|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001928116
007 cr mnu|||uuuuu
008 080221s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001937
Johnson, Hazel-Ann Michelle.
Service with a smile :
b antecedents and consequences of emotional labor strategies
h [electronic resource] /
by Hazel-Ann Michelle Johnson.
[Tampa, Fla] :
University of South Florida,
ABSTRACT: Organizations across the United States and in many parts of the globe are increasingly focused on providing their customers with an excellent service experience by implementing organizational emotion display rules (Hochschild, 1983). These display rules dictate the requisite employee emotions for a particular encounter (Ekman, 1973). However, over the course of a work day display rules may call for expressions that contradict an employee's genuine emotions, thus prompting a discrepancy between felt emotions and required emotions -- emotional dissonance (Hochschild, 1983). Emotional labor involves employee efforts to reduce emotional dissonance in order to adhere to organizational display rules (Hochschild, 1983; Grandey, 2000). Hochschild (1983) identified two emotional labor strategies that may be used by employees --^ surface acting (managing observable expressions to obey display rules) and deep acting (corresponds to managing feelings in order to actually feel the emotion required by the display rules). This study examined emotional intelligence, affectivity and gender as potential antecedents of an employee's choice of emotional labor strategy in order to meet organizational display rules. I also investigated the differential impact of the emotional labor strategies on the individual outcomes of emotional exhaustion and job satisfaction, and service performance.Correlation and moderated regression analyses as well as structural equation modeling were employed to test the proposed hypotheses. Two hundred and twenty-three employee-supervisor pairs completed surveys to examine the research hypotheses. Correlation results indicate that emotional intelligence, affectivity and gender related to the emotional labor strategies in the expected directions.^ ^Similarly, deep acting and surface acting displayed differential relationships with emotional exhaustion, job satisfaction and service performance. Moderated regression analyses suggest that females were more likely to report negative outcomes when engaging in surface acting. Structural equation modeling results indicate that affectivity predicted choice of the emotional labor strategies, which in turn predicted the outcomes of emotional exhaustion, job satisfaction and service performance.
Dissertation (Ph.D.)--University of South Florida, 2007.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 89 pages.
Adviser: Paul E. Spector, Ph.D.
Emotions at work.
t USF Electronic Theses and Dissertations.
Service With a Smile: Antecedents and Consequences of Emotional Labor Strategies by Hazel-Anne Michelle Johnson A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Psychology College of Arts and Sciences University of South Florida Major Professor: Paul E. Spector, Ph.D. Walter C. Borman, Ph.D. Michael T. Brannick, Ph.D. Ellis L. Gesten, Ph.D. Kristen L. Salomon, Ph.D. Date of Approval: April 6, 2007 Keywords: deep acting, surface acting, service perf ormance, emotions at work, employee well-being Copyright 2007 Hazel-Anne M. Johnson
Dedication To my dear mother, Icilma Johnson, for her constant encouragement and faith in me, and to my wonderful fianc, Courtney Marcus, for his lo ve, patience and unwavering support as I undertook this academic journey.
Acknowledgements Thank you to everyone who has inspired and encourag ed me over the years of this dissertation journey. Specifically, I would like t o thank my supervisory committee, Walter Borman, Ph.D., Michael Brannick, Ph.D., Elli s Gesten, Ph.D. and Kristen Salomon, Ph.D. for their valuable insights and cont ributions to my dissertation. I am so very grateful for all the opportunities and profess ional guidance provided by my mentor, Paul Spector, Ph.D. Thank you for being there for me when I needed you and for granting me the space to develop my academic potent ial! I am also blessed to have had Tammy Allen, Ph.D. as a mentor. Thank you for your advice and constant support! I need to thank Laura Pierce and the Psychology Depar tment staff for their assistance in ensuring that I adhered to all the policies and dea dlines during my tenure in graduate school. Also, I am grateful to my research assista nts, Herrica Telus, Kyle Groff and Jeremy Bauer, for their yeomans work. Finally, I would like to thank all of my classmates who have made my years of graduate schoo l some of the best, especially those in the Center for Occupational Health Psychology! Special thanks are due to Elena Lopez, Haitham Khoury, Charlie Ottinot, Burcu Rodop man, Kimberly OBrien, and Xian Xu, all of whom I am proud to call my friends!
i Table of Contents List of Tables iii List of Figures iv Abstract v Introduction 1 Emotional Labor Strategies 5 Antecedents 11 Consequences 23 Current Study 27 Method 29 Participants 29 Measures 30 Emotional Labor and Emotion Regulation 30 Emotional Intelligence 32 Positive and Negative Affectivity 33 Emotional Exhaustion 34 Job Satisfaction 34 Service Performance 34 Demographic Information 35 Procedure 36 Analyses 36 Results 38 Discussion 44 References 52 Appendices 76 Appendix A: Deep Acting and Surface Acting Measures (Brotheridge & Lee, 2003; Grandey, 2003; Gross & John, 2003) 77 Appendix B: Emotional Intelligence Scale (Wong & La w, 2002) 79 Appendix C: PANAS (Watson, Clark & Tellegen, 1988) 80 Appendix D: Emotional Exhaustion (Maslach & Jackson 1986) 81 Appendix E: Job Satisfaction from the Michigan Orga nizational Assessment Questionnaire (Camman, Fichman, Jenkins & Klesh, 19 79) 82
ii Appendix F: Affective Delivery Measure (McLellan, S chmit, Amundson & Blake, 1998 as modified by Grandey, 2003) 83 Appendix G: Task Performance Measure (Williams & An derson, 1991) 84 Appendix H: Demographics 86 Appendix I: Employee Letter 87 Appendix J: Supervisor Letter 89 About the Author End Page
iii List of Tables Table 1 Summary of Results for Study Hypotheses 64 Table 2 Descriptive Statistics and Cronbachs Alpha for Scale Variables 66 Table 3 Intercorrelations Between Study Variables 67 Table 4 Moderated Regression Analyses 6 8 Table 5 Summary of Fit Statistics for Hypothesized Model 69
iv List of Figures Figure 1. Overall hypothesized model of relationshi ps between antecedents, emotional labor strategies and outcomes. 71 Figure 2. Graph of the moderating effect of gender on the relationship between faking and emotional exhaustion. 72 Figure 3. Graph of the moderating effect of gender on the relationship between faking and job satisfaction. 73 Figure 4. Hypothesized structural model of antecede nts, emotional labor strategies and outcomes with path coefficients. 74 Figure 5. Hypothesized structural and measurement m odel of antecedents, emotional labor strategies and outcomes with path c oefficients. 75
v Service With a Smile: Antecedents and Consequences of Emotional Labor Strategies Hazel-Anne Michelle Johnson ABSTRACT Organizations across the United States and in many parts of the globe are increasingly focused on providing their customers w ith an excellent service experience by implementing organizational emotion display rule s (Hochschild, 1983). These display rules dictate the requisite employee emotions for a particular encounter (Ekman, 1973). However, over the course of a work day display rule s may call for expressions that contradict an employees genuine emotions, thus pro mpting a discrepancy between felt emotions and required emotions emotional dissonan ce (Hochschild, 1983). Emotional labor involves employee efforts to reduce emotional dissonance in order to adhere to organizational display rules (Hochschild, 1983; Gra ndey, 2000). Hochschild (1983) identified two emotional labor strategies that may be used by employees surface acting (managing observable expressions to obey display ru les) and deep acting (corresponds to managing feelings in order to actually feel the emo tion required by the display rules). This study examined emotional intelligence, affecti vity and gender as potential antecedents of an employees choice of emotional la bor strategy in order to meet organizational display rules. I also investigated the differential impact of the emotional
vi labor strategies on the individual outcomes of emot ional exhaustion and job satisfaction, and service performance. Correlation and moderated regression analyses as we ll as structural equation modeling were employed to test the proposed hypothe ses. Two hundred and twentythree employee-supervisor pairs completed surveys t o examine the research hypotheses. Correlation results indicate that emotional intelli gence, affectivity and gender related to the emotional labor strategies in the expected dire ctions. Similarly, deep acting and surface acting displayed differential relationships with emotional exhaustion, job satisfaction and service performance. Moderated r egression analyses suggest that females were more likely to report negative outcome s when engaging in surface acting. Structural equation modeling results indicate that affectivity predicted choice of the emotional labor strategies, which in turn predicted the outcomes of emotional exhaustion, job satisfaction and service performance.
1 Service With A Smile: Antecedents and Consequences of Emotional Labor Strategies Organizations across the United States and in many parts of the globe are increasingly focused on providing their customers w ith an excellent service experience. One of the ways in which they seek to do so is thro ugh organizational emotion display rules (Hochschild, 1983); these rules dictate the r equisite employee emotions for a particular encounter (Ekman, 1973). As the service-oriented e conomy provides the customer with an interactive experience, organizational display r ules are implemented to ensure that the experience is pleasant and satisfying. These displ ay rules can be formally transmitted through training manuals (Rafaeli & Sutton, 1987) o r informally encouraged by the organizational culture (Van Maanen & Kunda, 1989). For example, studies have demonstrated that when employees were dressed in th eir work uniforms they were more likely to express positive emotions to customers (R afaeli, 1989; Van Maanen & Kunda, 1989). The uniforms remind the employees of the in formal display norms that exist within their organizational culture. It is expected that pleasant and friendly employees transmit positive emotions to their customers through emotional contagion (Hatfie ld, Cacioppo & Rapson, 1992; Pugh, 2002). Emotional contagion is the tendency to aut omatically mimic and synchronize facial expressions, vocalizations, postures, and mo vements with those of another person and, consequently to converge emotionally (Hatfiel d, Cacioppo & Rapson, 1992, p. 153). Given that organizations are trying to orche strate a satisfying service experience, employees that infect customers with their positi ve emotions would be ideal. However,
2 the automatic nature of emotional contagion can bac kfire when customers present employees with negative emotions; it is in this sce nario when emotion regulation becomes necessary and potentially taxing (Pugh, 200 2). For instance, Grandey, Dickter and Sin (2004) demonstrated that call-center employ ees appraisal of verbal aggression from customers predicted their method of emotion re gulation. Employees who reported feeling more stressed with verbally aggressive cust omers faked their emotions more than employees who perceived them as less stressful. According to Wharton and Erickson (1993), there are three main types of display rules integrative, differentiating and masking. Integrative emotions are hedonically positive, create good feelings in others and encour age harmony among people (e.g., love, happiness, compassion); conversely, differentiating emotions are hedonically negative and tend to drive people apart (e.g., fear, hate, a nger). Emotional masking involves displays of neutrality and restraint with respect t o either integrative or differentiating emotions (Cropanzano, Weiss & Elias, 2004). In gen eral, organizations require that employees adhere to integrative emotion display rul es. Yet, over the course of a work day display rules may call for expressions that con tradict an employees genuine emotions, thus prompting a discrepancy between felt emotions and required emotions, this discrepancy has been a focus of research atten tion over the last two decades (e.g., Hochschild, 1983; Rafaeli & Sutton, 1987; Morris & Feldman, 1996; Brotheridge & Grandey, 2002). Hochschild (1983) termed this disc repancy emotional dissonance, that is, the separation of felt emotion from emotion express ed to meet external expectations, and she contended that it is harmful to the physical an d psychological well-being of employees.
3 Emotional dissonance is an unpleasant state, so emp loyees seek to reduce this discrepancy by utilizing a variety of emotion regul ation strategies (Grandey, 2000). Employee efforts to resolve emotional dissonance in order to adhere to organizational display rules have been termed emotional labor (Hoc hschild, 1983). Emotional labor is the expression of organizationally desired emotions by service agents during service encounters (Hochschild, 1979, 1983; Ashforth & Hump hrey, 1993). Grandey (2000) has defined emotional labor as the process of regulati ng both feelings and expressions for organizational goals (p. 97). Emotional labor has also been regarded as a type of impression management, because it is a deliberate a ttempt by the individual to manipulate his or her behavior toward others in ord er to foster both certain social perceptions of himself or herself and a certain int erpersonal climate (Gardner & Martinko, 1988; Grove & Fisk, 1989). Essentially, emotional labor involves the emotion regulation strategies employed to reduce the discre pancy between felt and organizationally-mandated emotions. Hochschild (1983) identified two strategies that ma y be used by employees to manage their emotions: Surface acting, which corres ponds to managing observable expressions to obey display rules, and deep acting, which corresponds to managing feelings in order to actually feel the emotion requ ired by the display rules. As surface acting only modifies the outward expression, the em ployee is likely to continue to experience the uncomfortable state of emotional dis sonance. On the other hand, deep acting brings the felt emotion in line with the exp ressed emotion so this strategy should serve to reduce emotional dissonance. This study w ill focus on these two emotion regulation strategies, deep acting and surface acti ng. First, I will examine the constructs
4 of deep acting and surface acting, and then I will discuss individual difference antecedents (emotional intelligence, affectivity an d gender) to these emotional labor strategies. Finally, I will examine the impact of the emotional labor strategies on emotional exhaustion, job satisfaction and customer service performance.
