Available online at www.sciencedirect.

com

Journal of Vocational Behavior 71 (2007) 446459

www.elsevier.com/locate/jvb

Psychological strain and emotional labor

among police-ocers: A diary study
Benjamin van Gelderen a, Ellen Heuven b,*, Marc van Veldhoven c,
Marcel Zeelenberg c, Marcel Croon c
b

a
Dutch Police Force, Politie Midden en West Brabant, The Netherlands
Department of Social and Organizational Psychology, Utrecht University, Heidelberglaan 1,
P.O. Box 80.140, 3508 TC, Utrecht, The Netherlands
c
Faculty of Social and Behavioural Sciences, Tilburg University, The Netherlands

Received 27 August 2007

Available online 14 September 2007

Abstract
The authors examined the relationship between psychological strain, emotional dissonance and
emotional job demands during a working day of 65 Dutch (military) police ocers, using a 5-day
diary design. We hypothesized that emotional dissonance partly mediated the relationship between
psychological strain at the start and at the end of a work shift. We also tested the mediating role of
emotional dissonance between emotional job demands and psychological strain at the end of a work
shift. Results of structural equation modeling analyses showed that psychological strain at the start
of a work shift had a positive eect on the experience of emotional dissonance and psychological
strain at the end of a work shift. Emotional dissonance partly mediated the relationship between psychological strain at the start and psychological strain at the end of a work shift. Results are discussed
in light of conservation of resources theory (Hobfoll, S. E. (1988). The ecology of stress. New York:
Hemisphere Publishing Corporation).
2007 Elsevier Inc. All rights reserved.
Keywords: COR theory; Emotional dissonance; Emotional labor; Police ocers; Psychological strain

Corresponding author. Fax: +31 30 2534718.

E-mail address: [email protected] (E. Heuven).

0001-8791/$ - see front matter 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jvb.2007.09.001

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

447

1. Introduction
Emotional labor refers to how employees regulate their emotions as part of the work
role, and their consequences of doing so (Hochschild, 1983). Most emotional labor
research has been conducted within the human service industry, in which human interactions play an important part of the job. Emotional labor is mainly instigated by implicit or
explicit display rules that state which emotions are appropriate for employees to express.
As a result of obeying these rules, employees who engage in emotion work may experience
emotional dissonance, which refers to a discrepancy between felt and displayed emotions
(Hochschild, 1983). Emotional dissonance is considered to be an important predictor of
impaired psychological well-being in the form of emotional exhaustion and depersonalization (Ashfort & Humphrey, 1993; Brotheridge & Lee, 1998, 2003; Grandey, 2000; Heuven
& Bakker, 2003; Zapf, 2002).
The present study among police ocers examined how psychological strain at the start of
a work shift may increase the likelihood of emotional dissonance during the shift. The innovation of our research was found in the addition of psychological strain at the start of a
work shift. We came to this based on the theoretical assumptions regarding the conservations of resources (COR) theory (Hobfoll, 1989). Following COR theory, we argued that
a higher level of psychological strain at the start of a work shift will aect the experience
of emotional dissonance; a higher level of psychological strain at the start of a work shift
will leave less energy for regulating ones emotions during work. Also, the mediating role
of emotional dissonance, mediating psychological strain at the start and at the end of a
work shift, was explored. The idea was that this initial state may steam through the experiences of the police ocers during the rest of the work shift and may increase the level of
emotional dissonance and psychological strain at the end of a work shift. Past research did
examine the relationship between strain, task-regulation and task-performance (e.g.,
Schellekens, Sijtsma, Vegter, & Mijeman, 2000), but put less emphasis on the eects of initial strain. This hitherto unstudied temporal pattern oered an interesting addition in studying the eects of emotional dissonance within emotional labor research. In addition, we
examined the mediating role of emotional dissonance in the relationship between emotional
job demands and psychological strain at the end of a work shift. It was expected that emotional demanding situations would contribute to the arise of emotional dissonance, which in
turn would lead to a higher level of psychological strain at the end of a work shift.
1.1. Emotion regulation among police ocers
Police ocers are a very relevant and highly intriguing population to study. In performing police work many emotionally demanding interactions take place. On a daily basis
police ocers are confronted with human sorrow in the form of violence, aggression, accidents, crime victims and death (e.g., Brown & Campbell, 1990). Since police ocers constantly have to show the right emotions in order to keep up a professional appearance and
achieve organizational goals, the management of emotions holds a central role in conducting police work. It is therefore not surprising that the expression and suppression of a wide
variety of emotions is an important element in the performance of police work.
Most service jobs require employees to suppress negative emotions and express positive
emotions. Expressing these socially desired emotions are assumed a key dimension of
delivering a high-quality service and reach a higher level of customer satisfaction (Tsai

