429 (2011) © 2011 John Wiley & Sons, Ltd.

M. A. Steinhardt et al. Stress, Burnout and Depressive Symptoms

prior to inclusion in the analysis, yielding standardizedcoefficients for all effects (reported below and inFigure 1). The mediation analysis approach requiredus to drop seven teachers with partially missing datafrom the analysis, resulting in a final sample size ofn= 260. Bivariate relationships, however, were almostidentical between this complete‐data subsample andthe full sample of n= 267.

Results

Descriptive analysis

For all data analyses, years of teaching experience wasretained as a continuous variable. Multiple‐categorydemographic variables were collapsed into binary vari-ables in order to construct appropriately sized groups foranalysis: award (0= did not receive a teaching excellenceaward, 1 = received an award), high school (0= teachesprimary or middle school, 1 = teaches high school),female (0=male, 1 = female), minority (0=nonminority,1=minority) and advanced degree (0= bachelor’s degree,1=master’s or doctorate).

Table II displays the means, standard deviationsand correlations for all study variables. Each of theburnout subscales was significantly correlated with theother subscales, the highest correlation being betweenthe emotional exhaustion and depersonalizationsubscales. Chronic work stress was strongly correlatedwith both emotional exhaustion and depressive symp-toms. Mean scores for teacher burnout and depressivesymptoms were slightly higher than those in similarstudies on teacher burnout (Bakker et al., 2000;Brenninkmeyer, Van Yperen, & Buunk, 2001; Grayson& Alvarez, 2008).

The strongest association among the control variableswas between winning an award and years of teachingexperience; also, teachers with an advanced degree had

Table II. Means, standard deviations (SD) and correlations for all varia

Variable M SD CWS Burnout

EE DP

Chronic work stress (CWS) 6.55 4.63 —

Emotional exhaustion (EE) 2.64 1.22 0.63*** —

Depersonalization (DP) 1.31 1.14 0.38*** 0.65*** —

Reduced personal

accomplishment (RPA)

1.03 0.81 0.29*** 0.43*** 0.46***

Depressive symptoms (DS) 0.53 0.45 0.47*** 0.61*** 0.49***

Years taught — — −0.03 −0.16** −0.15*Award — — −0.10 −0.18** −0.20**High school — — 0.07 0.08 0.11

Female — — 0.16* 0.14* −0.03Minority — — 0.04 −0.03 −0.04Advanced degree — — −0.05 −0.12 0.00

*p< 0.05, **p< 0.01, ***p< 0.001.

Stress and Health 27: 420–429 (2011) © 2011 John Wiley & Sons, Ltd.

more years of teaching experience andweremore likely tohave won an award. Control variables were not stronglyrelated to chronic work stress, burnout or depressivesymptoms; although there were some significant correla-tions, they were small in size. More experienced andaward‐winning teachers reported less burnout anddepressive symptoms, teachers in high school reportedgreater depressive symptoms and reduced personalaccomplishment and females reported greater chronicwork stress and emotional exhaustion (Table II).

Path analysis

In terms of the standardized direct effects shown inFigure 1, teachers who experienced high degrees of stresswere more burned out in terms of emotional exhaustion(b = 0.61, p < 0.001), depersonalization (b = 0.38,p < 0.001) and reduced personal accomplishment(b= 0.28, p< 0.001). Of the three components ofburnout, emotional exhaustion was moderately andpositively related to depressive symptoms (b=0.38,p < 0.001), whereas depersonalization (b = 0.13,p < 0.05) and reduced personal accomplishment(b=0.11, p<0.05) had only small positive relationshipswith depressive symptoms. After controlling for the threecomponents of burnout, as well as the set of demographiccontrols, the direct relationship between chronic workstress and depressive symptoms was small (b=0.13,p<0.05). Variables included as controls (not shown inFigure 1) had small to negligible relationships withdepressive symptoms; only high school had a significantrelationship (b=0.10, p<0.05), indicating that highschool teachers experienced slightly more depressivesymptoms than primary and middle school teachers.