5 Emotional Labor Strategies Grandey (2000) recommended the utilization of emoti on regulation theory as a framework to guide emotional labor research. Emoti on regulation involves the processes by which individuals influence which emot ions they have, when they have them and how they experience and express these emot ions (Gross, 1998b, p. 275). Gross (1998a) proposed a process model of emotion t hat begins with emotional cues that lead to emotional response tendencies (behavioral, experiential, physiological), which can then lead to emotional responses. Accordingly, thi s model posits that emotion regulation is comprised of two processes, where the first proc ess is antecedent-focused in which an individual regulates the situation or appraisal tha t precedes emotion; this is analogous to deep acting. The second process, response-focused involves modification of the observable signs of emotion in a manner consistent with surface acting (Grandey, 2000). According to Gross and John (2002), it is essential that emotions are viewed as multi-componential processes concerning changes in subjective experience, expressive behavior, and physiological responding. Therefore, emotion regulation entails efforts to modify these three components. Grosss (1998a, 199 8b) process model differentiates emotion regulation strategies along the timeline of the unfolding emotional response. Mainly, there is a distinction between antecedent-f ocused and response-focused emotion regulation strategies. Antecedent-focused strategi es occur before changes in the three components full activation of emotional response tendencies, changes in behavior, and
6 peripheral physiological responding. On the other hand, response-focused strategies are attempts to curtail an emotional response that is a lready underway. Gross (1998a, 1998b) proposed four antecedent-focus ed strategies: Situation selection (approach or avoidance of a particular si tuation), situation modification (tailoring a situation to alter its emotional impac t), attentional deployment (selective focus on other aspects of the situation), and cogni tive change (reappraisal of the meaning of the situation). The main response-focused strat egy, response modulation, involves efforts to influence emotion response tendencies th at have already been elicited (Gross, 1998a, 1998b). Typically adherence to integrative display rules via response modulation involves the faking of positive emotions, suppressi on of negative emotions, or intensification of an authentic emotion (Grandey & Brauburger, 2002). As Grandey (2000) indicated, some of Gross emotion regulation strategies are more clearly applicable to the service context, namely, the atte ntional deployment (positive refocus), cognitive change (reappraisal) and response modulat ion strategies. Of the antecedentfocused emotion regulation strategies, situation se lection and situation modification are the least likely to be employed by customer service employees as they do not often have the ability or autonomy necessary to avoid or to mo dify the situation. To some degree, attentional deployment is applicable, but may be di fficult for service employees because the source of the differentiating or negative emoti on may be the customer and it would inappropriate for them to divert their focus from t he customer. However, one aspect of attentional deployment, positive refocus, may be ut ilized if the employee is able to successfully focus on a positive aspect of the situ ation without ignoring the customer. For these reasons, this study will focus on the emo tion regulation strategies of positive
7 refocus and reappraisal (analogous to deep acting) as well as the faking and suppression of emotion (analogous to surface acting). Gross and John (2002) posited that because reapprai sal (deep acting) involves emotional cues and impacts emotional response tende ncies before they are fully activated, it does not require significant cognitiv e effort. On the other hand, suppression (surface acting) requires continuous monitoring and modification of activated emotional response tendencies. Consequently, suppression (su rface acting) entails greater cognitive costs than reappraisal. Research by Richards and G ross (1999) demonstrated that suppression (surface acting) impaired female partic ipants incidental memory for information presented during suppression. Therefor e, suppression (surface acting) may not be the best choice for an employee who needs to remember critical information obtained during a service interaction. For instanc e, an irate client calls a financial services representative and while screaming in an a busive fashion, provides information that the representative needs to alleviate the clie nts frustration. However, if the representative engaged in suppression (surface acti ng) due to the clients angry display some of the important information provided during t he transaction may be forgotten, which potentially makes that representative a targe t for further abuse by the dissatisfied client. While suppression (surface acting) effectively decr eases expressive behavior, it does not reduce subjective experience of the emotio n and in fact leads to increases in physiological responding. Conversely, reappraisal (deep acting) serves to decrease expressive behavior as well as subjective experienc e and is not associated with increased physiological responding (Gross, 1998a). Overall, reappraisal (deep acting) is the
8 emotion regulation strategy that produces the outco mes most in accordance with the integrative display rules of most organizations; sp ecifically, individuals who habitually engage in reappraisal (deep acting) feel and exhibi t more positive emotion and less negative emotion. Individuals who suppress (surfac e act) have a contrasting result they feel and exhibit less positive emotion, while they actually feel more negative emotion than habitual reappraisers (Gross & John, 2003). G ross and John (2003) posit that suppressors (surface actors) experience greater neg ative emotion because of inauthenticity caused by the discrepancy between fe lt emotions and expressive behaviors, also known as emotional dissonance. Laboratory research has indicated that suppression (surface acting) extracted cognitive costs such as distraction and reduced res ponsiveness during conversation that led to increased physiological responding in the co nversation partner (Butler, Egloff, Wilhelm, Smith, Erickson & Gross, 2003). Expressio n of emotions during a social interaction conveys relevant information to the int eraction partner about the other partys emotions, intentions, and orientation to the relati onship. Disruption of the accurate transfer of emotions contributes to the collapse of social interactions (Keltner & Kring, 1998). That is, emotional displays are usually met with a prescribed range of emotions and when our interaction partners response deviate s significantly from that range; it becomes socially awkward. Imagine having just desc ribed to a companys service representative how their product caused you grievou s bodily injury, to which they respond with gales of laughter entirely inappropr iate and very socially inept! While this is example is a little extreme, it serves to i llustrate the social consequences of inappropriate emotional responses. Engaging with a n individual who does not provide
9 the appropriately contingent responses is stressful and according to Butler et al. (2003) actually causes increased physiological responding for the interaction partner of the individual suppressing their emotions. Given that organizations implement display rules to facilitate a pleasant service experience for the ir customers, employees who habitually suppress (surface act) may actually produce a negat ive experience by increasing customers blood pressure! Most emotional labor research has been concerned wi th its potentially negative impact on service employees (e.g., Brotheridge & Gr andey, 2002; Totterdell & Homan, 2003). In particular, field research has demonstra ted a clear link between surface acting and burnout (Brotheridge & Grandey, 2002; Johnson & Spector, in press), while deep acting has been positively associated with service performance (Grandey, 2003; Totterdell & Holman, 2003); therefore, it is clear that the choice of emotion regulation strategy influences both individual and organizatio nal outcomes. The strategy that employees choose to address emoti onal dissonance can have negative effects, for instance, surface acting may lead to feelings of misalignment and inauthenticity that can decrease an employees sens e of well-being (Sheldon, Ryan, Rawsthorne, & Ilardi, 1997). Conversely, regulatio n through deep acting in a good faith type of emotional labor may result in a sens e of accomplishment depending on the employees level of identification with the organiz ation (Ashforth & Humphrey, 1993). Emotions research has shown that the inhibition of negative emotions over time can be associated with a variety of physical illnesses, su ch as asthma (Florin, Freudenberg & Hollaender, 1985), cardiovascular disease (Guyton & Hall, 1997) and cancer (Gross, 1989; Greer & Watson, 1985). Therefore, while deep acting and surface acting enable an
10 employee to successfully achieve organizational goa ls, they may also contribute to detrimental effects to that employees health and p sychological well-being. However, it must be noted that the expression of positive emoti ons may cause physiological changes that result in increased well-being for employees ( Zajonc, 1985), so positive display rules may lead to positive emotions in employees in a way that might be beneficial. This study aims to examine potential antecedents of an employees choice of emotional labor strategy in order to meet organizat ional display rules. In particular, emotional intelligence, affectivity and gender are thought to influence the selection of an emotional labor strategy. I will also investigate the differential impact of the emotional labor strategies on individual outcomes of emotiona l exhaustion, job satisfaction, and service performance, an especially important outcom e for organizations. To this end, Figure 1 presents a model that illustrates the prop osed links between the antecedents, emotional labor strategies and outcomes. The remai nder of this introduction will cover the various linkages in this model. Based on the p receding discussion of the emotional labor strategies, it is expected that they will be differentially related to proposed antecedents and the proposed outcomes.
11 Antecedents Emotional Intelligence Emotional intelligence (EI) is arguably one of the most passionately debated constructs in the field of social sciences. Yet, t o date, there has been little consensus on what it is, what it measures and its unique contrib ution to the prediction of meaningful outcomes. We can generally define emotional intell igence as an ability (or constellation of abilities) involving emotions in the self and ot hers (Mayer & Salovey, 1997). While this suffices as a general definition, more specifi c definitions of emotional intelligence depend on the research stream in question. Two maj or research streams on emotional intelligence have emerged; the ability models propo sed by Mayer and Salovey (1997) and the mixed models proffered by Goleman (1995) and Ba r-On (1997) that present broader definitions of emotional intelligence involving a r ange of emotion-related skills and traits. The main source of the controversy around the emoti onal intelligence construct stems from the disparity between the definitions presente d by both camps; that is, Mayer and Salovey (1997) view it as a form of intelligence th at only involves emotion-related abilities, while Goleman (1995) and Bar-On (1997) i nclude dimensions of personality and social competence. Consequently, a schism has developed between the proponents of the ability models and the mixed models such tha t measures of emotional intelligence as an ability do not converge with measures of emot ional intelligence that encompass personality dimensions. The breadth of the mixed m odel approach to emotional intelligence has led to the criticism that emotiona l intelligence is nothing more than the
12 re-packaging of old constructs (e.g., Landy, 2005). In addition, the measures of the mixed model approach tend to substantially overlap with existing measures which casts doubt on their ability to provide incremental predi ction of meaningful outcomes. For these reasons, this study will utilize the ability model proposed by Mayer and Salovey (1997) as this more precise model has received more empirical support, while the support for the mixed models often derives from anecdotes a nd resides within proprietary databases less subject to peer review (Landy, 2005) Salovey and Mayers (1990) initial definition of em otional intelligence is widely recognized as the origin of research on the ability model of emotional intelligence (Ashkanasy & Daus, 2005). Salovey and Mayer (1990) defined emotional intelligence as an individuals ability to monitor feelings and emo tions in the self and others, to discriminate among emotions, and to use information about emotions to guide ones thinking and actions. Therefore, individuals high in emotional intelligence are capable of understanding and expressing their own emotions, re cognizing emotions in others and regulating affect, as well as the use of emotions t o engage in adaptive behaviors (Salovey & Mayer, 1990). Further work on the construct led Mayer and Salovey (1997) to propose an emotional intelligence framework that is compris ed of four branches (1) the perception, appraisal and expression of emotion, (2 ) emotional facilitation of thought, (3) understanding of emotion, and (4) managing of emoti on in self and others. This framework is a multidimensional hierarchy in which perception of emotions, as the most basic skill, serves as a precursor to the remaining three sets of skills. As such, emotion management is the most complex branch that depends on successful utilization of the other abilities in this hierarchy.
13 The first dimension, or branch, concerns an individ uals ability to accurately identify emotions in the self and others and to acc urately express emotions. The second branch involves the assimilation of emotions into m ental processes, that is, emotions may serve as memory aids for judgments about feelings. Alternatively, problem-solving approaches may be influenced by current emotional s tates; for instance, happy moods facilitate inductive reasoning and creativity (Palf ai & Salovey, 1993). The third branch focuses on the ability to understand emotions and t he complexity of emotions and their progressions. Finally, the fourth branch is concer ned with the regulation of emotion in the self and others; for instance, the ability to c alm down after feeling anger or to alleviate the fears of another person (Mayer & Salo vey, 1997). Support for the Ability Model. Mayer, Caruso and Salovey (1999) conceptualize emotional intelligence as a new form of intelligenc e, which serves to broaden the coverage of the intelligence construct space. They present three standard criteria that must be met by an intelligence in order to be consi dered scientifically legitimate and then go on to demonstrate how emotional intelligence mee ts these criteria. First, the conceptual criterion mandates that an intelligence reflects intellectual performance instead of a preferred way of behaving or a persona lity trait, and should clearly measure the particular concept; in this case, emotion-relat ed abilities. The correlational criterion maintains that an intelligence should describe a s et of closely-related abilities that are similar to, but distinct from already-established i ntelligences (Mayer et al., 1999, p. 270; Carroll, 1993). The final criterion concerns the d evelopmental nature of intelligence, as it is supposed to improve with age and experience.
14 According to Mayer et al. (1999) emotional intellig ence is an ability that enables individuals to utilize emotional knowledge to solve emotional problems. The solutions to these emotional problems can be objectively verifie d by expert or group consensus. Darwin (1872/1965) established the universality of emotions, such that all humans can recognize and express at least six basic emotions. Subsequently, there has been universal agreement on emotional information thereby lending credence to the notion that there can be consensus about the correct answer to an emotion al problem. Mayer, Salovey, Caruso and Sitarenios (2001) discuss a method of measuring emotional intelligence (via the MSCEIT v. 2.0) in which they utilize expert consens us, such that they gather emotions experts from various countries and across demograph ic characteristics, and have them provide their correct answers to the emotional prob lems posed in the measure. The group of experts is better able to assess the correct ans wer because their training enables them to more accurately determine the consensus than mem bers of the non-expert group. In order to meet the correlational criterion, Mayer et al. (1999) demonstrated that emotional intelligence is correlated to, yet distin ct from, other types of intelligence such as verbal intelligence, as measured by the Army Alp ha intelligence test (Yerkes, 1921). Studies by Mayer et al. (1999, 2001) showed that ad ults exhibited higher levels of emotional intelligence than adolescents, and that t he pattern of relationships between EI and related variables remained the same between the adolescent and adult group. Therefore emotional intelligence can be considered developmental in nature and consequently meets the third criterion for an intel ligence. Moreover, Brackett and Mayer (2003) demonstrated the criterion-related validity of emotional intelligence, measured by
15 the MSCEIT v. 2.0, through its ability to predict n egative behaviors in a group of college males after controlling for personality and verbal SAT scores. The ability model of emotional intelligence (Salove y & Mayer, 1990; Mayer & Salovey, 1997) has spurred the development of a num ber of measures that unlike the MSCEIT v.2.0 are based on selfor peer-reports (e. g., Jordan, Ashkanasy, Hrtel & Hooper, 2002; Wong & Law, 2002). In fact, Ashkanas y and Daus (2005) classify such measures as a new stream of research on emotional i ntelligence. However, as Conte and Dean (2006) point out, self-report measures based o n the ability model may be best characterized as measures of self-perceptions of em otional abilities rather than as measures of EI abilities (Barchard & Hakstian, 2004 ). Spector and Johnson (2006) suggest that self-report measures of emotional inte lligence may reflect emotional selfefficacy rather than emotional intelligence itself, but these are empirical questions that should be addressed in future research. Wong and Law (2002) developed a self-report measure of emotional intelligence (Wong and Law Emotional Intelligence Scale, WLEIS) that derives its four dimensions from the Salovey and Mayer (1990) conceptualization of emotional intelligence. Consequently, Wong and Laws (2002) dimensions are (1) appraisal and recognition of emotion in the self (self-emotional appraisal); (2) appraisal and recognition of emotion in others (others emotional appraisal); (3) regulatio n of emotion in the self (regulation of emotion); and (4) use of emotion to facilitate perf ormance (use of emotion). Contrary to the Mayer and Salovey ability measures of emotional intelligence, the WLEIS does not assess an individuals ability to solve emotional p roblems. Instead, it measures selfperceptions of emotional intelligence or emotional self-efficacy.
