Emotional dissonance

Journal of Managerial Psychology
Emotional dissonance in call centre work
Jürgen Wegge Rolf Van Dick Christiane von Bernstorff

Article information:

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

To cite this document:
Jürgen Wegge Rolf Van Dick Christiane von Bernstorff, (2010),"Emotional dissonance in call centre work",
Journal of Managerial Psychology, Vol. 25 Iss 6 pp. 596 - 619
Permanent link to this document: http://dx.doi.org/10.1108/02683941011056950 Downloaded on: 03 February 2015, At: 16:41 (PT)
References: this document contains references to 60 other documents.
To copy this document: [e-mail]
The fulltext of this document has been downloaded 2485 times since 2010*

Users who downloaded this article also downloaded:
Michel Rod, Nicholas J. Ashill, (2013),"The impact of call centre stressors on inbound and outbound callcentre agent burnout", Managing Service Quality: An International Journal, Vol. 23 Iss 3 pp. 245-264 http:// dx.doi.org/10.1108/09604521311312255 Keith Townsend, (2007),"Recruitment, training and turnover: another call centre paradox", Personnel
Review, Vol. 36 Iss 3 pp. 476-490 http://dx.doi.org/10.1108/00483480710731383
Bill Halliden, Kathy Monks, (2005),"Employee-centred management in a call centre", Personnel Review,
Vol. 34 Iss 3 pp. 370-383 http://dx.doi.org/10.1108/00483480510591480

Access to this document was granted through an Emerald subscription provided by 378296 []

For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.

About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation. *Related content and download information correct at time of download.

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0268-3946.htm JMP
25,6

Emotional dissonance in call centre work

596

Technical University of Dresden, Dresden, Germany

Ju¨rgen Wegge
Rolf Van Dick
Goethe University Frankfurt, Frankfurt, Germany, and

Received May 2009
Revised November 2009
Accepted November 2009

Christiane von Bernstorff

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

University of Dortmund, Dortmund, Germany
Abstract
Purpose – The purpose of this paper is to investigate new hypotheses regarding potential correlates and underpinnings of emotional dissonance experienced in call centre work. It is argued that prior attempts to measure emotional dissonance are incomplete because such measures often do not specify which emotions are actually not shown (e.g. faked, suppressed, veiled) during work.
Design/methodology/approach – A field study with 161 call centre agents was conducted.
Positive affectivity (PA), negative affectivity (NA) of agents and customer verbal aggression were conceptualized as correlates of emotional dissonance, whereas job satisfaction, health disorders and burnout were assessed as indicators of agents’ work motivation and well-being. To investigate the emotional underpinnings of emotional dissonance the Frankfurt Emotion Work Scales (FEWS) was used and, in addition, agents were asked to report frequency, intensity and “not showing” of 15 separate emotions.
Findings – The results show that emotional dissonance was associated with lower work motivation and well-being. Moreover, NA and customer aggression correlated positively whereas PA correlated negatively with emotional dissonance. Emotional dissonance measured with the FEWS was significantly related to the frequency of longing, the intensity of anger and the not showing of boredom, affection and anger.
Originality/value – The findings support the construct validity of the FEWS. However, based on correlations with agents’ self-rated ability to perform on a high level and interactions between NA and customer aggression that emerged only when emotion-specific dissonance measures were analyzed, this paper suggests combining emotion-specific dissonance measures with the FEWS in future research. Keywords Emotional dissonance, Call centres, Motivation (psychology), Personal health
Paper type Research paper

Journal of Managerial Psychology
Vol. 25 No. 6, 2010 pp. 596-619 q Emerald Group Publishing Limited
0268-3946
DOI 10.1108/02683941011056950

In many service jobs it is expected that service providers (e.g. nurses, clerks, call centre agents) express positive emotions such as joy or interest in interactions with customers even in situations in which they actually feel no specific emotion or in which they feel negative emotions such as boredom (Diefendorff and Richard, 2003). The term
“emotional dissonance” describes this discrepancy between expressed and felt emotions and prior studies have shown that emotional dissonance is an important phenomenon in service work that is, for example, linked to burnout (Zapf, 2002).
The authors are grateful to Shay Tzafrir and two anonymous reviewers for their helpful comments on earlier versions of this paper.

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

However, only few studies have examined potential antecedents of emotional dissonance at work (Bono and Vey, 2005) and, therefore, the first aim of this research is to gather more evidence in this respect. More specifically, our study investigates the personality of employees (positive and negative affectivity) and the aggressiveness of customers as potential antecedents of emotional dissonance. Our second goal is to shed further light on the specific emotional underpinnings of emotional dissonance. It will be argued that prior attempts to measure emotional dissonance are incomplete because these measures often do not specify which emotions are actually not shown (e.g. faked, suppressed, veiled). To investigate if a more complete measure of emotional dissonance is beneficial, we compare traditional measures of emotional dissonance with newly constructed, emotion-specific measures of emotional dissonance. In the following, we discuss prior research on emotion work in general and then focus on what is known about emotional dissonance in call centre work, deriving the specific hypotheses we tested in our field study.
Emotional dissonance as one facet of emotion work
It is not our aim to present a comprehensive review of the literature on emotion work
(synonymous: emotional labour) but we will provide a brief overview. Emotion work is a rather broad construct that typically includes the experience of emotional dissonance as one sub-facet. Several approaches have been developed to describe the various elements that are involved in emotion work (for recent reviews see Bono and Vey, 2005;
Grandey, 2000; Humphrey et al., 2008; Zapf, 2002; Zapf and Holz, 2006). Some researchers, for example, conceptualize emotion work with a clear focus on the volitional display of normatively appropriate emotions. Morris and Feldman (1996, p.
987) defined emotional labour as the “effort, planning and control needed to express organizationally desired emotions during interpersonal transactions”. In this definition, the observable behaviour of the service provider (e.g. nurse, clerk) is emphasized. Other definitions (see Bono and Vey, 2005 for a comparison of several definitions) focus more on the presence of emotional display rules at work and / or the various cognitive processes (e.g. goal setting, planning, monitoring) that regulate the presentation of emotions (e.g. Diefendorff and Gosserand, 2003; Diefendorff et al., 2005;
Gosserand and Diefendorff, 2005). Thus, even though there is no complete agreement on how to define emotion work in general, most researchers describe it as a multidimensional construct that comprises several facets such as attentiveness to emotional display rules and emotions of other people, the more or less frequent volitional display of emotions, or cognitive processes that are involved in producing
(e.g. faking) or not showing specific emotions. Based on the work of Hochschild (1983), these processes are usually described as surface acting (e.g. managing only the visible aspects of emotions by faking or suppressing feelings) or as deep acting (e.g. when people try to influence what they really feel by reframing the situation or by taking the perspective of a client in order to fulfil their prescribed work role). In line with this differentiation, emotional dissonance experiences are sometimes also referred to as
“surface acting” (Tschan et al., 2005) because successful deep acting might lead to the resolution of a dissonance experience. In sum, emotional dissonance is a concept that can be found in most theories of emotion work and in our research we understand emotional dissonance as a strain reaction towards a specific stressor, namely the

Emotional dissonance in call centre work
597

JMP
25,6

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

598

experience of dissonance between an expressed and a felt emotion (see also Jex et al.,
1992 for a clarification of the terms stress, stressor and strain).
Potential benefits of emotion work such as increased service quality for customers or higher income for service providers (e.g. waiters) and potential problems such as emotional exhaustion due to emotion work in general or emotional dissonances in particular are discussed for some time (Goffman, 1959; Grandey, 2000; Hochschild,
1983). Reliable measures for these specific phenomena, however, have been developed only recently. To our knowledge, Zapf et al. (1999) presented the first attempt to develop and test a comprehensive self-report instrument with respect to emotion work in service jobs: the Frankfurt Emotion Work Scales (FEWS). The FEWS comprise several scales that measure, among other things, the requirement to display positive emotions during work, the necessity of displaying unpleasant emotions towards customers, sensitivity requirements (e.g. whether empathy or knowledge about clients’ current feelings are needed in the job), requirements to show sympathy to the sorrows and problems of the client, emotion work control (control with regard to the display of emotions), interaction control (e.g. control regarding available time and completion of an interaction) and emotional dissonance, the mismatch between felt emotions and organizationally desired expressions of emotions.
Based on this work and related, similar approaches of other researchers (e.g.
Bu¨ssing and Glaser, 1999; Morris and Feldman, 1996, 1997), recent research in service jobs has shown that emotion work is a double edged sword (see research on flight attendants, police officers, nurses, employees in hotels, kinder gardens, restaurants, shoe retail shops, travel agencies, and call centres, e.g. Brotheridge and Lee, 2003;
Diefendorff and Richard, 2003; Dormann and Zapf, 2004; Fischbach, 2003; Heuven and
Bakker, 2003; Schaubroeck and Jones, 2000; Totterdell and Holman, 2003; Zapf and
Holz, 2006). On the one hand, there is some evidence that the frequent display of positive emotions correlates positively with job satisfaction and feelings of personal accomplishment. On the other hand, there is evidence that employees who experience emotional dissonance are typically less satisfied with their job and report higher emotional exhaustion. In the review of Zapf (2002) the average correlation between emotional dissonance and emotional exhaustion was r¼0.32 (k¼12) and in the meta-analysis of Bono and Vey (2005) similar correlations were found for emotional dissonance (r¼0.30, k¼11) and scales assessing surface acting (r¼0.36, k¼7). Thus, even though empirical research on emotion work is still in the beginning and some aspects of emotion work might function more as a challenge stressor than as a hindrance stressor (LePine, Podsakoff and LePine et al., 2005), there is some evidence that especially emotional dissonance (as a form of surface acting) can lower well-being and lead to health complaints.
Emotional dissonance in call centre work
Based on previous research, it is clear that emotional dissonance is a salient strain phenomenon in call centre work, too (Grandey et al., 2004; Grebner et al., 2003; Holman,
2003; Holman et al., 2002; Kinman, 2009; Lewig and Dollard, 2003; Zapf et al., 1999,
Zapf et al., 2003). Even though there is no direct face-to-face contact with customers, there are typically strong demands to be friendly to customers of call centres. Zapf et al.
(2003) found that – compared to employees working in similar service jobs – call centre agents are asked to express less negative emotions. Moreover, the performance

