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 001469386
007 cr mnu|||uuuuu
008 040524s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000287
Johnson, Hazel-Anne M.
The story behind service with a smile
h [electronic resource] :
the effects of emotional labor on job satisfaction, emotional exhaustion, and affective well-being /
by Hazel-Anne M. Johnson.
[Tampa, Fla.] :
University of South Florida,
Thesis (M.A.)--University of South Florida, 2004.
Includes bibliographical references.
Text (Electronic thesis) 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 82 pages.
ABSTRACT: The present study examines the process of emotional labor as performed by customer service employees. This research investigates some of the consequences of performing emotional labor such as emotional exhaustion, affective well-being, and job satisfaction, and attempts to determine which individual and organizational variables play moderating roles in these relationships. One hundred and seventy-six participants from 10 customer service organizations, ranging from retail stores to call centers, completed a 126-item survey. Correlation and regression analyses were conducted to test the proposed hypotheses. Results indicate that gender, emotional intelligence and autonomy are key moderator variables in the relationship between emotional labor and emotional exhaustion, affective well-being, and job satisfaction.
Adviser: Spector, Paul
service employee outcomes.
emotions at work.
t USF Electronic Theses and Dissertations.
The Story Behind Service With A Smile: The Effects of Emotional Labor on Job Satisfaction, Emotional Exhaustion, and Affective Well-Being by Hazel-Anne M. Johnson A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Psychology College of Arts and Sciences University of South Florida Major Professor: Paul Spector, Ph.D. Kristen Salomon, Ph.D. Ellis Gesten, Ph.D. Date of Approval: April 1st, 2004 Keywords: emotional dissonance, customer service, emotions at work, occupational health, service employee outcomes Copyright 2004, Hazel-Anne M. Johnson
i Table of Contents List of Tables iii List of Figures iv Abstract v Introduction 1 Emotional Labor 4 A Conceptual Model of Emotional Labor 7 Consequences of Emotional Labor11 Individual Factors 13 Organizational Factors 19 Method 23 Participants 23 Measures 24 Emotional Labor 24 Job Satisfaction 25 Emotional Exhaustion 25 Emotional Intelligence 25 Affective Well-Being 26 Affectivity 27 Autonomy 27 Self-Monitoring 27 Supervisory Support 28 Coworker Support 28 Demographics 28 Procedure 29 Results 30 Discussion 37 References 46 Appendices 65 Appendix A: Emotional Labor Scale (Brotheridge & Lee, 1998) 66 Appendix B: Job Satisfaction from the Michigan Organizational Assessment Questionnaire (Cammann, Fichman, Jenkins & Klesh, 1979) 67
ii Appendix C: Emotional Exhaustion (Maslach & Jackson, 1986) 68 Appendix D: Emotional Intelligence (Wong & Law, 2002) 69 Appendix E: Job-related Affective Well-Being Scale (Van Katwyk, Fox, Spector & Kelloway, 1999) 70 Appendix F: PANAS (Watson, Clark & Tellegen, 1988) 71 Appendix G: Job Autonomy (JDS: Hackman & Oldham, 1975) 72 Appendix H: Self-Monitoring Scale (Snyder & Gangestad, 1986) 73 Appendix I: Supervisory Support (Greenhaus, Parasuraman & Wormley, 1990) 74 Appendix J: Coworker Support (Caplan, Cobb, French, Van Harrison & Pinneau, 1980) 75 Appendix K: Demographic Information 76
iii List of Tables Table 1 Number of Participants from each Organization 52 Table 2 Descriptive Statistics and CronbachÂ’s Alpha for Scale Variables 53 Table 3 Intercorrelations Between Study Variables 54 Table 4 One-way ANOVAs for Ethnicity 55 Table 5 Results of Moderated Regression Analyses 56
iv List of Figures Figure 1. Conceptual framework of emotion regulation performed in the work setting. 60 Figure 2. Graph of the proposed moderating effect of emotional intelligence on the relationship between emotional labor and affective well-being 61 Figure 3. A subset of the emotion regulation framework proposed by Grandey (2000). 62 Figure 4. Graph of the moderating effect of autonomy on the relationship between emotional labor and affective well-being. 63 Figure 5. Graph of the moderating effect of autonomy on the relationship between emotional labor and emotional exhaustion. 64