16 According to Wong and Laws (2002) theoretical fram ework, employees who are high in emotional intelligence should be able to ef fectively engage in emotion regulation to satisfy organizational display rules with greate r ease and effectiveness. Employees high on the first two dimensions will be more aware of their emotions and those of their customers. Consequently, they will recognize the n eed to engage in emotional labor in order to satisfy display rules and contribute to a positive service experience for the customer. Individuals high on the third and fourth dimensions should be skilled emotion laborers because they possess the ability to quickl y adapt to the conflict between felt and expressed emotions. Accordingly, employees high in emotional intelligence are more likely to utilize deep acting because it is the mor e effective strategy to produce the emotions required by the display rules. While emot ional intelligence is a multidimensional construct, it is prudent to consid er the overall abstraction, as it is the driver of the abilities within each dimension (Ct 2005; Law, Wong & Mobley, 1998; Wong & Law, 2002). Ct (2005) provides some initial findings that sup port the relationship between emotional intelligence and deep acting. He reports that individuals with high emotional intelligence were more likely to engage in deep act ing during interpersonal interactions. Cts (2005) findings are among the first to direc tly link emotional intelligence to the emotional labor strategies, and serve as a good sta rting point for further exploration of the relationship between these two constructs. Given t hat the emotional labor strategies are posited to have differential effects on individual well-being and performance, emotional intelligence is thought to be a vital characteristi c that enables an individual to appropriately match the strategy to the situation ( Feldman Barrett & Gross, 2001).
17 H1a: Emotional intelligence will be positively rela ted to deep acting. H1b: Emotional intelligence will be negatively rela ted to surface acting. Affectivity Affective traits serve as predispositions to partic ular emotional responses (Weiss & Cropanzano, 1996). Positive affectivity indicate s the extent to which a person feels enthusiastic and optimistic, whereas negative affec tivity corresponds to pessimism and aversive mood states (Watson & Tellegen, 1985; Gran dey, 2000). Morris and Feldman (1996) contend that an individuals predisposition to experience positive or negative affect will influence emotional dissonance. That i s, if the organizationally prescribed emotions conflict with an employees affectivity (p ositive or negative), then emotional dissonance will occur more often, therefore, indivi duals whose display rule requirements are congruent with their affective states should ex perience fewer negative outcomes. Brotheridge and Lee (2003) posited that affectivity corresponds to both the range and intensity of emotions displayed, and the use of sur face or deep acting. Individuals with high levels of affectivity may have greater trouble concealing their feelings with surface acting and realigning their feelings through deep a cting, than individuals with low affectivity (Brotheridge & Lee, 2003). Therefore, an individual who is high in positive affectivity may not fit well in a job that required the expression of negative emotions, such as a bill collector. Conversely, an individua l high in negative affectivity may not be the best choice for the job of a customer service r epresentative. Research has consistently found a positive relation ship between negative affectivity and surface acting (Brotheridge & Grand ey, 2002; Brotheridge & Lee, 2003; Johnson, 2004; Gosserand & Diefendorff, 2005); such that it appears that high negative-
18 affect individuals are more likely to fake or suppr ess their emotions than to modify their feelings in order to stick to display rules. Howev er, findings on affectivity and deep acting have not been so clear only two known stud ies report findings. Johnson (2004) found that positive affectivity and deep acting wer e positively associated, and while deep acting and negative affectivity were negatively rel ated, this relationship was nonsignificant. Similarly, Gosserand and Diefendorff (2005) reported a positive relationship between positive affectivity and deep acting, and a negative, albeit non-significant, relationship between negative affectivity and deep acting. As there are so few findings about two constructs that exist within the same nom ological network, it is important that further research examines the relationship between affectivity and the emotional labor strategies, especially deep acting. It is expected that individuals high in positive af fectivity are more inclined to employ deep acting to meet positive display rules b ecause modification of their currently negative or neutral state to a more disposition-app ropriate positive state, should reduce emotional dissonance and its attendant negative con sequences. On the other hand, individuals with high negative affectivity should b e more prone to engage in surface acting to meet positive display rules, because such a strategy only modifies the expression of emotion, not the experience of emotio n. These hypotheses are proposed in the context of integrative organizational display r ules that mandate the expression of positive emotion and suppression of negative emotio n. H2a: Positive affectivity will be positively relate d to deep acting. H2b: Positive affectivity will be negatively relate d to surface acting. H3a: Negative affectivity will be negatively relate d to deep acting.
19 H3b: Negative affectivity will be positively relate d to surface acting. Gender Hochschilds initial (1983) work on emotional labor focused on female flight attendants. She noted that women significantly out number men in the service industry. Her initial concern was that, due to their numerica l superiority in service work, the negative aspects of emotional labor were disproport ionately affecting women. Women have maintained their numerical superiority within service occupations as the Current Population Survey (Bureau of Labor Statistics, 2005 ) estimates that two-thirds of employees in service occupations are indeed female. However, contrary to Hochschilds (1983) original concern, Whartons (1993) research has demonstrated that women who perform jobs requiring emotional labor are signific antly more satisfied than men who perform the same type of job. This contradictory f inding implies that perhaps women are better socialized to handle the interpersonal deman ds of emotion management in service work, and this competency may lead them to have a m ore positive experience than their male counterparts. Rafaeli (1989) also posits sexrole socialization as an explanation for higher display of positive emotions by female conve nience store clerks. Alternatively, more positive expression of emotion may be due to w omens superiority at expressing emotions, that is, both male and female clerks may be trying to adhere to the positive organizational display rules, but females are more successful due to their superior ability to express emotions (Rafaeli, 1989; Hall & Halberst adt, 1981). Research by Bulan, Erickson and Wharton (1997) demo nstrated that effectiveness in working with people was more important to job su ccess for women than for men. This perceived effectiveness in working with people was associated with positive feelings
20 about work for women, but not for men. Bulan et al (1997) suggested that the peopleoriented nature of service work is more closely rel ated to womens traditional care-taking role, so the authenticity felt by women involved in such roles contributes to their positive feelings about service work. Along those lines, Pu gh (2002) referred to service work as gendered, that is, stereotyped feminine behaviors l ike friendliness, deference, and flirting are all hallmarks of good service (Hall, 1993; Hoch schild, 1983). Interestingly, Gross and John (1998) found that wom en scored higher than men on the three core dimensions of emotional expressiv ity positive expressivity, negative expressivity and impulse intensity (strength of res ponse tendencies). On the other hand, they found that men reported more masking of their emotions than women, in essence, men reported more suppression of the type employed in surface acting. Subsequent research by Gross and John (2003) also demonstrated that men suppress more than women. Furthermore, Pugh (2002) pointed out that w omen are likely to display more positive and negative emotions in the service encou nter. While organizational display rules support the display of positive emotions, dis play of negative emotions is often a sanction-worthy event (Van Maanen & Kunda, 1989). Due to females greater tendency to display stronger positive and negative emotions than males they may have to engage in more emotion regulation in order to adhere to the i ntegrative display rules. Research has demonstrated a relationship between ge nder and emotional dissonance, such that women reported more cases in which they felt differently than they expressed (Kruml & Geddes, 1998). It is possible th at while more satisfied, women may have higher levels of stress or psychological ailme nts that are related to their more frequent and successful emotion regulation. The co ntradictory research on the effects of
21 emotional labor on women may be explained by whethe r they engage in deep acting or surface acting. Deep acting may enable women to ex perience positive emotions, which correspond to outcomes such as job satisfaction. C onversely, surface acting while producing the appropriate expressive behavior, does not alleviate emotional dissonance which can lead to negative outcomes such as emotion al exhaustion and poor service performance (Grandey, 2003). In a recent study by Johnson and Spector (in press) women reported significantly more deep acting than men while surface acting was associated with more negative outcomes for women th an for men. Specifically, women experienced more emotional exhaustion and lower aff ective well-being with increased surface acting. On the other hand, men actually ex perienced lower emotional exhaustion and slightly higher affective well-being with incre ased surface acting. Women may be more inclined to choose deep acting as their emotio n regulation strategy, therefore when they do engage in surface acting the outcomes are m ore negative than for men because such regulation contradicts their preferred strateg y. While gender differences in emotion expression have been well-documented (LaFrance & Banaji, 1992), it is important to exami ne gender differences in preferred emotion regulation strategy as service performance often hinges upon successful emotion regulation. Gender role socialization may better e quip women to adhere to organizational display rules as they often coincide with societal display rules; that is, women are expected to, and usually do, display more warmth an d liking cues (Bem, 1974; Rafaeli, 1989) that are consistent with the integrative disp lay rules implemented by organizations. Consequently, when faced with emotional dissonance in a service encounter, women should be more likely to engage in deep acting in o rder to produce the authentic emotion
22 required by the display rules. However, given that socialization arguably predisposes women to feel and display integrative emotions, eng aging in surface acting should be more detrimental than for men because the discord c reated by surface acting magnifies their lack of adherence to both sets of display rul es (societal and organizational). Therefore, it is also expected that women who engag e in surface acting will experience more negative outcomes than men who choose this met hod of emotion regulation. H4a: Females will be more likely than males to enga ge in deep acting. H4b: Males will be more likely than females to enga ge in surface acting.
23 Consequences Emotional exhaustion Emotional exhaustion is a specific stress-related r eaction, and it is considered a key component of burnout (Maslach, 1982). Emotiona l exhaustion is the state of depleted energy caused by excessive emotional deman ds made on people interacting with customers or clients (Saxton, Phillips & Blakeney, 1991), and involves feelings of being emotionally overextended and exhausted by ones wor k (Maslach, Jackson & Leiter, 1996, p. 4). Research by Wharton (1993) has shown that although jobs requiring emotional labor do not place employees at greater r isk of emotional exhaustion than other jobs, all else being equal, emotional labor does re sult in negative consequences under some circumstances. Kruml and Geddes (2000) demons trated that surface acting (which they conceptualized as dissonance) was more strongl y related to emotional exhaustion than deep acting (conceptualized as effort). Broth eridge and Grandey (2002) found that surface acting was positively related to emotional exhaustion while deep acting showed almost no relationship. In an experience sampling study of call-center employees, Totterdell and Holman (2003) demonstrated that surf ace acting was more positively associated with emotional exhaustion than deep acti ng. In addition, Grandey, Fisk and Steiner (2005) also found that surface acting was p ositively related to emotional exhaustion. Finally, Johnson and Spectors (in pre ss) recent findings also support this notion, that is, surface acting was positively rela ted to emotional exhaustion, while deep acting exhibited a negative relationship with emoti onal exhaustion. Therefore, it is likely
24 the choice of emotional labor strategy that influen ces an employees level of emotional exhaustion. H5a: Deep acting will be negatively related to emot ional exhaustion. H5b: Surface acting will be positively related to e motional exhaustion. As addressed in the previous section, recent resear ch has demonstrated that gender serves to moderate the impact of surface act ing on individual outcomes (Johnson & Spector, in press). Therefore, it is also expect ed that women who engage in surface acting will experience more negative outcomes than men who choose this method of emotion regulation. H5c: Gender moderates the relationships between sur face acting and emotional exhaustion. Females who surface act will experienc e more emotional exhaustion than males. Job satisfaction Job satisfaction is an attitudinal variable that ga uges how an employee feels about his or her job. Early research on the relationship between emotional labor and job satisfaction found both positive (Adelmann, 1995; W harton, 1993) and negative relationships (Abraham, 1998; Morris & Feldman, 199 7). These findings may be explained by the emotional labor strategy employed, for instance, surface acting may lead to feelings of inauthenticity and consequently job dissatisfaction. Conversely, if an employee engages in deep acting this may lead to fe elings of personal accomplishment and by extension, job satisfaction (Kruml & Geddes, 2000). In fact, Wolcott-Burnam (2004) found that job satisfaction was negatively r elated to surface acting and positively related to deep acting. Grandey et al. (2005) also report a negative relationship between
25 surface acting and job satisfaction. Further, a re cent meta-analysis by Bono and Vey (2005) indicates that the type of emotional labor m atters; that is, surface acting was negatively related to job satisfaction, while deep acting displayed a non-significant relationship with job satisfaction. H6a: Deep acting will be positively related to job satisfaction. H6b: Surface acting will be negatively related to j ob satisfaction. Based on recent finding that gender moderates the i mpact of surface acting on individual outcomes (Johnson & Spector, in press), it is expected that women who engage in surface acting will experience more negat ive outcomes than men who employ this emotional labor strategy. H6c: Gender moderates the relationships between sur face acting and job satisfaction. Females who surface act will experie nce lower job satisfaction than males. Service performance Employee performance encompasses voluntary behavior s that are relevant to organizational goals (Campbell, McCloy, Oppler & Sa ger, 1993). In particular, employee service performance involves behaviors tha t serve and help customers (Liao & Chuang, 2004). Display rules make emotion regulati on a critical part of service performance because conforming to them requires emp loyee planning and effort (Pugh, 2002). Emotional labor, when it serves to induce t he appropriate feelings in customers, should result in good service performance (Ashforth & Humphrey, 1993). Typically, positive emotional expressions lead to better servi ce performance. However, insincere emotional expressions, if perceived as such by cust omers, result in poor service
26 performance (Rafaeli & Sutton, 1987). Given that d eep acting modifies the employees feelings to approximate the expressed emotion, this type of display is less likely to be perceived as disingenuous. Conversely, surface act ing involves just the modification of expressed emotion, so it is still possible for the negative feelings to leak out through other channels of nonverbal communication or tone of voic e (Grandey, 2000; Ekman & Friesen, 1969). Grandey (2003) found that affectiv e delivery (expression of positive emotions in a service interaction) was positively r elated to deep acting and negatively related to surface acting In addition, Totterdell and Holman (2003) report th at deep acting was positively associated with display of po sitive emotions and service performance while surface acting did not demonstrat e such a relationship. Similarly, Wolcott-Burnam (2004) reported that deep acting was positively related to coworker ratings of service performance. Therefore, choice of emotional labor strategy should predict service performance, such that, service per formance will be positively related to deep acting and negatively related to surface actin g. H7a: Deep acting will be positively related to serv ice performance. H7b: Surface acting will be negatively related to s ervice performance.