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

of agents is often monitored by the organization, for example, by making test calls or by recording calls, so that deviations from this norm can be easily detected (Holman,
2003). As customers are often unfriendly to agents (Grandey et al., 2004; Totterdell and
Holman, 2003) and because having the same – often scripted – communication with a hundred and more customers a day might easily yield boredom, there are strong requirements to suppress negative emotions in call centre work.
Prior research on emotion work in call centres has also revealed that there are several important variables that have an impact on the amount and intensity of emotional regulation processes at work, for instance the frequency and duration of interactions or the degree of monitoring of interactions (Holman, 2003; Zapf, 2002). In the following, we briefly discuss relevant research leading to the selection of those factors that were examined as correlates of emotional dissonance in our study. We will also develop arguments for a more precise analysis of emotional dissonance that takes into consideration the specific type and hedonic tone of emotions that have to be faked or suppressed by agents.
Positive and negative affectivity as potential determinants of emotional dissonance According to the meta-analysis of Thoresen et al. (2003) individuals high on negative affectivity (NA) typically report lower job satisfaction whereas individuals high on positive affectivity (PA) are generally more satisfied. NA is defined as having a negative outlook on life which is often accompanied by the experience of negative emotions, health complaints and burnout. PA reflects a general positive outlook on life and is accompanied by experiencing positive emotions and self-efficacy with respect to problem-solving. In view of the fact that NA is consistently correlated with self-reports of health complaints but not correlated with actual, long-term health status, it was suggested to consider NA as a nuisance factor in health research that should be controlled for (Watson and Pennebaker, 1989). However, there are strong arguments that the correlations usually obtained do not solely reflect a bias but also a substantial role of dispositional variables in the stress process (Spector et al., 2000). It is plausible, for example, that individuals high in NA are hyper-responsive to the environment so that they experience exaggerated negative strain responses to identical stressors (e.g.
Wegge, 2006). Therefore, several researchers have already suggested that the experience of emotional dissonance might be influenced by these dispositions, too. To date, however, little empirical support exists for this assumption. Bono and Vey (2005) found only two studies and calculated an r ¼ 20.13 for PA-emotion work and r ¼ 0.19 for
NA-emotion work relationships. In line with this finding are further results presented by
Brotheridge and Lee (2003; r ¼ 0.17 for NA and surface acting; r ¼ 20.22 for PA and surface acting), Deery, Iverson and Walsh (2002; r ¼ 20.12 for PA and emotional exhaustion), Grebner et al. (2003; r ¼ 0.15 for NA and emotional dissonance) and
Grandey et al. (2004; r ¼ 0.46 for NA and stress appraisal with respect to aggressive customers). NA and PA predispose individuals to be in bad or good moods, and in call centre work the expression of positive emotion is often desired and the expression of negative emotions is typically not tolerated. Therefore, high NA individuals have probably more negative emotions to suppress than low NA individuals. Moreover, high
PA individuals are often in a good mood and, as a consequence, they probably have to suppress less negative emotions than low PA individuals. Thus, we propose:

Emotional dissonance in call centre work
599

JMP
25,6

H1a. Negative affectivity is positively correlated with emotional dissonance in the work of call centre agents.
H1b. Positive affectivity is negatively correlated with emotional dissonance in the work of call centre agents.

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

600

Customer behaviour as a potential determinant of emotional dissonance
Based on a newly developed instrument that assesses various forms of customer-related stressors such as disproportionate customer expectations or verbally aggressive customers, Dormann and Zapf (2004) found in three samples of service workers that verbal customer aggression (e.g. customers shouting at service providers) is a strong stressor that is positively correlated with burnout and experiences of emotional dissonance. As verbally aggressive customers are a common problem in the work of call centre agents (e.g. Grandey et al., 2004), similar relationships should be observed in call centre work, too. Thus, it can be assumed that the verbal aggressiveness of customers is correlated positively with emotional dissonance because aggressive customers express and probably also induce more negative emotions in agents that – according to common emotional display rules in call centres – should not be shown. However, to our knowledge, there is no empirical data regarding this question. Thus, we propose to find a similar effect in the realm of call centre work:
H2. Verbal customer aggression is positively correlated with emotional dissonance among call centre agents.
Potential consequences of emotional dissonance
In view of the first findings regarding the impact of emotional dissonance on job satisfaction, burnout and health disorders which were briefly reviewed above, it can be expected that similar relationships should be observed in this study. This expectation is further based on several theoretical accounts explaining such relationships, e.g. the theory of action regulation (Zapf, 2002), or theories regarding the use of volition in emotion regulation (e.g. Baumeister et al., 1998). According to these theories, the experience of emotional dissonance might affect work motivation and health negatively because emotional dissonance functions as a regulation problem requiring extra effort and draining emotional resources in the long run (Zapf, 2002). Moreover, the volitional presentation of positive emotions and the continuous self-control of one’s own feelings (e.g. by suppression of negative emotions) is very demanding. Volitional emotion regulation consumes energies and, as a consequence, performance can decrease (Baumeister et al., 1998). Taken together, we propose that emotional dissonance is positively correlated with burnout (emotional exhaustion, depersonalization) and health disorders of agents. Moreover, a negative correlation with job satisfaction and achieving one’s own performance goals (personal accomplishments) should be observed. This leads to:
H3. Emotional dissonance is positively correlated with (a) emotional exhaustion,
(b) depersonalization and (c) health disorders and negatively correlated with
(d) job satisfaction and (e) judgments of personal accomplishment among call centre agents.

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

Positive and negative emotions as potential underpinnings of emotional dissonance Reviews of the emotion work literature (e.g. Bono and Vey, 200; Grandey, 2000; Zapf,
2002) consistently note that the relationships between indicators of emotion work and stress and well-being of service providers across studies are often contradictory. In some studies, for example, negative correlations with job satisfaction were found whereas other studies reported zero correlations or even positive correlations. One possible explanation for these inconsistencies might be that emotion work is a double-edged sword: some facets of emotion work (e.g. expressing positive emotions) are rather unproblematic (e.g. Diefendorff and Richard, 2003; Zapf and Holz, 2006) whereas other aspects (e.g. emotional dissonance) are negatively related to work motivation and well-being. As different researchers use different indicators of emotion work, the pattern of results discussed above is not surprising. This is particularly true because emotion work in general and also the experience of emotional dissonance is embedded in social and cultural contexts in complex ways.
While we do agree with this general reasoning, we believe that there might be an additional explanation for the inconsistent findings with respect to the consequences of emotional dissonance at work. Our suggestion is based on the observation that widely used measures of emotional dissonance such as the FEWS scale do not specify which specific emotion a service provider has to fake or suppress at work. The corresponding scale comprises five items that ask employees to indicate how often they “. . . show positive feelings or negative feelings when they feel indifferent”, “. . . suppress emotions to be experienced as neutral”, “. . . display emotions that are not in correspondence with real feelings”, “. . . display emotions that are not in accordance with those that are actually felt with respect to customers”. A fifth item requires indicating the degree in which one’s own job resembles a job where it is important to hide real feelings in interaction with customers. When answers to these items are summed up to compute a summary score for emotional dissonance, there is no information to the specific emotions and the hedonic tone of the emotions which are suppressed or faked during work.
Now, the question is whether it is sensible to ignore the specific type and hedonic tone of these emotions. Are the determinants and consequences of faking or suppressing positive and negative emotions the same? In view of many findings from emotional psychology, this assumption seems problematic for at least two reasons.
First, discrete emotions like anger, envy, guilt or anxiety are triggered by specific events and the consequences of these emotions typically differ substantially
(Ashkanasy et al., 2002; Payne and Cooper, 2001; Wegge, 2004). Anxiety, for example, is often aroused when facing uncertain, existential threat; anger is aroused if we experience a threat to our identity or an offense against our interests and goals. If we now acknowledge that some negative emotions (e.g. anger) typically activate people and that other negative emotions (e.g. hopelessness, boredom) often deactivate, it follows that specific emotions and their immediate suppression might have also different consequences. Second, the fact that a positive (negative) emotion has to be suppressed implies that a positive (negative) emotion was actually aroused before, at least to some notable degree. Therefore, if positive emotions are aroused, the immediate consequences of these emotions (e.g. a rise of attention or interest) are probably still active in this situation at least for a short time – even if the prescribed work role

Emotional dissonance in call centre work
601

JMP
25,6

602

requires no expression of this emotion. The same is true for negative emotions like anger or guilt. Thus, a clear separation of emotional dissonance linked to positive versus negative emotions seems warranted. We know from prior work in call centres that mainly the expression of negative emotions such as anger and boredom is problematic. Thus, it can be expected that a measure of emotional dissonance which does not consider the hedonic tone of suppressed emotions mainly reflects dissonance experiences in which negative emotions are not expressed. Therefore, we propose:

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

H4a. Emotional dissonance assessed with a measure that is indifferent with respect to the hedonic tone of emotions (e.g., FEWS) is mainly based on dissonance experiences that are linked with suppressing negative emotions in call centre work. In line with results of Grandey et al. (2004) and Brotheridge and Lee (2003) who found that affective dispositions might have an impact on dissonance in this kind of service work, we further propose that NA moderates the relationship between customer aggression and emotional dissonance experiences. As such, the role of individual differences in emotional labour (see Bono and Vey, 2005; Diefendorff et al., 2005; Zapf and Holz, 2006) and in particular moderating effects of dispositional variables are not well examined. Here, we solely focus on NA as not showing negative emotions is probably the main problem in call centre work. Based on the studies cited above, we assume that for agents high in NA a dissonance between their emotions should be most prominent if customers are perceived as aggressive as this probably induces and yields very strong negative emotions in call centre agents. Of course, such an interaction should be more pronounced if a specific measure of emotional dissonance is used that assesses emotional dissonance experiences linked to negative emotions. This leads to:
H4b. The association between customer aggression and emotional dissonance is moderated by NA as the dissonance inducing impact of customer aggression is stronger in individuals high in NA. This interaction is most pronounced for emotional dissonance experiences linked to negative emotions in call centre work. Method
Sample and procedure
We collected data from eight call centres in Germany located near to our universities
(thus using an opportunity sampling strategy) and altogether distributed 300 standardized questionnaires[1]. Call centre agents were informed about the option to participate in this study by written advertisements in the entrance area of their call centre. Agents filled out the questionnaires during leisure time. Participation was confidential and voluntary. A total of 161 usable questionnaires were returned within two weeks after the distribution in boxes that were positioned in the entrance area of call centres (response rate 53 percent). Of the respondents, 62 percent were female, average age was 32.6 years (SD ¼9.7 years), and mean professional experience was 2.3 years (SD ¼2.6 years). A total of 69 percent of agents had a permanent contract, 31 percent a short term contract. Of the agents, 119 (72.6 percent) indicated that they would perform mainly inbound tasks[2] (i.e. receiving calls from customers), 39 agents also would perform outbound tasks (i.e. dialling up customers).

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

Measures
Dispositional differences regarding affective experiences – positive affectivity (PA) and negative affectivity (NA) – were assessed with the PANAS questionnaire comprising ten positive and ten negative mood states (Watson et al., 1988). Participants were asked to rate how intensively they experience these moods in general using scales from 1¼ “not present” to 5¼ “very strong” with a validated German version of this instrument (Krohne et al., 1996). Verbal customer aggression was assessed as a further potential antecedent of emotional dissonance. For this purpose, a corresponding sub-scale of a broader instrument measuring customer-related social stressors presented by Dormann and Zapf (2004) was utilized. This scale comprises five items
(e.g. “customers often shout at us”; “some customers argue all the time”). We used the
FEWS, Version 4.0 to measure emotional dissonance. The corresponding scale consists of five items referring to the suppression of felt emotions as well as to the display of emotions not actually felt (e.g. “I have to suppress emotions at work in order to be experienced as neutral to others”). This scale does not distinguish in detail which emotions are suppressed or overplayed (see above).
Job satisfaction was measured with five items drawn from the study of Wegge and
Neuhaus (2002;, e.g. “How satisfied are you with your job in general?”). This scale is a translation of the Hackman and Oldham’s (1975) five-item job satisfaction scale and has shown high reliability in prior studies. To measure employees’ well-being two instruments were used. The Maslach Burnout Inventory in the German translation
(MBI-D) from Bu¨ssing and Perrar (1992) assesses the three components of burnout:
(1) Emotional exhaustion (eight items).
(2) Depersonalization (five items).
(3) Personal accomplishment (eight items, in the German version high values indicate good performance).
In addition, a list of 21 different types of health disorders developed by Mohr (1986) was presented and agents indicated how often they have experienced these disorders over the last six months. For all scales described above, participants were provided five-point answering scales with endpoints 1 ¼ ”not at all correct” to 5 ¼ ”totally correct”;
1 ¼ “never” to 5 ¼ “very often”; and 1 ¼ “not at all” to 5 ¼ “very much”, respectively.
Finally, for a closer inspection of the emotional underpinnings of dissonance measured with the FEWS, further items assessing several emotions were presented.
This questionnaire is built on the EMO-16, a self-report instrument developed and validated by Schmidt-Atzert and Hu¨ppe (1996) in 12 experiments. It comprises a list of
16 specific emotions which can be clearly distinguished on the basis of several emotion theories. After dropping “sexual arousal” the remaining list of 15 emotions – dislike, anger, envy, boredom, anxiety, unrest, sadness, longing, shame, guilt, happiness, pride, sympathy, affection, surprise – was presented three times in different parts of the questionnaire. The first time agents rated each emotion and were instructed to report
“how intensive” they experience these emotions in their work (from 1 ¼ “not present” to 6 ¼ “very strong”). The second time the same list was presented with the instruction to report “how often” these emotions are experienced at work (from 1 ¼ “never” to
5 ¼ “very often”). Finally, the third time the list was presented with an instruction to assess emotional dissonance linked to each emotion. Agents were asked how often they do not show this emotion during work after it had been elicited. The instruction further

Emotional dissonance in call centre work
603

JMP
25,6

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

604

explained that it is not important why an emotion would not be shown, and that not showing means suppressing or hiding (masking) an experienced emotion, also by faking a desired emotion in this situation. Answers were given on a scale from
1 ¼ “never” to 5 ¼ “very often”. For some of the following analyses, data with respect to each separate emotion will be used. However, in order to derive summary scales for the emotional dissonance experiences we conducted a principal components factor analysis with Varimax rotation. This analysis suggested a three factor solution
(Eigenvalues . 1: 6.8, 2.6, and 1.2, respectively) explaining 71 percent of the variance.
Based on rotated factor loadings a scale summarizing the two negative emotions anger and dislike (negative customer related emotions or Emodis_ncr), a scale summarizing all other negative emotions (e.g. envy, boredom, etc., Emodis-neg), and a scale comprising the positive emotions (happiness, pride, sympathy, affection, surprise,
Emodis_pos) were computed[3].
Results
Table I presents means, standard deviations, correlations of the main variables and reliability coefficients. All scales were sufficiently reliable.
In order to test H1, zero-order correlations between PA, NA and emotional dissonance measured with the FEWS were examined (see Table I). As expected, NA is positively correlated (r ¼ 0.17) with the experience of emotional dissonance at work, whereas PA correlates negatively (r ¼ 2 0.16) with emotional dissonance. The size of these correlations is in accordance with results from prior studies (e.g. Bono and Vey,
2005) and the direction of these relationships is consistent with the idea that individuals with low NA and high PA fulfil typical emotional requirements of call centre work most effectively. Beyond these general associations, it is also possible in our study to examine this hypothesis with more emotion-specific measures of emotional dissonance. As shown in Table I, the correlations between NA and the two emotion-specific measures of emotional dissonance linked to negative emotions
(Emodis_ncr; Emodis_neg) are also positively correlated (r ¼ 0.24 and r ¼ 0.22, both p , 0.01). However, there is no significant correlation between PA and dissonance linked to positive emotions (Emodis_pos). This finding, together with the low correlations between the four emotional dissonance measures, already indicates that both types of measure might reflect somehow different phenomena. Taken together,
H1 was supported. It should also be noted that the significant and opposite correlations between PA, NA and indicators of work motivation ( job satisfaction, personal accomplishment) and well-being (burnout, health complaints) are in accordance with prior results (Thoresen et al., 2003).
As predicted in H2, the correlation between verbally aggressive customers and experiencing emotional dissonance at work measured with the FEWS is also significant (r ¼ 0.18). Thus, verbally aggressive customers can be considered as a potential antecedent of emotional dissonance among call centre agents, too. With respect to the emotion-specific assessments of emotional dissonance (e.g. Emodis_ncr), this relationship, however, is not substantial. Thus, for H2 we find mixed support in our data if we assume that this relationship should hold for all measures of emotional dissonance. Of course, this pattern of findings once again indicates that both types of emotional dissonance measures might reflect somehow different phenomena.

9.7
0.48
0.85
0.63
0.72
0.97
1.14
1.26
0.914
0.80
0.72
0.75
0.82
0.61

SD
–
2 0.11
0.20 *
2 0.23 * *
2 0.26 * *
2 0.09
0.02
0.11
0.16 *
0.05
2 0.10
2 0.14
2 0.40 * *
0.41 * *

1

–
20.16 *
20.04
0.08
20.04
0.01
20.07
0.02
20.11
20.12
0.13
0.25 * *
20.06

2

4

5

6

7

8

9

10

11

12

13

14

(0.94)
20.20 *
(0.88)
20.32 * *
0.30 * *
(0.85)
*
20.16
0.17 *
0.18 *
(0.78)
20.10
0.24 * *
0.13
0.37 * *
(0.82)
0.07
0.30 * *
0.87 * *
(0.93)
0.02
0.22 * *
0.05
0.14
0.07
20.04
0.21 * *
0.20 * * (0.80)
*
*
*
*
*
*
*
0.44
2 0.18
20.31
20.28
2 0.12
2 0.04
0.05
(0.85)
*
*
*
*
*
*
*
*
*
20.38
0.38
0.38
0.36
0.21
0.13
0.05 20.37 * *
(0.91)
*
*
*
*
*
*
*
*
*
*
*
20.51
0.54
0.43
0.29
0.29
0.17
0.13 20.46 * *
0.56 * *
(0.87)
*
*
*
*
*
*
*
*
*
*
*
20.41
0.45
0.53
0.31
0.16
0.07
0.04 20.35
0.37 * *
0.64 * *
(0.75)
0.57 * * 2 0.28 * * 20.30 * * 20.22 * *
0.02
0.11
0.07
0.32 * * 2 0.32 * * 2 0.38 * * 20.40 * * (0.77)

3

Notes: afemale ¼ 1, male ¼ 2; * p , 0.05, * *p , 0.01; n between 161 and 140 due to missing data, Cronbach’s alphas are on the diagonal

