Content uploaded by Ij.H. Van Emmerik
Author content
All content in this area was uploaded by Ij.H. Van Emmerik
Content may be subject to copyright.
UNCORRECTED PROOF
How job demands, resources, and burnout predict
objective performance: A constructive replication
ARNOLD B. BAKKER
1
, HETTY VAN EMMERIK
2
, & PIM VAN RIET
2
1
Department of Work & Organizational Psychology, Erasmus University Rotterdam, Rotterdam, The
Netherlands &
2
Department of Social & Organizational Psychology, University Utrecht, Utrecht, The
Netherlands
Abstract
The present study uses the Job Demands-Resources model (Bakker & Demerouti, 2007) to examine
how job characteristics and burnout (exhaustion and cynicism) contribute to explaining variance in
objective team performance. A central assumption in the model is that working characteristics evoke
two psychologically different processes. In the first process, job demands lead to constant
psychological overtaxing and in the long run to exhaustion. In the second process, a lack of job
resources precludes actual goal accomplishment, leading to cynicism. In the present study these two
processes were used to predict objective team performance. A total of 176 employees from a
temporary employment agency completed questionnaires on job characteristics and burnout. These
self-reports were linked to information from the company’s management information system about
teams’ (N71) objective sales performance (actual sales divided by the stated objectives) during the 3
months after the questionnaire data collection period. The results of structural equation modeling
analyses did not support the hypothesis that exhaustion mediates the relationship between job
demands and performance, but confirmed that cynicism mediates the relationship between job
resources and performance suggesting that work conditions influence performance particularly
through the attitudinal component of burnout.
Keywords: Burnout, Job Demands-Resources model, objective performance
Generally, it is assumed that burnout negatively affects job performance, although the
evidence is still limited (Demerouti & Bakker, 2006). Moreover, research has usually relied
on subjective assessments of job performance (Taris, 2006) and studies on the impact of
burnout on objective performance are virtually lacking (Halbesleben & Buckley, 2004).
Using the Job Demands-Resource (JD-R) model as a guiding framework, the goal of the
present study was to examine how job demands and resources are related to teams’
objective financial performance focusing on the mediating role of burnout (Bakker &
Demerouti, 2007; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). Specifically, the
purpose of the present study was a constructive replication of Bakker, Demerouti, and
Verbeke (2004). Following Lykken (1968), a constructive replication extends the general-
izability of the research after which it is modeled by avoiding exact duplication.
5
10
15
20
25
30
35
Correspondence: Arnold B. Bakker, Institute of Psychology, Erasmus University Rotterdam, Woudestein, T12-47,
P.O. Box 1738, 3000 DR Rotterdam, The Netherlands. Tel: 31 10 408 8853. Fax: 31 10 408 9009. E-mail:
bakker@fsw.eur.nl
ISSN 1061-5806 print/ISSN 1477-2205 online # 2008 Taylor & Francis
DOI: 10.1080/10615800801958637
Anxiety, Stress, & Coping,
July 2008; 21(3): 116
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:44
UNCORRECTED PROOF
Consequently, we used different demands and resources as predictors, another measure of
burnout, and objective financial performance at the team level as the dependent variable
(instead of colleague ratings of performance).
Burnout
Burnout was originally conceived as a work-related syndrome that most often occurs among
individuals who work with other people (Maslach & Jackson, 1986). However, research of
the past decade has shown that the two core burnout dimensions emotional exhaustion
and cynicism can be observed in virtually any occupational group (Bakker, Demerouti, &
Schaufeli, 2002; Maslach, Schaufeli, & Leiter, 2001). Emotional exhaustion refers to a
general feeling of extreme chronic fatigue, caused by continuous exposure to demanding
working conditions. Cynicism is defined as a callous, distanced and cynical attitude toward
the work itself or the people with whom one works.
Of these two burnout dimensions, emotional exhaustion appears to be the central
variable in the burnout process (Baba, Jamal, & Tourigny, 1998; Shirom, 2005). A number
of studies have indeed shown that exhaustion is more strongly related to important outcome
variables (such as personnel turnover and absenteeism) than the other burnout dimensions
(Lee & Ashforth, 1993, 1996; Leung & Lee, 2006; Wright & Bonett, 1997). In addition,
exhaustion is always part of the different definitions of burnout (Shirom, 2005). Leiter’s
(1993) process model of burnout proposes that cynicism should be seen as a consequence
of emotional exhaustion. Accordingly, feelings of exhaustion arise from stressful working
conditions, whereby employees are repeatedly confronted with high job demands (such as
work pressure or high emotional demands) and as a consequence, they can develop a
cynical attitude as a coping strategy to distance themselves emotionally and mentally from
work (e.g., Bakker, Schaufeli, Sixma, Bosveld, & Dierendonck, 2000; Taris, LeBlanc,
Schaufeli, & Schreurs, 2005).
Burnout and Performance
Job performance refers to employees’ behaviors that are supposed to contribute to the
effectiveness of the organization and to overall organizational performance (Campbell,
1990). Singh, Goolsby, and Rhoads (1994) offer three possible explanations for the
influence of burnout on performance. First, burnout is characterized by a reduction of the
available energy and the amount of effort that is invested to perform well. Another reason is
that employees with burnout get trapped in a negative, vicious cycle, in which they are not
inclined to search for support or are not motivated to change their situation. The
consequence is that performance declines (Bakker et al., 2004). A final explanation for the
impairment of performance is that burnout undermines employees’ self-confidence in their
ability to solve work-related problems.
However, although a negative relationship between burnout and performance seems
apparent, empirical evidence for the influence of burnout on job performance is scarce
(Demerouti & Bakker, 2006). On the basis of six studies (total N2000), Schaufeli and
Enzmann (1998) found that burnout correlated only weakly with self-reported job
performance. On average, emotional exhaustion and cynicism explained 5 and 4%,
respectively, of the variance in job performance. Taris (2006) reviewed 16 studies on the
relationship between burnout and ‘‘objective’’ performance (mainly supervisor reports),
and found a meta-analytical correlation (r .22) between exhaustion and in-role
40
45
50
55
60
65
70
75
2 A. B. Bakker et al.
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:45
UNCORRECTED PROOF
performance. Taris also found that the evidence for the relationship between cynicism
(depersonalization) and performance was inconclusive.