In terms of specific indirect effects, the indirecteffect of chronic work stress on depressive symptomsvia emotional exhaustion was moderate and positive(b= 0.23), with all three types of confidence intervals

bles (n=267)

DS Control

RPA Years

taught

Award High

school

Female Minority

—

0.39*** —

−0.15* −0.20** —

−0.26** −0.16** 0.44** —

0.13* 0.15* 0.00 −0.05 —

−0.05 0.09 −0.03 0.07 −0.11 —

−0.02 0.05 −0.10 −0.09 −0.13* −0.18** —

0.02 −0.11 0.29** 0.21** 0.11 −0.12 0.01

425

Stress, Burnout and Depressive Symptoms M. A. Steinhardt et al.

indicating a significant effect. However, indirect effectsvia depersonalization (b= 0.05) and reduced personalaccomplishment (b= 0.03) were negligible, with allthree types of confidence intervals indicating lack ofsignificance. Pairwise comparisons of the three indirecteffects indicated that the mediating effect of emotionalexhaustion was significantly stronger than the mediat-ing effects of either depersonalization or reducedpersonal accomplishment. Taking together the directand indirect effects via burnout, the total effect of stresson depressive symptoms was strong and positive(b= 0.44, p< 0.001). The overall model accounted for43% of the total variance in depressive symptoms.

DiscussionThe literature suggests that high levels of work stressprecipitate depression (Melchior et al., 2007; Wang,2005). The current cross‐sectional study provides furthersupport for this relationship such that chronic workstress has a direct relationship with depressive symptomsand an indirect effect on depressive symptoms viaburnout, in a sample of highly experienced public schoolteachers. Although we expected that each of the threecomponents of burnout would carry a unique mediatingeffect, only emotional exhaustion served as a significantmediator between chronic work stress and depressivesymptoms after controlling for the other two compo-nents of burnout. The study also supports therelationship between work stress and teacher burnout.Of the three components of burnout, chronic workstress was most related to higher levels of emotionalexhaustion, followed by depersonalization and reducedpersonal accomplishment.

The connection between work stress and burnoutsuggests that the alleviation of stressors or the adoptionof effective coping strategies could aid in reducing theprevalence of burnout in the teacher population. Forexample, organizational stressors have repeatedly beenfound to determine levels of emotional exhaustion,especially when support is lacking (Leiter, 1991a,1991b), suggesting that the coping strategy of enhancedsupport could moderate the negative impact of chronicstress on emotional exhaustion. The fact that emo-tional exhaustion is defined as the stress component ofthe burnout syndrome helps explain why teacherstypically exhibit higher levels of emotional exhaustionthan the other two components (Maslach et al., 2001;Schaufeli & Enzmann, 1998).

The positive relationships between each of the threecomponents of burnout and depression support theassertion that burnout may be a phase in the develop-ment of depression, with the probability of a depressivedisorder rising with the level of burnout experienced(Ahola et al., 2005; Wang, 2005). Of the threecomponents of burnout, emotional exhaustion was moststrongly connected to depressive symptoms, withdepersonalization and reduced personal accomplishmentrelated to a much lesser degree. This finding aligns with

426

the work of Bakker and Schaufeli (2000), in whichemotional exhaustion and depression share symptoms,such as decreased energy, low motivation and negativeattitudes (Schaufeli & Enzmann, 1998).