v The Story Behind Service With A Smile: The Effects of Emotional Labor on Job Satisfaction, Emotional Exhaustion, and Affective Well-Being Hazel-Anne M. Johnson ABSTRACT The present study examines the process of emotional labor as performed by customer service employees. This research investigates some of the consequences of performing emotional labor such as emotional exhaustion, affective well-being, and job satisfaction, and attempts to determine which individual and organizational variables play moderating roles in these relationships. One hundred and seventy-six participants from 10 customer service organizations, ranging from retail stores to call centers, completed a 126-item survey. Correlation and regression analyses were conducted to test the proposed hypotheses. Results indicate that gender, emotional intelligence, and autonomy are key moderator variables in the relationship between emotional labor and emotional exhaustion, affective well-being, and job satisfaction. Females are more likely to experience negative consequences when engaging in surface acting. Individuals high in emotional intelligence experienced positive outcom es as emotional labor increased, and the converse is true for those low in emotional intelligence. Autonomy serves to alleviate negative outcomes primarily at the higher levels of emotional labor. [158 words]
1 The Story Behind Service With A Smile: The Effects of Emotional Labor on Job Satisfaction, Emotional Exhaustion, and Affective Well-Being Emotional labor is the expression of organizationally desired emotions by service agents during service encounters (Hochschild, 1979, 1983; Ashforth & Humphrey, 1993). Morris and Feldman (1996) define emotional labor as Â“the effort, planning, and control needed to express organizationally desired emotions during interpersonal transactionsÂ” (p. 987). Grandey (2000) has defined emotional labor as Â“the process of regulating both feelings and expressions for organizational goalsÂ” (p. 97). Emotional labor has been regarded as a type of impression management, because it is a deliberate attempt by the individual to direct his or her behavior toward others in order to foster both certain social perceptions of himself or herself and a certain interpersonal climate (Gardner & Martinko, 1988; Grove & Fisk, 1989). Hochschild (1983) has defined emotional dissonance as the separation of felt emotion from emotion expressed to meet external expectations, and contends that it is harmful to the physical and psychological well being of employees. When an employee is required to express organizationally desired emotions that contradict genuinely felt emotions, emotional dissonance may be experienced. This is considered a form of individual-role conflict, such that an individualÂ’s response conflicts with role expectations regarding the display of emotions (Zapf, 2002).
2 This paper will examine the process of emotional labor as performed by customer service employees. In particular, this research investigates some of the hypothesized consequences of performing emotional labor such as emotional exhaustion, affective well-being, and job satisfaction, and attempts to determine which individual and organizational variables play moderating roles in these relationships. Rafaeli and Sutton (1987) suggest that displayed emotions can serve as control moves which, as defined by Goffman (1969) are an individualÂ’s strategic manipulation of emotional expressions designed to influence the behavior of others. Sutton and Rafaeli (1988) propose that displayed emotions will not only be seen as characteristic of the individual, but will be ascribed to the organization as well. Consequently, organizations generally require emotional labor to ensure the display of positive emotions because it is expected that regulated emotional expression will increase sales through the re inforcement provided to the customer in the form of positive socially desirable emotions. However, in the case of convenience stores, Sutton and Rafaeli (1988) found that where a premium is not placed on warm friendly service, sales volume does not increase when it is provided.In this situation, customers are more concerned with the speed of the transaction so a neutral demeanor actually served to help clerks influence the behavior of their customers and therefore provide faster service that led to increased sales. Therefore, the authors suggested that a warm emotional front may promote sales when customers expect that it should and will be a central part of the service provided by organizations such as at Disneyland (Van Maanen & Kunda, 1989). Hence, emotional labor requirements that are appropriate in one service
3 environment may be inappropriate and dysfunctional in another environment (Morris & Feldman, 1996). This is also seen in the case of bill collectors whose emotional labor environment is the direct opposite of a warm emotional front. It is this negative emotional front, if rendered successfully that will increase payments by debtors, and hence increase profitability of the bill collection agency.