27 Current Study This study contributes to the emotional labor lite rature by advancing the understanding of some antecedents and consequences of emotional labor strategies. First, the examination of affectivity and emotional intell igence as antecedents to the emotional labor strategies is somewhat novel as only a few ot her studies have done so (see Gosserand & Diefendorff, 2005; Ct, 2005). In add ition, the simultaneous examination of these variables allows for the investigation of their unique predictive contributions to the emotional labor process. Secondly, this study compares three measures of emotional labor strategies in an effort to broaden the scope of the construct and provide more details about how the emotional labor strategies impact the individual and the organization. Brotheridge and Lees (2003) Emotional Labour Scale (ELS) has been the most widely used measure of emotional labor research and I util ized their deep acting and surface acting subscales. This study also uses the Emotion Regulation Questionnaire developed by Gross and John (2003) that assesses the reapprai sal and suppression processes that are analogous to deep acting and surface acting (Grande y, 2000). Grandeys (2003) measure of antecedentand response-focused emotion regulat ion was also used in this study. While Grandeys (2003) measure is based in part on Brotheridge and Lees (2003) ELS, she provides a number of unique items that warrant comparison to the other measures. The inclusion of these three measures allows for a unique comparison that can help further refine the measurement of the emotional lab or strategies.
28 A third contribution of this study is the examinati on of the link between emotional labor strategies and service performance. The emot ional labor concept evolved from the notion that organizations wanted their service empl oyees to manage their emotions for a wage and that this practice would be detrimental to employee well-being (Hochschild, 1983). As such, most of the emotional labor resear ch has investigated the personal consequences of managing ones emotions to adhere t o organizational display rules, and far fewer studies have examined the impact of emoti onal labor strategies on service performance a very important organizational outco me (for exceptions see Grandey, 2003; Gosserand & Diefendorff, 2005). Finally, as emotional labor is an intra-individual process, most studies have solely employed a self-report approach (for exceptions see Grandey, 2003 and Gosserand & Diefendorff, 2005). In order to minimize the poten tial of shared biases between reports of emotional labor and performance, this study obta ined supervisor reports of employee service performance. Moreover, tapping into the su pervisor as an alternative source of data should provide better measurement of service p erformance as it is perceived by others (Borucki & Burke, 1999). For this study the service performance construct was assessed with measures of affective delivery and ta sk performance in an effort to adequately cover the relevant performance criterion space.
29 Method Participants A sample of 280 employees and 223 supervisors parti cipated in this study. A criterion for participation in the study was that p articipants must have engaged in a significant amount of customer interaction as a par t of their job, so this sample should be representative of customer service employees across a number of different organizations. For this study, 595 employee-supervisor survey pack ets were distributed to full-time employees in undergraduate classes at the Universit y of South Florida. This data collection effort yielded 280 usable employee surve ys and 223 usable supervisor surveys for respective response rates of 47 percent and 38 percent. The employee sample was 74 percent female and had a n overall mean age of 22, with a range from 18 to 60. Average tenure for the employees was one year and eight months and ranged from three weeks to ten years and eight months. Approximately 64 percent of the employee sample was White, 15 percen t Hispanic, 13 percent Black, 3 percent Asian, and 5 percent Other. The supervisor s were more evenly split according to gender with women accounting for 57 percent and men for 43 percent of the sample. The mean age for supervisors was 36 and ranged from 19 to 62. On average, tenure was longer for supervisors (about six years) and ranged from one month to almost 40 years. Ethnicity varied less among supervisors with 72 per cent White, 9 percent Hispanic, 7 percent Black, 4 percent Asian and 8 percent Other.
30 Measures Measures of Emotional Labor and Emotion Regulation (Brotheridge & Lee, 2003; Grandey, 2003; Gross & John, 2003; Appendix A ) Three established scales were utilized in this stud y in order to broadly cover the emotional labor strategies of deep acting and surfa ce acting. These three scales are used to gain a better understanding of the emotional lab or strategies and to replicate previous findings. Deep acting can be achieved through reap praisal and positive refocus, while surface acting can operate through suppression and emotive faking. Three items from Brotheridge and Lees (2003) Emotional Labour Scale (ELS) cover the positive refocus construct, and six items from Gross and Johns (200 3) Emotion Regulation Questionnaire (ERQ) cover the reappraisal construct for a total of nine items that assess the deep acting construct. Four items from the ERQ as well as two items from the ELS measure the suppression component of surface acting, while emotive faking is captured by five items from Grandey (2003) for a total of 11 items measuring the surface acting construct. For the Emotional Labour Scale (ELS), the dimension s of interest are measured with a five-point Likert response scale (1 = never 5 = always ). Participants are asked to answer items in response to the stem question, On an average day at work, how often do you do each of the following when interacting with customers? Higher average scores on each of the subscales represent higher levels of the dimension being assessed. The three items in the deep acting subscale assess how much an employee has to modify
31 feelings to comply with display rules. The followi ng represents a sample item from the deep acting subscale, Make an effort to actually f eel the emotions that I need to display to others. The surface acting dimension consists of three items that measure the extent to which the employee has to express emotions that are not felt and suppress feelings that conflict with display rules. Two of the items from the surface acting dimension address suppression, while the other item addresses emotive faking. The following represents a sample item from the surface acting subscale, Hide my true feelings about a situation. Brotheridge and Lee (2002) reported acceptable coef ficient alphas for the deep acting and surface acting subscales ( = 0.89, =0.86). Grandeys (2003) antecedent-focused and response-fo cused emotion regulation measure consists of three items to measure antecede nt-focused emotion regulation, which corresponds to deep acting, and five items to measu re response-focused emotion regulation, which corresponds to surface acting. T he three deep acting items closely parallel Brotheridge and Lees (2003) deep acting s ubscale so they will not be used in this study. The emphasis here is on the antecedent -focused emotion regulation items that address the emotive faking construct. Items assess the extent to which employees have to engage in these behaviors to effectively perform th eir job. A sample item would be I put on an act in order to deal with customers. A five-point Likert scale is used where one corresponds to never and five corresponds to always Grandey (2003) reported acceptable coefficient alphas for the deep acting a nd surface acting subscales ( = 0.79, = 0.88). Gross and Johns (2003) Emotion Regulation Question naire (ERQ) assesses individual differences in expressive suppression an d cognitive reappraisal with a seven-
32 point Likert response scale (1 = strongly disagree ; 7 = strongly agree ). Items measure emotional experience, or feelings, and emotional ex pressions, in the form of speech, gestures and behaviors. The suppression subscale c onsists of four items (e.g., I control my emotions by not expressing them .). The reappraisal subscale is comprised of six items, for instance, When I want to feel more positive emotion (such as joy/amusement), I change what Im thinking about . Gross and John (2003) reported acceptable alpha s for the reappraisal and suppression subscales ( = 0.79, = 0.73). As discussed, the refocus (Brotheridge & Lee, 2003) and reappraisal measures (Gross & John, 2003) were both utilized to cover th e deep acting construct for the correlational analyses. However, in the interest o f parsimony, these two measures were combined to represent an overall deep acting constr uct for examination of the hypothesized model with the structural equation mod eling. Similarly, the emotive faking (Grandey, 2003) and suppression (Brotheridge & Lee, 2003; Gross & John, 2003) measures were employed separately to cover the surf ace acting construct in the correlational analyses, but combined to represent a n overall surface acting construct in the hypothesized model. All four separate measures of the emotional labor strategies demonstrated acceptable internal consistency reliab ilities (refocus, = .88; reappraisal, = .74; faking, = .90; suppression, = .69). In addition, the composite measures of de ep acting (refocus and reappraisal) and surface acting (faking and suppression) also demonstrated acceptable alphas ( = .76 and .81, respectively). Emotional Intelligence Scale (Wong & Law, 2002; App endix B) This scale measures individual differences in the ability to i dentify and regulate emotions in the self and others. The scale consists of 16 items in a si x-point Likert format where one
33 corresponds to strongly disagree and six corresponds to strongly agree High average scores should correspond to high levels of emotiona l intelligence. A sample item would be I am sensitive to the feelings and emotions of others. The internal consistency reliability for this scale was 0.87. This measure of emotional intelligence demonstrates good convergence with two measures of emotional int elligence, the Trait Meta-Mood Scale (Salovey, Mayer, Goldman, Turvey & Palfai, 19 95) and the EQ-I (Bar On, 1997). When correlated with the Big Five personality dimen sions this scale demonstrated smaller correlations in comparison to the EQ-I, thu s demonstrating its discriminant validity. Also in support of its discriminant vali dity, this measure had minimal correlations with a measure of IQ by Eysenck (1990) In contrast to the Trait Meta-Mood Scale, this measure was able to explain incremental variance in predicting life satisfaction above the Big Five dimensions. Positive Affectivity Negative Affectivity Scale (PA NAS: Watson, Clark & Tellegen, 1988; Appendix C) The PANAS measures both positive and negative af fectivity using a five-point Likert format that ranges from very slightly or not at all to extremely Higher scores on positive or negative affectivity correspo nd to higher levels of positive and negative traits, respectively. For each of the 20 items, participants are asked to choose a response that best indicates how they feel on avera ge. The items consisted of ten emotion words for each type of affectivity, for instance, p ositive affectivity items include interested and excited while negative affectivity items include distressed and upset Watson et al. (1988) report acceptable internal con sistency reliabilities for both the positive and negative affectivity scales ( = 0.88, = 0.87) that are almost identical to the alphas ( = 0.87, = 0.87) obtained for these measures in this study.
34 Emotional Exhaustion (Maslach & Jackson, 1986; Appe ndix D) Nine items comprise the emotional exhaustion subscale of the M aslach Burnout Inventory. The measure assesses how often respondents report feeli ng the symptoms of emotional exhaustion at work. A sample item is, I feel emot ionally drained at work. The scale employs a seven-point Likert format that ranges fro m never to every day Higher mean scores on this measure suggest high levels of emoti onal exhaustion ( = 0.93). Job Satisfaction Subscale of Michigan Organization al Assessment Questionnaire (Cammann, Fichman, Jenkins & Klesh, 1979; Appendix E) This measure consists of three items that assess overall job satisfaction an d demonstrated an acceptable alpha in this study ( = 0.89). A six-point Likert response scale is use d where one corresponds to strongly disagree and six corresponds to strongly agree A higher mean score indicates overall satisfaction with the job. A sample item i s, All in all, Im satisfied with my job. Service Performance Measures. (McLellan, Schmit, Amundson & Blake, 1998 as modified by Grandey, 2003; Williams & Anderson, 199 1; Appendices F & G). Two established scales were utilized in this study in o rder to broadly cover the construct of service performance. The affective delivery measur e was adapted by Grandey (2003) from a secret shopper service rating measure deve loped by McLellan et al. (1998). Six items capture positive affective delivery by requir ing supervisors to address qualities such as service employees display of friendliness and warmth during interactions with customers. A sample item would be, This person tr eats customers with courtesy, respect and politeness. This measure utilizes a five-poin t Likert format (1 = strongly disagree to 5 = strongly agree ) and displays acceptable internal consistency reli ability ( = .87).
35 Grandey (2003) also reports a satisfactory alpha fo r the self-report version of this measure of affective delivery ( = .88). The task performance measure consists of seven ite ms that assess the extent to which an employee exhibits prescribed task performa nce. A seven-point Likert response scale is used where one corresponds to strongly disagree and seven corresponds to strongly agree Supervisors responded with respect to their empl oyees general service performance, for example My staff member adequatel y completes assigned tasks. Rank (2006) reports an adequate alpha for this meas ure ( = .91). The internal consistency reliability for the task performance me asure was acceptable ( = .80). As with the emotional labor strategy measures, the aff ective delivery and task performance measures were utilized separately for the correlati onal analyses, but combined to form a composite service performance measure for examinati on of the hypothesized model. The composite service performance measure also demonstr ated a satisfactory alpha ( = .86). Demographic Information (Appendix H). Five items were included to assess the gender, ethnicity, age, job tenure and type of serv ice job of respondents. The tenure, job type and age items were open-ended, while responden ts chose either male or female for gender and Asian, Black, Hispanic, White or Other f or ethnicity.
36 Procedure Participants were recruited during undergraduate c lasses and through the Psychology department research participant pool. I n order to ensure an adequate number of participants, undergraduate students were also r ecruited to distribute the survey packets to full-time customer service employees. T he survey packets contained separate envelopes for the employee and the supervisor, insi de of each was a letter that described the study and instructions on how to complete the s urveys. Employees and supervisors returned the surveys via postal mail in postage-pai d business reply envelopes. The surveys were coded in order to match the employee a nd supervisor surveys. Participants in some classes received credit toward their course s for each completed employee survey and for each completed supervisor survey returned. In order to award credit each student was associated with the code on the survey packet, so that credit could be assigned to the appropriate student once those surveys were returne d. Anonymity was maintained as the researcher was unable to identify the employee or s upervisor to whom the student gave the packet, only that the surveys associated with t he student were returned. Most participants received a pen as a small gift in exch ange their time. Analyses Structural equation modeling (SEM) was used to tes t the proposed model with LISREL 8.5 being used to obtain the covariance matr ices necessary to test the model and maximum likelihood (ML) estimation was employed for all procedures. The structural equation modeling provided path coefficients to ass ess the relationships posited in the
37 model. The covariances were not estimated for the error terms. Correlation analyses were used to test the majority of hypotheses, while moderated regression analyses were conducted for the hypotheses that involved gender a s a moderator of the relationship between the emotional labor strategies and the outc ome variables.