In the current study, we wanted to replicate the study of Bakker et al. (2004) among
human service professionals. In addition to employee ratings of their own job demands,
resources, and burnout (as measured with the Oldenburg Burnout Inventory), Bakker et al.
collected colleague ratings of in-role and extra-role performance. The results showed that
exhaustion was the most important predictor of in-role performance, whereas cynicism
(called disengagement in the study) was the most important predictor of extra-role
performance. Both dimensions of burnout explained 8% of the variance in performance.
The Job Demands-Resources (JD-R) Model
In the present study, we used the Job Demands-Resources (JD-R) model (Bakker &
Demerouti, 2007; Demerouti et al., 2001) to examine how job characteristics and burnout
contribute to explaining variance in objective team performance. A central assumption in
the JD-R model is that working characteristics may evoke two psychologically different
processes. In the first process, demanding aspects of work (i.e., work overload and
complaining customers) lead to constant psychological overtaxing and in the long run to
exhaustion (e.g., Lee & Ashforth, 1996; Wright & Cropanzano, 1998).
According to Hockey’s (1993) control model of demand management, employees will use
a performance protection strategy when confronted with high job demands. In order to
maintain the desired performance level, they will mobilize extra energy to compensate
fatigue through mental effort. This implies that when people become exhausted under the
influence of environmental demands, they will not be able to perform well because their
energy resources are diminished. Indeed, Veldhuizen, Gaillard, and De Vries (2003), using
office tasks in order to simulate a working day, found that exhausted participants had
problems investing sufficient energy to their tasks. Moreover, exhausted participants’
performance results decreased as they reacted more slowly and produced a smaller number
of correct responses. Further, exhausted participants seemed unable to perform particularly
well in the evening, although they tried to invest more effort than their non-exhausted
counterparts. This implies that the impact of job demands on performance could be
mediated by feelings of (enhanced) exhaustion (cf. Hockey, 1993). Therefore, we
formulated the following hypothesis (see also Figure 1).
Hypothesis 1: Emotional exhaustion fully mediates the relationship between job demands
and objective performance.
In the second process proposed by the JD-R model, a lack of job resources precludes
actual goal accomplishment, which causes failure and frustration (Bakker, Demerouti, De
Boer, & Schaufeli, 2003). When organizations do not provide their employees with
sufficient job resources, withdrawal and decreased commitment will be the end result
(Bakker, Demerouti, & Schaufeli, 2003; Demerouti et al., 2001). These outcomes can be
interpreted as self-protecting mechanisms that prevent the development of employee
frustrations caused by not achieving work-related goals (cf. Hackman & Oldham, 1980).
Indeed, consistent with this line of reasoning, Hobfoll’s (1989) conservation of resources
theory argues that people strive to obtain and maintain (job) resources, and that situations
are experienced as stressful when loss or threat of loss occurs, or when the acquiring of job
resources fails. To reduce their level of stress, employees will try to limit losses. One way to
achieve this is to develop a detached attitude to the job (Wright & Bonett, 1997). We
propose that this detached attitude can diminish job performance (cf. Maslach, 1993). The
80
85
90
95
100
105
110
115
120
125
Job demands, resources, and burnout 3
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:45
UNCORRECTED PROOF
prediction is that cynicism is the most important consequence of lacking job resources (cf.
Bakker & Demerouti, 2007; Demerouti et al., 2001), which, in turn, has a negative
association with job performance (see Figure 1).
Hypothesis 2: Cynicism fully mediates the relationship between job resources and
objective (financial) performance.
Context of the Present Study: Performance as an Outcome
It was not before the 1990s that scholars formulated precise definitions of performance
(Campbell, 1990; Roe, 1999). When trying to define performance, researchers either
referred to the process of performance, or to the outcome of performance. According to the
former perspective, performance is the process by which people try to achieve a given work
goal (Roe, 1999). This definition focuses on the actions (or behaviors) that people
undertake to achieve their performance or what individuals do in their work situation, like
reading scientific literature, writing research proposals, and conducting studies in case of
researchers (Demerouti & Bakker, 2006). The second definition views performance as the
congruence between the work goal and the outcome of the process by which people try to
accomplish that work goal (Roe, 1999). For scientists, publication of a scientific article is an
example of an outcome of individual behaviors. Conceptually, this differentiation is helpful
to explicate what we talk about when studying performance. Practically, and in many cases
even psychometrically, it is difficult to separate behaviors from their outcome because the
behaviors are mainly enacted in order to achieve the outcome.
In the present study, performance was operationalized through an objective assessment of
team effectiveness, which is clearly an outcome-related indicator of performance. This kind
of performance is also known in the literature as in-role performance, defined as those
officially required outcomes and behaviors that directly serve the goals of the organization
(Motowildo & Van Scotter, 1994). Among other things, in-role performance includes
meeting company objectives (Behrman & Perreault, 1982).
130
135
140
145
150
Job
Resources
Team Cohesion
Harmony
Autonomy
Coaching
Job
Demands
Work Pressure
Emotional
Demands
Work-Home
Conflict
Colleague
Support
Exhaustion
Performance
Supervisor
Support
Cynicism
+
+
-
-
-
-
Figure 1. The Job Demands-Resources model.
4 A. B. Bakker et al.
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:45
UNCORRECTED PROOF
We used the JD-R model to examine how burnout affected the performance of groups of
employees who worked in autonomous teams of a temporary employment agency. Every
team included at least one desk manager, one sales executive, and one team manager. The
goal of each team was to realize the yearly sales objectives by placement of temporary
employees. We used the extent to which the sales objectives (of the teams) were met as the
operationalization of job performance. This operationalization is an objective measurement
of team effectiveness. Previous studies on performance most frequently used individual level
outcomes (e.g., Bakker et al., 2004) or aggregated individual level outcomes to the group or
unit level (see Taris, 2006).
There is scarce empirical evidence for the relationship between burnout and team
performance. The PsychInfo database revealed only one empirical study (Keijsers,
Schaufeli, Le Blanc, Zwerts, & Reis Miranda, 1995). This study reported a negative
correlation between intensive care nurses’ feelings of burnout and subjective unit
performance. In contrast, burnout was positively related to objective unit performance.
Regarding the temporary employment agencies of the present study, a negative association
between burnout and performance is to be expected. Team members are jointly responsible
for the realization of their yearly sales objectives (work goal interdependence), and are
strongly dependent on each others’ efforts to meet these objectives (task interdependence).
A number of theories presuppose that interdependence is one of the most important
predictors of team effectiveness (e.g., Gladstein, 1984; O’Leary Kelly, Martocchio, &
Frink, 1994; Shea & Guzzo, 1987). Team members who perform insufficiently would have
a negative impact on the performance of their team. For example, within an employment
agency, a desk manager has to perform many ad hoc tasks, and this demands flexibility and
mental energy. If the desk manager suffers from burnout, he or she will be less able to
comply with these task demands. The placement of the temporary employees will most
probably be affected by these processes and will be less efficient, which, in turn, negatively
affects the realization of the company’s objectives.
