h i g h l i g h t s
Interpersonal relationships are differentially related to the burnout dimensions.
Relations with students are most strongly related to the three burnout dimensions.
Supervisors play a crucial role in giving teachers more autonomy.
More autonomy leads to less dissatisfaction with non-teaching-related workload.
Our data support the sequential process model of burnout.
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 9 October 2013
Received in revised form
7 July 2014
Accepted 10 July 2014
Available online 30 July 2014
According to the intensication thesis, external pressures from policymakers, supervisors, parents, and
experts fuel an ever-expanding teaching role and associated workload. Against that background, we
examined how four interpersonal relationships (students, colleagues, supervisors, and parents),
teaching-related and non-teaching-related workload (e.g., paperwork), and autonomy are related to
teacher burnout (emotional exhaustion, depersonalization, personal accomplishment). The participants
were 1878 Flemish teachers 45 years of age or older. The data were analyzed by means of structural
equation modelling. Interpersonal relationships were differentially related to burnout. Teaching-related
and non-teaching-related workload were both related to emotional exhaustion. Autonomy was most
strongly related to non-teaching-related workload.
2014 Elsevier Ltd. All rights reserved.
Keywords:
Teacher burnout
Teacher autonomy
Workload
Social support
Intensication
1. Introduction
Burnout is a persistent, negative, work-related state of psychological exhaustion that results from a mist between personal intentions and motivations on the one hand and actual on-the-job
experiences on the other (Schaufeli & Enzmann, 1998). It includes a
wide range of psychological (e.g., chronic fatigue, low self-esteem,
lack of self-condence, depression) and physiological (e.g., headaches, muscle pains, hypertension) symptoms. Burnout impedes
the attainment of professional goals (Maslach, 2003), depletes
* Corresponding author. Vrije Universiteit Brussel, Sociology Department, Pleinlaan 2, 1050 Brussel, Belgium. Tel.: 32 474481384; fax: 32 26148140.
E-mail addresses: Filip.Van.Droogenbroeck@vub.ac.be (F. Van Droogenbroeck),
Bram.Spruyt@vub.ac.be (B. Spruyt), Christophe.Vanroelen@vub.ac.be (C. Vanroelen).
http://dx.doi.org/10.1016/j.tate.2014.07.005
0742-051X/ 2014 Elsevier Ltd. All rights reserved.
100
pupils, colleagues, parents and principals, (2) teaching and nonteaching related workload, and (3) perceived autonomy on
burnout among senior teachers (aged 45e65) in Flanders (the
Dutch-speaking part of Belgium, N: 1878).
The focus on senior teachers is warranted, because the consequences of suffering from burnout for older employees in terms of
their (re)integration in the labour market prove to be dramatic.
Teachers suffering from burnout have a higher chance of leaving
the profession and retiring early (Leung & Lee, 2006; Martin, Sass,
& Schmitt, 2012; Rudow, 1999; Skaalvik & Skaalvik, 2011; Van
Droogenbroeck & Spruyt, 2014). This is especially problematic,
because in several countries, including Belgium, the teacher work
force is ageing (OECD, 2005, 2012). According to the OECD (2012)
this tendency will put increasing pressure on the teacher labour
market and may lead to increasing teacher shortages. Since policymakers attempt to motivate employees e and more specically
teachers e to work longer, it becomes crucial to understand the
determinants of burnout among senior teachers.
The remainder of this article is organized as follows. First, we
situate our research question against the background of changes
the teaching profession went through during the last two decades.
This section justies our focus on the impact of the quality of
different interpersonal relationships, types of workload, and autonomy. Next, we review the existing literature on burnout among
teachers and locate our own study against that background. This
section is followed by a description of our data and the methods we
used. Subsequently, we present the results of our study. In the
concluding section we elaborate on the implications of our ndings.
A second e and related e issue concerns the scope of education's tasks within larger society which continued to broaden.