448

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

& Huang, 2002). However, in comparison with most other service occupations, a typical
characteristic of police work is, that police work calls for constantly switching between
more dierent types of emotional expressions depending on the situation. Next to the
expression of positive or neutral emotions, police ocers sometimes are required to suppress positive emotions and express negative emotions (Rafaeli & Sutton, 1991). For
example police ocers are required to express anger when correcting an oender, while
at the successive moment they should be able to simulate sympathy for a crime victim.
1.2. Emotional dissonance and psychological strain
Police work is demanding in many ways. The experience of emotional dissonance by
police ocers is clearly contributing to this. The causes and consequences of emotional
dissonance is the topic of several theoretical approaches, that will be addressed in this section. First, emotional dissonance can be seen as a dependent variable caused by dierent
factors that pressure the employee to identify to his work role and withholds him or her to
feel the appropriate emotions that ought to be displayed (Ashfort & Humphrey, 1993).
Other authors treated emotional dissonance as one of the dimensions of emotion work
itself (Kruml & Geddes, 2000; Morris & Feldman, 1997) or as a feeling of unease arising
when the felt emotion is, according to internal standards, evaluated as a threat to the persons own identity (Jansz & Timmers, 2002). According to Zapf (2002), emotional dissonance can also be seen as a regulation problem and a work stressor. The work situation
prescribes a certain emotion expression from the employee, irrespective of whether the
actual emotion is felt. The organizational prescription of emotions that ought to be
expressed, is also referred to as feeling rules or display rules (Ekman, 1973). In conducting
police work, a police ocer has to pay attention to the elements of the job itself, for example writing down a statement of a victim. Simultaneously, the police ocer also has to regulate his or her emotional expression, for example showing sympathy for the victim. When
this emotion is not actually felt, emotional dissonance will be experienced. Emotional dissonance is expected to be related to emotional exhaustion and depersonalization (Zapf,
2002). Dierent empirical studies showed a positive relationship between emotional dissonance and elements of burnout (Bakker & Heuven, 2003, 2006; Brotheridge & Grandey,
2002; Zapf, Vogt, Seifert, Mertini, & Isic, 1999). Emotional dissonance could thus be
assumed an important predictor of the experience of psychological strain.
Hypothesis 1. The experience of emotional dissonance during a shift will be positively
related to psychological strain at the end of a work shift.
1.3. Psychological strain at the start of a work shift
A police ocer who starts a work shift, might for dierent reasons already have an
increased level of psychological strain. This increased level of psychological strain can
for example be a result of inadequate sleep or personal problems in private life. It was
expected that this initial level of psychological strain at the start of a work shift played
an important role in experiencing emotional dissonance during work and psychological
strain at the end of a work shift.
The Conservation of Resource (COR) approach oers an integrative stress theory that
considers external environmental processes as well as internal ones (Hobfoll, 1988, 1989).

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

449

Resources can be dened as the things that are important to people or that help them to
attain valuable things (Hobfoll, 1988). Some examples of resources are the amount of
energy, endurance, time for adequate sleep, free time, personal health and motivation to
get things done. When a police ocer for example has no time for adequate rest, this
can result in lower energy levels at the start of a work shift. This might increase the level
of psychological strain at the start of a work shift.
One of the main principles of the COR theory is that starting a job with a low level of
resources will lead to a future loss spiral (Hobfoll, 1989, 2001). According to COR theory,
it could be assumed that those police ocers with higher psychological strain at the start
of a work shift were more vulnerable to losses throughout the work shift. Resources can
be important to cope with the dierent job demands and stressors during the rest of the shift,
by buering the impact of job demands on burnout (Bakker, Demerouti, & Euwema, 2005).
Expending eort in the form of emotional dissonance would draw further on the resources of
the individual police ocer and, in line with COR theory, contributed to the described loss
spiral during the work shift. This eort expenditure could lead to a further depletion of the
total amount of available resources. Starting with lower resources, it is more dicult preventing losses occurring during the shift. This presumption was also supported by a study of
Demerouti, Bakker, and Bulters (2004), which showed that work pressure, workhome interference and exhaustion, predict each other over time. Police ocers who have less energy at
the start of their work shift, are thus more likely to lose energy than colleagues who start their
working day freshly. The importance of having enough resources during work, is also in line
with the Demand-Control model (DCM) (Karasek, 1979) and more specically, the Job
Demands-Resources Model (JD-R) (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001).
The Demand-Control model stated that stress is a result of a high work pressure combined
with a lack of resources like job autonomy and regulation possibilities (Karasek, 1979). The
JD-R model assumed that a combination of high job demands and limited resources will lead
to the development of burnout. As a result of higher psychological strain at the start of a
work shift, energy-related resources were expected to be low. Combined with mental
demands in the form of emotional dissonance, a higher likelihood of psychological strain
at the end of a work shift was expected. The level of psychological strain at the start of a work
shift, might thus partly predicted the level of psychological strain that was experienced at the
end of a work shift.
Moreover, it was also assumed that psychological strain at the start of a shift will not only
lead to more psychological strain at the end of a shift, but also made police ocers more vulnerable to the experience of emotional dissonance. This assumption is consistent with the
COR theory, stating that people low on resources are more vulnerable to resource losses
and are likely to adopt a defensive and protective attitude (Hobfoll, 2001). Also, there is less
energy left for emotion regulation. People who feel more exhausted are less able to put much
eort in regulating their emotions (Aspinwall & Taylor, 1997). In the end, a lack of resources
may lead to an impaired psychological well-being (Hobfoll, 2001; Maslach, Schaufeli, &
Leiter, 2001).
Because of the described loss spiral and energy-related losses in the form of emotional dissonance occurring during the work shift, emotional dissonance was supposed to partially
mediate the relationship between psychological strain at the start and at the end of a work shift.
Hypothesis 2. Emotional dissonance will partly mediate the relationship between psychological strain at the start of a work shift and psychological strain at the end of a work shift.