Method
Participants and Procedure
A survey study was conducted in an organization of temporary employment agencies. All
508 employees were approached for the present study. The management informed the
employees with a letter that the study would consist of a qualitative and a quantitative part.
For the qualitative part, 15 explorative interviews were conducted. Based on the results,
specific job demands and job resources were chosen to be included in the questionnaire. For
the quantitative part, the questionnaire and an accompanying letter was sent to employees’
home addresses. It was emphasized that the information would be treated confidentially
and that the results would be reported anonymously to the management. Two reminders
from the human resource management department were sent to increase the response rate.
In total, 290 employees responded (57%). For the present study, we used the data of those
176 respondents who were still employed within the company during the follow-up study of
financial performance (the 3 months after the questionnaire data collection). It should be
noted that a high personnel turnover is not uncommon in this sector.
We performed a multivariate analysis of variance (MANOVA) to examine the extent to
which there were differences between our final sample (N 176) and the dropouts
(N114) on all model variables. The MANOVA produced a multivariate significant effect,
F(11, 278) 3.13, p B.001. Univariate analyses showed that the two groups differed
155
160
165
170
175
180
185
190
195
Job demands, resources, and burnout 5
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:45
UNCORRECTED PROOF
significantly regarding work pressure, F(1, 288)10.27, p B.01; harmony F(1, 288)
17.62, pB.001; exhaustion, F(1, 288) 7.63, p B.01; and cynicism, F(1, 288)10.32,
pB.001. The research sample scored less favorable than the dropouts on each of these
variables: work pressure M3.16 vs. M2.88; harmony M 4.00 vs. M4.28; exhaus-
tion M 1.69 vs. M 1.39; and cynicism M 1.28 vs. M .95. There were no differences
between both groups regarding the background variables, or the other job demands and
resources. Taken together, these results suggest that our research group experienced more
strain than the dropouts.
The research sample included 97 desk managers (55%), 29 sales executives (17%) and
50 branch managers (28%) working in one of 71 teams. The mean number of respondents
per team was 2.5 (SD 1.2, range 16), whereas in the whole company this was five
employees. Sixteen (25%) of these 71 teams represented all three job categories. The vast
majority of the sample was female (85%). Age ranged between 21 and 50 years, with a mean
of 27 years (SD 4.5). The educational level was mainly upper secondary education (53%)
and higher vocational training (43%). The average organizational tenure was 2.7 years
(SD2.6), and mean number of working hours was 38 hours per week (SD4.5). Almost
all employees (94%) had a permanent employment contract. The correlational analyses
showed no significant relationships between the demographic characteristics and the model
variables. Therefore, demographic characteristics were excluded from all further analyses.
Measurement Instruments
Job Demands. Three demands were included in the questionnaire: work pressure, emotional
demands, and workhome conflict. All items were answered on a scale from 1never to
5always. Work pressure was assessed with the five-item scale developed by Bakker et al.
(2004). An example is: ‘‘How often does it occur that you have to work extra hard to finish
your work?’’ Emotional demands were measured with three items of the scale developed by
Van Veldhoven and Meijman (1994). One item was deleted, because the reliability analysis
indicated that it was unsound. The remaining two items are: ‘‘Does your work put you in
emotional situations?’’, and ‘‘Is your job emotionally demanding?’’ Finally, workhome
conflict was measured with three items of the SWING (Geurts et al., 2005; see also
Demerouti, Bakker, & Bulters, 2004). An example is: ‘‘How often does it happen that you
find it difficult to fulfill your domestic obligations because you are constantly thinking about
your work?’’
Job Resources. Six job resources were included in the questionnaire: social support from
colleagues, team cohesion, harmony, autonomy, supervisory coaching, and supervisor
support. All items were answered on a scale from 1never to 5always, unless otherwise
indicated. Social support was measured with the three-item scale developed by Bakker et al.
(2003). An example is: ‘‘If necessary, can you ask your colleagues in your team for help?’’
Team cohesion was also assessed with three items, based on the ‘‘Substitutes for Leadership
Scales’’ (Podsakoff & MacKenzie, 1994). Two examples are: ‘‘Together, my colleagues
and I constitute a cohesive team’’, and ‘‘In my team we are well-tuned to work together’’
(1completely disagree, 5completely agree). Harmony was measured with a four-item
scale, based on Jehn (1994, 1995). An example is: ‘‘How often do colleagues within your
team get angry with each other?’’ (reverse coded). Autonomy was measured with a short
scale developed by Bakker et al. (2004) that comprises three items referring to decision
authority (i.e., freedom of action in accomplishing the formal work task). An example item
200
205
210
215
220
225
230
235
240
6 A. B. Bakker et al.
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:45
UNCORRECTED PROOF
is: ‘‘Can you decide yourself how you execute your work?’’ Supervisory Coaching was
assessed with a four-item scale derived from the ‘‘Leadership Practices Inventory’’ (Posner
& Kouzes, 1988, 1994). An example is: ‘‘Does your manager clearly envision future
opportunities?’’ (1not at all, 5to a high extent). Finally, supervisor support (the team
manager) was measured with five items from the scale of LeBlanc (1994), for example, ‘‘My
supervisor uses his/her influence to help me to perform well’’.
Burnout was measured with the scales from the Dutch version (Schaufeli & Van
Dierendonck, 2000) of the Maslach Burnout Inventory-General Survey (MBI-GS;
Schaufeli, Leiter, Maslach, & Jackson, 1996). We used the scales to assess the two core
dimensions of burnout: emotional exhaustion and cynicism. The exhaustion scale consists
of five items, including ‘‘I feel burned out from my work’’ and ‘‘I feel tired when I get up in
the morning and have to face another day on the job’’ (0 never, 6always). Cynicism was
measured with four of the five items from the original scale. An example item is: ‘‘I have
become more cynical about whether my work contributes anything’’. Item 4 (‘‘I just want to
do my job and not be bothered’’) was omitted, as suggested by Schaufeli and Van
Dierendonck (2000) and Schutte, Toppinnen, Kalimo, and Schaufeli (2000). They have
shown that this item does not load on the intended factor, and thus creates problems with
factorial validity.