1. Age
32.6
2. Sexa
1.37
3. Positive affectivity (PA)
3.34
4. Negative affectivity (NA) 1.67
5. Aggressive customers
2.06
6. Emodis (FEWS)
3.12
7. Emodis_ncr (e.g. anger)
2.68
8. Emodis_neg (e.g., envy)
2.30
9. Emodis_pos (e.g. pride)
2.37
10. Job satisfaction
3.62
11. Health disorders
2.24
12. Emotional exhaustion
2.23
13. Depersonalization
2.14
14. Personal accomplishment 3.40

M

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

Emotional dissonance in call centre work
605

Table I.
Means, standard deviations, reliability coefficients, and zero-order correlations

JMP
25,6

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

606

For testing H3a to H3e, we again examined correlations between emotional dissonance and various outcome variables (Table I). Our hypothesis is fully corroborated by the data when we focus on emotional dissonance measured with the FEWS. Emotional dissonance assessed with the FEWS is positively correlated with emotional exhaustion
(r ¼ 0.29; H3a), depersonalization (r ¼ 0.31, H3b), health disorders (r ¼ 0.36, H3c) and negatively correlated with job satisfaction (r ¼ 2 0.28, H3d ) and personal accomplishment (r ¼ 20.22, H3e). Again, these relationships are similar if we examine the correlations between the emotion-specific dissonance measures and emotional exhaustion and health disorders. However, for job satisfaction and personal accomplishment there are no significant correlations. Thus, our results strongly support
H3a to H3e if we base this analysis on the general FEWS scale but we only find mixed support if we base this analysis on emotion-specific measures of emotional dissonance.
To explore these differences in more detail, we also computed the partial correlations between the single emotional dissonance experiences, the FEWS-scale and the other variables controlling for PA and NA (see Table II). Here we focus on the correlations that were found with respect to job satisfaction and personal accomplishments. Interestingly, there is no single significant correlation between the
15 emotion-specific dissonance measures and overall job satisfaction when controlling for PA and NA. Perhaps even more surprising, we found four positive significant correlations between these dissonance measures and personal accomplishment
(indicating high performance if dissonance is also high). Thus, call centre agents reporting that they accomplish a lot at work also report that they do experience often dissonance at work with respect to the emotions envy, fear, sadness and surprise. We will discuss this point later on in more detail.
In order to test H4a we conducted a series of regressions with emotional dissonance
(FEWS) as criterion and the 15 specific work emotions as predictors. NA and PA were always controlled for in the first step. In the first regression analysis with intensity of the 15 emotions as predictors and in the second regression analysis with frequency of the 15 emotions as predictors, we investigated whether the reported intensity and frequency of emotions is related to the criterion emotional dissonance assessed with the
FEWS. As the FEWS were designed to measure dissonance experiences that are rather independent from the intensity or frequency of emotions at work, only weak correlations should be found. Accordingly, in both stepwise regressions only two significant relationships were found: the intensity of anger (b ¼ 0.37, p , 0.01,
R 2 ¼ 0.06) and the frequency of longing (b ¼ 0.21, p , 0.03, R 2 ¼ 0.04) correlated significantly with emotional dissonance of agents. In the third stepwise regression “not showing-values” of the 15 emotions were entered as predictors. Here, stronger correlations were to be expected. As already presented in Table II, eleven of the 15 correlations remain significant when controlling for PA and NA. Of course, the 15 specific dissonance measures are also correlated to each other (2 0.24 , r , 0.79).
Thus, it is warranted to control for these intercorrelations within a regression. In this regression analysis we found that only three predictors remain significant: not showing of boredom (b ¼ 0.22, p , 0.01), not showing of affection (b ¼ 2 0.20, p , 0.01) and not showing of anger (b ¼ 0.21, p , 0.02); together, these three variables explain 15 percent of variance in emotional dissonance.
In order to analyze whether these relationships are additive in nature, a further hierarchical regression was conducted in which all prior significant variables were

2.93
3.15
2.00
2.63
2.22
2.49
2.32
2.35
2.10
2.12
2.45
2.44
2.47
2.32
2.21

1.5
1.2
1.5
1.4
1.5
1.4
1.5
1.6
1.6
1.4
1.2
1.3
1.2
1.2
1.2

SD

Job satisfaction 20.08
20.05
0.01
20.10
20.01
0.00
20.02
20.02
20.03
20.06
20.09
0.03
0.04
0.09
0.10

Emodis
(FEWS)
0.26 * *
0.31 * *
0.22 * *
0.32 * *
0.23 * *
0.23 * *
0.22 * *
0.20 *
0.24 * *
0.26 * *
2 0.08
0.07
0.02
2 0.17 *
2 0.07

Aggressive customers 0.06
0.06
2 0.02
0.06
0.00
0.06
0.03
2 0.05
0.03
0.07
0.12
2 0.03
0.06
0.02
0.01

0.08
0.16
0.07
0.17 *
2 0.02
0.11
0.14
0.01
0.04
0.10
0.02
0.08
0.02
0.00
2 0.05

Health disorder 0.20 *
0.19 *
0.10
0.07
0.10
0.08
0.12
0.08
0.13
0.12
0.08
0.10
0.11
0.12
0.05

Emotional exhaustion 0.07
0.01
20.03
0.10
20.02
0.00
20.03
20.07
0.06
0.06
0.05
20.06
0.06
20.05
0.02

Depersonalization

Notes: n ¼ 135 to 143 due to single missing values, correlations are adjusted for 2 covariates (PA, NA); *p , 0.05, * *p , 0.01

Dislike
Anger
Envy
Boredom
Fear
Unrest
Sadness
Longing
Shame
Guilt
Joy
Pride
Sympathy
Affection
Surprise

M

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

0.11
0.05
0.17 *
0.01
0.19 *
0.13
0.25 * *
0.16
0.13
0.15
20.01
0.04
20.05
0.16
0.17 *

Personal accomplishment Emotional dissonance in call centre work
607

Table II.
Means, standard deviations, and partial correlations between specific emotional dissonance experiences and other variables
(controlled for PA and
NA)

JMP
25,6

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

608

entered after controlling for PA and NA. The results of this analysis are documented in
Table III. All five emotion variables identified in the previous regressions are significant predictors of emotional dissonance assessed with the FEWS and explain about 22 percent of variance over and above PA and NA. Based on these findings, we can conclude that H4a finds some support in the data as a substantial amount of variance in emotional dissonance measured with the FEWS is associated with negative emotions, in particular with not showing anger and boredom at work (two emotions representing also different arousal levels). However, it is noteworthy that the intensity of anger, the frequency of longing and a low dissonance regarding feelings of affection towards customers are significant predictors of emotional dissonance measured with the FEWS, too. We come back to this point in the discussion.
To test H4b we conducted four moderated regressions following the procedures recommended by Aiken and West (1991) using standardized (centred) predictor scores.
In all analyses, age and gender were entered as controls in the first step. In the second step, negative affectivity and customer aggressiveness were entered. Finally, the interaction term computed as the cross-product of both predictors was entered. The incremental variance explained in this third step provides an estimate of the size of the hypothesized interaction effect of NA and customer aggression. Finally, interaction effects were analyzed by plotting the resulting regression lines for high (one SD above mean) and low (one SD below mean) values for the two predictor variables.
The results of the first analysis with emotional dissonance measured with the
FEWS revealed that regressing the dependent variable onto the two predictors resulted in a significant increase in R 2 of 0.06 (F(2,140) ¼ 4.32, p , 0.02). Adding the interaction term, however, did not lead to a significant amount of incremental variance
(DR 2 ¼ 0.012).
We predicted that such an interaction should be more pronounced for dissonance linked to negative emotions. Thus, we conducted three further moderated regressions with the emotion-specific dissonance scales (emodis_ncr, emodis_neg, and emodis_pos).
In support of H4b it was found that the expected interaction was significant when analyzing customer related negative emotional dissonance experiences (see Figure 1).
Adding the interaction term led to a significant amount of explained variance
(DR 2 ¼ 0.045, the final equation is significant with F (5,139) ¼ 3.57, p , 0.01, total
R 2 ¼ 0.11; b (NA) ¼ 0.37, p , 0.01; b (customer aggression) ¼ 0.13, p , 0.23, and
Step
1
2

Table III.
Regression of emotional dissonance measured with the FEWS (4.0) on the Intensity (I),
Frequency (F) and
Dissonance-value (D) of specific emotions

PA
NA
PA
NA
Longing (F)
Anger (I)
Boredom (D)
Affection (D)
Anger (D)

B

SE B

Beta

t

p

20.19
0.13
20.05
20.04
0.18
0.14
0.14
20.19
0.14

0.09
0.13
0.09
0.13
0.07
0.06
0.06
0.07
0.07

2 0.17
0.09
2 0.04
2 0.03
0.21
0.21
0.22
2 0.23
0.18

21.90
1.03
20.47
20.31
2.43
2.29
2.54
22.88
2.06

0.06
0.31
0.64
0.75
0.02
0.02
0.01
0.01
0.04

Notes: R 2 ¼ 0.04 for step 1, DR 2 ¼ 0.22 for step 2 ( p , 0.001), R 2 for the final model ¼ 0.26;
PA and NA are entered as controls in step 1