450

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

1.4. Emotional job demands and psychological strain

Police ocers are confronted with a high amount of demanding situations. Demanding situations can for example consist of a high workload or dicult client interactions.
Police ocers frequently have to deal with intense interactions with civilians or suspects. Continuously dealing with emotionally charged interactions has been traditionally proposed as a primary source of burnout complaints among human service
employees in general (Lee & Ashforth, 1996; Maslach & Jackson, 1984). A work environment with demanding situations may lead to an increase of emotional exhaustion, a
component of burnout (Leiter & Maslach, 1988). Research of Dorman and Zapf (2004)
showed that next to the inuence of emotional labor on burnout, other customerrelated social aspects can also be important. Verbal aggressive clients and unpleasant
customers turned out to form a serious component in predicting burnout (Dorman
& Zapf, 2004). In addition, dealing with these kind of emotional job demands in police
work, was expected to be directly related to psychological strain at the end of a work
shift.
Hypothesis 3. Emotional job demands will be positively related to psychological strain at
the end of a work shift.
Besides emotional job demands, job related stressors are also supposed to form a
predictor of burnout complaints. Especially among police ocers, they were found
to be an even more important predictor than emotional job demands (Brown &
Campbell, 1990). The experience of emotional job demands was expected to inuence the level of emotional dissonance. Next to the primary task of dealing with
dicult clients and emotional demanding situations, police ocers might have to
regulate their emotions as part of their job. This in turn, made the arise of emotional dissonance more likely. Within emotional labor research this is a common
assumption based on the fact that emotional demanding situations will trigger more
emotional reactions that require the police ocers to fake or suppress certain emotions (e.g., Bakker & Heuven, 2006; Grandey, 2000; Heuven, Bakker, Schaufeli, &
Huisman, 2006).
In addition, emotion work can be seen as a secondary task in support of the main
task of dealing with emotional demanding situations (Zapf, 2002). A police ocer has
to deal with the demanding situation and can at the same time experience feelings of
emotional dissonance. A study conducted by (Zapf, Seifert, Schmutte, Mertini, &
Holz, 2001), showed the importance of a combined eect of emotion work variables
and other stressors. Emotional job demands combined with emotional dissonance can
result in even higher levels of burnout. It was thus expected that emotional job
demands are positively related to psychological strain, partly through the mediating
role of emotional dissonance.
Hypothesis 4. Emotional dissonance will mediate the relationship between emotional job
demands and psychological strain at the end of a work shift.
Fig. 1 summarizes the combined hypothesized relationships and describes the dierent
paths from psychological strain at the start of a work shift (Time 1) to the dependent variable psychological strain at the end of a work shift (Time 2).

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

Emotional Job
Demands

.102*

.159**

Emotional Dissonance
.
Psychological Strain T1

451

Psychological Strain T2
.

.405**
.388**

Fig. 1. Signicant (unstandardized) regression coecients in nal Time-invariant model. Note. *p < .05, **p < .01.