One potential mechanism by which emotionalexhaustion and depressionmay be linked is via individualpersonality factors such as neuroticism. Depression ispart of the neurotic personality (McCrae & John, 1992),and emotional exhaustion and neuroticism are correlated(Cano‐Garcia, Padilla‐Munoz, & Carrasco‐Ortiz, 2005;Piedmont, 1993; Tönjes, Dickhäuser, & Kröner, 2008).Neurotic individuals expressmore negative emotions andgreater stress reactions, leaving themmore susceptible toboth burnout and psychopathology (Watson, Clark,& Harkness, 1994). A meta‐analytic study (Pfenning &Husch, 1994 in Schaufeli & Enzmann, 1998) found thatanxiety, a factor of neuroticism, correlated most highlywith emotional exhaustion, followed by depersonaliza-tion and reduced personal accomplishment. Similarly,Deary et al. (1996) found that neuroticism correlatedmost strongly with emotional exhaustion, followed bydepersonalization and reduced personal accomplish-ment. Neuroticism also may moderate the relationshipbetween daily hassles and emotional exhaustion, withneurotic police officers experiencing higher levels ofemotional exhaustion than their less neurotic colleagues(Hills & Norvell, 1991). Given the evidence linkingneuroticism to both emotional exhaustion and depres-sion, it is possible that neuroticism is an underlying causeof both.

Although chronic work stress was related to eachcomponent of burnout and each component of burnoutwas in turn related to depressive symptoms, onlyemotional exhaustion served as a significant mediatorbetween chronic work stress and depressive symptoms.This finding strengthens the notion that emotionalexhaustion is the central component that makes burnout‘real’ to the teacher and serves as the emotional vectortranslating feelings of stress into depression. It isimportant to note, however, that emotional exhaustionalone does not embody the burnout syndrome (Maslach,1993). Rather, burnout is a self‐perpetuating syndromein which all three components must be present. Whenstress accumulates to an unbearable level, teachers beginto feel emotionally exhausted and withdraw themselvesfrom their work. Depersonalizing as a copingmechanismenables the teacher to continue in the profession, albeitfunctioning at a minimum (Holt, Fine, & Tollefson,1987). Given that teachers list their sense of achievementas a primary reward for teaching (Engelking, 1986),functioning at a minimum may in turn exacerbatewithdrawal and a sense of exhaustion. Because this self‐perpetuating process is difficult to stop once it begins,the current study’s finding points to the need to addressthe prolonged exposure to stressful situations, frequentlyaccompanied by insufficient recovery, that teachersoften experience and that initially lead to emotionalexhaustion.

Stress and Health 27: 420–429 (2011) © 2011 John Wiley & Sons, Ltd.

M. A. Steinhardt et al. Stress, Burnout and Depressive Symptoms

From a practical standpoint, interventions targetedto reduce burnout and the depressive symptoms ofteachers might be most effective if designed to focusprimarily on stress prevention and stress‐relatedgrowth and secondarily on stress management. Aproactive stress prevention approach would assistteachers in building personal resources and broadeningtheir coping skills in order to diminish the connectionbetween chronic stress and emotional exhaustion. Inthis approach, personal and environmental resources(e.g. locus of control, self‐esteem, coping strategies,positivity, support from colleagues and administrators)would be identified, strengthened and used to helpreduce initial perceptions of stress, build resilience(O’Leary & Ickovics, 1995) and enhance job engagement(Maslach & Leiter, 2008). Further, it has been suggestedthat growth occurs as a result of stressful situations whenindividuals have a sufficient foundation of personalresources yet sufficient distress to merit an examinationof current beliefs and feelings in the context of paststressors and adaptations (Carver, 1998). As such,depressive symptoms may serve as a catalyst to disruptand then help constructively reshape the basic beliefsabout oneself and one’s job (Dolbier, Smith Jaggars,& Steinhardt, 2010).

In addition to proactive interventions that focus onstress prevention and stress growth, secondary interven-tions could help teachers deal more effectively withprolonged exposure to chronic emotional and interper-sonal stressors, accompanied by insufficient recovery,burnout and decreased health (Wood & McCarthy,2002). Such approaches include more traditional reactivestress management programmes, which help burned‐outteachers cope more effectively in an attempt to bufferagainst negative health outcomes, including depressivesymptoms. Both teachers and administrators should bemade aware of the characteristics of burnout and how toidentify burnout in themselves and others. If support canbe provided to teachers experiencing burnout, it ispossible that the burnout can be managed prior to thedevelopment of depression or other negative healthoutcomes.