Emotional dissonance in call centre work

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

609
Figure 1.
Interaction between verbal customer aggression and emotional dissonance linked to “anger” and
“dislike” for agents high or low on negative affectivity

b (interaction) ¼ 20.27, p , 0.01). However, in contrast to our expectations, the dissonance inducing impact of aggressive customers is higher for individuals low in NA.
For individuals high in NA even a small decrease in not showing the negative emotions anger and dislike is found if customers are perceived as more aggressive. Thus, it seems that dispositional differences in affect play a role in the experience of emotional dissonance mainly if interaction with friendly customers. For dissonance related to the other negative emotions no significant interaction was found b (interaction) ¼ 20.06, p , 0.54) but a significant interaction emerged for dissonance related to positive emotions (see Figure 2, DR 2 ¼ 0.025, the final equation is significant with F
(5,139) ¼ 2.91, p , 0.02, total R 2 ¼ 0.09; b (NA) ¼ 0.12, p , 0.16; b (customer aggression) ¼ 0.02, p , 0.74, and b (interaction) ¼ 0.14, p , 0.05). This interaction indicates that aggressive customer behaviour is associated with higher (respective lower)

Figure 2.
Interaction between verbal customer aggression and emotional dissonance for positive emotions for agents high or low on negative affectivity

JMP
25,6

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

610

values for not showing positive emotions if agents are high (respectively low) in NA. One potential explanation for this observation might be that agents high in NA sympathise with aggressive customers (e.g. because they can understand the negative affect of customers very well) but hide these positive feelings because this is in conflict with their work role. Taken together, H4b found some support in the data.
Discussion
Previous research has documented that emotional dissonance is an important strain in call centre work that has an impact on work motivation and well-being of agents beyond other typical problems such as time pressure or low autonomy. Therefore, the first purpose of the present study was to generate more information regarding potential correlates of this phenomenon. With respect to this aim, the results of this study contribute to the literature in at least two ways.
First, and in line with recent studies investigating strain in call centres (Grebner et al., 2003; Grandey et al., 2004), we found that inter-individual differences in negative and positive affectivity of agents are related to emotional dissonance at work. As several reliable correlations between emotional dissonance on the one hand and PA,
NA, health disorders, and burnout on the other hand were found in this study, the impact of these dispositions on strain in call centre work definitely warrants greater attention in future work.
Second, our results support and extend recent work on customer-related stressors.
Dormann and Zapf (2004) found that the experience of emotional dissonance in three service jobs (travel agency, flight attendants, shoe store clerks) is positively correlated with facing demanding customers, in particular verbally aggressive customers. The correlation between the same measures found in this sample of call centre agents was also significant though smaller in size. Nevertheless, as the correlations between aggressive customers and the three burnout components in the present study are similar to the ones found by Dormann and Zapf (2004), it can be concluded that verbally aggressive customers are indeed a potential source of emotional dissonance and burnout in call centre work, too.
Taken together, these findings are largely consistent with our current knowledge.
However, our results also yield new and perhaps more controversial insights that, in our view, deserve attention for both practice and theory. As most prior attempts to measure emotional dissonance did not specify and examine which specific emotions are actually not shown during work, it was our intention to shed some more light on the specific emotional underpinnings of emotional dissonance. For this purpose, we used three scales assessing the intensity, frequency and dissonance (in terms of not showing by suppressing or faking) associated with 15 specific positive and negative emotions that are distinguished in emotional psychology. In view of the predominant emotional requirements of call centre work (being friendly to customers; not expressing negative emotions etc.), it was suggested that traditional measures of emotional dissonance probably tap mainly into those emotional dissonance experiences that are linked to suppressing of negative emotions in interactions with customers.
As expected, we found that the FEWS dissonance measure is mainly linked to negative emotions at work (e.g. dislike, boredom, envy, fear, unrest, shame), not to positive emotions such as pride or joy (see Table II). Results from regression

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

analyses (controlling for PA and NA and the frequency and intensity of other emotions) showed that emotional dissonance assessed with the FEWS is based mainly on not showing the three emotions anger, boredom, and affection to customers. This finding nicely illustrates the construct validity of the FEWS. Not showing anger and boredom (e.g. induced by having the same, scripted calls all the day) is often explicitly encouraged by the management of call centres and therefore should be reflected in a summative measure of emotional dissonance. However, the finding that low dissonance with respect to not showing affection to customers is a predictor of dissonance measured with the FEWS is surprising. How can we explain that employees who feel affection to customers and do not hide this feeling very often (i.e. reporting low emotional dissonance for this emotion) have at the same time high emotional dissonance measured with the FEWS? It might be that most agents know that expressing affection towards customers is an unprofessional behaviour. Support for this idea can be found in the work of Rafaeli (1989) on behaviour of supermarket cashiers. She found that cashiers reduced smiling to customers when time pressure was increasing (see also Rafaeli and Sutton, 1987).
Thus, it could be argued that the violation of this emotion rule (requiring the suppression of positive emotions towards customers) is measured with the
FEWS-scale. In our view, this finding poses interesting questions for future studies.
Of course, as this observation is new, further studies should seek to replicate it.
Further evidence for the usefulness of separating emotional dissonance experiences related to specific negative and positive emotions was gained through a closer inspection of the correlations with important outcome variables (e.g. health disorders, emotional exhaustion, see Table II). Whereas customer aggression, job satisfaction and depersonalization are not significantly related to any emotion-specific dissonance measures (though FEWS-dissonance measures are), we found that emotion-specific dissonance measures were significantly correlated with health disorders and self-rated performance of call centre agents in an interesting way. Based on this data, not showing boredom is the most prominent (only) strain phenomenon in call centre work associated significantly with health disorders. Surprisingly, for personal accomplishment we even found some positive correlations. Whereas the overall
FEWS-dissonance sore was negatively correlated (r ¼ 2 0.22) with personal accomplishments (indicating that any dissonance might yield negative performance), agents reporting high dissonance with respect to envy, fear, sadness and surprise also reported high personal accomplishment. Thus, having dissonance experiences with respect to these specific emotions seem to be rather functional for the successful fulfilment of work roles in call centre work.
Additional evidence for the potential benefits of using emotion-specific dissonance measures in addition to traditional dissonance scales was gathered when analyzing the moderating role of NA in the reaction towards more or less aggressive customers. We predicted that the dissonance-inducing effect of aggressive customers is most prominent when agents have high values in NA as this constellation should reflect the arousal of very intensive negative emotions (that should not be expressed in call centre work). Moreover, it was proposed that these relationships should be most prominent if dissonance is assessed in relation to specific, negative emotions (e.g. anger, dislike). In support of this idea, we found that based on the FEWS-dissonance scale such interactions where not observable

Emotional dissonance in call centre work
611

JMP
25,6

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

612

whereas the analysis of emotion-specific measures of dissonance yielded significant interactions. The pattern of this interaction (see Figure 1) was, however, not consistent with our corresponding hypothesis. The findings indicate instead that agents low in NA react with a more pronounced increase in dissonance if customers become more aggressive and those agents high in NA show, if at all, a decrease in emotional dissonance if customers become more aggressive. These results are interesting. Whereas agents low in NA behave in an understandable way (they have more problems with aggressive customers), agents high in NA are less affected by the specific level of customer aggression. It might be that agents high in NA show less differences across levels of customer aggression as they are always in a bad mood (and have to hide this) or that these individuals have more problems also with friendly customers (e.g. they provoke negative reactions in customers that, in turn, elicit stronger negative emotions in themselves). Moreover, it might also be that agents high in NA understand the negative reactions of aggressive customers better
(see Tamir and Robinson, 2004) and, therefore, simply do not hide their anger and dislike if they have to serve an aggressive customer.
We observed one further interaction with respect to NA as aggressive customer behaviour was also associated with dissonance linked to not showing positive emotions – in particular if NA is high. Why do agents high in NA have to suppress positive emotions when dealing with aggressive customers? We already speculated above that those agents high in NA might sometimes sympathize with these customers because they can understand their negative affect well or even might share it (e.g. if customers complain about the organization). Of course, as this phenomenon was discovered in this study further research is called for to see whether this observation can be replicated.
Taken together, in view of our findings we strongly recommend combining the
FEWS scale with other instruments that assess emotional dissonance in terms of various specific emotions in future studies. If we consider that the FEWS dissonance scale and our comprehensive list measuring dissonance with respect to
15 different work emotions only shared 15 percent of the variation, it seems likely that the FEWS does not cover the whole range of emotional dissonance experiences at work. Of course, we also controlled for PA and NA in this analysis so that further correlations with other emotional variables were perhaps not easy to find.
Nevertheless, we believe that a more comprehensive understanding of this phenomenon might be gained if dissonance linked to several negative and positive emotions is assessed in addition. A similar proposition is made by Glomb and Tews
(2004) who developed and tested a new scale based on the assessment of specific affective states: the Discrete Emotions Emotional Labour Scales (DEELS). This instrument considers 14 specific emotions, such as anger, fear and anxiety or happiness, liking and enthusiasm. Respondents have to judge the frequency of these emotions with three different instructions, assessing how often these emotions are expressed in a genuine (non-dissonant) way, how often these emotions are expressed when they are not actually felt (faking) and how often these feelings are kept to oneself even if they are aroused (suppression). Compared to the measures used in this study, this instrument allows a finer assessment of dissonance experiences because faking and suppression of emotions is clearly distinguished. Obviously, the list of emotions assessed by Glomb and Tews (2004) is also different as they used