1.5. The current research

Study hypotheses were tested using a diary study. More specically, subjects received a
questionnaire in which questions could be answered at dierent occasions during the day,
during consecutive days or directly after the measured experiences take place. The main
advantage of a diary study is, that experiences of the respondents can be examined in their
natural context. Also the time between the experiences and documentation of these experiences is shorter than with a regular survey (Bolger, Davis, & Rafaeli, 2003). This made
the measurement more accurate, which can be especially advantageous in measuring psychological strain. In addition, a diary study makes it possible not only to measure dierences between persons but also to measure intra-individual dierences. This oered the
opportunity to look at the development of measures over a certain time frame. The diary
study technique used in this study is the Experience Sampling Method (ESM). Within the
ESM questions are being answered during the work performance (Bolger et al., 2003).
2. Methods
2.1. Participants
One hundred and fty Dutch police ocers and 70 Dutch military police ocers were
asked to participate. The police ocers were working in a police district in the south of the
Netherlands. The work of these participants mainly consisted of community policing or
oering rst aid police service to civilians. The military police ocers were working at
the Dutch National Airport and most of their work consisted of police duties. All the participants were conned to tasks in which there was a high interaction rate with recipients.
Participation on this study was condential and voluntary. Of the 150 police ocers 28.7%
(N = 43) completed the diary study questionnaires. Of the military police ocers the
response rate was 31% (N = 22).
2.2. Procedure
Participants were approached in person or by telephone. They were informed about the
study and were asked to participate. Participants were told that a study about the regulation of emotions and its eects on psychological strain among police ocers was being
conducted. They were asked to ll in a 5-day diary study. In this diary study questions
had to be answered at the beginning and at the end of a shift during ve consecutive shifts.
Those who were willing to participate were given or sent a dairy study booklet and a

452

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

return envelope. With this envelope the participants were able to condentially return the
questionnaire to the University.
2.3. Measures
In the diary questionnaires, participants responded to scales assessing psychological
strain, emotional job demands and emotional dissonance. The scales are discussed below
and the exact questions are found in the Appendix A. All questions were to be answered
on 7-point scales, with endpoints labeled no thats not correct (1), and yes thats correct (7).
2.3.1. Psychological strain time 1 (5 items, a = .68); psychological strain time 2 (5 items,
a = .63)
Psychological strain was measured twice, at the beginning and at the end of a work
shift. For measuring psychological strain, the same measure was used on both occasions.
In measuring psychological strain elements of exhaustion and disfunctional attitudes concerning work were taken into account. These two dimensions can be considered as the
most important elements of burnout (Demerouti et al., 2001). Psychological strain was
measured using questions, adopted from the Utrecht Burnout Scale (UBOS) (Schaufeli
& van Dierendonck, 2000), the Oldenburg Burnout Inventory OLBI (Demerouti et al.,
2004) and the Checklist Individual Strenght (CIS-20) (Vercoulen, Alberts, & Bleijenberg,
1999).
2.3.2. Emotional job demands (7 items, a = .77)
The emotional job demands scale was a self developed scale based on prominent categories of civilians and suspects with whom the police ocers have to deal with, during
their duties. The categories were based on common emotional demanding interactional
experiences of Dutch police ocers. Emotional job demands were measured at the end
of a shift.
2.3.3. Emotional dissonance (4 items, a = .84)
Emotional dissonance was measured at the end of a shift, using the emotional dissonance scale of Erickson and Wharton (1997). Questions concerned the suppression or faking of feelings.
2.4. Statistical analyses
The conceptual model described in Fig. 1, hypothesizes certain causal relationships
among the core variables in this study. It makes clear that psychological strain at the start
of each shift (T1) has an eect on emotional dissonance experienced during that shift. Psychological strain at the start of a shift also has an eect on psychological strain measured
at the end of the shift (T2). The model stipulates further that rst, the relationship between
the two measures of psychological strain are partially mediated by emotional dissonance.
Second, the relationship between emotional job demands and psychological strain at the
end of a shift is partially mediated by emotional dissonance.
The same conceptual model is assumed to apply at each of the ve shifts for which data
on the core variables are available. It is important to note that the conceptual model only