38 Results Table 1 summarizes the results for each of the stud y hypotheses. The means, standard deviations, observed and possible ranges, as well as the Cronbachs alphas for each scale variable are presented in Table 2. All of the scales demonstrated internal consistency reliability from = 0.69 to 0.93, where an alpha of 0.70 is the mini mum considered acceptable (Nunnally & Bernstein, 1994). Table 3 displays the zero order correlations among all study variables. Both forms of deep acting (refocus and reappraisal) and both forms of surface acting (faki ng and suppression) displayed opposing relationships with all variables thereby s upporting the majority of the study hypotheses involving correlation analyses. Four of the hypotheses dealt with the relationship between the emotional labor strategies and the proposed antecedents, emotional intelligence, positive and negative affectivity and gender. Hypothesis 1a was fully su pported as emotional intelligence was positively related to both the refocus and reapprai sal forms of deep acting, indicating that individuals with higher perceived emotional intelli gence were more likely to refocus and reappraise in order to obey display rules than thos e with lower perceived emotional intelligence. Hypothesis 1b received partial suppo rt as the negative relationship between emotional intelligence and surface acting was signi ficant for faking but not for suppression. Thus, individuals with lower perceive d EI reported faking their emotions more often than those with higher perceived emotion al intelligence. As predicted in Hypothesis 2, positive affectivity displayed opposi ng relationships with the emotional
39 labor strategies such that employees predisposed to experience positive affect were significantly more likely to refocus and reappraise in order to obey positive display rules than to fake or suppress their emotions. Conversel y, employees high in negative affectivity were significantly more likely to utili ze surface acting (faking and suppression) to adhere to positive display rules th an to deep act (refocus). The reappraisal form of deep acting was not significant ly related to negative affectivity, so Hypothesis 3 was only partially supported. Hypothe sis 4 proposed that choice of emotional labor strategy differed by gender, and in deed females were significantly more likely to report engaging in deep acting (refocus a nd reappraisal) than males with the opposite being true for surface acting (faking and suppression) where males are more likely to report faking their emotions than women i n order to positive obey display rules. The final three hypotheses dealt with the relations hips between the emotional labor strategies and the three proposed outcomes of emotional exhaustion, job satisfaction and customer service performance. Hypotheses 5a an d 5b were supported as higher levels of deep acting (refocus and reappraisal) wer e associated with lower levels of emotional exhaustion, while higher levels of surfac e acting (faking and suppression) corresponded with higher levels of emotional exhaus tion. Job satisfaction was positively related to both forms of deep acting (refocus and r eappraisal), but negatively related to both forms of surface acting (faking and suppressio n) providing support for Hypotheses 6a and 6b. Hypotheses 7a and 7b were partially sup ported as affective delivery was significantly related to deep and surface acting; h owever task performance was only significantly related to deep acting. Specifically affective delivery was positively related
40 to refocus and reappraisal and negatively related t o faking and suppression; while task performance was positively related to refocus and r eappraisal. With respect to the moderated regression analyses, both of the hypotheses were fully supported for the faking aspect of surface ac ting. For Hypothesis 5c, emotional exhaustion was regressed on faking, the proposed mo derator, gender and the interaction between faking and gender. As shown in Table 4, bo th the R2 and the b weight for the interaction were significant ( b = -.55, p < .05). For women, as faking increased there was a sharper increase in emotional exhaustion than for men (Figure 2). Similarly, at high levels of faking, females reported lower job satisf action than males at the same level of faking, and the converse occurred at low levels of faking, thereby supporting Hypothesis 6c (see Figure 3). In the interest of parsimony, composite measures of the emotional labor strategies and service performance were utilized to examine th e overall hypothesized model. As expected, the two measures of deep acting, reapprai sal and refocus were positively related ( r = .30, p < .05), as were the two measures of surface acting faking and suppression ( r = .24, p < .05). These composite measures of deep acting a nd surface acting also exhibited acceptable internal consisten cy reliabilities ( = .76; = .81). Similarly, the measures of affective delivery and t ask performance were also positively related ( r = .36, p < .05) and the alpha for the composite service per formance measure was satisfactory ( = .81). Table 5 presents overall goodness of fit measures s uch as chi-square, root mean square error of approximation (RMSEA), expected cro ss-validation index (ECVI), normed fit index (NFI), Tucker-Lewis Index (TLI) an d the comparative fit index (CFI)
41 for the hypothesized model. The chi-square test ex amines the null hypothesis that the proposed model holds exactly in the population, and there was a significant chi-square which indicates that the proposed model is not comp letely plausible. The RMSEA examines the error of approximation in the populati on and addresses the question of how well the proposed model with optimally chosen param eter values would fit the population covariance matrix. MacCallum, Browne and Sugawara (1996) suggest that RMSEA values between 0.05 and 0.08 indicate good to reaso nable fit, and values between 0.08 and 0.10 represent mediocre fit, while those values above 0.10 indicate poor fit. The RMSEA value for the hypothesized model is 0.068, th erefore this model demonstrates good to reasonable fit. The ECVI assesses the likelihood that the model wil l cross-validate across similarly sized samples from the same population (B rowne & Cudeck, 1989). There is no predetermined range of appropriate values for the E CVI, but the lower the values the better the potential for replication. The hypothes ized model has a somewhat low value for the ECVI, thereby suggesting that this model ma y have a chance for replication. The NFI, TLI, and CFI are measures of practical fit and values greater than 0.90 are considered to indicate acceptable fit. For the hyp othesized model, the NFI value approached acceptable fit, while the TLI and CFI va lues can be considered acceptable. Power was estimated by utilizing the sample size an d degrees of freedom for each proposed model. For the hypothesized model the deg rees of freedom were 480, so with the sample size of 198, the power estimate approxim ated 0.90 ( df = 100, N = 200). In addition to the overall goodness of fit measures for the hypothesized structural model, specific parameter estimates for most of the paths denoted by the study
42 hypotheses were obtained during the SEM analyses. Specifically, the relationships between gender and the emotional labor strategies w ere not included in the structural model as gender is a measured variable and structur al models are only comprised of latent variables. The hypothesized structural mode l is comprised of paths from the antecedents (emotional intelligence, positive affec tivity and negative affectivity) to the emotional labor strategies (deep acting and surface acting), which then have paths to the consequences (emotional exhaustion, job satisfactio n, and service performance). Of the twelve paths in the structural model, eight of them were significant, which further bolsters the support for this hypothesized model of the antecedents and consequences of the emotional labor strategies (see Figure 4). Contrary to correlation results for Hypothesis 1, e motional intelligence did not have significant paths to deep acting or surface ac ting. However, the positive path between positive affectivity and deep acting ( = .43) further corroborates the significant findings of Hypothesis 2a. On the other hand, posi tive affectivity did not significantly relate to surface acting thereby contradicting the findings of Hypothesis 2b. In support of Hypothesis 3, negative affectivity did significantl y relate to deep acting and surface acting in the expected directions ( = -.23; = .27). In terms of the outcomes, Hypothesis 5a and 5b were also supported by the SEM findings, as deep acting negatively related to emotional exhaustion ( = -.60) while surface acting positively related to emotional exhaustion ( = .35). Job satisfaction was also differentially related to the emotional labor strategies, such that it was positively related to deep acting ( = .73) and negatively related to surface acting ( = -.19) further supporting Hypotheses 6a and 6b. Finally, only deep acting was significantl y related to service performance ( =
43 .35) thereby lending further support to Hypothesis 7a, but not to Hypothesis 7b. For completeness, Figure 5 displays the indicators for the measurement as well as for the structural model.
44 Discussion The purpose of this study was to investigate antece dents and consequences of the emotional labor strategies of deep and surface acti ng in a sample of customer service employees. Specifically, the individual difference variables of emotional intelligence, affectivity and gender were examined as antecedents of the emotional labor strategies, while emotional exhaustion, job satisfaction and cu stomer service performance were examined as outcomes of the emotional labor process Indeed the majority of proposed hypotheses received support clearly demonstrating t he differential relationships of the emotional labor strategies of deep acting and surfa ce acting with the antecedents and consequences of interest in this study. This is th e first study to assess the different forms of deep acting and surface acting and the findings suggest that a finer-grained analysis of these emotional labor strategies may be warranted i n future research. Antecedents All of the proposed antecedents were related to the emotional labor strategies in the expected directions. First, the finding that e motional intelligence was differentially related to the emotional labor strategies supports and extends the research of Ct (2005). That is, individuals high in perceived emotional in telligence were more likely to report employing deep acting (refocus and reappraisal) tha n surface acting (faking) when engaging in emotional labor. Given the differentia l relationships of the emotional labor strategies with individual well-being and performan ce, the choice of deep acting as the more effective strategy indicates that emotional in telligence may be a vital characteristic
45 that enables an individual to appropriately match t he strategy to the situation (Feldman Barrett & Gross, 2001). One part of emotional inte lligence is knowledge of emotions, and the choice of deep acting over surface acting i s advantageous as deep acting is linked with improved employee well-being and service perfo rmance. That emotional intelligence was not significantly related to the s uppression form of surface acting is interesting, as it suggests that individuals lower in perceived emotional intelligence, may find it easier to comply with the portion of the or ganizational display rule that encourages suppression of negative emotion than with the produ ction of positive expressions via refocus, reappraisal or emotive faking. Positive affectivity was positively related to deep acting which suggests that high positive-affect individuals are more likely to atte mpt to feel the requisite positive emotions dictated by organizational display rules w hen they experience negative emotions at work. This preference for deep acting over surface acting is likely because these individuals are generally predisposed to expe rience positive emotions, so on the occasions when they experience negative emotions th at conflict with display rules, they are more likely to try to change their feelings via deep acting than to provide fake expressions through surface acting. Overall, integ rative organizational display rules mandate the expression of positive emotion and supp ression of negative emotion, so they are likely to be favorable to individuals high in p ositive affectivity because this approximates their natural state of being, that is in positive mood states. These findings make a unique contribution to the emotional labor l iterature because few studies have examined the relationship between positive affectiv ity and the emotional labor strategies despite their proximity with the same nomological n etwork.
46 On the other hand, individuals high in negative aff ectivity were more likely to surface act in order to obey integrative organizati onal display rules, because these rules call for expressions that conflict with their gener al negative mood states. In this case, these individuals are more prone to negative moods, so surface acting which only modifies the expression and not the feeling would b e chosen more frequently than deep acting which modifies the feeling as well as the co rresponding expression. An interesting contribution of this study is the re lationship between gender and the emotional labor strategies which corresponds wi th the research of Johnson and Spector (in press). That is, women are more likely than men to report engaging in deep acting, while men are more likely to report engagin g in surface acting than women. The correspondence of integrative organizational displa y rules with societal display rules that encourage women to display more warmth and positive emotions than men may contribute to womens preference for deep acting (B em, 1974; Rafaeli, 1989). Thus, when faced with negative emotion during a service e ncounter, women are more likely to refocus or reappraise in order to produce the authe ntic positive emotion required by the organizational display rules. As such, emotive fak ing and suppression appear to be more detrimental for women than for men because the disc ord created by surface acting magnifies their lack of adherence to both sets of d isplay rules (societal and organizational). Gender As A Moderator Johnson and Spector (in press) found that gender se rved as a moderator of the relationship between surface acting and a number of personal outcomes. This studys results are consistent with their findings, such th at gender moderated the relationships
47 between surface acting and emotional exhaustion and job satisfaction. That is, women who reported faking their emotions reported more em otional exhaustion and less job satisfaction than men who reported faking their emo tions. Faking emotions perpetuates emotional dissonance, which for women contributes t o inauthenticity in two sets of roles, organizational and societal; thereby this is likely a more taxing strategy for women than for men. This heightened role inauthenticity likel y contributes to the increased emotional exhaustion and reduced job satisfaction reported by women in this study. As a practical implication, female service employees who surface a ct may experience more negative outcomes, and therefore should be encouraged to avo id surface acting where possible. Consequences The emotional labor strategies were related to all the proposed consequences in the hypothesized directions. The finding that both forms of surface acting (faking and suppression) were positively related to emotional e xhaustion is consistent with previous research (Brotheridge & Grandey, 2002; Johnson & Sp ector, in press). However, this is the first study to find significant negative relati onships between the forms of deep acting (refocus and reappraisal) and emotional exhaustion perhaps suggesting that the process of deep acting, which brings the employees emotions i n line with the positive display rules, may actually contribute to reduced emotional exhaus tion as employees are now experiencing positive emotions and thereby less lik ely to be emotionally exhausted. An implication of this finding would be that employers should provide training in deep acting as it appears to alleviate some of the negat ive consequences of emotional dissonance. For job satisfaction, there was a nega tive relationship with surface acting and a positive relationship with deep acting, such that employees who reported faking or
48 suppressing their emotions reported lower levels of job satisfaction than those who reported attempting to feel the requisite emotions through refocus or reappraisal. The surface acting-job satisfaction relationship corres ponds to recent meta-analytic findings by Bono and Vey (2005), and this is the first known study to establish the deep acting-job satisfaction relationship. Employees who genuinely try to experience the posi tive emotion that they are expected to display to customers ultimately provide better service performance (affective delivery and task performance), as rated by their s upervisors. Conversely, surface actors were rated poorly on affective delivery by their su pervisors. These findings are similar to those of Grandey (2003) who found that deep acting was positively and surface acting was negatively related to affective delivery, as ra ted by coworkers. In essence, the genuine nature of the emotions expressed by deep ac tors should contribute to a better service experience than the faked and suppressed em otions of surface actors. Hypothesized Model The overall hypothesized model has acceptable fit w hich provides added support for the study hypotheses. That is, the good fit of the hypothesized model lends further support to the linkages proposed between the antece dents and consequences of the emotional labor strategies. In particular, eight o f twelve paths in the structural model were significant; however, the non-significant path s in the model deserve further scrutiny. While the paths from emotional intellige nce to the emotional labor strategies were non-significant, they were in the expected dir ections. Similarly, as expected, the non-significant path between positive affectivity a nd surface acting was positive. However, given the existence of the emotional intel ligence and affectivity constructs
49 within the general nomological network of affect an d emotions, it is possible that multicollinearity served to attenuate the path coef ficients between these predictors and the emotional labor strategies. Finally, the non-s ignificant path between surface acting and service performance may be due to the weak corr elations that exist between these variables. In fact, neither form of surface acting (faking and suppression) was significantly related to task performance, which ma y indicate that supervisors do not tie an employees fake expressions of positive emotion to the more general tasks captured in the task performance measure. Limitations and Conclusions The cross-sectional nature of this study does serve as a limitation in that causality can not be inferred about the relationships in the model as the data were collected at one point in time. Future research should adopt a long itudinal design where the antecedents, emotional labor strategies and outcome variables ar e assessed at various points in time so that inferences can be made about the causal nature of these relationships. Alternatively, experience sampling methodology, in which employees respond to questions about their feelings, expressions and emotion regulation severa l times throughout the workday, holds promise for the future of emotional labor research (Beal, Trougakos, Weiss & Green, 2006). Another potential limitation of this study is the c oncern that self-report methodology will lead to the overstatement of relat ionships between variables, however this is somewhat offset by the supervisor reports o f service performance. It must be noted that the use of self-report measures may be a ppropriate in this instance because this study seeks to assess individual behaviors, attitud es and perceived outcomes.