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

another classification of human emotions. Thus, shame, boredom, envy, longing, guilt, surprise and sympathy are not assessed. Instead, the DEELS considers affective states such as irritation, concern, aggravation, hate and enthusiasm.
However, it is interesting that Glomb and Tews also found that distinguishing emotional dissonance (faking, suppression of emotion) for positive and negative emotions is worthwhile as the consequences of these phenomena are different. In a mixed sample of various service jobs emotional exhaustion, for example, was more strongly correlated with suppressing negative emotions than with suppressing positive emotions at work. This indicates, once again, that considering both discrete positive and negative emotions in assessing emotional dissonance provides a better coverage of the emotional dissonance construct space. As we discovered the DEELS after finishing data collection in our study, we were unfortunately not able to use this scale. In our view, this is recommendable as a clear distinction between the suppression and faking of specific emotions is warranted. However, when using this scale researchers should perhaps also consider adding important emotions that are relevant in the specific work context under investigation (e.g. boredom in call centre work). Practical implications
In our view, one finding can be emphasized for practical reasons. Agents with low NA show a good person-job fit as these individuals report significantly less emotional dissonance. In call centre work, the expression of positive emotion is desired and the expression of negative emotions is typically not tolerated (Zapf et al., 2003). Therefore, individuals who are predisposed to be in a bad mood have more problems in this kind of work. Whereas the positive relation between NA and emotional dissonance obtained in our study is perhaps not really surprising, the negative relation between PA and emotional dissonance deserves attention. Consistent with the study of Zellars et al.
(2006) who found positive effects of PA on strain, our results show that PA can be considered as a personal resource protecting individuals from experiencing stress. For theoretical and empirical reasons PA and NA should be largely independent from each other (Thoresen et al., 2003). This finding was once again corroborated in this study and, therefore, it can be concluded that considering both affective dispositions in personnel selection of agents is fruitful (see also Grant and Langan-Fox, 2007 for a recent discussion on the role of personality in the stress-strain process). Of course, there are other possibilities to increase work motivation and well-being in call centres, e.g. by increasing the motivating potential of work and organizational identification (Holman,
2003; Wegge, van Dick, Fisher, Wecking and Moltzen, 2006), by improving the
(interaction) control of agents (Zapf, 2002), or by promoting a positive work climate and reducing work overload (Wegge, van Dick, Fisher, West and Dawson, 2006).
Nevertheless, the impact of affective dispositions in this emotional loaded work environment should not be underestimated.
Limitations and directions for future research
This study has several limitations. First, caution in interpreting the results is warranted due to the use of single-source self-reports. In general, the use of self-reports was necessary because most concepts involved here (e.g. emotional dissonance, perceptions of customers) are very difficult to measure by other means. However, it

Emotional dissonance in call centre work
613

JMP
25,6

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

614

might be fruitful to add more behavioural indicators in future studies (e.g. indicators for customer aggressiveness, absence and turnover from personnel files) and to measure independent and dependent variables at different points in time. As the data collection was anonymous and the reliabilities good for all scales, some problems common to self-reports (e.g. social desirability, low reliability of single item measures) probably did not have a strong impact. Moreover, by assessing PA and NA, we were able to control for the impact of affective dispositions on reporting more specific stress and strain phenomena at work. As there were several findings indicating that these dispositions indeed have an impact on these measures, we strongly recommend considering both dispositions in future studies. Of course, the influence of common method variance could not be controlled in our study and therefore, further replications with other methods (e.g. multi-source reports; experience sampling methods) are desirable. Having said this, the findings of interactions cannot be accounted for by common method biases as those influences are already controlled for in the first steps of the regression analyses (McClelland and Judd, 1993). This makes us more confident in our findings.
A further limitation pertains to the cross-sectional design of our study. Only experimental research allows for testing causal hypotheses regarding the impact of, for example, aggressive customers on emotional dissonance and burnout of employees.
Thus, the direction of causality of the variables examined in this study is merely that: a plausible inference. It might be, for example, that agents using depersonalization to detach themselves from customers create more hostile callers. However, in designing this study we relied on the available empirical evidence that is mainly organized along the well-validated stress-strain model in which, for example, emotional dissonance is considered as a state that causes burnout and health disorders. Therefore, we assume that several of the propositions derived from our correlational study will, at least in part, also prove valid in a longitudinal test. Third, we should be also aware of the fact that we used an opportunity sampling strategy so that our data is not representative of the population of employees working in call centres. In addition, potentially influential antecedents (e.g. identification with the organization) and consequences of emotional dissonance (e.g. call productivity of agents) were not measured. Future research should address this lacuna by examining these variables.
Despite these limitations, we believe that our study highlights an important gap in the current perception of emotion work in general and emotional dissonance in particular. In most approaches, insights from general theories of emotion have been not incorporated yet (see Grandey, 2000; Payne and Cooper, 2001). As a consequence, a closer inspection of the emotional underpinnings of emotional dissonance at work that distinguishes between dissonances linked to faking and suppressing positive emotions and faking and suppressing negative emotions should be a real addition to the firm platform that has already been constructed. In our view, extending the current approaches could be very stimulating both for further theory development and for solving practical problems such as personnel selection or training design in call centres. Of course, similar extensions and benefits might also prove fruitful with respect to other aspects of emotion work, for example, regarding the crossover of emotions (Ha¨rtel and Page, 2009) and emotion regulation of supervisors (Boss and
Sims, 2008). Thus, we strongly recommend addressing these questions in future studies. Notes
1. We checked the overall differences between these call centres for all of our variables.
However, as there were no substantial differences between the eight call centres and as we do not have enough power for a multi-level analysis, we decided that we would not differentiate between these organizations further.

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

2. A comparison of agents working in – versus outbound with respect to dependent variables showed only one significant difference as job satisfaction was higher for agents working outbound ( p , 0.05). In view of the minor differences and because our sample of agents working outbound is rather small, we do not consider this variable in the following analysis.
The same reasoning applies to the form of agents contracts (short term versus permanent).
3. It should be noted that surprise is an arousal-related emotion that can be both positive and negative. In our study, surprise empirically aligned with the other positive emotions. We also calculated correlations between the three sets of variables and found that the correlations between “not showing” items and “frequency” as well as “intensity” judgments of the same emotions were rather low (0.11 , r , 0.31 for intensity judgements; 2 0.13 , r , 0.32 for frequency judgments). Thus, respondents had no problems to differentiate between “not showing” an emotion and the frequency or intensity of the same emotion at work. However, correlations between frequency and intensity judgments were substantially higher
(0.45 , r , 0.82) indicating that respondents did not differentiate as well between frequency and intensity of emotions.

References
Aiken, L.S. and West, S.G. (1991), Multiple Regression: Testing and Interpreting Interactions,
Sage, Newbury Park, CA.
Ashkanasy, N.M., Zerbe, W.J. and Ha¨rtel, C.E. (2002), Managing Emotions in the Workplace,
Sharpe, Armonk, NY.
Baumeister, R.F., Bratslavsky, E., Muraven, M. and Tice, D.M. (1998), “Ego depletion: is the active self a limited resource?”, Journal of Personality and Social Psychology, Vol. 74, pp. 1252-65.
Bono, J.E. and Vey, M.A. (2005), “Toward understanding emotional management at work: a quantitative review of emotional labour research”, in Ha¨rtel, C.E.J., Zerbe, W. and
Ashkanasy, N.M. (Eds), Emotions in Organizational Behaviour, Erlbaum, Mahwah, NJ, pp. 213-34.
Boss, A.D. and Sims, H.P. (2008), “Using emotion regulation and self-leadership for recovery”,
Journal of Managerial Psychology, Vol. 23, pp. 135-50.
Brotheridge, C.M. and Lee, R.T. (2003), “Development and validation of the emotional labour scale”, Journal of Occupational and Organizational Psychology, Vol. 76, pp. 365-79.
Bu¨ssing, A. and Glaser, J. (1999), “Interaktionsarbeit. Konzept und Methode der Erfassung im
Krankenhaus” (“Interaction-work. Construct and method of assessment in hospitals”),
Zeitschrift fu¨r Arbeitswissenschaft, Vol. 53, pp. 164-73.
Bu¨ssing, A. and Perrar, K.-M. (1992), “Die Messung von Burnout. Untersuchung einer deutschen
Fassung des Maslach Burnout Inventory (MBI-D)” (“Measurement of burnout: analysis of a German version of the MBI”), Diagnostica, Vol. 38, pp. 328-53.
Deery, S., Iverson, R. and Walsh, J. (2002), “Work relationships in telephone call centres: understanding emotional exhaustion and employee withdrawal”, Journal of Management
Studies, Vol. 39, pp. 471-96.