4 It was in her 1983 book that Arlie Russell Hochschild first coined the term emotional labor to refer to Â“the management of feeling to create a publicly observable facial and bodily displayÂ” (p. 7). Hochschild (1983) examined the development of emotions into a marketplace commodity, and how it is incumbent upon the service employee to manage their emotions as a part of the job. Her chief contention was that emotion management was detrimental to service employees because their emotions were now mandated by the organization. Hochschild (1983) identified two methods that employees use to manage their emotions: Surface acting, which corresponds to managing observable expressions, and deep acting, which corresponds to managing feelings. She also introduced to the service context, the notion of display rules, which are shared norms about appropriate emotional expression (Ekman, 1973). It is adherence to these display rules that can result in negative consequences for service workers such as burnout or job stress (Hochschild, 1983). Morris and Feldman (1996) proposed that emotional labor is comprised of four dimensions: Attentiveness to display rules, frequency of emotional display, variety of emotions to be expressed, and emotional dissonance. Display rules are generally a function of societal, occupational, and organizational norms (Rafaeli & Sutton, 1989). The more attentiveness to display rules that is required, the more psychological energy and physical effort the service job will demand from employees. Within this dimension there are two sub-dimensions, and The longer the emotional displays the more likely they will become less scripted; consequently, longer emotional displays
5 require greater attention and emotional stamina (Hochschild, 1983). Cordes and Dougherty (1993) have shown that the longer the interaction, the more burnout the employee is likely to suffer. Conversely, research on convenience store clerks (Sutton & Rafaeli, 1988; Rafaeli, 1989) suggested that short interactions with customers involved highly scripted interaction formats, such as a simple thank you and maybe a smile. This implies that short interactions require less emotional effort. Typically, surface acting will not produce intense emotions, so deep acting is required because the employee must actively call to mind thoughts, images and memories that will aid in expressing the required emotion. Therefore, work roles requiring display of intense emotions entail more deep acting and thus greater effort on the part of the role occupants (Morris & Feldman, 1996). The of emotional display has been the most studied dimension of emotional labor, and still remains an important indicator because the more often an organization requires socially appropriate emotional displays, the greater the demand for emotional labor. Emotions displayed within organizations can be classified as positive, negative, or neutral (Wharton & Erickson, 1993). If employees are required to change their emotions frequently then this requires more active planning and monitoring of their behavior, hence more emotional labor. The wider the range of emotions to be expressed, the more emotional labor the employee will have to perform, such that of emotion is the third dimension. Morris and FeldmanÂ’s (1996) fourth dimension, as already described, is the conflict between genuinely felt emotions and organizationally prescribed emotions (Middleton, 1989). Emotional dissonance makes
6 emotional labor more difficult because when conflicts between genuinely felt emotions and organizationally desired emotions exist, greater control and management of behavior is necessary. For instance, this may occur when salespeople have to sell products to which they are not fully committed. Lack of commitment to a product may require considerable emotional effort to display the positive emotions necessary to effectively sell this product.
7 Grandey (2000) integrates previous models of emotional labor (Hochschild, 1983; Ashforth & Humphrey, 1993; Morris & Feldman, 1996) to provide a comprehensive theoretical model (see Figure 1). This model encom passes situational cues, the individual and organizational factors that affect the emotion regulation process and the long-term consequences of emotional labor. She proposes that the emotional labor processes of surface acting and deep acting correspond to the description of emotional labor as emotional regulation, and can serve as a means to operationalize emotional labor. Grandey (2000) provides three reasons for the operationalization of emotional labor as surface and deep acting. First, surface and deep acting can have both positive and negative outcomes, therefore researchers can explain negative outcomes such as burnout, as well as positive outcomes such as customer service and increased personal accomplishment. Next, if these two processes have differential outcomes, then organizational training and stress management programs can be modified accordingly. Lastly, conceptualizing emotional labor as surface and deep acting links this model of emotional labor to an established theoretical model of emotion regulation, which facilitates expansion of this research area. Grandey (2000) utilizes emotion regulation theory as a framework to guide emotional labor research. Emotion regulation involves Â“the processes by which individuals influence which emotions they have, when they have them and how they experience and express these emotionsÂ” (Gross, 1998b, p. 275). GrossÂ’s (1998b) model posits that emotion regulation is comprised of two processes, where the first process is
8 in which an individual regulates the situation or appraisal that precedes emotion; this is analogous to deep acting. The second process, involves modification of the observable signs of emotion in a manner consistent with surface acting (Grandey, 2000). The method that employees choose to address emotional dissonance can have negative effects, for instance, surface acting may lead to feelings of misalignment and inauthenticity that can decrease an employeeÂ’s sense of well-being (Sheldon, Ryan, Rawsthorne, & Ilardi, 1997). Conversely, regulation through deep acting in a Â“good faithÂ” type of emotional labor may result in a sense of accomplishment depending on the employeeÂ’s level of identification with the organization (Ashforth & Humphrey, 1993). Emotion research has shown that the inhibition of negative emotions over time can be associated with a variety of physical illnesses, such as high blood pressure and cancer (Gross, 1989; King & Emmons, 1990; Smith, 1992). Therefore, while deep acting and surface acting enable an employee to successfully achieve organizational goals, they may also contribute to detrimental effects to that employeeÂ’s health and psychological well-being. However, it must be noted that the expression of positive emotions 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. GrandeyÂ’s (2000) model proposes two major situational antecedents for emotional labor, customer interaction expectations, and emotional events. Customer interaction expectations can be subsumed under the frequency and duration of interactions, the variety of emotional expressions required, and display rules of the organization.