50 Nevertheless, future research in this area could im prove on the methods used here by including some physiological measures of distress, perhaps during the performance of emotional labor. In fact, the combination of the e xperience sampling methodology with physiological measures of distress would make for a very interesting examination of the emotion regulation process. The participants for this sample were employed at a variety of service organizations, so the diversity of sources should c ontribute to the generalizability of these findings. Given that organizational display rules may differ across organizations, collecting data from a number of organizations faci litates the examination of emotional labor processes across varied organizational contex ts. Overall, this study provides a valuable contribution to the literature on emotions in the workplace, and in particular it serves to clarify some of the antecedents and conse quences of emotional labor strategies. These results also have practical implications for service organizations. Given that all service employees will experience emotional dissona nce at some point, it is important to recognize that surface acting has less favorable ou tcomes than deep acting for both the individual and the organization; therefore deep act ing should be encouraged where possible. The findings on the antecedents suggests that servi ce organizations should look for employees who are high in positive affectivity, low in negative affectivity and high in perceived emotional intelligence, as they are more likely to employ deep acting, which is the emotional labor strategy that related favorably to the outcomes of emotional exhaustion, job satisfaction and service performanc e. In conclusion, this study provides useful information to organizations in the service industry, as well as to researchers
51 because the negative consequences associated with p erformance of emotional labor can have immense personal and organizational costs. Un derstanding the emotional labor process and how it can result in negative consequen ces for employees is the first step in attempting to ameliorate the sometimes negative asp ects of service work and reduce the related personal and organizational costs.
52 References Abraham, R. (1998). Emotional dissonance in organiz ations: Antecedents, consequences, and moderators. Genetic, Social, and General Psychology Monographs, 124, 229-246. Adelmann, P.K. (1995). Emotional labor as a potent ial source of job stress. In S. L. Sauter & L. R. Murphy (Eds.), Organizational risk factors for job stress (pp. 371381). Washington, DC: American Psychological Assoc iation. Ashforth, B.E., & Humphrey, R.H. (1993). Emotional labor in service roles: The influence of identity. Academy of Management Review, 18, 88-115. Ashkanasy, N. M., & Daus, C.S. (2005). Rumors of t he death of emotional intelligence in organizational behavior are vastly exaggerated. Journal of Organizational Behavior, 26 441-452. Barchard, K. A., & Hakstian, A. R. (2004). The natu re and measurement of emotional intelligence abilities: Basic dimensions and their relationships with other cognitive ability and personality variables. Educational and Psychological Measurement, 64, 437-462. BarOn, R. (1997). BarOn EQ-i technical manual. Toronto, Canada: Psychological Assessment Resources. Beal, D. J., Trougakos, J. P., Weiss, H. M. & Green S. G. (2006). Episodic processes in emotional labor: Perceptions of affective delivery and regulation strategies. Journal of Applied Psychology, 91(5) 1053-1065.
53 Bem, S.L. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42 155-162. Bono, J. E., & Vey, M. A. (2005). Toward understand ing emotional management at work: A quantitative review of emotional labor rese arch. In N. Ashkanasy and C. Haertel (Eds.), Understanding Emotions in Organizational Behavior (pp. 213233). Mahwah, NJ: Erlbaum. Borucki, C.C. & Burke, M.J. (1999). An examination of service-related antecedents to retail store performance. Journal of Organizational Behavior, 20 943-962. Brackett, M., & Mayer, J.D. (2003). Convergent, di scriminant, and incremental validity of competing measures of emotional intelligence. Personality and Social Psychology Bulletin, 29 1147-1158. Brotheridge, C.M., & Grandey, A.A. (2002). Emotion al labor and burnout: Comparing two perspectives of people work. Journal of Vocational Behavior, 60 17-39. Brotheridge, C.M., & Lee, R.T. (2002). Testing a c onservation of resources model of the dynamics of emotional labor. Journal of Occupational Health Psychology, 7(1) 57-67. Brotheridge, C.M., & Lee, R.T. (2003). Development and validation of the Emotional Labour Scale. Journal of Occupational and Organizational Psycholo gy, 76(3) 365-379. Bulan, H. F., Erickson, R. J. & Wharton, A. S. (199 7). Doing for others on the job: The affective requirements of service work, gender, and emotional well-being. Social Problems, 44(2) 235-256.
54 Bureau of Labor Statistics (2005). Current Populat ion Survey. Retrieved January 9, 2006, from http://www.bls.gov/cps/cpsaat9.pdf Butler, E.A., Egloff, B., Wilhelm, F.H., Smith, N.C ., Erickson, E.A., & Gross, J.J. (2003). The social consequences of expressive supp ression. Emotion, 3(1) 4867. Cammann, C., Fichman, M., Jenkins, D., & Klesh, J. (1979). The Michigan Organizational Assessment Questionnaire Unpublished manuscript, University of Michigan, Ann Arbor. Campbell, J.P., McCloy, R.A., Oppler, S.H. & Sager, C.E. (1993). A theory of performance. In N. Schmitt & W.C. Borman (Eds.), Personnel selection in organizations : 35-70. San Francisco: Jossey-Bass. Carroll, J.B. (1993). Human cognitive abilities: A survey of factor-analy tic studies New York: Cambridge University Press. Conte, J.M., & Dean, M.A. (2006). Can emotional in telligence be measured? In K. Murphy (Ed.) A Critique of Emotional Intelligence: What are the Problems and How Can They be Fixed ? Mahwah, NJ: Erlbaum. Ct, S. (2005, April). Do emotionally intelligent people manage their emotions wisely? In Robin Gosserand and James Diefendorff (Chairs), Toward a Better Understanding of Emotion Regulation at Work Symposium conducted at the 20th annual conference of the Society for Industrial and Organizational Psychology, Los Angeles, CA. Cropanzano, R., Weiss, H.M., & Elias, S.M. (2004). The impact of display rules and emotional labor on psychological well-being at work In P.L. Perrew and D.C.
55 Ganster (Eds.) Emotional and physiological processes and positive intervention strategies Research in Occupational Stress and We ll-Being (3), 45-89. Darwin, D. (1965). The expression of emotions in man and animals Chicago: University of Chicago Press. (Original work publish ed 1872). Ekman, P. & Friesen, W.V. (1969). Non-verbal leaka ge and clues to deception. Psychiatry, 32 88-106. Ekman, P. (1973). Cross culture studies of facial expressions. In P. Ekman (Ed.) Darwin and facial expression: A century of research in rev iew 169-222. New York: Academic Press. Eysenck, H. J. (1990). Check your own IQ (2nd ed.). Harmondsworth, England; Penguin. Feldman Barrett, L.F., & Gross, J.J. (2001). Emoti onal intelligence: A process model of emotion representation and regulation. In T.J. May ne and G.A. Bonanno (Eds.), Emotions: Current issues and future directions (pp. 286 310). New York: Guilford Press. Florin, I., Freudenberg, G., & Hollaender, J. (1985 ). Facial expressions of emotion and physiologic reactions in children with bronchial as thmas. Psychosomatic Medicine, 47 382-393. Gardner, W. L., & Martinko, M. J. (1988). Impressi on management in organizations. Journal of Management, 14, 321-338. Gosserand, R.H., & Diefendorff, J.M. (2005). Emoti onal display rules and emotional labor: The moderating role of commitment. Journal of Applied Psychology, 90(6) 1256-1264.
56 Grandey, A. A. (2000). Emotion regulation in the w orkplace: A new way to conceptualize emotional labor. Journal of Occupational Health Psychology, 5(1), 95-110. Grandey, A. A. (2003). When the show must go on: Surface acting and deep acting as determinants of emotional exhaustion and peer-rated service delivery. Academy of Management Journal, 46(1) 86-96. Grandey, A.A. & Brauburger, A.L. (2002). The emoti onal regulation behind the customer service smile. In R.G. Lord, R. Klimoski & R. Kanfer (Eds.), Emotions in the workplace (pp. 260-294). San Francisco: Jossey-Bass. Grandey, A.A., Dickter, D.N. & Sin, H.-P. (2004). T he customer is not always right: Customer aggression and emotion regulation of servi ce employees. Journal of Organizational Behavior, 25 397-418. Grandey, A.A., Fisk, G.M. & Steiner, D.D. (2005). M ust service with a smile be stressful? The moderating role of personal control for American and French employees. Journal of Applied Psychology, 90 (5) 893-904. Greer, S., & Watson, M. (1985). Towards a psychobi ological model of cancer: Psychological considerations. Social Science and Medicine, 20 773-777. Gross, J. (1989). Emotion expression in cancer ons et and progression. Social Science and Medicine, 28, 1239-1248. Gross, J. (1998a). Antecedentand response-focuse d emotion regulation: Divergent consequences for experience, expression, and physio logy. Journal of Personality and Social Psychology, 74(1) 224-237.
57 Gross, J. (1998b). The emerging field of emotion r egulation: An integrative review. Review of General Psychology, 2(3), 271-299. Gross, J.J., & John, O.P. (1998). Mapping the doma in of emotional expressivity: Multimethod evidence for a hierarchical model. Journal of Personality and Social Psychology, 74, 170-191. Gross, J.J., & John, O.P. (2002). Wise emotion reg ulation. In L.F. Barrett and P. Salovey (Eds.), The wisdom in feeling: Psychological processes in e motional intelligence (pp.297-318). New York: Guilford Press. Gross, J.J., & John, O.P. (2003). Individual diffe rences in two emotion regulation processes: Implications for affect, relationships, and well-being. Journal of Personality and Social Psychology, 85(2) 348-362. Gross, J.J., Sutton, S.K., & Ketelaar, T. (1998). Relations between affect and personality: Support for the affect-level and affective-reactivi ty views. Personality and Social Psychology Bulletin, 24 279-288. Grove, S. J., & Fisk, R. P. (1989). Impression man agement in services marketing: A dramaturgical perspective. In R. A. Giacalone & P. Rosenfeld (Eds.), Impression management in the organization 427-438. Hillsdale, NJ: Erlbaum. Guyton, A.C., & Hall, J.E. (1997). Human physiology and mechanisms of disease (6th ed.). Philadelphia: W.B. Saunders Company. Hall, E.J. (1993). Smiling, deferring, and flirtin g: Doing gender by giving good service. Work and Occupations, 20 452-471. Hall, J.A., & Halberstadt, A.G. (1981). Sex roles and nonverbal communication skills. Sex Roles, 7 273-287.
58 Hatfield, E., Cacioppo, J.T., & Rapson, R.L. (1992) Primitive emotional contagion. In M.S. Clark (Ed.), Emotion and social behavior 151-177. Thousand Oaks, CA: Sage Publications. Hochschild, A. R. (1979). Emotion work, feeling ru les, and social structure. American Journal of Sociology, 85, 551-575. Hochschild, A. R. (1983). The managed heart: Commercialization of human feeli ng Berkeley: University of California Press. Johnson, H. M. (2004). The story behind service wit h a smile: The effects of emotional labor on job satisfaction, emotional exhaustion, an d affective well-being. Unpublished masters thesis, University of South Fl orida, Tampa, Florida. Johnson, H. M. & Spector, P. E. (in press). Service with a smile: Do emotional intelligence, gender and autonomy moderate the emot ional labor process? Journal of Occupational Health Psychology Jordan, P.J., Ashkanasy, N.M., Hrtel, C.E.J., & Ho oper, G.S. (2002). Workgroup emotional intelligence: Scale development and relat ionship to team process effectiveness and goal focus. Human Resource Management Review, 12 195214. Keltner, D., & Kring, A. (1998). Emotion, social f unction, and psychopathology. Review of General Psychology, 2 320-342. Kruml, S., & Geddes, D. (1998, August). Exploring the dimensions of emotional labor: The heart of Hochschilds work Paper presented at the First Conference of Emotions in Organizational Life, San Diego, CA.