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

453

species the relationships among the variables from the same shift, but lets the relationships among variables across shifts unspecied.
On the basis of the conceptual model described above, a structural model (Kline, 2005)
was dened for all variables from the ve shifts. Next to the hypothesized relationships,
the rst analysis took the relationship between all the mentioned variables into account.
In the next step, the model was further simplied to test the described conceptual model.
For each shift a path model consisting of three regression equations was formulated. In the
rst equation emotional job demands was regressed on psychological strain at T1. In the
second equation emotional dissonance was regressed on psychological strain at T1 and
emotional job demands. Finally, in the third equation, psychological strain at T2 was
regressed on psychological strain at T1, emotional job demands and emotional dissonance
(Baron & Kenny, 1986). Since the conceptual model did not specify anything about the
relationship among variables measured at dierent shifts, the corresponding covariance
parameters were left free to vary.
In a rst analysis, the complete path model (Model S) was tted to the data without
imposing any constraint on the regression parameters across the shifts. Since the covariances between variables measured at dierent shifts were left free to vary, model S is a saturated model against which the t of more restrained submodels can be tested by means of
conditional likelihood tests. In a second analysis, the same model was tted under the constraint that corresponding unstandardized regression parameters were set equal across
shifts (Model T). In this way, it was tested whether the parameter estimates in the structural model were invariant over shifts. Since the variances of the variables involved may
slightly uctuate over time, equality of unstandardized regression coecients does not
imply equality of standardized regression coecients. Based on the results of the second
analysis, a third analysis was carried out in which the non-signicant shift-invariant
regression parameters were set equal to zero (Model U). It was hoped that in this way
a parsimonious path model could be obtained, whose results were easy to interpret.
Finally, the model was further simplied by removing the arrow between psychological
strain at T1 and emotional job demands. All analyses were carried out using AMOS 6.0
(Arbuckle, 2003). The t of model T was assessed by a conditional likelihood ratio test
against the saturated model S. The t of model U was assessed by a conditional likelihood
ratio test against model T.
Before carrying out the SEM analyses, missing values in the data were regression
imputed using the Missing Value Analysis procedure from SPSS. From a total of 1290
scores, 141 were missing (11%). A missing value was replaced by a score dened by its
expected values under a regression model in which only the variables observed at the same
shift were used as predictors to which a randomly selected regression residual was added.
3. Results
Table 1 shows the means, standard deviations and productmoment correlations of all
the major variables. The variables were measured on ve dierent shifts. Results showed
however that the measures stayed constant over time and no intra-individual development
on the measures during the ve shifts took place. This means that the relationships
between the measured variables stayed constant, independent of the specic work shift.
Although the t of the saturated model S, having zero degrees of freedom, could not be
assessed, this model acted as a baseline model, against which the t of the restricted time-

454

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

Table 1
Means, standard deviations, and correlation matrix among variables for each shift

Shift 1

Shift 2

Shift 3

Shift 4

Shift 5

*
**

Variable

SD

Emotional job demands

Emotional dissonance
Psychological strain time 1
Psychological strain time 2

2.20
1.84
1.86
2.49

1.12
.97
.84
.86

.47**
.28*
.25

.59**
.47**

.47**

Emotional job demands

Emotional dissonance
Psychological strain time 1
Psychological strain time 2

2.41
1.79
1.92
2.68

1.23
1.06
.93
1.00

.14
.12
.04

.43**
.42**

.58**

Emotional job demands

Emotional dissonance
Psychological strain time 1
Psychological strain time 2

2.02
1.84
2.10
2.89

1.16
1.16
1.04
1.13

.01
.09
2.97*

.34*
.45**

.51**

Emotional job demands

Emotional dissonance
Psychological strain time 1
Psychological strain time 2

2.71
1.90
2.02
2.66

1.50
1.24
.91
1.11

.31*
.08
.06

.20
.33*

.48**

Emotional job demands

Emotional dissonance
Psychological strain time 1
Psychological strain time 2

2.21
1.90
2.20
2.74

1.25
1.47
1.18
1.22

.19
.07
.07

.07
.30*

.54**

p < .05.
p < .01.

or shift-invariant model T could be tested. The conditional likelihood was v2 = 36.505

with df = 24 and p = .049, RMSEA = .09, Pclose = .139. Model U was obtained by setting the non-signicant regression parameters in model T equal to zero. This was the case
for two of the six coecients: psychological strain at T1 had no signicant eect on emotional job demands and emotional job demands itself had no signicant eect on psychological strain at T2. The conditional likelihood ratio test of model U against model T
yielded v2 = 1.182 with df = 2 and p = .554, for model U it was also found that
RMSEA = .084, Pclose = .176.
We simplied the model further by removing the arrow from psychological strain T1 to
emotional job demands, the analysis results in v2 = 36.669 with df = 25 so that p = .062.
The conditional likelihood ratio test of this model against the original one was equal to
0.164 with df = 1 so that p = .67, showing that the more restrictive model did not yielded
a signicantly worse t. The outcomes of both conditional likelihood ratio tests indicated,
that the results obtained under model U could be used as a basis for substantive conclusions. Fig. 1 depicts the nal parsimonious model together with the unstandardized regression parameters for the signicant eects.
Emotional dissonance was positively related to psychological strain at time 2 (b = .159,
p < .01) which gives support to the rst hypothesis. Psychological strain at time 1 was positively related to emotional dissonance (b = .405, p < .01) and psychological strain at time
2 (b = .388, p < .01), but no signicant relationship was found with emotional job
demands. Emotional dissonance is a mediator between psychological strain at T1 and
at T2, although a strong direct eect of psychological strain at T1 on its measurement