9 Emotional events influence the amount of emotional labor that an employee must perform, because if an event results in an emotional response that is contrary to the organizationally prescribed emotion then that employee has to engage in emotional labor to perform effectively. Two of the long-term consequences of emotional labor identified by Grandey (2000), burnout and job satisfaction deal with individual well-being. Burnout is a stress outcome that is comprised of three dimensions, emotional exhaustion, depersonalization, and reduced personal accomplishment. Burnout will occur if an employee is emotionally invested in interactions with customers and has little recourse to recuperate from the drain on emotional resources (Jackson, Schwab, & Schuler, 1986). Job satisfaction provides an estimation of how an employee feels about his or her job. In general, research has shown that customer service employees with a high level of emotion regulation tend to be less satisfied with their jobs, but there has been some research that may contradict this finding. For instance, Adelmann (1995) found that wait staff, who have high levels of emotion regulation, and expressed genuine emotions at work were more satisfied than those who displayed fake emotions. It is possible that some of these emotions were produced through deep acting, and hence a product of emotional labor. However, this is one of the few studies that have found a positive relationship between emotional labor and job satisfaction. Typically, the research shows that there is a negative relationship between emotional labor and job satisfaction for employees who engage in surface acting and consequently experience emotional dissonance (Abraham, 1998; Morris & Feldman, 1997). Perhaps it is the manner in which employees engage in
10 emotional labor (deep acting versus surface acting) that influences their level of job satisfaction. Customer service performance is perhaps the most desired outcome of emotional labor. Emotion management, when it serves to induce the appropriate feelings in customers should result in good customer service performance (Ashforth & Humphrey, 1993). Typically, positive emotional expressions lead to better customer service performance. However, insincere emotional expressions, if perceived as such by the customer, will negatively impact customer service (Rafaeli & Sutton, 1987). Employee withdrawal from customer interaction is one of the emotional labor outcomes least desired by organizations, because an employee who leaves a service encounter to cope with their emotions may negatively impact that customerÂ’s impression of the organization (Parasuraman, Zeithaml, & Berry, 1985). The level of emotion management required may ultimately cause an employee to leave the organization, either through absence or turnover, due to poor person-job fit. GrandeyÂ’s (2000) model also presents a number of personal and organizational factors that relate to emotional labor. She suggests five individual difference and personality variables that should be examined in relation to emotion management; gender, emotional expressivity, emotional intelligence, self-monitoring, and affectivity. Future research should examine organizational factors that may influence the level of emotional labor necessary; autonomy, and supervisor and coworker support are starting points proposed by Grandey (2000).