Performance was operationalized through an objective assessment of team effectiveness
using a standardized measurement of sales. Information about each team’s sales
performance during the 3 months after the questionnaire data collection period was drawn
from the management information system. As the size of the teams was not constant, we
divided the actual sales by the stated objectives (the larger the team, the larger the stated
objective). This resulted in a standardized objective measure of effectiveness that signified
to what extent the stated sales objectives had been realized.
Analyses
The model in Figure 1 was tested in two steps with structural equation modeling (SEM)
analyses using AMOS 6.0 (Arbuckle, 2005). Maximum likelihood estimation methods and
the covariance matrix of the above-mentioned scales were used. To test the fit of alternative
models to the data, the traditional chi-square, the goodness-of-fit index (GFI) and the root
mean square error of approximation (RMSEA) were assessed. As a rule of thumb, a
GFI.90 and RMSEAB.08 indicate a reasonable fit of the model to the data (Browne &
Cudeck, 1989). As recommended by Marsh, Balla, and Hau (1996), the non-normed fit
index (NNFI), the incremental fit index (IFI), and the comparative fit index (CFI) were also
assessed. These values should meet the criterion of .90 (Hoyle, 1995).
In the first step, we tested the measurement model. The second step involved the test of
our theoretical model with structural paths (N 176). For this analysis, a team score on
objective financial performance (a team variable) was allocated to each individual team
member. The latent exogenous job demands and resources factors were operationalized by
three and six observed variables, respectively (see above). Further, the model consisted of
two endogenous latent variables that were included as mediators: emotional exhaustion and
cynicism. Both emotional exhaustion and cynicism were operationalized by one indicator
(the respective scale of the MBI-GS) to ensure a parsimonious model. We controlled for
measurement errors by equalling the error variance of emotional exhaustion to the product
of its variance and the quantity 1 minus the internal consistency (Jo
¨
reskog & So
¨
rbom,
1993). The same procedure was followed for cynicism. The endogenous variable
‘‘performance’’ was included in the model as an observed variable.
245
250
255
260
265
270
275
280
285
Job demands, resources, and burnout 7
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:45
UNCORRECTED PROOF
Results
Descriptive Statistics
Table I presents the means, standard deviations, reliabilities, and correlations between the
model variables. All constructs were measured reliably. It can be seen that the specific job
demands correlated most strongly with emotional exhaustion and that the specific resources
correlated most strongly with cynicism. In addition, performance correlated primarily with
resources (positive) and cynicism (negative).
Measurement Model
Our initial measurement model showed a poor fit to the data, x
2
(47)193.14, GFI.84,
RMSEA.13, NNFI .68, CFI .77, and IFI .78. The modification indices revealed
high covariations between the errors of several job resources indicators. Specifically, the
AMOS output suggested that the fit of the measurement model to the data could be
improved considerably if two separate job resources factors would be distinguished. One job
resource factor included the three job resources referring to the interpersonal relationships
with colleagues in the team, namely colleague social support, team cohesion, and harmony.
This factor will be referred to as ‘‘colleague resources’’. The second job resource factor
included the three resources referring to the relationship with the supervisor, namely
supervisor support, supervisory coaching, and autonomy. This factor will be referred to as
‘‘supervisor resources’’.
The modified measurement model (see Figure 2) showed an acceptable fit to the data, x
2
(42)90.21, GFI.92, RMSEA.08, NNFI .88, CFI .92, and IFI .93. All indica-
tors loaded well on the respective factors. For the job demands, the factor loadings ranged
from .41 to .73. The factor loadings of the colleague resources social support, harmony, and
team cohesion were .65, .75, and .80, respectively; the factor loadings of the supervisor
resources autonomy, supervisory coaching, and supervisor support were .48, .90, and .84,
respectively.
Test of the Job Demands-Resources (JD-R) Model: Level 1 Analyses
In the next step, we tested the JD-R model including the hypothesized relationships and the
correlations between the latent factors job demands, colleague resources, and supervisor
resources. Using the chi-square difference test, this model was compared with simpler
nested models that specify alternative relationships. As the dependent variable in the
present study (objective financial performance) was measured at the aggregate, team level,
we allocated the team score on objective financial performance to each of the employees
working in that team. The results of the SEM analyses showed that the hypothesized model
fits reasonably well to the data, x
2
(48)96.89, GFI.92, RMSEA.08, NNFI .90,
CFI.93, and IFI.93. In line with expectations, the results revealed a positive
relationship between emotional exhaustion and cynicism (b.39, pB.01).
In order to test the mediation hypotheses formulated in Hypotheses 1 and 2, we followed
Baron and Kenny’s (1986) approach, according to which there are four steps in establishing
a significant mediation effect. First, there must be a significant relationship between the
predictor and the outcome. Second, the predictor must be significantly related to the
mediator. Third, the mediator should be significantly related to the outcome variable.
Finally, there is a significant mediation effect when the relationship between the predictor
and the outcome becomes non-significant after the inclusion of the mediator. Before testing
290
295
300
305
310
315
320
325
330
AQ1
AQ1
AQ1
8 A. B. Bakker et al.
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:45
UNCORRECTED PROOF
Table I. Means, standard deviations, reliability coefficients, and correlations between the variables.
Variable MSD 1234567891011
1. Work pressure 3.16 .71 (.81)
2. Emotional demands 1.83 .66 .24** (.71)
3. Workhome conflict 1.92 .67 .30** .29** (.70)
4. Social support colleagues 3.99 .74 .01 .07 .28** (.81)
5. Harmony 4.00 .59 .12 .06 .20** .47** (.88)
6. Team cohesion 3.67 .72 .05 .08 .18** .52** .60** (.80)
7. Autonomy 3.73 .74 .14 .17* .32** .31** .32** .32** (.79)
8. Supervisory coaching 3.58 .77 .04 .11 .26** .31** .35** .38** .38** (.84)
9. Supervisor support 3.42 .89 .09 .09 .27** .36** .24** .24** .42** .76** (.90)
10. Exhaustion 1.69 .94 .40** .35** .61** .20* .27** .18** .24** .20** .17 (.90)
11. Cynicism 1.28 .87 .15* .08 .37** .27** .37** .38** .44** .38** .31** .46** (.83)
12. Objective (team) performance (%) 22.00 4.00 .03 .04 .11 .01 .27** .29** .04 .20* .05 .12 .26**
Note. N 176 employees divided over 71 teams. Reliabilities (Cronbach’s alphas) are shown in the diagonal.
*pB.05, **pB.01.