59 Kruml, S., & Geddes, D. (2000). Catching fire with out burning out: Is there an ideal way to perform emotional labor? In N. M. Ashkanasy, C. E. J. Haertel, & W. J. Zerbe. (Eds.) Emotions in the workplace: Research, theory and pra ctice (pp. 177-188). Westport, CT: Quorum Books. LaFrance, M., & Banaji, M. (1992). Toward a recons ideration of the gender-emotion relationship. In M. Clarke (Ed.), Review of personality and social psychology (pp. 178-2001). Beverly Hills, CA: Sage. Landy, F.J. (2005). Some historical and scientific issues related to research on emotional intelligence. Journal of Organizational Behavior, 26 411-424. Law, K.S., Wong, C.S., & Mobley, W.H. (1998). Towa rd a taxonomy of multidimensional constructs. Academy of Management Review, 23(4) 741-755. Liao, H. & Chuang, A. (2004). A multilevel investi gation of factors influencing employee service performance and customer outcomes. Academy of Management Journal, 47 41-58. MacCallum, R.C., Browne, M.W. & Sugawara, H.M. (199 6). Power analysis and determination of sample size for covariance structu re modeling. Psychological Methods, 1(2) 130-149. Maslach, C. (1982). Burnout: The cost of caring Englewood Cliffs, NJ: Prentice Hall. Maslach, C., & Jackson, S. (1986). Maslach burnout inventory manual (2nd ed.). Palo Alto, CA: Consulting Psychologists Press. Maslach, C., Jackson, S. & Leiter, M. (1996). Maslach burnout inventory manual (3rd ed.). Palo Alto, CA: Consulting Psychologists Press
60 Mayer, J.D., & Salovey, P. (1997). What is emotion al intelligence? In P. Salovey & D. J. Sluyter (Eds.), Emotional development and emotional intelligence (pp. 3-31). New York: Basic Books. Mayer, J.D., Caruso, D.R., & Salovey, P. (1999). E motional intelligence meets traditional standards for an intelligence. Intelligence, 27(4) 267-298. Mayer, J.D., Salovey, P., Caruso, D.R., & Sitarenio s, G. (2001). Emotional intelligence as a standard intelligence. Emotion, 1 232-242. McLellan, R.A., Schmit, M.J., Amundson, M., & Blake R. (1998, April). Secret shopper ratings as an individual-level criterion for valida tion studies. Paper presented at the 13th Annual Conference of the Society for Industrial an d Organizational Psychology, Dallas, TX. Morris, J. A., & Feldman, D. C. (1996). The dimens ions, antecedents, and consequences of emotional labor. Academy of Management Review, 21(4), 986-1010. Morris, J. A., & Feldman, D. C. (1997). Managing e motions in the workplace. Journal of Managerial Issues, 9(3) 257-274. Palfai, T.P. & Salovey, P. (1993). The influence o f depressed and elated mood on deductive and inductive reasoning. Imagination, Cognition, and Personality, 13 272-280. Pugh, S.D. (2002). Emotional regulation in individ uals and dyads: Causes, costs, and consequences. In R.G. Lord, R. Klimoski & R. Kanfe r (Eds.), Emotions in the workplace (pp. 147-182). San Francisco: Jossey-Bass. Rafaeli, A. (1989). When clerks meet customers: A test of variables related to emotional expression on the job. Journal of Applied Psychology, 74(3) 385-393.
61 Rafaeli, A., & Sutton, R. I. (1987). Expression of emotion as part of the work role. Academy of Management Review, 12, 23-37. Rank, J. (2006). Leadership predictors of proactiv e organizational behavior: Facilitating personal initiative, voice behavior, and exceptiona l service performance. Unpublished doctoral dissertation, University of So uth Florida, Tampa. Richards, J.M., & Gross, J.J. (1999). Composure at any cost? The cognitive consequences of emotion suppression. Personality and Social Psychology Bulletin, 25 1033-1044. Salovey, P. & Mayer, J. D. (1990). Emotional intel ligence. Imagination, Cognition, and Personality, 9 : 185-211. Salovey, P., Mayer, J. D., Goldman, S. L., Turvey, C., & Palfai, T. (1995). Emotional attention, clarity and repair: exploring emotional intelligence using the Trait Meta-Mood Scale. In J. W. Pennebaker (Ed.), Emotion, disclosure and health Washington, DC: American Psychological Association. Saxton, M. J., Phillips, J. S., & Blakeney, R. N. ( 1991). Antecedents and consequences of emotional exhaustion in the airline reservations service sector. Human Relations, 44, 583-602. Sheldon, K. M., Ryan, R. M., Rawsthorne, L. J. & Il ardi, B. (1997). Trait self and true self: Cross-role variation in the big-five personal ity traits and its relations with psychological authenticity and subjective well-bein g. Journal of Personality and Social Psychology, 73(6), 1380-1393. Spector, P.E., & Johnson, H.-A., M. (2006). Improving the definition, measurement and application of emotional intelligence. In K. Murphy (Ed.) A Critique of Emotional
62 Intelligence: What are the Problems and How Can The y be Fixed? Mahwah, NJ: Erlbaum. Totterdell, P. & Holman, D. (2003). Emotion regulat ion in customer service roles: Testing a model of emotional labor. Journal of Occupational Health Psychology, 8(1) 55-73. Van Maanen, J., & Kunda, G. (1989). Real feelings : Emotional expression and organizational culture. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior, 11, 43-103. Greenwich, CT: JAI Press. Watson, D., Clark, L. A., & Tellegen, A. (1988). D evelopment and validation of brief measures of positive and negative affect: The PANA S scales. Journal of Personality and Social Psychology, 54, 1064-1070. Watson, D., & Tellegen, A. (1985). Toward a consen sual structure of mood. Psychological Bulletin, 98(2) 219-235. Weiss, H., & Cropanzano, R. (1996). Affective even ts theory: A theoretical discussion of the structure, causes, and consequences of affectiv e experiences at work. Research in Organizational Behavior, 18, 1-74. Wharton, A. S. (1993). The affective consequences of service work. Work and Occupations, 20, 205-232. Wharton, A. S., & Erickson, R. J. (1993). Managing emotions on the job and at home: Understanding the consequences of multiple emotiona l roles. Academy of Management Review, 18, 457-486. Wolcott-Burnam, S. B. (2004). Examining emotional labor from an interactionist perspective: The impact of work conditions on the r elationship between emotional
63 labor and outcomes. (Doctoral dissertation, Centra l Michigan University, 2004). Dissertation Abstracts International, 65(5-B) 2681. Wong, C., & Law, K. S. (2002). The effects of lead er and follower emotional intelligence on performance and attitude: An explor atory study. Leadership Quarterly, 13(3) 243-274. Yerkes, R.M. (1921). Psychological examining in th e United States. Memoirs of the National Academy of Sciences (15:890 pp. entire). Zajonc, R. B. (1985). Emotion and facial efference : An ignored theory reclaimed. Science, 228, 15-21.
64 Table 1 Summary of Results for Study Hypotheses Hypothesis Result H1a: Emotional intelligence will be positively rela ted to deep acting (refocus and reappraisal). Supported H1b: Emotional intelligence will be negatively rela ted to surface acting (faking and suppression). Partially supported H2a: Positive affectivity will be positively relate d to deep acting (refocus and reappraisal). Supported H2b: Positive affectivity will be negatively relate d to surface acting (faking and suppression). Supported H3a: Negative affectivity will be negatively relate d to deep acting (refocus and reappraisal). Partially supported H3b: Negative affectivity will be positively relate d to surface acting (faking and suppression). Supported H4a: Females will be more likely than males to enga ge in deep acting (refocus and reappraisal). Supported H4b: Males will be more likely than females to enga ge in surface acting (faking and suppression). Supported H5a: Deep acting (refocus and reappraisal) will be negatively related to emotional exhaustion. Supported
65 Table 1 (continued) Summary of Results for Study Hypotheses Hypothesis Result H5b: Surface acting (faking and suppression) will b e positively related to emotional exhaustion. Supported H5c: Gender moderates the relationships between sur face acting (faking and suppression) and emotional exhaustion. Females who surface act will experience more emotional exhaustion than males. Supported H6a: Deep acting (refocus and reappraisal) will be positively related to job satisfaction. Supported H6b: Surface acting (faking and suppression) will b e negatively related to job satisfaction. Supported H6c: Gender moderates the relationships between sur face acting (faking and suppression) and job satisfaction. Females who surface act will experience lower job satisfaction than males. Supported H7a: Deep acting (refocus and reappraisal) will be positively related to service performance (affective delivery and task pe rformance). Supported H7b: Surface acting (faking and suppression) will b e negatively related to service performance (affective delivery and task pe rformance). Partially supported
66 Table 2 Descriptive Statistics and Cronbachs Alpha for Sca le Variables Scale N Mean SD Alpha Range (Observed) Range (Possible) Refocus 277 3.71 0.78 0.88 1.00 5.00 1 5 Reappraisal 277 4.99 0.90 0.74 1.00 7.00 1 7 Deep acting 277 8.70 1.36 0.76 2.00 12.00 2 12 Faking (SA) 277 3.20 0.82 0.90 1.00 5.00 1 5 Suppression (SU) 275 3.35 1.11 0.69 1.00 6.50 1 7 Surface acting 275 6.56 1.53 0.81 2.57 11.11 2 12 Positive affectivity 279 3.57 0.67 0.87 1.10 5.00 1 5 Negative affectivity 277 1.82 0.63 0.87 1.00 4.20 1 5 Emotional intelligence 277 4.75 0.55 0.87 3.19 6. 00 1 6 Emotional exhaustion 278 2.31 1.45 0.93 0.00 5.78 0 6 Job satisfaction 279 4.51 1.08 0.89 1.00 6.00 1 6 Affective delivery 222 4.58 0.46 0.87 2.83 5.00 1 5 Task performance 221 6.45 0.59 0.80 3.57 7.00 1 7 Service performance 220 11.03 0.89 0.80 7.05 12.0 0 2 12
67 Table 3 Intercorrelations Between Study Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1. Refocus 2. Reappraisal .30** 3. Deep acting .78** .84** 4. Faking -.17** .00 -.10 5. Suppression -.27** -.12* -.23** .24** 6. Surface acting -.29** -.09 -.22** .71** .85** 7. Emotional intelligence .34** .20** .33** -.23** -.09 -.19** 8. Positive affectivity .33** .14** .29** -.25** -.18** -.27** .49** 9. Negative affectivity -.13** -.09 -.13** .33** .18** .31** -.34** -.15** 10. Emotional exhaustion -.22** -.11* -.20** .46** .18** .37** -.27** -.35** .37** 11. Job satisfaction .28** .19** .28** -.35** -.27** -.38** .31** .41** -.27** -.61** 12. Affective delivery .12* .17** .18** -.12* -.14* -.16** .13* .08 -.08 -.19** .40** 13. Task performance .12* .12* .15** -.02 -.07 -.06 .08 .08 -.06 -.14** .21** .36** 14. Service performance .14* .16** .19** -.08 -.12* -.13* .12* .09 -.08 -.19** .22** .81** .89** 15. Gender -.20** -.14** -.20** .14** .21** .23** -.04 .03 -.02 .01 -.01 -.15** -.10* -.10 p < .10, ** p < .05
68 Table 4 Moderated Regression Analyses Emotional Exhaustion Job Satisfaction Independent variable R2 b R2 b H5c .23** H6c .14** Gender 1.65** -1.45** Faking 1.49** -1.02** Gender Faking -.55** .46** H5c .03** H6c .08** Gender .46 -.30 Suppression .43* -.40** Gender Suppression -.15 .11 p < .10, ** p < .05
69 Table 5 Summary of Fit Statistics for Hypothesized Model Model 2 df RMSEA ECVI NFI TLI CFI Hypothesized 916.88 480 0.068 5.48 0.84 0.90 0.91
70 Figure 1 Overall hypothesized model of relationships betw een antecedents, emotional labor strategies and out comes. Emotional intelligence Positive affectivity Gender Emotional exhaustion Job satisfaction Service Performance Deep acting Negative affectivity Surface acting H1a (+) H1b () H2a (+) H2b () H3a () H3b (+) H4a () H4b (+) H5a () H5b (+) H6a (+) H6b () H7a (+) H7b () H5c H6c
71 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Low Faking High Faking FakingEmotional exhaustion Female Male Figure 2. Graph of the moderating effect of gender on the rel ationship between faking and emotional exhaustion.
72 0 1 2 3 4 5 6 Low Faking High Faking FakingJob satisfaction Female Male Figure 3. Graph of the moderating effect of gender on the rel ationship between faking and job satisfaction.