11 Emotional exhaustion is a specific stress-related reaction, and it is considered a key component of burnout (Maslach, 1982). Emotional exhaustion is the state of depleted energy caused by excessive emotional demands made on people interacting with customers or clients (Saxton, Phillips & Blakeney, 1991), and involves Â“feelings of being emotionally overextended and exhausted by oneÂ’s workÂ” (Maslach, Jackson & Leiter, 1996, p. 4). Research by Wharton (1993) has shown that although jobs requiring emotional labor do not place employees at greater risk of emotional exhaustion than other jobs, all else being equal, emotional labor does result in negative consequences under some circumstances. Emotional labor leads to increased emotional exhaustion among employees with low job autonomy, longer job tenure, and who work longer hours. Specifically, employees with low job autonomy are constrained by organizational display rules with little latitude to deviate from these rules. In the case where their feelings do not match the display rules they may engage in surface acting in order to display the appropriate emotions. Research by Kruml and Geddes (2000) supports this notion because they found that employees who engage in surface acting were more emotionally exhausted than those who adhered to display rules by deep acting. The duration of emotional labor, whether in job tenure or hours worked, requires either emotional dissonance (surface acting) or emotional effort (deep acting) both of which may lead to emotional exhaustion. Rafaeli and Sutton (1987) consider emotional dissonance to be a form of role conflict, because it involves a clash between the needs and principles of the employee and the requirements of others within the same role (Kahn, 1964). Research
12 suggests that a key antecedent of emotional exhaustion is role conflict (Jackson et al., 1986), consequently, having to engage in emotional labor that results in emotional dissonance may lead to higher levels of emotional exhaustion. #$% % "& #' %% "& ( Wharton (1993) posits that employees who find emotional labor jobs more satisfying were probably attracted to such jobs because they possess personal qualities especially suited to working with the public. Therefore, if organizations choose frontline service employees based on their interpersonal skills and individuals seek jobs compatible with their personality, the Â“fitÂ” betw een job demands and personal qualities may be high in these positions, thereby leading to increased job satisfaction (Diener, Larsen & Emmons, 1984). The research on the relationship between emotional labor and job satisfaction has found both positive (Adelmann, 1995; Wharton, 1993) and negative relationships (Abraham, 1998; Morris & Feldman, 1997). These findings may be explained by the method of emotional labor undertaken, for instance, surface acting may lead to feelings of inauthenticity and consequently job dissatisfaction. Conversely, if an employee engages in deep acting this may lead to feelings of personal accomplishment and by extension, job satisfaction (Kruml & Geddes, 2000). )$%%*& )' % %*&
13 + HochschildÂ’s initial (1983) work on emotional labor focused on female flight attendants and she noted that women significantly outnumber men in the service industry. Her initial concern was that, due to their numerical superiority in service work, the negative aspects of emotional labor were disproportionately affecting women. However, research has shown that women who perform emotional labor are significantly more satisfied than men who perform the same type of job (Wharton, 1993). This suggests that women may be socialized to handle the interpersonal demands of emotion management in service work, and this competency may lead them to have a more positive experience than their male counterparts. Perhaps this socialization may lead women to engage in deep acting to adhere to positive display rules, which should result in less emotional dissonance. In fact, Grandey (2000) suggests that men may need more training in emotion management in a service setting. However, research has demonstrated a relationship between gender and emotional dissonance, such that women reported more cases in which they felt differently than they expressed (Kruml & Geddes, 1998). It is possible that while more satisfied, women may have higher levels of stress or psychological ailments that are related to their successful suppression of feelings. The contradictory research on the effects of emotional labor on women may be explained by whether they engage in deep acting or surface acting. Deep acting may enable women to experience positive emotions, which may in turn result in increased affective well-being. However, surface acting while producing the appropriate expressive behavior, will lead to emotional dissonance which can lead to stress.
14 -,% % "&%. % "&. % "& -,% % "&..% "&%.. %"& + Emotional intelligence has been defined as Â“the ability to perceive, appraise, and express emotion accurately; the ability to access and generate feelings when they facilitate cognition; the ability to understand affect-laden information and make use of emotional knowledge; and the ability to regulate emotions to promote emotional and intellectual growth and well-beingÂ” (Salovey, Bedell, Detweiler, & Mayer, 2000, p. 506). Mayer and Salovey (1997) propose an emotional intelligence framework that is comprised of four dimensions, (1) the appraisal and expression of emotion in the self, (2) appraisal and recognition of emotion in others, (3) regulation of emotion in the self, and (4) use of emotion to facilitate performance. The first dimension concerns an individualÂ’s ability to recognize and express their emotions, individuals high on this dimension are more easily aware of their emotions than the average individual. Individuals high on the second dimension tend to be more sensitive to the feelings and emotions of others. The ability to regulate emotion in the self facilitates an expeditious recovery from psychological distress. Finally, the fourth dimension relates to an