Job demands, resources, and burnout 9
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:46
UNCORRECTED PROOF
whether exhaustion was a mediator of the relationship between job demands and
performance (Hypothesis 1), we first checked for the prerequisite conditions. The results
of the hypothesized model showed that the coefficient of the path between emotional
exhaustion and objective performance was not significant, b.04, tB1, ns indicating that
Hypothesis 1 was not supported.
In the next step, we checked the prerequisites before testing whether cynicism mediates
the relationship between job resources and performance (Hypothesis 2). Preliminary results
showed that colleague resources and supervisor resources were both related to the mediator
cynicism, b.30 (pB.01) and b .22 (pB.05), respectively. In addition, cynicism was
significantly related to performance, b.30 (pB.01). However, while colleague
resources was related to performance, b.21, pB.05, supervisor resources was not,
b.03, ns. These results allowed proceeding with the test of Hypothesis 2, but only for
colleague resources.
We compared the proposed mediation model with a model including an additional direct
path from colleague resources to performance in order to see which model fitted better to
the data (Frazier, Tix, & Barron, 2004). The results showed that the alternative partial
mediation model did not fit better to the data than the proposed model, Delta x
2
(1)3.04,
ns, and that the direct path of colleague resources to performance became non-significant
(t1.80, ns). This indicates that cynicism mediated the relationship between colleague
resources and performance. The results of the Sobel test confirmed that colleague resources
had an indirect effect on performance through cynicism (z2.28, pB.05). Thus,
Hypothesis 2 was supported as far as resources from colleagues are concerned. These
findings suggest that employees who received support from their colleagues and worked in
close harmony were less cynical about their job, and performed better than employees who
335
340
345
350
355
Colleague
Resources
Team Cohesion
Harmony
Autonomy
Coaching
Job
Demands
Work Pressure
Emotional
Demands
Work-Home
Conflict
Colleague
Support
Exhaustion
Supervisor
Support
Cynicism
.38
-.30
Supervisor
Resources
-.22
-.28
-.31
.51
-.34
.89
.45
.41
.73
.65
.80
.75
.48
.84
.90
Exhaustion
.94
.83
Cynicism
Performance
Figure 2. Maximum likelihood estimates of the revised JD-R model, N176 employees and N 71 teams.
Note. All parameters are significant at the pB.01 level, except the coefficient of the path from supervisor resources
to cynicism, which is significant at the pB.05 level.
10 A. B. Bakker et al.
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:46
UNCORRECTED PROOF
had non-harmonious mutual relationships and displayed a detached attitude toward their
work.
Based on these results, the proposed model was modified regarding one pathway: the
path from emotional exhaustion to performance was deleted. The fit of this model to the
data was acceptable, x
2
(49)97.03, GFI.92, RMSEA .08, NNFI.90, CFI .93,
and IFI.93. The results are summarized in Figure 2. The JDR model explains 79% of
the variance in exhaustion, 47% of the variance in cynicism, and 8% of the variance in
objective financial performance.
Level 2 Analyses
The previously described SEM analyses were performed at the individual level (N176) by
allocating team scores on the (aggregate level) performance variable to each of the
employees working in the teams. This procedure may have led to an artificial increase in the
statistical power. In addition, the estimation of the effect of burnout on objective financial
performance may have been affected. We therefore decided to conduct additional level 2
analyses.
As the dependent variable was at the aggregate level, we aggregated all the independent
variables and we assessed the inter-rater agreement kappa. Kappa indicates the percentage
agreement, the closer to 100% agreement, the higher the reliability. Inter-rater reliability
kappa was calculated using multi-rater kappa (an unweighted form of kappa) using the
MKAPPASC.SPS macro in SPSS. Although there are different methodologies for the
assessment of observer agreement developed, in the present study we used Kappa because
of the small number of employees per team (the mean number of respondents per team was
2.5 with SD 1.2 and range16). Kappa can handle cases where there are only two raters
and when there are more than two raters or when there are more than three categories to be
rated. According to Altman (1991), interpretation of Kappa is: poor agreementB.20, fair
agreement .20 to .40, moderate agreement.40 to .60, good agreement.60 to .80, and
very good agreement.80 to 1.00. Kappas for the scales included in the present study
generally indicated fair agreement: social support from colleagues.39, team cohesion
.26, harmony.32, autonomy .35, supervisory coaching .29, supervisor support .38,
work pressure.15, emotional demands.39, workhome conflict.26, exhaustion .28,
and cynicism.34.
Additional analyses for both burnout components exhaustion and cynicism were
performed to examine the correlation with objective financial performance at the team
level (N71). The results showed that both exhaustion (r .27, p.02) and cynicism
(r.47, p B.001) were significantly and negatively associated with objective financial
performance. Note that the pattern of the correlations at the team level was consistent with
the pattern at the individual level.
Discussion
The present study is one of the first studies that examined the relationship between working
conditions, burnout, and objective (financial) performance. On the basis of the JD-R model
(Bakker & Demerouti, 2007; Demerouti et al., 2001), two central hypotheses were
formulated: exhaustion mediates the relationship between job demands and performance,
and cynicism mediates the relationship between job resources and performance. SEM
analyses found only support for the mediating role of cynicism. Colleague resources
(e.g., team cohesion and harmonic relationships with colleagues) were most clearly related
360
365
370
375
380
385
390
395
400
Job demands, resources, and burnout 11
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:46
UNCORRECTED PROOF
to objective team performance, because these resources are motivating and prevent
employees from developing a cynical, detached attitude. Therefore, the conclusion is that
motivation, and not strain, plays a prominent role when it comes to objective performance.
This is a relevant finding, since exhaustion (and not cynicism) is considered to be the core
symptom of burnout (Baba et al., 1988; Shirom, 2005), and exhaustion is known to have
the strongest relationship with self-reported job performance (Schaufeli & Enzmann,
1998).
How can we explain that previous studies found most support for a relationship between
emotional exhaustion (not cynicism) and in-role performance, including the study we
replicated (Bakker et al., 2004)? First of all, it is conceivable that we had a restriction in
range regarding the exhaustion variable. For one thing, the levels of exhaustion among the
young participants were relatively low even though they scored somewhat higher than the
dropouts. This may have limited the statistical possibility to find a relationship between
exhaustion and performance in the present study.
Second, given these low levels of exhaustion, it is unlikely that our participants were
unable to perform well because their energy resources were diminished, as in the study of
Veldhuizen et al. (2003). To this extent, it would be interesting to replicate the present study
among a heterogeneous group of employees, including a wide range of exhaustion levels.