Interventions designed to reduce the syndrome ofburnout should have a primary focus on the emotionalexhaustion component, given that it is the stresscomponent of burnout and a significantly strongermediator of depressive symptoms than either deperson-alization or reduced personal accomplishment. Indeed,programmes that have been shown to decrease burnoutin medical residents (McCue & Sachs, 1991), health careworkers (Hyman, 1993) and members of humanresource management teams (Golembiewski, Hilles,& Daly, 1987) reported a reduction in the subcomponentemotional exhaustion. These programmes addressedburnout as a syndrome, emphasizing practices that helpindividuals to de‐stress as well as to successfully managestressful situations by harnessing effective coping strat-egies and building social support networks. Undoubtedly,

Stress and Health 27: 420–429 (2011) © 2011 John Wiley & Sons, Ltd.

individuals who utilize more problem‐focused copingskills and receive more social support from theirsupervisors and colleagues are less likely to experiencehigh levels of emotional exhaustion (Ito & Brotheridge,2003; McCarthy, Kissen, Yadley, Wood, & Lambert,2006; Wilk & Moynihan, 2005).

With regard to demographics, previous studies haveshown that age, gender, ethnicity, education level, gradelevel taught and level of work experience may affectburnout and depression. As such, the present studycontrolled for these six factors in the estimation of theimpacts of chronic work stress and burnout ondepression. Importantly, teachers with more teachingexperience or who had received a teaching awardreported stress levels similar to the entire sample yetwere more likely to report lower levels of depressivesymptoms, suggesting that they were handling the stressmore effectively. Finally, teachers that taught at the highschool level were found to have more depressivesymptoms than elementary and middle school teachers.Thisfinding alignswith that of Beer andBeer (1992), whofound that high school teachers experienced more stress,burnout and depression than grade school teachers.Future qualitative research should help determine thereason for the heightened depressive symptoms that highschool teachers experience.

The implications for this study should be viewed inlight of several limitations. Firstly, the study was cross‐sectional in nature, and therefore causal inference andtemporal precedence cannot be determined. It is possiblethat other variables account for some of the observedrelationships or that depressive symptoms reciprocallycontribute to chronic work stress and teacher burnout.Future research should use a prospective design to furtherexamine the nature of the predictive relationshipsbetween chronic work stress, burnout and depressivesymptoms across time. Likewise, a prospective designwould allow for assessment of the stress levels teachersexperience over the course of the academic year, as thereis some evidence that it may vary (Travers & Cooper,1996). Better procedures for assessing the chronic natureof stressors (e.g. exposure, appraised as stressful,unrelenting) are needed. Secondly, the use of self‐reportsurvey data has inherent limitations, such as the potentialfor untruthful or inaccurate responses due to lack of self‐awareness. Although the use of anonymous surveysenhanced the strength of the study, future researchshould include objective observations of chronic workstressors, burnout and verifiable depression. The study’sresponse rate of 26% was only moderate, and informa-tion is not available about the non‐respondents.Additionally, the rather selective sampling procedure ofasking one teacher to select five other teachers tocomplete the survey resulted in a sample that did notrepresent any definable population. Our sample washighly experienced with an average of 18 years ofteaching, and 22% had received a prestigious teachingexcellence award; therefore, the results may not be

427

Stress, Burnout and Depressive Symptoms M. A. Steinhardt et al.

generalized to a sample of novice teachers. Finally, inaddition to assessing predictors of job burnout anddepressive symptoms, further research should also assessthe predictors of job engagement and positive healthoutcomes. Although a necessary and important perspec-tive, a deficit‐oriented approach provides a limited view

428

of individuals and the possible resources that may assistthem in adapting to stressful situations and lead topositive outcomes (Tedeschi & Kilmer, 2005). Never-theless, the study findings have important implicationsfor addressing burnout and depressive symptoms in theteaching profession.