Today, education is expected to contribute to the solution of
numerous social problems (e.g., drug abuse prevention, civic education, health education, etc.). This so called educationalization of
society implies that schools are expected to take over tasks that
were once carried out by parents (Smeyers & Depaepe, 2008). Some
of these are experienced by teachers as improper to their assumed
job description (Kelchtermans, 1999). Moreover, the broadening of
education's scope also increased the number of other professionals
(e.g., equal opportunities teachers, student counsellors, special
needs teachers, education inspectors) that work with pupils and
teachers. Teachers thus nd themselves in a contradictory situation
in which they face greater responsibility for their work but have
less control over the way they wish to work (Ball, 2003).
Although the literature often emphasizes the negative impact of
intensication, different authors stress that it may also have positive effects. Being confronted with intensication, for example,
some teachers look for further training or collaborative lesson
planning as coping strategies which may result in re-skilling and
re-professionalization (e.g., action research, teacher-as-researcher
movements) (Ballet & Kelchtermans, 2009; Woods, 1999). More
generally, intensication does not directly ow from educational
policy into the classroom. Intensication is a nuanced, mediated
and not completely deterministic process (Ballet et al., 2006). For
example, teachers lter imposed policy changes through their
personal ideologies and perspectives and have a certain freedom in
how they implement them in the classroom. Ballet et al. (2006)
distinguished several working conditions in schools that mediate
the impact of intensication. First, allowing teachers to co-decide
how work is organized can induce job satisfaction and fosters
feelings of collegiality and a sense of community. Participation in
decision-making is crucial to mitigate the impact of intensication
by enabling teachers to identify problematic external pressures and
tackling them collectively. At the same time, however, participating
in school policy-decisions can also increase the feeling of being
overburdened especially when a lot of time has to be spent on
making decisions without direct benet for the pupils. Secondly,
supervisors play a central role in translating policy arrangements
and demands from the local community or parents to the specic
school context. Supervisors, for example, can act as a buffer by
postponing the implementation of a new curriculum if they feel
that this would cause too much strain on the teachers. This implies
that supervisors also face intensication which makes their buffering role more difcult. They have to deal with more (administrative) tasks and have to focus on the more managerial aspects of
their job. Thirdly, positive interpersonal relationships at school can
also mediate the impact of intensication. Positive relations with
colleagues are very important in building shared values and norms
and create a collective goal orientation. In this regard, Hargreaves
(1994) noticed signicant changes in teacher culture. He points to
a shift from individualism to collaboration, from hierarchies to
teams, from supervision to mentoring, from in-service training to
professional development and from authority towards parents to a
contract with parents. Indeed, teachers have to interact with and
are made accountable to an increasing numbers of actors. Not only
do they have to maintain relationships with students, but also with
colleagues, supervisors, parents, and experts. On the one hand,
positive interpersonal relationships can function as a buffer against
the impact of intensication. On the other hand, Hargreaves identied contrived collegiality, which entails enforced collaboration
resulting from a drive towards managerialism and accountability,
rather than the interests of the students. This creates stress and
contributes to the alienation of individuals who are opposed to
such developments.
101
102
Fig. 1. Theoretical structural model. The variables RELCOL, RELSTU, and RELPAR were grouped together to increase the clarity of the image. Note: SPC support with policy changes;
RELSUP relationship with supervisor; RELCOL relationship with colleagues; RELSTU relationship with students; RELPAR relationship with parents; AUT autonomy;
EE emotional exhaustion; DP depersonalization; PA personal accomplishment.
103
104
1
The MLM method estimates the SatorraeBentler scaled c2 and a scaling
correction factor instead of the regular c2. See http://www.statmodel.com/chidiff.
shtml for more information on the adjusted c2 difference test when using the
MLM estimator.