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

455

at T2 remains present. This is in support of H2. No direct relation was found between
emotional job demands and psychological strain on time 2 (b = .038, p > .05). The third
hypothesis was rejected. Emotional job demands were positively related to emotional dissonance (b = .102, p < .05). Emotional dissonance completely mediated the relationship
between emotional job demands and psychological strain at time 2. Moreover, the relationship between emotional demands and psychological strain at T2 was entirely due to
the mediating role of emotional dissonance. This was in support of H4.
4. Discussion
First, the results showed that during the ve work shifts the relations between all examined variables stayed constant over time. This implied that within person variations over
time were not signicantly present.
Second, results suggested that police ocers who started the work shift with more
psychological strain were more vulnerable to the experience of emotional dissonance
than the police ocers who started the shift with less psychological strain. This eect
supported the COR theory (Hobfoll, 1989), stating that people who have an increased
level of psychological strain at the start of a shift are more vulnerable to future losses.
This result may implicate that in studying emotional dissonance, psychological strain at
the start of the work shift may partly explain the individuals sensitivity to experiencing
emotional dissonance. It might be claried by the fact that people who feel more
exhausted are less able to put much eort in regulating their emotions (Aspinwall &
Taylor, 1997). Furthermore, emotional dissonance costs energy and might prevent
the acquisition of other energy-related resources. As one of the underlying elements
in anticipatory coping (Aspinwall & Taylor, 1997), it is less likely that a lack of
resources will lead to much eort in regulating ones emotions. So, not being able to
put enough eort in the regulation process makes the experience of emotional dissonance worse and the arise of it even more likely. Also the experience of psychological
strain at the start of a work shift can lead to a more negative attitude. When people
feel negative emotions, they seem to put top priority and energy in feeling better and
thereby abandon self-control (Tice, Baumeister, & Zhang, 2004). This makes the regulation of emotions even more dicult and will make police ocers more vulnerable to
the experience of emotional dissonance.
Emotional dissonance appeared to mediate the relationship between emotional job
demands and psychological strain at the end of a work shift, but the relationship was
not very strong. More interestingly though, was the eect of psychological strain at the
start of a work shift and the mediating eect of emotional dissonance. Emotional dissonance partly mediated the relationship between psychological strain at the start of a work
shift and psychological strain at the end of a work shift. In contrast, no direct relationship
was found between emotional job demands and psychological strain at the end of a work
shift. It appeared, that emotional dissonance as a work stressor has a greater impact on the
arise of psychological strain. An explanation for not nding the same eect of emotional
job demands might have been, that job demands form a basic part of the work of police
ocers. It is something they choose to do and are professionally trained for. This is in line
with ndings of Brown and Campbell (1990) and Biggam, Power, MacDonald, Carcary,
and Moodie (1997), who supported that, the elemental contents of police work were
not experienced as the main stressors for police ocers.

456

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

The partly mediating eect of emotional dissonance between psychological strain at the
start and at the end of a work shift showed, that emotional dissonance oered a contribution in increasing the level of psychological strain during a work shift. The experience of
emotional dissonance as a resource loss, resulted in more psychological strain at the end of
a shift. This nding is in line with the COR theory (Hobfoll, 2001) as well as the JD-R
model (Demerouti et al., 2001). Police ocers with a higher level of psychological strain
at the start of a shift, possessed less energy-related resources. This resulted in more
resource losses and a lesser change of reaching gains during the work shift, nally resulting
in a higher level of psychological strain.
4.1. Strength, limitations and implications for further research
The present study oered an interesting insight in the concept of emotional labor
among police ocers. It emphasized the important role of psychological strain at the start
of a work shift in studying the implications of emotional dissonance. Moreover, an important strength of this study was the use of a diary study, thus measuring the eects in its
natural context and on the moment the measured variables were in real experienced.
The results of this study oered some interesting ndings for the practice of police
work. Emotional dissonance at work clearly contributed to the level of psychological
strain of police ocers. The training of police ocers, to not just deal with the work itself,
but also to understand the simultaneously arising emotional regulatory demands, might
benet their overall psychological well-being. Second, early identication and assistance
of police ocers with higher levels of psychological strain in general, oers an opportunity
to make this category less vulnerable to energy losses during their work shift.
Some limitations of this study should be noted. First, the rather small number of
respondents who contributed to this diary study can aect the generalizability of our
ndings.
Second, the inuence of energy-related gains during a work shift were not taken into
account. Only demands like job demands and emotional dissonance that were felt during
the shift, were taken into consideration. It is interesting to expand future research, with
taking into account some important gains that might arise during a work shift. Even
for recipients who start their shift with more psychological strain and are experiencing
emotional dissonance, it should be theoretically possible to reach gains that high, that
the eect on psychological strain at the end of a shift, might be more in favor of the individual employee. It could be questioned in what way the balance would end up, when the
amount of psychological strain during the shift is signicantly decreased, as a result of
greater gains and what gains would be important for police ocers to turn back the negative eects of psychological strains.
Third, emotional dissonance was measured looking at the suppression and faking of
emotions in general. The use of discrete emotions and a further renement of the dierent
forms of emotional dissonance, might show separate eects of suppressing and faking
these emotions and can have unequal results on psychological strain.
In addition, in testing the eects of emotional job demands, a further renement of psychological strain into emotional exhaustion and depersonalization, ought to be made.
Also, the important role of emotional labor and emotion regulation for police ocers
was partially demonstrated. In interacting with recipients future research should expand
beyond emotional dissonance in testing the eects of emotion regulation techniques and