73 Figure 4 Hypothesized structural model of antecedents, em otional labor strategies and outcomes with path coefficients. p < .05 Emotional intelligence Positive affectivity Emotional exhaustion Job satisfaction Service Performance Deep acting Negative affectivity Surface acting .10 .10 .43* .16 .23* .2 7* .60* .35* .73* .19* .35* .01
74 Figure 5 Hypothesized structural and measurement model of antecedents, emotional labor strategies and outcom es with path coefficients. p < .05 Emotional intelligence Positive affectivity Emotional exhaustion Job satisfaction Deep acting Negative affectivity Surface acting .10 .10 .43* .16 .23* .27* .60* .35* .73* .19* .35* .01 .21* DA1 DA2 DA3 .32* .10 SA1 .86* .86* SA2 SA3 SA4 .46* .23* EX1 EX2 EX3 1.32* 1.37* 1.35* JS1 .97* JS2 1.03 JS3 SP1 SP2 SP4 SP3 .39* .40* .42* .32* EI1 EI2 EI3 EI4 EI6 .45* .27* .49* .60* .30* NA1 NA2 NA3 NA4 PA1 PA2 PA3 PA4 .57* .72* .57* .57* .63* .62* .64* .62* .60* SP5 EI5 Service performance .36* .92 .65 .84* .43* 3 8 .87
76 Appendix A Deep Acting and Surface Acting Measures (Brotheridge & Lee (2003); Grandey (2003); Gross & John (2003)) On an average day at work, how frequently do you do each of the following when interacting with customers? Never Rarely Sometimes Often Always 1 Make an effort to actually feel the emotions that I need to display to others. (D) 1 2 3 4 5 2 Try to actually experience the emotions that I must show. (D) 1 2 3 4 5 3 Really try to feel the emotions I have to show as p art of my job. (D) 1 2 3 4 5 4 Resist expressing my true feelings. (S) 1 2 3 4 5 5 Hide my true feelings about a situation. (S) 1 2 3 4 5 6 Put on an act in order to deal with customers in an appropriate way. (S) 1 2 3 4 5 7 Fake a good mood when interacting with customers. ( S) 1 2 3 4 5 8 Put on a show or performance when interacting w ith customers. (S) 1 2 3 4 5 9 Just pretend to have the emotions I need to display for my job. (S) 1 2 3 4 5 10 Put on a mask in order to display the emotions I need for the job. (S) 1 2 3 4 5 We would like to ask you some questions about your emotional life, in particular, how you control (tha t is, regulate and manage) your emotions. Although some of the following questions may seem similar to one ano ther, they differ in important ways. Strongly disagree Disagree Slightly disagree Neutral Slightly agree Agree Strongly agree 1 When I want to feel more positive emotion (such as joy/amusement), I change what Im thinking about (D) 1 2 3 4 5 6 7 2 When I want to feel more negative emotion (such as sadness/anger), I change what Im thinking about (D) 1 2 3 4 5 6 7 3 When Im faced with a stressful situation, I make m yself think about it in a way that helps me stay calm. (D) 1 2 3 4 5 6 7
77 4 When I want to feel more positive emotion, I change the way Im thinking about the situation. (D) 1 2 3 4 5 6 7 5 I control my emotions by changing the way I think about the situation Im in. (D) 1 2 3 4 5 6 7 6 When I want to feel more negative emotion, I change the way Im thinking about the situation. (D) 1 2 3 4 5 6 7 7 When I am feeling positive emotions, I am careful not to express them. (S) 1 2 3 4 5 6 7 8 I control my emotions by not expressing them (S) 1 2 3 4 5 6 7 9 When I am feeling negative emotions, I make sure not to express them. (S) 1 2 3 4 5 6 7 10 I keep my emotions to myself. (S) 1 2 3 4 5 6 7
78 Appendix B Wong and Law Emotional Intelligence Scale (Wong & Law, 2002) Please circle the one number for each question that comes closest to reflecting your opinion About it. Strongly disagree Disagree Slightly disagree Slightly agree Agree Strongly agree 1 I have a good sense of why I have certain feelings most of the time 1 2 3 4 5 6 2 I have good understanding of my own emotions 1 2 3 4 5 6 3 I really understand what I feel 1 2 3 4 5 6 4 I always know whether or not I am happy 1 2 3 4 5 6 5 I always know my friends emotions from their behav ior 1 2 3 4 5 6 6 I am a good observer of others emotions 1 2 3 4 5 6 7 I am sensitive to the feelings and emotions of othe rs 1 2 3 4 5 6 8 I have good understanding of the emotions of people around me 1 2 3 4 5 6 9 I always set goals for myself and then try my best to achieve them 1 2 3 4 5 6 10 I always tell myself I am a competent person 1 2 3 4 5 6 11 I am a self-motivated person 1 2 3 4 5 6 12 I would always encourage myself to try my best 1 2 3 4 5 6 13 I am able to control my temper and handle difficult ies rationally 1 2 3 4 5 6 14 I am quite capable of controlling my own emotions 1 2 3 4 5 6 15 I can always calm down quickly when I am very angry 1 2 3 4 5 6 16 I have good control of my own emotions 1 2 3 4 5 6
79 Appendix C PANAS (Watson, Clark & Tellegen, 1988) This scale consists of a number of words that descr ibe different feelings and emotions. Please check one response for each item that best indicates how you feel on average. Very slightly or not at all A little Moderately Quite a bit Extremely 1 Interested (P) 1 2 3 4 5 2 Distressed (N) 1 2 3 4 5 3 Excited (P) 1 2 3 4 5 4 Upset (N) 1 2 3 4 5 5 Strong (P) 1 2 3 4 5 6 Guilty (N) 1 2 3 4 5 7 Scared (N) 1 2 3 4 5 8 Hostile (N) 1 2 3 4 5 9 Enthusiastic (P) 1 2 3 4 5 10 Proud (P) 1 2 3 4 5 11 Irritable (N) 1 2 3 4 5 12 Alert (P) 1 2 3 4 5 13 Ashamed (N) 1 2 3 4 5 14 Inspired (P) 1 2 3 4 5 15 Nervous (N) 1 2 3 4 5 16 Determined (P) 1 2 3 4 5 17 Attentive (P) 1 2 3 4 5 18 Jittery (N) 1 2 3 4 5 19 Active (P) 1 2 3 4 5 20 Afraid (N) 1 2 3 4 5
80 Appendix D Emotional Exhaustion (Maslach & Jackson, 1986) Please circle the one number that indicates how often you experience each of the following. Never A few times a year or less Once a month or less A few times a month Once a week A few times a week Every day 1 I feel emotionally drained at work. 0 1 2 3 4 5 6 2 I feel used up at the end of the day. 0 1 2 3 4 5 6 3 I feel fatigued when I get up in the morning and ha ve to face another day on the job. 0 1 2 3 4 5 6 4 Working with people is really a strain on me. 0 1 2 3 4 5 6 5 I feel burned out from my work. 0 1 2 3 4 5 6 6 I feel frustrated on my job. 0 1 2 3 4 5 6 7 I feel I am working too hard on my job. 0 1 2 3 4 5 6 8 Working with people directly puts too much stress o n me. 0 1 2 3 4 5 6 9 I feel like I am at the end of my rope. 0 1 2 3 4 5 6
81 Appendix E Job Satisfaction from the Michigan Organizational A ssessment Questionnaire (Cammann, Fichman, Jenkins & Klesh, 1979) Please circle the one number for each question that comes closest to reflecting your opinion about it. Strongly disagree Disagree Slightly disagree Slightly agree Agree Strongly agree 1 In general, I do not like my job. (R) 1 2 3 4 5 6 2 All in all, I am satisfied with my job. 1 2 3 4 5 6 3 In general, I like working here. 1 2 3 4 5 6
82 Appendix F Affective Delivery Measure Supervisor (McLellan, Schmit, Amundson & Blake, 1998 as modifi ed by Grandey, 2003) We are trying to get a supervisors perspective on employeecustomer interactions. Please consider the custome r interactions of the person who gave you this form, and answer each part below as truthfully as possible. Never Rarely Sometimes Often Always 1 This person seems sincere when dealing with the pub lic. 1 2 3 4 5 2 Customers seem to like interacting with this person 1 2 3 4 5 3 This person shows friendliness and warmth to most c ustomers 1 2 3 4 5 4 This person treats customers with courtesy, respect and politeness 1 2 3 4 5 5 This person smiles and communicates expressively wi th customers 1 2 3 4 5 6 This person shows enthusiasm when dealing with cust omers. 1 2 3 4 5 Affective Delivery Measure Employee We are trying to get an employees perspective on c ustomer interactions. Please consider your customer intera ctions, and answer each part below as truthfully as possibl e. Never Rarely Sometimes Often Always 1 I am sincere when dealing with the public. 1 2 3 4 5 2 Customers seem to like interacting with me. 1 2 3 4 5 3 I show friendliness and warmth to most customers. 1 2 3 4 5 4 I treat customers with courtesy, respect, and polit eness. 1 2 3 4 5 5 I smile and communicate expressively with customers 1 2 3 4 5 6 I show enthusiasm when dealing with customers. 1 2 3 4 5
83 Appendix G Task Performance Measure Supervisor (Williams & Anderson, 1991) Using the choices below, please rate the person who gave you this form. For each item, please circle the number that expresses your agreement or disagreement best. Concerning his/her customer service performance, my staff member Strongly disagree Disagree moderately Disagree slightly Neither agree nor disagree Agree slightly Agree moderately Strongly agree 1 Adequately completes assigned tasks. 1 2 3 4 5 6 7 2 Fulfills responsibilities specified in his/her job description. 1 2 3 4 5 6 7 3 Performs tasks that are expected of him/her. 1 2 3 4 5 6 7 4 Meets formal performance requirements of the job. 1 2 3 4 5 6 7 5 Engages in activities that will directly affect his /her performance evaluation. 1 2 3 4 5 6 7 6 Neglects aspects of the job he/she is obligated to perform. 1 2 3 4 5 6 7 7 Fails to perform essential duties. 1 2 3 4 5 6 7
84 Appendix G (continued) Task Performance Measure Employee (Williams & Anderson, 1991) For each item, please circle the number that expresses your agreement or disagreement best. Concerning my customer service performance, I Strongly disagree Disagree moderately Disagree slightly Neither agree nor disagree Agree slightly Agree moderately Strongly agree 1 Adequately complete assigned tasks. 1 2 3 4 5 6 7 2 Fulfill responsibilities specified in my job description. 1 2 3 4 5 6 7 3 Perform tasks that are expected of me. 1 2 3 4 5 6 7 4 Meet formal performance requirements of the job. 1 2 3 4 5 6 7 5 Engage in activities that will directly affect my performance evaluation. 1 2 3 4 5 6 7 6 Neglect aspects of the job I am obligated to perform. 1 2 3 4 5 6 7 7 Fail to perform essential duties. 1 2 3 4 5 6 7
85 Appendix H Demographic Information __________ Female __________ Male _____ Asian _____ Black _____ Hispanic _____ White _____ Other How long have you worked for this company (in month s)? ____________ Age in years: ___________ Indicate your type of service job: ________________ _____________________
86 Appendix I Employee Letter Dear Participant; Thank you so much for agreeing to participate in th is study! Your assistance with this project is greatly appreciated and will be extremel y valuable! My name is Hazel-Anne M. Johnson and I am a graduate student in the Ph.D. program in Industrial and Organizational Psychology at the University of Sout h Florida. For my dissertation research project I am surveying individuals who hav e service-related jobs, like yours, that require the management of emotions as a part of the job. Please be candid when you complete the survey the re are no right or wrong answers. You are free to participate in this study or to withdraw at any time. Your participation or withdrawal does not have any assoc iated risks. Your privacy and research records will be kept conf idential to the extent of the law. Authorized research personnel, employees of the Dep artment of Health and Human Services, and the USF Institutional Review Board, i ts staff, and any other individuals acting on behalf of USF, may inspect the records fr om this research project. The results of this study may be published. However, the data obtained from you will be combined with data from others in the publication. The publ ished results will not include your name or any other information that would personally identify you in any way. If you have questions about your rights as a person who is taking part in a research study, you may contact the Division of Research Integrity and Compliance of the University of South Florida at (813) 974-5638. USF Psychology Students : If you are completing this survey for extra credi t your name will be temporarily linked to the survey number on a separate sheet in order to award the points. This sheet will be kept confidential and o nce the extra credit points have been awarded this sheet will be destroyed. Why should you fill out this survey? o You are helping to collect data for my research pro ject! o While you will not directly benefit from participat ing, you will help to advance scientific knowledge and understanding about people in customer service positions similar to yours! Instructions There are two surveys in this packet the Employee Survey and the Supervisor Survey. These surveys are numbered in order to match your s urvey to your supervisors survey. I do not ask for your name so this survey will not be directly linked to you. This survey should take no more than ten minutes to complete, w hile the supervisor survey should take no more than five minutes to complete.
87 All you need to do is complete this short survey an d ask your supervisor to complete the other survey based on his/her observations, experie nces and conversations with you on your present job. Please do not discuss these questions with your sup ervisor before you both have completed the surveys Once you complete the survey, please return it in t he enclosed postage-paid business reply envelope within two weeks. If you have any q uestions, concerns, or would like a summary of the studys results, please contact me a t firstname.lastname@example.org Thanks in advance for your help, I greatly apprecia te it! Sincerely, Hazel-Anne M. Johnson, M.A. Psychology Department, PCD 4118G University of South Florida
88 Appendix J Supervisor Letter Dear Participant; Thank you so much for agreeing to participate in th is study! Your assistance with this project is greatly appreciated and will be extremel y valuable! My name is Hazel-Anne M. Johnson and I am a graduat e student in the Ph.D. program in Industrial and Organizational Psychology at the Uni versity of South Florida. You can contact me at email@example.com or 813.974.2492. This survey is a part of my dissertation research project on individuals in cus tomer service jobs. One of your employees has agreed to participate in this research study, and has authorized you to answer questions about his/her jo b behaviors in this survey. Please complete this Supervisor Survey with regards to your employee who is participating in this study. Answer the questions based on your obs ervations, experiences, and conversations with this employee on his/her present job. I do not ask for your name, so the information you provide will be completely anonymous. Your privacy and research records will be kept conf idential to the extent of the law. Authorized research personnel, employees of the Department of Health and Human Services, and the USF Institutional Review Bo ard, its staff, and any other individuals acting on behalf of USF, may inspect th e records from this research project. The results of this study may be published. Howeve r, the data obtained from you will be combined with data from others in the publication. The published results will not include your name or any other information that would perso nally identify you in any way. Please be candid when you complete the survey the re are no right or wrong answers. You are free to participate in this study or to withdraw at any time. Your participation or withdrawal does not have any assoc iated risks. If you have questions about your rights as a person who is taking part in a research study, you may contact the Division of Research Integrity and Compliance of th e University of South Florida at (813) 974-5638. Why should you fill out this survey? o You are helping to collect data for my research pro ject! o While you will not directly benefit from participat ing, you will help to advance scientific knowledge and understanding about people in the customer service industry! Instructions This survey should take no more than five minutes t o complete. All you need to do is complete this short survey based on your observatio ns, experiences and conversations
89 with the employee who gave you this survey. Please do not discuss these questions with your employee before you both have completed t he surveys Once you complete the survey, please return it in t he enclosed postage-paid business reply envelope within two weeks. If you have any q uestions, concerns, or would like a summary of the studys results, please contact me a t firstname.lastname@example.org Thanks in advance for your help, I greatly apprecia te it! Sincerely, Hazel-Anne M. Johnson, M.A. Department of Psychology, PCD 4118G University of South Florida
About the Author Hazel-Anne Michelle Johnson was born in Barbados, w here she attended secondary school at Harrison College, and began her undergrad uate education at the Cave Hill campus of the University of the West Indies. In Ma y 2001, she graduated from the University of Florida with a Bachelor of Science de gree in Psychology with a minor in Business Administration ( Honors ). In August 2001, Hazel-Anne began the Industrial Organizational doctoral program at the University o f South Florida, earning the Master of Arts degree in Industrial-Organizational Psychology in May 2004. Hazel-Anne has taught several undergraduate courses at the Univers ity of South Florida, for which she has received a number of teaching awards. Hazel-An nes research interests involve emotions in the workplace, as well as workplace men toring and diversity. Her masters thesis research has been published in the Journal o f Occupational Health Psychology. She is now eagerly anticipating the start of a prod uctive and rewarding career as a university professor!