Indeed, it is conceivable that the relationship between emotional exhaustion and job
performance is better explained by a curvilinear than a linear relationship (Gardner &
Cummings, 1988). A curvilinear relationship assumes the existence of an optimal activation
level of effort to perform well, with impaired performance under conditions of very low and
very high levels of exhaustion, and increased performance under conditions of average levels
of exhaustion. However, additional analyses of the present data do not support this
explanation.
Third, it is conceivable that previous studies overestimated the relationship between
exhaustion and performance because of common-method variance problems. Indeed, most
previous studies used a questionnaire design and did not include objective indicators of
performance. Finally, it should be noted that we did find a correlation between exhaustion
and performance at the team level. However, this correlation disappeared when the data
were analyzed at the employee level, and when exhaustion had to compete with cynicism in
the prediction of performance.
In Lykken’s (1968) terminology, the present study is a constructive replication of Bakker
et al. (2004) because we used several other job demands and resources, a different measure
of burnout, and objective financial performance at the team level as the dependent variable
(instead of colleague ratings of performance). However, our theoretical framework and
some of the hypotheses were the same. The findings of the present study are generally in
line with Bakker et al. (2004) supporting the two independent pathways leading to
performance as proposed by the JD-R model (Bakker & Demerouti, 2007). Thus, our
findings suggest that job resources may start a motivational process because job resources
had a negative impact on cynicism, and indirectly contributed to objective performance. In
contrast, job demands seem to evoke a health impairment process, in which increased job
demands coincide with increased levels of exhaustion.
Our findings are also consistent with Campbell’s (1990) theory, which gives a central role
to motivation as a determinant of job performance. Eight percent of the variance in
objective financial performance was explained by cynicism. This 8% is considerable if we
consider the following arguments. First, in Campbell’s theory besides motivation two
other determinants are distinguished: declarative and procedural knowledge. Sternberg
405
410
415
420
425
430
435
440
445
450
12 A. B. Bakker et al.
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:46
UNCORRECTED PROOF
(1994) proposes a number of additional determinants, including learning styles and
personality differences. Muchinsky (1993) has argued that situational factors may act as
determinants of performance as well. In short, in addition to motivation, job performance
may be explained by several other factors. Second, cynicism is an indirect indicator of
motivation as defined by Campbell. He assumes that motivation is the combined effect of
three choices: (a) the choice to put in effort; (b) the choice for the amount of effort; and (c)
the choice to continue effort. In contrast, cynicism refers to a negative job attitude, which
will generally coincide with the choice to put in less effort.
In the present study, the indicator of job performance was an indirect indicator as well
because team effectiveness referred to the job performance of all team members. If we had
been able to use more direct measures for motivation and job performance, the association
between cynicism and performance might have been stronger. As a matter of fact,
additional analyses showed a considerably stronger association between cynicism and
performance at the team level as compared with the individual level. Nevertheless, if we
compare the amount of variance (8%) that cynicism explained in the objective (financial)
performance with previous studies, the conclusion is that the present study explained more
variance than previous studies (see Schaufeli & Enzmann, 1998; Taris, 2006).
Limitations of the Present Study
In the present study, we related employees’ individual burnout scores to team effectivity. As
the teams in our sample represented only 50% of all employees, a considerable amount of
information could not be included in the analysis. It is therefore unclear whether the finding
that cynicism explained 8% of the variance in performance is an under- or over-estimation.
Furthermore, some teams were represented by one person only. As performance is the
result of the combined effort of all team members, we believe that using the information
regarding working conditions and burnout of only one member is not necessarily a problem,
but represents a conservative test of our hypotheses. Nevertheless, it is evident that more
employees per team would have been preferred.
Another limitation pertains to the design of the present study. The present study is cross-
sectional as far as the questionnaire data are concerned; all these data were collected at one
point in time. A consequence of this is that we cannot be confident about the causal
direction of the pathways included in the JD-R model. For example, it is conceivable that
exhaustion is not only a consequence, but also a cause of job demands (Demerouti et al.,
2004). Future longitudinal studies with a cross-lagged panel design are therefore desirable.
Note, however, that longitudinal studies have their own challenges, including attrition. Our
dropout analysis showed that the attrition in the present study was not random. The results
revealed that our participants scored higher than the dropouts on work pressure, lower on
harmony, and higher on exhaustion and cynicism. It is unclear how these differences have
affected our findings, but it is conceivable that the research sample showed a restriction of
range in the model variables, which may have made it more difficult to find effects. Finally,
the vast majority of the participants were female. It is therefore important to replicate the
current findings in more heterogeneous samples, including men and women, younger and
older employees, and different occupations.
Practical Implications
The results of the present study are valuable for human resource management policies
within this organization. Motivation (cynicism) appeared to be an important explanatory
455
460
465
470
475
480
485
490
495
Job demands, resources, and burnout 13
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:46
UNCORRECTED PROOF
variable for team performance. This suggests that it is profitable to invest in an increase of
the resources at the level of interpersonal relationships (i.e., between colleagues in the
teams). The resources ‘‘cohesion’’ and ‘‘harmony’’ appeared to be most strongly associated
with team performance, through cynicism. This is consistent with the assumptions of Shea
and Guzzo (1987), who argue that interdependence is one of the most important predictors
of team effectiveness. In addition, these findings are in line with the theoretical model of
Guzzo and Campbell (1990) which assumes that the provision of social resources by the
organization is an important determinant of objective team performance. Organizations
may mobilize such resources by investing in team development. Training sessions could
focus on the development of effective team communication and use roleplay to promote
team decision making and to facilitate collaboration between team members.
Acknowledgements
We thank Garry Hall for proofreading the manuscript.
References
Altman, D. G. (1991). Practical statistics for medical research. London: Chapman and Hall.
Arbuckle, J. L. (2005). Amos 6.0 user’s guide. Chicago, IL: SPSS, Inc.
Baba, W., Jamal, M., & Tourigny, L. W. (1998). Work and mental health: A decade of Canadian research. Canadian
Psychology, 39,94107.
Bakker, A. B., & Demerouti, E. (2007). The Job Demands-Resources model: State of the art. Journal of Managerial
Psychology, 22, 309328.
Bakker, A. B., Demerouti, E., De Boer, E., & Schaufeli, W. B. (2003). Job demands and job resources as predictors
of absence duration and frequency. Journal of Vocational Behavior, 62, 341356.
Bakker, A. B., Demerouti, E., & Schaufeli, W. B. (2002). The validity of the Maslach Burnout Inventory-General
Survey: An internet study. Anxiety, Stress, and Coping, 15, 245260.