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

457

psychological strain among police ocers more general. Testing these eects in relationship with the psychological eects of the responses of interaction partners might be
benecial.
Notwithstanding the described limitations, it should be noted however, that this study
oered an interesting extension in emotional labor research and especially in examining
the eects of emotional dissonance. This study, showed on a day-to-day basis, the importance for police ocers to be aware of taking adequate rest and recovery. Recovery and
loading the batteries can lead to a desired energy level that is needed to make a good performance during the next work shift. These insights can thus contribute in creating and
maintaining a healthier police force.
Appendix A
Items for each measured variable
Emotional dissonance (a = .84)
Today I had the feeling that in contact with suspects/civilians. . .. . .
1. Could not be myself.
2. Was forced to fake certain feelings.
3. Appeared dierent than I felt.
4. Suppressed my own feelings to fake a neutral appearance.
Emotional job demands (a = .77)
During this shift I came in contact with. . .. . .
1. Condescending suspects/civilians.
2. Mentally disturbed suspects/civilians.
3. Unhygienic or contagiously diseased suspects/civilians.
4. Verbal intimidating suspects/civilians.
5. Physically threatening suspects/civilians.
6. Obtrusive suspects/civilians.
7. Suspects/civilians under inuence of alcohol or drugs.
Psychological strain at time 1 (a = .68)
On this moment I feel. . .. . .
1. I am tired now.
2. I feel exhausted.
3. The things I do at work do not aect me at all.
4. I plan to work on the automatic pilot today.
5. I question the things I do.
Psychological strain at time 2 (a = .63)
On this moment I feel. . .. . .
1. I am tired now.
2. I feel exhausted.
3. The things I did today at work did not aect me at all.
4. Today I worked on the automatic pilot.
5. I question the things I did today.
Note. All questions were to be answered on 7-point scales (1 = no that is not correct, 7 = yes that is correct).

458

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

References
Arbuckle, J. L. (2003). Amos users guide version 5.0. Chicago: Smallwaters Corp..
Ashfort, B. E., & Humphrey, R. H. (1993). Emotional labor in service roles: The inuence of identity. Academy of
Management Review, 18, 88115.
Aspinwall, L. G., & Taylor, S. E. (1997). A stitch in time: Self-regulation and proactive coping. Psychological
Bulletin, 121, 417436.
Bakker, A. B., Demerouti, E., & Euwema, M. C. (2005). Job resources buer the impact of job demands on
burnout. Journal of Occupational Health Psychology, 10, 170180.
Bakker, A. B., & Heuven, E. (2003). Emotional dissonance and burnout among cabin attendants. European
Journal of Work and Organizational Psychology, 12, 81100.
Bakker, A. B., & Heuven, E. (2006). Emotional dissonance, burnout, and in-role performance among nurses and
police ocers. International Journal of Stress Management, 13, 423440.
Baron, R. M., & Kenny, D. A. (1986). The moderatormediator variable distinction in social psychology
research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology,
51, 11731182.
Biggam, F. H., Power, K. G., MacDonald, R. R., Carcary, W. B., & Moodie, E. (1997). Self-perceived
occupational stress distress in a Scottish police force. Work & Stress, 11, 118133.
Bolger, N., Davis, A., & Rafaeli, E. (2003). Diary methods: Capturing life as it is lived. Annual Review
Psychology, 54, 579616.
Brotheridge, C. M., & Lee, R. T. (1998). On the dimensionality of emotional labour: Development and validation of
the Emotional Labour Scale. Paper presented at the rst conference on emotions in organizational life, San
Diego.
Brotheridge, C. M., & Grandey, A. A. (2002). Emotional labor and burnout: comparing two perspectives of
people work. Journal of Vocational Behaviour, 60, 1739.
Brotheridge, C. M., & Lee, R. T. (2003). Development and validation of the emotional labor scale. Journal of
Occupational and Organizational Psychology, 76, 365379.
Brown, J. M., & Campbell, E. A. (1990). Sources of occupational stress in the police. Work & Stress, 4,
305318.
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.
Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demands-resource model of
burnout. Journal of Applied Psychology, 86, 499512.
Dorman, C., & Zapf, D. (2004). Customer-related social stressors and burnout. Journal of Occupational Health
Psychology, 9, 6182.
Ekman, P. (1973). Cross culture studies of facial expression. In P. Ekman (Ed.), Darwin and facial expression: A
century of research in review.
Erickson, R. J., & Wharton, A. S. (1997). Inauthenticity and depression: Assessing the consequences of
interactive service work. Work and Occupations: An international Sociological Journal, 24, 188213.
Grandey, A. A. (2000). Emotion regulation in the workplace: A new way to conceptualize emotional labor.
Journal of Occupational Health Psychology, 5, 95110.
Heuven, E., & Bakker, B. (2003). Emotional dissonance and burnout among cabin attendants. European Journal
of Work and Organizational Psychology, 12, 81100.
Heuven, E., Bakker, A. B., Schaufeli, W. B., & Huisman, N. (2006). The role of self-ecacy in performing
emotion work. Journal of Vocational Behavior, 69, 222235.
Hobfoll, S. E. (1988). The ecology of stress. New York: Hemisphere Publishing Corporation.
Hobfoll, S. E. (1989). Conservation of resources: A new attempt at conceptualizing stress. American Psychologist,
44, 513524.
Hobfoll, S. E. (2001). The inuence of culture, community and the nested self in the stress process: Advancing
conservation of resource theory. Applied Psychology: An International Review, 50, 337421.
Hochschild, A. R. (1983). The managed heart. Berkeley: University of California Press.
Jansz, J., & Timmers, M. (2002). Emotional dissonance: When the experience of an emotion jeopardizes an
individuals identity. Theory and Psychology, 12, 7995.
Karasek, R. A. (1979). Job demands, job decision latitude, and mental strain: Implications for job redesign.
Administrative Science Quarterly, 24, 285308.
Kline, R. B. (2005). Principles and practice of structural equations modelling. New York: The Guildford press.