REFERENCES

Ahola, K., Honkonen, T., Isometsa, E., Kalimo, R.,

Nykyri, E., Aromaa, A., & Lonnqvist, J. (2005). The

relationship between job‐related burnout and depres-

sive disorders‐results from the Finnish Health 2000

Study. Journal of Affective Disorders, 88, 55–62. doi:

10.1016/j.jad.2005.06.004

Alreck, P.L., & Settle, R.B. (2004). The survey research

handbook (3rd edn). New York, NY: McGraw‐Hill.

Bakker, A.B., & Schaufeli, W.B. (2000). Burnout conta-

gion processes among teachers. Journal of Applied Social

Psychology, 30(11), 2289–2308.

Bakker, A.B., Schaufeli, W.B., Demerouti, E., Janssen,

P.P.M., Van der Hulst, R., & Brouwer, J. (2000).

Using equity theory to examine the difference

between burnout and depression. Anxiety, Stress,

and Coping, 13, 247–268.

Baron, R.M., & Kenny, D.A. (1986). The moderator–

mediator distinction in social psychological research:

Conceptual, strategic and statistical considerations.

Journal of Personality and Social Psychology, 51, 1173–1182.

Beer, J., & Beer, J. (1992). Burnout and stress,

depression, and self‐esteem of teachers. Psychological

Reports, 71, 1331–1336.

Blase, J.J. (1982). A social–psychological grounded

theory of teacher stress and burnout. Educational

Administration Quarterly, 18, 93–113.

Blase, J.J. (1986). A qualitative analysis of sources of teach-

er stress: Consequences for performance. American

Educational Research Journal, 23(1), 13–40.

Blau, P.M. (1964). Exchange and power in social life.

New York, NY: John Wiley & Sons.

Borg, M.G., & Riding, R.J. (1991). Occupational stress

and satisfaction in teaching. British Educational

Research Journal, 17(3), 263–281.

Brenninkmeyer, V., Van Yperen, N.W., & Buunk, B.P.

(2001). Burnout and depression are not identical twins:

Is decline of superiority a distinguishing feature?

Personality and Individual Differences, 30, 873–880.

Briggs, N. (2006). Estimation of the standard error and

confidence interval of the indirect effect in multiple

mediator models (Unpublished PhD dissertation).

The Ohio State University, OH, USA.

Cano‐Garcia, F.J., Padilla‐Munoz, E.A., & Carrasco‐Ortiz,

M.A. (2005). Personality and contextual variables in

teacher burnout. Personality and Individual Differences,

38, 929–940. doi: 10.1016/j.paid.2004.06.018

Carver, C.S. (1998). Resilience and thriving: Issues,

models, and linkages. Journal of Social Issues, 54(2),

245–266.

Chan, D.W. (2003). Hardiness and its role in the stress‐

burnout relationship among prospective Chinese

teachers inHongKong.Teaching andTeacher Education,

19, 381–395. doi: 10.1016/j.tate.2006.04.005

Chang, M. (2009). An appraisal perspective of teacher

burnout: Examining the emotional work of teachers,

Educational Psychology Review, 21, 193–218.

Cherniss, C. (1980). Staff burnout: Job stress in the

human services. Beverly Hills, CA: Sage.

Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A

global measure of perceived stress. Journal of Health and

Social Behavior, 24(4), 385–396. doi: 10.2307/2136404

Deary, I.J., Blenkin, H., Agius, R.M., Endler, N.S.,

Zealley, H., & Wood, R. (1996). Models of job‐

related stress and personal achievement among

consultant doctors. British Journal of Psychology, 87

(1), 3–29.

Demir, A., Ulusoy, M., & Ulusoy, M.F. (2003). Investi-

gation of factors influencing burnout levels in the

professional and private lives of nurses. International

Journal of Nursing Studies, 40(8), 807–827. doi: 10.1016/

S0020‐7489(03)00077‐4

Dolbier, C.L., Smith Jaggars, S., & Steinhardt, M.A.

(2010). Stress‐related growth: Pre‐intervention cor-

relates and change following a resilience intervention.