Bakker, A. B., Demerouti, E., & Schaufeli, W. B. (2003). Dual processes at work in a call centre: An application of
the Job Demands-Resources model. Journal of Work and Organizational Psychology, 12, 393417.
Bakker, A. B., Demerouti, E., & Verbeke, W. (2004). Using the Job Demands-Resources model to predict burnout
and performance. Human Resource Management, 43,83104.
Bakker, A. B., Schaufeli, W. B., Sixma, H. J., Bosveld, W., & Dierendonck, D. (2000). Patient demands, lack of
reciprocity, and burnout: A five-year longitudinal study among general practitioners. Journal of Organizational
Behavior, 21, 425441.
Baron, R. M., & Kenny, D. A. (1986). The moderatormediator variable distinction in social psychological
research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51,
11731182.
Behrman, D. N., & Perreault, W. D. J. (1982). Measuring performance of industrial persons. Journal of Business
Research, 10, 355370.
Bhagat, R. S., Allie, S. M., & Ford, D. L. (1995). Coping with stressful events: An empirical analysis. In R.
Crandall, & P. L. Perrewe´ (Eds), Occupational stress: A handbook series in health psychology and behavioral medicine
(pp. 93112). Philadelphia, PA: Taylor & Francis.
Browne, M. W., & Cudeck, R. (1989). Single sample cross-validation indices for covariance structures. Multivariate
Behavioral Research, 24, 445455.
Campbell, J. P. (1990). Modeling the performance prediction problem in industrial and organizational psychology.
In M. D. Dunnette, & L. M. Hough (Eds), Handbook of industrial and organizational psychology (pp. 687732).
Palo Alto, CA: Consulting Psychologists Press.
Cropanzano, R., Rupp, D. E., & Byrne, Z. S. (2003). The relationship of emotional exhaustion to work attitudes,
job performance, and organizational citizenship behaviors. Journal of Applied Psychology, 88, 160169.
Demerouti, E., & Bakker, A. B. (2006). Employee well-being and job performance: Where we stand and where we
should go. In S. McIntyre, & J. Houdmont (Eds), Occupational health psychology: European perspectives on research,
education and practice, Vol. 1 (pp. 83111). Castelo da Maia, Portugal: ISMAI Publications.
Demerouti, E., Bakker, A. B., & Bulters, A. J. (2004). The loss spiral of work pressure, workhome interference
and exhaustion: Reciprocal relations in a three-wave study. Journal of Vocational Behavior, 64, 131149.
500
505
510
515
520
525
530
535
540
545
AQ2
14 A. B. Bakker et al.
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:46
UNCORRECTED PROOF
Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The Job Demands-Resources model of
burnout. Journal of Applied Psychology, 86, 499512.
Frazier, P. A., Tix, A. P., & Barron, K. E. (2004). Testing moderator and mediator effects in counseling psychology.
Journal of Counseling Psychology, 51, 115134.
Gardner, D. G., & Cummings, L. L. (1988). Activation theory and job design: Review and reconceptualization. In
B. M. Staw, & L. L. Cummings (Eds), Research in organizational behavior, Vol. 10 (pp. 81122). Greenwich, CT:
JAI Press.
Geurts, S. A. E., Taris, T. W., Kompier, M. A. J., Dikkers, J. S. E., Van Hooff, M. L. M., & Kinnunen, U. M.
(2005). Workhome interaction from a work psychological perspective: Development and validation of a new
questionnaire, the SWING. Work & Stress, 19, 319339.
Gladstein, D. (1984). Groups in context: A model of task group effectiveness. Administrative Science Quarterly, 29,
499517.
Guzzo, R. A., & Campbell, R. J. (1990). Conditions for team effectiveness in management. Paper presented at the
Academy of Management, San Francisco, CA.
Hackman, J. R., & Oldham, G. R. (1980). Work redesign. Reading, MA: Addison-Wesley.
Halbesleben, J. R. B., & Buckley, M. R. (2004). Burnout in organizational life. Journal of Management, 30, 859
879.
Harman, H. H. (1967). Modern factor analysis. Chicago: University of Chicago Press.
Hobfoll, S. E. (1989). Conservation of resources: A new attempt at conceptualizing stress. American Psychologist,
44, 513524.
Hockey, G. R. J. (1993). Cognitive-energetical control mechanisms in the management of work demands and
psychological health. In A. Baddely, & L. Weiskrantz (Eds), Attention: Selection, awareness, and control (pp. 328
345). Oxford: Clarendon Press.
Hoyle, R. H. (1995). The structural equation modeling approach: Basic concepts and fundamental issues. In R. H.
Hoyle (Ed.), Structural equation modeling, concepts, issues, and applications (pp. 115). Thousand Oaks, CA: Sage.
Jehn, K. A. (1994). Enhancing effectiveness: An investigation of advantages and disadvantages of value-based
intragroup conflict. The International Journal of Conflict Management, 3, 223228.
Jehn, K. A. (1995). A multi-method examination of the benefits and detriments of intragroup conflict.
Administrative Science Quarterly, 40, 256282.
Jo
¨
reskog, K. G., & So
¨
rbom, D. (1993). Lisrel 8: Structural equations modeling with the SIMPLIS Command Language.
Chicago: Scientific Software International.
Keijsers, G. J., Schaufeli, W. B., Le Blanc, P. M., Zwerts, C., & Reis Miranda, D. R. (1995). Performance and
burnout in intensive care units. Work & Stress, 9, 513527.
Lazaro, L., Shinn, M., & Robinson, P. E. (1985). Burnout, performance and job withdrawal behavior. Journal of
Health and Human Resources Administration, 7, 213234.
LeBlanc, P. (1994). De steun van de leiding: Een onderzoek naar het Leader Member Exchange model in de verpleging
[Support from management. A study on the Leader Member Exchange model among nurses]. Unpublished
doctoral thesis, Utrecht University, Utrecht.
Lee, R. T., & Ashforth, B. E. (1993). A further examination of managerial burnout: Toward an integrated model.
Journal of Organizational Behavior, 14,320.
Lee, R. T., & Ashforth, B. E. (1996). A meta-analytic examination of the correlates of the three dimensions of
burnout. Journal of Applied Psychology, 81, 123133.
Leiter, M. P. (1993). Burnout as a developmental process: Consideration of models. In W. B. Schaufeli, C.
Maslach, & T. Marek (Eds), Professional burnout: Recent developments in theory and research (pp. 237250).
Washington, DC: Taylor & Francis.
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, and Coping, 19, 129141.