B. van Gelderen et al. / Journal of Vocational Behavior 71 (2007) 446459

459

Kruml, S. M., & Geddes, D. (2000). Exploring the dimensions of emotional labor. Management Communication
Quarterly, 14, 849.
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, 123133.
Leiter, M. P., & Maslach, C. (1988). The impact of interpersonal environment on burnout and organizational
commitment. Journal of Organizational Behavior, 9, 297308.
Maslach, C., & Jackson, S. E. (1984). Patterns of burnout among a national sample of public contact workers.
Journal of Health and Human Resource Administration, 7, 189212.
Maslach, C., Schaufeli, W . B., & Leiter, M. P. (2001). Job burnout. Annual Review of Psychology, 52, 397422.
Morris, J., & Feldman, D. C. (1997). Managing emotions in the workplace. Journal of Managerial Issues, 9,
257274.
Rafaeli, A., & Sutton, R. I. (1991). Emotional contrast strategies as means of social inuence: Lessons from
criminal interrogators and bill collectors. Academy of Management Journal, 34, 749775.
Schaufeli, W. B., & van Dierendonck, D. (2000). Utrechtse Burnout Schaal (UBOS): handleiding. Lisse: Swets &
Zeitlinger.
Schellekens, J. M. H., Sijtsma, G. J., Vegter, E., & Mijeman, T. F. (2000). Immediate and delayed after-eects of
long lasting mentally demanding work. Biological Psychology, 53, 3756.
Tice, D., Baumeister, R. F., & Zhang, L. (2004). The role of emotion in self-regulation: Diering roles of positive
and negative emotion. In P. Philippot & R. S. Feldman (Eds.), The regulation of emotion (pp. 213226). New
Jersey: Lawrence Erlbaum Associates.
Tsai, W., & Huang, Y. (2002). Mechanisms linking employee aective delivery and customer behavioral
intentions. Journal of Applied Psychology, 87, 10011008.
Vercoulen, J. H. M. M., Alberts, M., & Bleijenberg, G. (1999). Kort instrumenteel: De checklist individual
strength (CIS). Gedragstherapie, 32, 131136.
Zapf, D. (2002). Emotion work and psychological well-being. A review of the literature and some conceptual
considerations. Human Resource Management Review, 12, 237268.
Zapf, D., Seifert, C., Schmutte, B., Mertini, H., & Holz, M. (2001). Emotion work and job stressors and their
eects on burnout. Psychology and Health, 16, 527546.
Zapf, D., Vogt, C., Seifert, C., Mertini, H., & Isic, A. (1999). Emotion work as a source of stress: The concept and
development of an instrument. European Journal of work and Organizational Psychology, 8, 371400.