3. Results
Table 1 shows the zero order correlations between the variables
and the Cronbach's alphas. The correlations between the three
burnout dimensions varied from moderate to strong. The strongest
correlation was found between emotional exhaustion and depersonalization (.74).2 The correlations between the four interpersonal
relationships were positive and varied from weak to moderate. The
strongest correlations were found between relationships with the
supervisor and the relationships with colleagues (.42) and the relationships with the parents and the relationships with the students (.48). Furthermore, autonomy was positively related to the
four interpersonal relationships and negatively related to
emotional exhaustion (.32) and depersonalization (.43). Autonomy was most strongly correlated with relationships with the
supervisor (.71). Teaching-related and non-teaching-related workload had a moderate negative correlation (.39).
Table 2 summarizes the results for the CFA and SEM analyses.
We rst tested a measurement model for burnout. Results of the
CFA revealed an acceptable t for the hypothesized threedimensional MBI structure. The incorporation of one crossloading and two error-covariances signicantly improved the t
(model 2). The second measurement model specied six aspects of
the school context, four interpersonal relationships (students, colleagues, supervisors, and parents), autonomy, and the support with
policy changes. The baseline model had a good t with the data. Fit
indices suggested a signicantly better t with one additional
error-covariance (see model 4). In the next step the two measurement models were combined, also revealing a good t of data
(model 5).
Based on previous empirical ndings, a theoretical model (Fig. 1)
was specied. Since previous research predicts direct paths between
all interpersonal relationships and the three burnout dimensions, we
grouped the interpersonal relationships with students, colleagues,
and parents in the gure for the sake of clarity. As argued in the
theoretical section (see Section 1.2.1) we expect several of these
direct paths to become non-signicant because we take into account
four interpersonal relationships at the same time and we also specify
the relationships between the three burnout dimensions. This
should cause a mediation of several of the direct paths from the
interpersonal relationships to the burnout dimensions.
For the SEM analysis we followed the method suggested by
Byrne (2011). The best tting model is determined rst, then the
most parsimonious model. Once the nal model is established, the
remaining non-signicant parameters are deleted.
The SEM analysis of the postulated model (Fig. 1) revealed a
satisfactory t to the data (model 6). The modication indices
suggested that a better t could be achieved (model 7) if one path
was added (relationships with students / teaching-related
workload). The c2 difference test for the SatorraeBentler scaled c2
(Dc2(1) 18.07, p < .001) and other t indices indicated that the
addition of this path leads to a statistically signicant improvement
of the model. In addition, the modication indices suggested adding one further path (relationships with colleagues / autonomy),
which equally resulted in a statistically signicant improvement of
the model (model 8). In the nal step non-signicant paths were
removed. The nal model revealed an acceptable t (model 9).
Fig. 2 presents the standardized regression estimates. Interestingly,
the relationship with parents was not signicantly related to any
indicator in the nal model. When this latent factor was removed,
2
Because of these strong correlations we controlled for multicollinearity between all the study variables. The highest Variance Ination Factor (VIF) was 2.4,
with a threshold of 10, multicollinearity can be excluded.
105
Table 1
Zero order correlations and Cronbach's alphas.
1 EE
2 DP
3 PA
4 RELSUP
5 RELCOL
6 RELPAR
7 RELSTU
8 AUT
9 SPC
10 Teaching-related workload
11 Non-teaching-related workload
12 Age
13 Gender
10
11
12
13
.74
e
.29
.38
e
.26
.31
.17
e
.24
.33
.18
.42
e
.26
.35
.24
.25
.25
e
.28
.36
.26
.23
.19
.48
e
.32
.43
.22
.71
.43
.30
.25
e
.31
.35
.20
.40
.30
.24
.18
.48
e
.40
.35
.19
.25
.17
.19
.23
.25
.24
e
.36
.31
.16
.33
.16
.21
.16
.39
.29
.39
e
.13
.05
.10
.01
.03
.02
.03
.00
.00
.13
.07
e
.06
.08
.11
.02
.11
.05
.03
.00
.05
.13
.06
.17
e
.89
.82
.81
.90
.80
.75
.78
.70
Cronbach's alpha
Note: all correlations are signicant (p < .001); EE emotional exhaustion; DP depersonalization; PA personal accomplishment. RELSUP relationship with supervisor;
RELCOL relationship with colleagues; RELPAR relationship with parents; RELSTU relationship with students; AUT autonomy; SPC support with policy changes.