Lykken, D. T. (1968). Statistical significance in psychological research. Psychological Bulletin, 70, 151159.
Marsh, H. W., Balla, J. R., & Hau, K. T. (1996). An evaluation of incremental fit indices: A clarification of
mathematical empirical properties. In G. A. Marcoulides, & R. E. Schumacker (Eds), Advanced structural
equation modeling, issues techniques (pp. 315353). Mahwah, NJ: Lawrence Erlbaum.
Maslach, C. (1993). Burnout: A multidimensional perspective. In W. B. Schaufeli, C. Maslach, & T. Marek (Eds),
Professional burnout: Recent developments in theory and research (pp. 1932). Washington, DC: Taylor & Francis.
Maslach, C., & Jackson, S. E. (1986). Maslach Burnout Inventory: Second edition. Palo Alto, CA: Consulting
Psychologists Press.
Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of Psychology, 52, 397422.
Motowildo, S. J., & Van Scotter, J. R. (1994). Evidence that task performance should be distinguished from
contextual performance. Journal of Applied Psychology, 79, 475480.
550
555
560
565
570
575
580
585
590
595
600
Job demands, resources, and burnout 15
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:46
UNCORRECTED PROOF
Muchinsky, P. M. (1993). Psychology applied to work. Pacific Grove, CA: Brooks/Cole.
Nowack, K. M., & Hanson, A. L. (1983). The relationship between stress, job performance, and burnout in college
student resident assistants. Journal of College Student Personnel, 24, 545550.
O’Leary Kelly, A. M., Martocchio, J. J., & Frink, D. D. (1994). A review of the influence of group goals on group
performance. Academy of Management Journal, 37, 12851301.
Parker, P. A., & Kulik, J. A. (1995). Burnout, self- and supervisor-related job performance, and absenteeism among
nurses. Journal of Behavioral Medicine, 18, 581599.
Podsakoff, P. M., & MacKenzie, S. B. (1994). An examination of the psychometric properties and nomological
validity of some revised and reduced substitutes for leadership scales. Journal of Applied Psychology, 79, 702713.
Posner, B. Z., & Kouzes, J. M. (1988). Development validation of the Leadership Practices Inventory. Educational
and Psychological Measurement, 48, 483496.
Posner, B. Z., & Kouzes, J. M. (1994). An extension of the Leadership Practices Inventory to individual
contributors. Educational and Psychological Measurement, 54, 959966.
Randall, M., & Scott, W. A. (1988). Burnout, job satisfaction, and job performance. Australian Psychology, 23, 335
347.
Roe, R. A. (1999). Work performance: A multiple regulation perspective. In C. L. Cooper, & I. T. Robertson
(Eds), International review of industrial and organizational psychology, Vol. 14 (pp. 231335). Chichester: Wiley.
Schaufeli, W. B., & Enzmann, D. T. (1998). The burnout companion to study and practice: A critical analysis. London:
Taylor & Francis.
Schaufeli, W. B., Leiter, M. P., Maslach, C., & Jackson, S. E. (1996). The MBI-General Survey. In C. Maslach, S.
E. Jackson, & M. P. Leiter (Eds), Maslach Burnout Inventory Manual (3rd edn, pp. 1926). Palo Alto, CA:
Consulting Psychologists Press.
Schaufeli, W. B., & Van Dierendonck, D. (2000). Utrechtse burnout schaal: Handleiding. [Utrecht burnout scale:
Manual]. Lisse, The Netherlands: Swets & Zeitlinger.
Schutte, N., Toppinnen, S., Kalimo, R., & Schaufeli, W. B. (2000). The factorial validity of the Maslach Burnout
Inventory-General Survey across occupational groups and nations. Journal of Occupational and Organizational
Psychology, 73,5366.
Shea, G. P., & Guzzo, R. A. (1987). Groups as human resources. In K. M. Rowland, & G. R. Ferris (Eds), Research
in personnel human resources management (pp. 323356). Greenwich, CT: JAI Press.
Shirom, A. (2005). Reflections on the study of burnout. Work & Stress, 19, 263270.
Singh, J., Goolsby, J. R., & Rhoads, G. K. (1994). Behavioral and psychological consequences of boundary
spanning burnout for customer service representatives. Journal of Marketing Research, 31, 558569.
Sternberg, R. J. (1994). The PRSVL model of personcontext interaction in the study of human potential. In M.
G. Rumsey, C. B. Walker, & J. H. Harris (Eds), Personnel selection and classification (pp. 317332). Hillsdale, NJ:
Lawrence Erlbaum.
Taris, T. W. (2006). Is there a relationship between burnout and objective performance? A critical review of 16
studies. Work & Stress, 20, 316334.
Taris, T. W, LeBlanc, P. M., Schaufeli, W. B., & Schreurs, P. J. G. (2005). Are there causal relationships between
the dimensions of the Maslach Burnout Inventory? A review and two longitudinal tests. Work & Stress, 19, 238
255.
Ursprung, A. W. I. (1986). Incidence and correlates of burnout in residential service settings. Rehabilitation
Counseling Bulletin, 29, 225239.
Van Veldhoven, M., & Meijman, T. F. (1994). Het meten van psychosociale arbeidsbelasting met een vragenlijst: De
vragenlijst beleving en beoordeling van de arbeid [The measurement of psychosocial work demands with a
questionnaire]. Amsterdam: NIA.
Veldhuizen, I. J. T., Gaillard, T., & De Vries, A. W. K. J. (2003). The influence of mental fatigue on facial EMG
activity during a simulated workday. Biological Psychology, 63,5978.
Wright, T. A., & Bonett, D. G. (1997). The contribution of burnout to work performance. Journal of Organizational
Behavior, 18, 491499.
Wright, T. A., & Cropanzano, R. (1998). Emotional exhaustion as a predictor of job performance and voluntary
turnover. Journal of Applied Psychology, 83, 486493.
Zedeck, S., Maslach, C., Mosier, K., & Skitka, L. (1988). Affective response to work and quality of family life:
Employee and spouse perspectives. Journal of Social Behavior Psychology, 3, 135157.
Zwerts, C., Schaufeli, W. B., Keijsers, G., LeBlanc, P. M., & Reis Miranda, D. (1995). Burnout en prestatie in
intensive care units [Burnout and performance in intensive care units]. Tijdschrift Sociale Gezondheidszorg, 73,
382389.
605
610
615
620
625
630
635
640
645
650
655
660
16 A. B. Bakker et al.
C:/3B2WIN/temp files/GASC296029_S100.3d[x] Monday, 3rd March 2008 21:40:47