Stress and Health, 26, 135–147. doi:10.1002/smi.1275

Dyrbye, L.N., Thomas, M.R., Eacker, A., Harper, W.,

Stanford, M., Power, D.V., . . . Shanafelt, T.D.

(2007). Race, ethnicity, and medical student well‐

being in the United States. Archives of Internal

Medicine, 167(19), 2103–2109.

Embich, J.L. (2001). The relationship of secondary special

education teachers’ roles and factors that lead to

professional burnout. Teacher Education and Special

Education, 24(1), 58–69.

Engelking, J.L. (1986). Teacher job satisfaction and

dissatisfaction. Spectrum, 4, 33–38.

Evers, W.J.G., Tomic, W., & Brouwers, A. (2004). Burnout

among teachers: Students’ and teachers’ perceptions

compared. School Psychology International, 25(2),

131–148.

Farber, B.A. (1991). Crisis in education: Stress and

burnout in the American teacher. San Francisco, CA:

Jossey‐Bass.

Fimian, M.J. (1984). The development of an instrument

to measure occupational stress in teachers: The

Teacher Stress Inventory. Journal of Occupational

Psychology, 57(4), 277–293.

Friedman, I.A. (1991). High‐ and low‐burnout schools:

School culture aspects of teacher burnout. Journal of

Educational Research, 84(6), 325–333.

Friesen, D., & Sarros, J.C. (1989). Sources of burnout

among educators. Journal of Organizational Behavior,

10(2), 179–188.

Golembiewski, R.T., Hilles, R., & Daly, R. (1987).

Some effects of multiple OD interventions on

burnout and work site features. The Journal of

Applied Behavioral Science, 23(3), 295–313. doi:

10.1177/002188638702300302

Stress and Health 27: 4

Grayson, J.L., & Alvarez, H.K. (2008). School climate

factors relating to teacher burnout: A mediator model.

Teaching and Teacher Education, 24, 1349–1363. doi:

10.1016/j.tate.2007.06.005

Hakanen, J.J., Bakker, A.B., & Schaufeli, W.B. (2006).

Burnout and work engagement among teachers.

Journal of School Psychology, 43, 495–513.

Hammen, C., & deMayo, R. (1982). Cognitive corre-

lates for teacher stress and depressive symptoms:

Implications for attributional models of depression.

Journal of Abnormal Psychology, 91(2), 96–101.

Hastings, R.P., & Bham, M.S. (2003). The relationship

between student behaviour patterns and teacher

burnout. School Psychology International, 24(1),

115–127.

Hills, H., & Norvell, N. (1991). An examination of

hardiness and neuroticism as potential moderators of

stress outcomes. Behavioral Medicine, 17(1), 31–38.

Holt, P., Fine, M.J., & Tollefson, N. (1987). Mediating

stress: Survival of the hardy. Psychology in the Schools,

24, 51–58.

Hyman, R.B. (1993). Evaluation of an intervention for

staff in a long‐term care facility using a retrospective

pretest design. Evaluation & the Heath Professions,

16(2), 212–224. doi: 10.1177/016327879301600206

Ingersoll, R.M., & Smith, T.M. (2003). The wrong

solution to the teacher shortage. Keeping Good

Teachers, 60(8), 30–33.

Ito, J.K., & Brotheridge, C.M. (2003). Resources, coping

strategies, and emotional exhaustion: A conservation

of resources perspective. Journal of Vocational

Behavior, 63(3), 490–509. doi: 10.1016/S0001‐

8791(02)00033‐7

Kyriacou, C. (2001). Teacher stress: Directions for

future research. Educational Review, 53(1), 27–35.

doi: 10.1080/00131910120033628

Kyriacou, C., & Sutcliffe, J. (1978). A model of teacher

stress. Educational Studies, 4(1), 1–6. doi: 10.1080/

0305569780040101

Lackritz, J.R. (2004). Exploring burnout among uni-

versity faculty: Incidence, performance, and demo-

graphic issues. Teaching and Teacher Education,

20(7), 713–729. doi: 10.1016/j.tate.2004.07.002

Lazarus, R.S. (1966). Psychological stress and the coping

process. New York, NY: McGraw‐Hill.

Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal,

and coping. New York, NY: Springer.

Leiter, M.P. (1991a). Coping patterns as predictors of

burnout: The function of control and escapist coping

patterns. Journal of Organizational Behavior, 12(2),

123–144. doi: 10.1002/job.4030120205

Leiter, M.P. (1991b). The dream denied: Professional

burnout and the constraints of service organizations.

Canadian Psychology, 32(4), 547–558. doi: 10.1037/

h0079040

20–429 (2011) © 2011 John Wiley & Sons, Ltd.

M. A. Steinhardt et al. Stress, Burnout and Depressive Symptoms

Lepore, S.J. (1998). Problems and prospects of the social

support‐reactivity hypothesis. Annals of Behavioral

Medicine, 20(4), 257–269.

MacKinnon, D.P., Lockwood, C.M., Hoffman, J.M.,

West, S.G., & Sheets, V. (2002). A comparison of

methods to test mediation and other intervening

variable effects. Psychological Methods, 7, 83–104.

Manthei, R., Gilmore, A., Tuck, B., & Adair, V. (1996).

Teacher stress in intermediate schools. Educational

Research, 38(1), 3–19.

Maslach, C. (1993). Burnout: A multidimensional per-

spective. In W.B. Schaufeli, C. Maslach, & T. Marek

(Eds.), Professional burnout: Recent development in the-

ory and research (pp. 19–32).Washington, DC: Taylor &

Francis.

Maslach, C., & Leiter, M.P. (2008). Early predictors of

job burnout and engagement. The Journal of Applied

Psychology, 93(3), 498–512. doi: 10.1037/0021‐

9010.93.3.498

Maslach, C., Jackson, S.E., & Leiter, M.P. (1996). Maslach

Burnout Inventory manual. Palo Alto, CA: Consulting

Psychologists Press.

Maslach, C., Schaufeli, W.B., & Leiter, M.P. (2001). Job

burnout. Annual Review of Psychology, 52, 397–422.

McCarthy, C.J., Kissen, D., Yadley, L., Wood, T., & Lambert,

R.G. (2006). Relationship of teachers’ preventive coping

resources to burnout symptoms. In R. Lambert, & C.

McCarthy (Eds.), Understanding teacher stress in an age

of accountability (pp. 179–196). Greenwich, CT: Informa-

tion Age Publishing.

McCarty, W.P., Zhao, J.S., & Garland, B.E. (2007).

Occupational stress and burnout between male and

female police officers: Are there any gender differ-

ences? Policing: An International Journal of Police

Strategies & Management, 30(4), 672–691. doi:

10.1108/13639510710833938

McCrae, R.R., & John, O.P. (1992). An introduction to

the Five‐Factor Model and its applications. Journal of

Personality, 60(2), 175–215. doi: 10.1111/j.1467‐

6494.1992.tb00970.x

McCue, J.D., & Sachs, C.L. (1991). A stress manage-

ment workshop improves residents’ coping skills.

Archives of Internal Medicine, 151(11), 2273–2277.

McEwen, B. (2002). The end of stress as we know it.

Washington, DC: National Academic Press.

Stress and Health 27: 420–429 (2011) © 2011 John W

Melchior, M., Caspi, A., Milne, B.J., Danese, A., Poulton,

R., & Moffitt, T.E. (2007). Work stress precipitates

depression and anxiety in young, working women and

men. Psychological Medicine, 37(8), 1119–1129.

Montgomery, C., & Rupp, A.A. (2005). A meta‐analysis

for exploring the diverse causes and effects of stress

in teachers. Canadian Journal of Education, 28(3),

458–486.

O’Leary, V.E., & Ickovics, J.R. (1995). Resilience and

thriving in response to a challenge: An opportunity

for a paradigm shift in women’s health. Women’s

Health, 1(2), 121–142.