Table 2
Summary of goodness-of-t for the conrmatory factor analyzes and structural models.
Actions
Measurement model
1. Burnout
- Model 1 baseline
- Model 2 adapted
1 Cross-loading 2 Error
covariances
2. School context
- Model 3 baseline
- Model 4 adapted
1 Error covariance
Model 5 combined
Structural model
- Model 6 baseline
- Model 7 adapted
- Model 8 adapted
Model 9 nal model
RELSTU on teaching-related
workload
RELCOL on AUT
Deletion non-signicant paths
MLM c2
(df, scaling correction)
RMSEA
Upper limit
RMSEA 90%
CFI
TLI
SRMR
483.98
(101, 1.292)
319.629
(98, 1.280)
.053
.057
.954
.945
.039
.041
.046
.973
.967
.031
.039
.042
.952
.945
.038
.037
.040
.958
.951
.037
.031
.033
.955
.950
.039
.034
.036
.940
.935
.045
.033
.035
.942
.936
.043
.033
.035
.942
.937
.042
.033
.035
.942
.937
.043
761.689
(260, 1.201)
702.166
(259, 1.200)
1508.918
(739, 1.151)
1965.025
(906, 1.124)
1934.125
(905, 1.124)
1920.608
(904, 1.124)
1934.937
(913, 1.123)
106
Fig. 2. Final structural model of four interpersonal relations, teaching-related and non-teaching-related workload, autonomy, support with policy changes, age, and gender with
teacher burnout. Only signicant standardized regression weights are included in the gure. Note: SPC support with policy changes; RELSUP relationship with supervisor;
RELCOL relationship with colleagues; RELSTU relationship with students; RELPAR relationship with parents; AUT autonomy; EE emotional exhaustion;
DP depersonalization; PA personal accomplishment.
107
108
Eurydice Network. (2012). Key data on education in Europe 2012. Education, Audiovisual and Culture Executive Agency e European Commission.
Friedman, I. A. (1995). Student behavior patterns contributing to teacher burnout.
The Journal of Educational Research, 88(5), 281e289.
Friedman, I. A. (2003). Self-efcacy and burnout in teaching: the importance of
interpersonal-relations efcacy. Social Psychology of Education, 6(3), 191e215.
Gavish, B., & Friedman, I. A. (2010). Novice teachers' experience of teaching: a dynamic aspect of burnout. Social Psychology of Education, 13(2), 141e167.
Grayson, J. L., Alvarez, H. K., & Grayson, J. L. A. H. K. (2008). School climate factors
relating to teacher burnout: a mediator model. Teaching and Teacher Education,
24(5), 1349e1363.
Greenglass, E. R., Burke, R. J., & Konarski, R. (1997). The impact of social support on
the development of burnout in teachers: examination of a model. Work & Stress,
11(3), 267e278.
Hakanen, J. J., Bakker, A. B., & Schaufeli, W. B. (2005). Burnout and work engagement
among teachers. Journal of School Psychology, 43(6), 495e513.
Hargreaves, A. (1994). Intensication. Teachers' work e better or worse?. Changing
teachers, changing times: Teachers' work and culture in the postmodern age (pp.
117e140) London: Cassell.
Heus, P. D., & Diekstra, R. F. W. (1999). Do teachers burn out more easily? A comparison of teachers with other social professions on work stress and burnout
symptoms. In R. Vandenberghe, A. M. Huberman, R. Vandenberghe, &
A. M. Huberman (Eds.), Understanding and preventing teacher burnout (pp.
269e284). Cambridge: Cambridge University Press.
Horn, J. E., & Schaufeli, W. B. (1998). Maslach burnout inventory: The Dutch educators
survey (MBI-NL-ES) psychometric evaluations. Manual (unpublished manuscript).
Utrecht University, Department of Social and Organizational Psychology.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for t indexes in covariance structure
analysis: conventional criteria versus new alternatives. Structural Equation
Modeling, 6(1), 1e55.