Emotional dissonance in call centre work
615

JMP
25,6

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

616

Diefendorff, J.M. and Gosserand, R.H. (2003), “Understanding the emotional labor process: a control theory perspective”, Journal of Organizational Behavior, Vol. 24, pp. 945-59.
Diefendorff, J.M. and Richard, E.M. (2003), “Antecedents and consequences of emotional display rule perceptions”, Journal of Applied Psychology, Vol. 88, pp. 284-94.
Diefendorff, J.M., Croyle, M.H. and Gosserand, R.H. (2005), “The dimensionality and antecedents of emotional labor strategies”, Journal of Vocational Behavior, Vol. 66, pp. 339-57.
Dormann, C. and Zapf, D. (2004), “Customer-related social stressors and burnout”, Journal of
Occupational Health Psychology, Vol. 9, pp. 61-82.
Fischbach, A. (2003), “Determinants of emotion work”, available at: http://webdoc.sub.gwdg.de/ diss/2003/fischbach/fischbach.pdf (accessed 25 November 2003).
Glomb, T.M. and Tews, M.J. (2004), “Emotional labor: a conceptualization and scale development”, Journal of Vocational Behavior, Vol. 64, pp. 1-23.
Goffman, E. (1959), The Presentation of Self in Everyday Life, Doubleday Anchor, New York, NY.
Gosserand, R.H. and Diefendorff, J.M. (2005), “Emotional display rules and emotional labor: the moderating role of commitment”, Journal of Applied Psychology, Vol. 6, pp. 1256-64.
Grandey, A.A. (2000), “Emotion regulation in the workplace: a new way to conceptualize emotional labor”, Journal of Occupational Health Psychology, Vol. 5, pp. 95-110.
Grandey, A.A., Dickter, D. and Sin, H.-P. (2004), “The customer is not always right: customer verbal aggression toward service employees”, Journal of Organizational Behavior, Vol. 25, pp. 397-418.
Grant, S. and Langan-Fox, J. (2007), “Personality and the occupational stressor-strain relationship”, Journal of Occupational Health Psychology, Vol. 12, pp. 20-33.
Grebner, S., Semmer, N., Lo Faso, L., Gut, S., Ka¨lin, W. and Elfering, A. (2003), “Working conditions, well-being, and job-related attitudes among call centre agents”, European
Journal of Work and Organizational Psychology, Vol. 12, pp. 341-65.
Hackman, J.R. and Oldham, G.R. (1975), “Development of the job diagnostic survey”, Journal of
Applied Psychology, Vol. 60, pp. 159-70.
Ha¨rtel, C.E.J. and Page, K.M. (2009), “Discrete emotional crossover in the workplace: the role of affect intensity”, Journal of Managerial Psychology, Vol. 24, pp. 237-53.
Heuven, E. and Bakker, A.B. (2003), “Emotional dissonance and burnout among cabin attendants”, European Journal of Work and Organizational Psychology, Vol. 12, pp. 81-100.
Hochschild, A.R. (1983), The Managed Heart, University of California Press, Berkeley, CA.
Holman, D. (2003), “Call centres”, in Holman, D., Wall, T.D., Clegg, C.W., Sparrow, P. and
Howard, A. (Eds), The New Workplace: A Guide to the Human Impact of Modern Working
Practices, Wiley, Chichester, pp. 115-34.
Holman, D., Chissik, C. and Totterdell, P. (2002), “The effects of performance monitoring on emotional labor and well-being in call centres”, Motivation and Emotion, Vol. 26, pp. 57-81.
Humphrey, R.H., Pollack, J.M. and Hawver, T. (2008), “Leading with emotional labour”, Journal of
Managerial Psychology, Vol. 23, pp. 151-68.
Jex, S.M., Beehr, T.A. and Roberts, C.K. (1992), “The meaning of occupational stress items to survey respondents”, Journal of Applied Psychology, Vol. 77, pp. 623-8.
Kinman, G. (2009), “Emotional labour and strain in ‘front-line’ service employees. Does mode of delivery matter?”, Journal of Managerial Psychology, Vol. 24, pp. 118-35.
Krohne, H.W., Egloff, B., Kohlmann, C.W. and Tausch, A. (1996), “Untersuchungen mit einer deutschen Version der ‘positive and negative affect schedule’ (PANAS)” (“Studies with a
German version of the PANAS”), Diagnostica, Vol. 42, pp. 139-56.

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

LePine, J.A., Podsakoff, N.P. and LePine, M.A. (2005), “A meta-analytic test of the challenge stressor – hindrance stressor framework: an explanation for inconsistent relationships among stressors and performance”, Academy of Management Journal, Vol. 48, pp. 764-75.
Lewig, K.A. and Dollard, M.F. (2003), “Emotional dissonance, emotional exhaustion, and job satisfaction in call centre workers”, European Journal of Work and Organizational
Psychology, Vol. 12, pp. 366-92.
McClelland, G.H. and Judd, C.M. (1993), “Statistical difficulties of detecting interactions and moderator effects”, Psychological Bulletin, Vol. 114, pp. 376-90.
Mohr, G. (1986), Die Erfassung psychischer Befindensbeeintra¨chtigungen bei Arbeitern
(Assessment of Strain in Employees), Lang, Frankfurt.
Morris, J.A. and Feldman, D.C. (1996), “The dimensions, antecedents, and consequences of emotional labor”, Academy of Management Review, Vol. 21, pp. 986-1010.
Morris, J.A. and Feldman, D.C. (1997), “Managing emotions in the workplace”, Journal of
Managerial Issues, Vol. 9, pp. 257-74.
Payne, R.L. and Cooper, G.L. (2001), Emotions at Work, Wiley, New York, NY.
Rafaeli, A. (1989), “When cashiers meet customers: an analysis of the role of supermarket cashiers”, Academy of Management Journal, Vol. 32, pp. 245-73.
Rafaeli, A. and Sutton, R.I. (1987), “Expression of emotion as part of the work role”, Academy of
Management Review, Vol. 12, pp. 23-37.
Schaubroeck, J. and Jones, J.R. (2000), “Antecedents of workplace emotional labor dimensions and moderators of their effect on physical symptoms”, Journal of Organizational Behavior,
Vol. 21, pp. 163-83.
Schmidt-Atzert, L. and Hu¨ppe, M. (1996), “Emotionsskalen EMO 16. Ein Fragebogen zur
Selbstbeschreibung des aktuellen emotionalen Gefu¨hlszustandes” (“Emotion-scales EMO
16. A questionnaire for self-report of actual emotions”), Diagnostica, Vol. 42, pp. 242-67.
Spector, P.E., Zapf, D., Chen, P.Y. and Frese, M. (2000), “Why negative affectivity should not be controlled in job stress research: don’t throw out the baby with the bath water”, Journal of
Organizational Behavior, Vol. 21, pp. 79-95.
Tamir, M. and Robinson, M.D. (2004), “Knowing good from bad: the paradox of neuroticism, negative affect and evaluative processing”, Journal of Personality and Social Psychology,
Vol. 87, pp. 913-25.
Thoresen, G.J., Kaplan, S.A., Barsky, A.P., de Chermont, K. and Warren, C.R. (2003), “The affective underpinnings of job perceptions and attitudes: a meta-analytic review and integration”,
Psychological Bulletin, Vol. 129, pp. 914-45.
Totterdell, P. and Holman, D. (2003), “Emotion regulation in customer service roles: testing a model of emotional labor”, Journal of Occupational Health Psychology, Vol. 8, pp. 55-73.
Tschan, F., Rochat, S. and Zapf, D. (2005), “It’s not only clients: studying emotion work with clients and co-workers with an event-sampling approach”, Journal of Occupational and
Organizational Psychology, Vol. 78, pp. 195-220.
Watson, D. and Pennebaker, J.W. (1989), “Health complaints, stress, and distress: exploring the central role of negative affectivity”, Psychological Review, Vol. 96, pp. 234-54.
Watson, D., Clark, L.A. and Tellegen, A. (1988), “Development and validation of brief measures of positive and negative affect: the PANAS scales”, Journal of Personality and Social
Psychology, Vol. 54, pp. 1063-70.
Wegge, J. (2004), “Emotionen in Organisationen [Emotions in organizations]”, in Schuler, H. (Ed.),
Enzyklopa¨die der Psychologie. Organisationspsychologie I – Grundlagen und
Personalpsychologie, Hogrefe, Go¨ttingen, pp. 673-749.

Emotional dissonance in call centre work
617

JMP
25,6

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

618

Wegge, J. (2006), “Communication via videoconference: emotional and cognitive consequences of affective personality dispositions, seeing one’s own picture and disturbing events”,
Human Computer Interaction, Vol. 21, pp. 273-318.
Wegge, J. and Neuhaus, L. (2002), “Emotionen bei der Bu¨roarbeit am PC: Ein Test der ‘affective events’ – Theorie” (“Emotions and office work with computers: a test of affective events theory”), Zeitschrift fu¨r Arbeits – und Organisationspsychologie, Vol. 46, pp. 173-84.
Wegge, J., Van Dick, R., Fisher, G.K., Wecking, C. and Moltzen, K. (2006), “Work motivation, organizational identification and well-being in call centre work”, Work and Stress, Vol. 20, pp. 60-83.
Wegge, J., Van Dick, R., Fisher, G.K., West, M. and Dawson, J.F. (2006), “A test of basic assumptions of affective events theory (AET) in call centre work”, British Journal of
Management, Vol. 17, pp. 237-54.
Zapf, D. (2002), “Emotion work and psychological well-being. A review of the literature and some conceptual considerations”, Human Resource Management Review, Vol. 12, pp. 237-68.
Zapf, D. and Holz, M. (2006), “On the positive and negative effects of emotion work in organizations”, European Journal of Work and Organizational Psychology, Vol. 15, pp. 1-28.
Zapf, D., Isic, A., Bechtoldt, M. and Blau, P. (2003), “What is typical for call centre jobs? Job characteristics, and service interactions in different call centres”, European Journal of
Work and Organizational Psychology, Vol. 12, pp. 311-40.
Zapf, D., Vogt, C., Seifert, C., Mertini, H. and Isic, A. (1999), “Emotion work as a source of stress.
The concept and development of an instrument”, European Journal of Work and
Organizational Psychology, Vol. 8, pp. 371-400.
Zellars, K.L., Perrewe´, P.L., Hochwarter, W.A. and Anderson, K.S. (2006), “The interactive effects of positive affect and conscientiousness on strain”, Journal of Occupational Health
Psychology, Vol. 11, pp. 281-9.
Further reading
Dormann, C. and Zijlstra, F.R.H. (2003), “Call centres: high on technology – high on emotions”,
European Journal of Work and Organizational Psychology, Vol. 12, pp. 305-10.
About the authors
Ju¨rgen Wegge is Professor of Work and Organisational Psychology at the TU Dresden
(Germany). He is the Vice President of the German Work- and Organizational Psychologist organized in the German Psychological Society (DGPs) and a fellow of AOM, SIOP, IAAP, and
EAWOP. Ju¨rgen serves on the editorial boards of five scientific journals: Journal of Occupational and Organizational Psychology, British Journal of Management, German Journal of Work- and
Organizational Psychology (ZAO), Journal of Personnel Psychology and the German Journal of
Human Resource Research. His main research interests are in the field of work motivation, age-diversity, leadership, occupational health and excellence in organizations. He has published three books and four special issues, 35 journal articles and 62 book chapters related to these topics, including articles in the Journal of Applied Psychology, Human Computer Interaction,
Work and Stress and European Journal of Work and Organizational Psychology. Ju¨rgen Wegge is the corresponding author and can be contacted at: [e-mail]
Rolf Van Dick is Professor of Social Psychology at the Goethe-University Frankfurt
(Germany) and currently serves as Associate Dean of the Department of Psychology. His research interests center on the application of social identity theory in organizational settings. In particular, he is interested in identity processes in teams and organizations which are highly diverse, he is applying identity research in the area of mergers and acquisitions and is currently

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

investigating relationships between leadership and identity. Rolf has served as Editor or
Associate Editor of the British Journal of Management and the European Journal of Work and
Organizational Psychology, and is currently editor of the Journal of Personnel Psychology. He has published more than 30 books and book chapters and 70 papers in academic journals including the Academy of Management Journal, Journal of Applied Psychology, Journal of Marketing, and
Journal of Personality and Social Psychology.
Christiane von Bernstorff (formerly Wecking) is currently working as a Consultant for
Organizational Development of Public Organizations in Berlin (Germany). Previously she worked with Ju¨rgen Wegge in the Department of Organizational Psychology at the University of
Dortmund. The main focus of her work lay on the conditions and consequences of emotional labour in call centre work. Results from her studies had been published, for example, in the
Journal of Occupational and Organizational Psychology and Work & Stress.