Piedmont, R.L. (1993). A longitudinal analysis of burnout

in the health care setting: The role of personal

dispositions. Journal of Personality Assessment, 61(3),

457–473.

Preacher, K.J., & Hayes, A.F. (2008). Asymptotic and

resampling strategies for assessing and comparing

indirect effects in multiple mediator models. Behavior

Research Methods, 40(3), 879–891. doi: 10.3758/

BRM.40.3.879

Radloff, L.S. (1977). The CES‐D Scale: A self‐report

depression scale for research in the general population.

Applied Psychological Measurement, 1(3), 385–401.

doi: 10.1177/014662167700100306

Russell, D.W., Altmaier, E., & Van Velzen, D. (1987).

Job‐related stress, social support, and burnout among

classroom teachers. The Journal of Applied Psychology,

72(2), 269–274.

Sakharov, M., & Farber, B.A. (1983). A critical study of

burnout in teachers. In B.A. Farber (Ed.), Stress and

burnout in the human service professions (pp. 65–81).

New York, NY: Pergamon Press.

Schaufeli, W.B., & Enzmann, D. (1998). The burnout

companion to study and practice: A critical analysis.

Philadelphia, PA: Taylor & Francis, Ltd.

Schaufeli, W.B., & Greenglass, E.R. (2001). Introduction

to special issue on burnout and health. Psychology &

Health, 16, 501–510.

Schwarzer, R., & Hallum, S. (2008). Perceived teacher

self‐efficacy as a predictor of job stress and burnout:

Mediation analyses. Applied Psychology, 57, 152–171.

doi: 10.1111,j.1464‐0597.2008.00359.x

Sobel, M.E. (1982). Asymptotic confidence intervals for

indirect effects in structural equations models. In

iley & Sons, Ltd.

S. Leinhart (Ed.), Sociological methodology (pp. 290–312).

San Francisco, CA: Jossey‐Bass.

Sobel, M.E. (1986). Some new results on indirect effects

and their standard errors in covariance structure

models. In N. Tuma (Ed.), Sociological methodology

(pp. 159–186). Washington, DC: American Socio-

logical Association.

Tedeschi, R.G., & Kilmer, R.P. (2005). Assessing

strengths, resilience, and growth to guide clinical

interventions. Psychological Research and Practice, 36,

230–237.

Tönjes, B., Dickhäuser, O., & Kröner, S. (2008). Goal

orientation and lack of achievement in teachers. German

Journal of Educational Psychology, 2, 151–160.

Travers, C.J., & Cooper, C.L. (1996). Teachers under

pressure: Stress in the teaching profession. London:

Routledge.

Wang, J. (2005). Work stress as a risk factor for major

depressive episode(s). Psychological Medicine, 35,

865–871. doi: 10.1017/S0033291704003241

Watson, D., Clark, L.A., & Harkness, A.R. (1994).

Structures of personality and their relevance to

psychopathology. Journal of Abnormal Psychology,

103(1), 18–31.

Weber, A., Weltle, D., & Lederer, P. (2002). Zur

Problematik krankheitsbedingter. Frühpensionierungen

von Gymnasiallehrkräften [Illness related early pension-

ing of high school teachers]. Versicherungsmedizin,

54(2), 75–83.

Wilk, S.L., & Moynihan, L.M. (2005). Display rule

“regulators”: The relationship between supervisors

and worker emotional exhaustion. The Journal of

Applied Psychology, 90(5), 917–927.

Williams, J., & MacKinnon, D.P. (2008). Resampling

and distribution of the product methods for testing

indirect effects in complex models. Structural

Equation Modeling, 15, 23–51.

Wood, T., & McCarthy, C.J. (2002). Understanding and

preventing teacher burnout. Washington, DC: ERIC

Clearinghouse on Teaching and Teacher Education.

(ERIC Document Reproduction Service No.

ED4777).

WHO (1992). The ICD‐10 Classification of Mental and

Behavioral Disorders: Clinical descriptions and diag-

nostic guidelines. Geneva: World Health Organization.

429