Kelchtermans, G. (1999). Teaching career: between burnout and fading away? Reections from a narrative and biographical perspective. In R. Vandenberghe, &
A. M. Huberman (Eds.), Understanding and preventing teacher burnout (pp.
176e191). Cambridge: Cambridge University Press.
Kelchtermans, G. (2005). Teachers' emotions in educational reforms: selfunderstanding, vulnerable commitment and micropolitical literacy. Teaching
and Teacher Education, 21(8), 995e1006.
Kinman, G., Wray, S., & Strange, C. (2011). Emotional labour, burnout and job
satisfaction in UK teachers: the role of workplace social support. Educational
Psychology, 31(7), 843e856.
Kokkinos, C. M. (2007). Job stressors, personality and burnout in primary school
teachers. British Journal of Educational Psychology, 77, 229e243.
Langballe, E. M. (2006). The factorial validity of the Maslach burnout inventorygeneral survey in representative samples of eight different occupational
groups. Journal of Career Assessment, 14(3), 370e384.
Lau, P. S. Y., Yuen, M. T., & Chan, R. M. C. (2005). Do demographic characteristics
make a difference to burnout among Hong Kong secondary school teachers?
Social Indicators Research, 71(1e3), 491e516.
Lee, R. T., & Ashforth, B. E. (1996). A meta-analytic examination of the correlates of the
three dimensions of job burnout. Journal of Applied Psychology, 81(2), 123e133.
Leiter, M. P., & Maslach, C. (1988). The impact of interpersonal environment on burnout
and organizational commitment. Journal of Organizational Behavior, 9(4), 297e308.
Leithwood, K., Menzies, T., Jantzi, D., & Leithwood, J. (1999). Teacher burnout: a
critical challenge for leaders of restructuring schools. In R. Vandenberghe, &
A. M. Huberman (Eds.), Understanding and preventing teacher burnout (pp.
85e114). Cambridge: Cambridge University Press.
Leung, D. Y. P., & Lee, W. W. S. (2006). Predicting intention to quit among Chinese
teachers: differential predictability of the components of burnout. Anxiety,
Stress & Coping, 19(2), 129e141.
Martin, N. K., Sass, D. A., & Schmitt, T. A. (2012). Teacher efcacy in student
engagement, instructional management, student stressors, and burnout: a
theoretical model using in-class variables to predict teachers' intent-to-leave.
Teaching and Teacher Education, 28(4), 546e559.
Maslach, C. (2003). Burnout: The cost of caring. Los Altos, CA: Malor Books.
Maslach, C., & Jackson, S. E. (1984). Burnout in organizational settings. In S. Oskamp
(Ed.), Applied social psychology annual: Applications in organizational settings.
Beverly Hills, CA: Sage Publications.
Maslach, C., & Jackson, S. E. (1986). Maslach burnout inventory (2nd ed.). Palo Alto,
CA: Consulting Psychologists Press.
Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of
Psychology, 52(1), 397e422.
Mazur, P. J., & Lynch, M. D. (1989). Differential impact of administrative, organizational, and personality-factors on teacher burnout. Teaching and Teacher Education, 5(4), 337e353.
n, L. K., & Muthe
n, B. O. (1998). Mplus user's guide (7th ed.). Los Angeles, CA:
Muthe
n & Muthe
n.
Muthe
Noor, N. M., & Zainuddin, M. (2011). Emotional labor and burnout among female
teachers: workefamily conict as mediator: emotional labor and burnout. Asian
Journal of Social Psychology, 14(4), 283e293.
OECD. (2005). Teachers matter: Attracting, developing and retaining effective teachers.
Education and training policy. OECD Publishing.
OECD. (2012). Education at a glance 2012: OECD indicators. OECD Publishing.
Park, H. I., Jacob, A. C., Wagner, S. H., & Baiden, M. (2013). Job control and
burnout: a meta-analytic test of the conservation of resources model. Applied
Psychology. http://dx.doi.org/10.1111/apps.12008.
109