To purchase reprints of this article please e-mail: [e-mail]
Or visit our web site for further details: www.emeraldinsight.com/reprints

Emotional dissonance in call centre work
619

Downloaded by UNIVERSITY OF ESSEX At 16:41 03 February 2015 (PT)

This article has been cited by:
1. Joe VellaEmotional Dissonance 1-2. [CrossRef]
2. Nadya Golfenshtein, Anat Drach-Zahavy. 2015. An attribution theory perspective on emotional labour in nurse-patient encounters: a nested cross-sectional study in paediatric settings. Journal of Advanced Nursing n/a-n/a. [CrossRef]
3. Erika Tsukamoto, Takeru Abe, Michikazu Ono. 2014. Inverse roles of emotional labour on health and job satisfaction among long-term care workers in Japan. Psychology, Health & Medicine 1-10. [CrossRef]
4. Catherine K. Lam, Frank Walter, Kan Ouyang. 2014. Display rule perceptions and job performance in a
Chinese retail firm: The moderating role of employees’ affect at work. Asia Pacific Journal of Management
31, 575-597. [CrossRef]
5. Jared Kenworthy, Cara Fay, Mark Frame, Robyn Petree. 2014. A meta-analytic review of the relationship between emotional dissonance and emotional exhaustion. Journal of Applied Social Psychology 44:10.1111/ jasp.2014.44.issue-2, 94-105. [CrossRef]
6. Michel Rod, Nicholas J. Ashill. 2013. The impact of call centre stressors on inbound and outbound call‐ centre agent burnout. Managing Service Quality: An International Journal 23:3, 245-264. [Abstract] [Full
Text] [PDF]
7. Aino Tenhiälä, Robert B. Lount. 2013. Affective reactions to a pay system reform and their impact on employee behaviour. Journal of Occupational and Organizational Psychology 86:10.1111/ joop.2013.86.issue-1, 100-118. [CrossRef]
8. John D. Kammeyer-Mueller, Alex L. Rubenstein, David M. Long, Michael A. Odio, Brooke R.
Buckman, Yiwen Zhang, Marie D. K. Halvorsen-Ganepola. 2013. A Meta-Analytic Structural Model of Dispositonal Affectivity and Emotional Labor. Personnel Psychology 66:10.1111/peps.2013.66.issue-1,
47-90. [CrossRef]
9. Alma M. Rodríguez-Sánchez, Jari J. Hakanen, Riku Perhoniemi, Marisa Salanova. 2013. With a little help from my assistant: buffering the negative effects of emotional dissonance on dentist performance.
Community Dentistry and Oral Epidemiology n/a-n/a. [CrossRef]
10. Hana Medler-Liraz, Ronit Kark. 2012. It takes three to tango: Leadership and hostility in the service encounter. The Leadership Quarterly 23, 81-93. [CrossRef]

References: Aiken, L.S. and West, S.G. (1991), Multiple Regression: Testing and Interpreting Interactions, Sage, Newbury Park, CA. Ashkanasy, N.M., Zerbe, W.J. and Ha¨rtel, C.E. (2002), Managing Emotions in the Workplace, Sharpe, Armonk, NY. Baumeister, R.F., Bratslavsky, E., Muraven, M. and Tice, D.M. (1998), “Ego depletion: is the active self a limited resource?”, Journal of Personality and Social Psychology, Vol pp. 1252-65. Bono, J.E. and Vey, M.A. (2005), “Toward understanding emotional management at work: a quantitative review of emotional labour research”, in Ha¨rtel, C.E.J., Zerbe, W Boss, A.D. and Sims, H.P. (2008), “Using emotion regulation and self-leadership for recovery”, Journal of Managerial Psychology, Vol Brotheridge, C.M. and Lee, R.T. (2003), “Development and validation of the emotional labour scale”, Journal of Occupational and Organizational Psychology, Vol Bu¨ssing, A. and Glaser, J. (1999), “Interaktionsarbeit. Konzept und Methode der Erfassung im Krankenhaus” (“Interaction-work Bu¨ssing, A. and Perrar, K.-M. (1992), “Die Messung von Burnout. Untersuchung einer deutschen Fassung des Maslach Burnout Inventory (MBI-D)” (“Measurement of burnout: analysis of Deery, S., Iverson, R. and Walsh, J. (2002), “Work relationships in telephone call centres: understanding emotional exhaustion and employee withdrawal”, Journal of Management Diefendorff, J.M. and Gosserand, R.H. (2003), “Understanding the emotional labor process: a control theory perspective”, Journal of Organizational Behavior, Vol Diefendorff, J.M. and Richard, E.M. (2003), “Antecedents and consequences of emotional display rule perceptions”, Journal of Applied Psychology, Vol Diefendorff, J.M., Croyle, M.H. and Gosserand, R.H. (2005), “The dimensionality and antecedents of emotional labor strategies”, Journal of Vocational Behavior, Vol Dormann, C. and Zapf, D. (2004), “Customer-related social stressors and burnout”, Journal of Occupational Health Psychology, Vol Fischbach, A. (2003), “Determinants of emotion work”, available at: http://webdoc.sub.gwdg.de/ diss/2003/fischbach/fischbach.pdf (accessed 25 November 2003). Glomb, T.M. and Tews, M.J. (2004), “Emotional labor: a conceptualization and scale development”, Journal of Vocational Behavior, Vol Goffman, E. (1959), The Presentation of Self in Everyday Life, Doubleday Anchor, New York, NY. Gosserand, R.H. and Diefendorff, J.M. (2005), “Emotional display rules and emotional labor: the moderating role of commitment”, Journal of Applied Psychology, Vol Grandey, A.A. (2000), “Emotion regulation in the workplace: a new way to conceptualize emotional labor”, Journal of Occupational Health Psychology, Vol Grandey, A.A., Dickter, D. and Sin, H.-P. (2004), “The customer is not always right: customer verbal aggression toward service employees”, Journal of Organizational Behavior, Vol Grant, S. and Langan-Fox, J. (2007), “Personality and the occupational stressor-strain relationship”, Journal of Occupational Health Psychology, Vol Grebner, S., Semmer, N., Lo Faso, L., Gut, S., Ka¨lin, W. and Elfering, A. (2003), “Working conditions, well-being, and job-related attitudes among call centre agents”, European Hackman, J.R. and Oldham, G.R. (1975), “Development of the job diagnostic survey”, Journal of Applied Psychology, Vol Ha¨rtel, C.E.J. and Page, K.M. (2009), “Discrete emotional crossover in the workplace: the role of affect intensity”, Journal of Managerial Psychology, Vol Heuven, E. and Bakker, A.B. (2003), “Emotional dissonance and burnout among cabin attendants”, European Journal of Work and Organizational Psychology, Vol Hochschild, A.R. (1983), The Managed Heart, University of California Press, Berkeley, CA. Holman, D. (2003), “Call centres”, in Holman, D., Wall, T.D., Clegg, C.W., Sparrow, P. and Howard, A Holman, D., Chissik, C. and Totterdell, P. (2002), “The effects of performance monitoring on emotional labor and well-being in call centres”, Motivation and Emotion, Vol Humphrey, R.H., Pollack, J.M. and Hawver, T. (2008), “Leading with emotional labour”, Journal of Managerial Psychology, Vol Jex, S.M., Beehr, T.A. and Roberts, C.K. (1992), “The meaning of occupational stress items to survey respondents”, Journal of Applied Psychology, Vol Kinman, G. (2009), “Emotional labour and strain in ‘front-line’ service employees. Does mode of delivery matter?”, Journal of Managerial Psychology, Vol Krohne, H.W., Egloff, B., Kohlmann, C.W. and Tausch, A. (1996), “Untersuchungen mit einer deutschen Version der ‘positive and negative affect schedule’ (PANAS)” (“Studies with a Lewig, K.A. and Dollard, M.F. (2003), “Emotional dissonance, emotional exhaustion, and job satisfaction in call centre workers”, European Journal of Work and Organizational McClelland, G.H. and Judd, C.M. (1993), “Statistical difficulties of detecting interactions and moderator effects”, Psychological Bulletin, Vol Mohr, G. (1986), Die Erfassung psychischer Befindensbeeintra¨chtigungen bei Arbeitern (Assessment of Strain in Employees), Lang, Frankfurt. Morris, J.A. and Feldman, D.C. (1996), “The dimensions, antecedents, and consequences of emotional labor”, Academy of Management Review, Vol