15 individualÂ’s ability to utilize their emotions to enhance personal performance. Based on this theoretical framework, employees that are high in emotional intelligence should be able to effectively engage in emotion regulation to satisfy organizational display rules. Goleman (1995) proposes that individuals high in emotional intelligence are adept at social interaction, and perhaps may generate positive feelings in their interaction partner; which is an ideal characteristic in service interactions. Therefore, employees who are high in emotional intelligence should be able to engage in the processes of emotional labor with greater ease and effectiveness. /% % %!& .% %!% %&%. %!%%%.%! %%%0)1& /% % "& .%"% %&%. "%%. %"%%&
16 $! Self-monitoring refers to the extent to which people monitor, control and modify their expressive behavior to meet standards of social appropriateness (Snyder, 1974). Research has indicated that high self-monitors pay more attention to situational cues about which emotions are appropriate, and also are more skilled at presenting emotions (Snyder, 1974; Riggio & Friedman, 1982). Therefore, employees who are high self-monitors should be more likely to comply with organizational display norms because they are more willing to monitor expressive behavior. That is, high selfmonitors may be more likely to engage in surface acting than deep acting, because they are proficient at monitoring and controlling their expressive behavior. In addition, this inclination of high self-monitors to comply with organizational display norms may result in less dissatisfaction with the emotional labor par t of their jobs due to their ability to regulate their expressive behavior. In fact, they may be more satisfied with the emotional labor component of their job because it rewards them for behavior in which they normally engage. Conversely, low self-monitors may be more prone to emotional exhaustion than other workers who perform emotional labor, because their expressive behavior is guided more by their affective states rather than by desire to comply with social standards, therefore to obey display rules they may have to engage in more effortful deep acting (Wharton, 1993). Consequently, high self-monitors should have to expend less emotional effort to display the organizationally prescribed emotions via surface acting. In fact, recent research has shown that high self-monitors engaged in more surface acting than low self-monitors (Brotheridge & Lee, 2002). 2$!% %&
17 2$!% %& 2$!% % & 2$!% % *&!%* %&%!%%* %&%* %!%& 2$!% % "&!%% "%&%!% "%&% "%!%& Affective traits serve as predi spositions to particular emotional responses (Weiss & Cropanzano, 1996). Positive affectivity indicates the extent to which a person feels enthusiastic and optimistic, whereas negative affectivity corresponds to pessimism and aversive mood states (Watson & Tellegen, 1985; Grandey, 2000). Morris and Feldman (1996) contend that positive and negative affectivity will influence emotional dissonance. That is, if the organizationally prescribed emotions conflict with an employeeÂ’s affectivity (positive or negative), then emotional dissonance will occur, therefore, individuals may want to ensure that their emotion work requirements are congruent with their affective states. Brotheridge and Lee (1998) posit that affectivity may correspond to both the range and intensity of emotions displayed, and the use of
18 surface 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 acting, than low-affect intense individuals (Brotheridge & Lee, 1998). Therefore, an individual who is high in positive a ffectivity may not fit well in a job that required the expression of negative emotions, such as a bill collector. 34 % % "& %"% %&% ."% %.%" %%& 35% % "&%. %"% %&."% %.%" %%&
19 6 ( According to Hackman and Oldham (1975), autonomy describes the level of independence and discretion available to an employee in the completion of their jobs. Job autonomy indicates the extent to which a service employee can modify the display rules to fit their own personality and interpersonal styles (Morris & Feldman, 1996). Lack of autonomy about which emotions are displayed can be a source of stress for service employees. For instance, display rules that required flight attendants express positive emotions to rude or threatening passengers must have been difficult to comply with (Hochschild, 1983). Rafaeli and Sutton (1989) suggest that individuals with more job autonomy regarding expressive behavior will express emotions that match their affective states regardless of the organizational display rules. This may lead to higher job satisfaction for service workers who have the autonomy to show their true nasty feelings to obnoxious customers. Indeed, research has indicated that there is a positive relationship between autonomy and job satisfaction (Hackman & Oldham, 1975). In addition, Wharton (1993) demonstrated that even in jobs with a high level of emotional labor, autonomy serves to alleviate the negative effects of such emotional labor. Therefore, high levels of job autonomy should result in a reduction in emotional exhaustion due to surface acting. 7% % %!&%%! %% %&+%%%%%!
20 % %! %%& 7% "&%" %%%& +%%%" % .%" % %& $ %8 Schneider and Bowen (1985) suggest that supervisor and coworker support should create a positive working environment, which by extension should minimize the need to engage in emotional labor when the display rules are positive. That is, if an employee is in a positive mood due to the environment, then less emotional effort is needed to display positive organizationally prescribed emotions. Social support is thought to enable individuals to cope better with job stressors and to increase their sense of personal control (Cohen & Wills, 1985). On the other hand, Beehr (1995) presents evidence of reverse buffering, that is, high levels of social support may contribute to a positive relationship between job stressors and individual strains. Therefore reverse buffering suggests that social support may not act to buffer the effects of job stressors on individuals. These contrary findings may have resulted from treating the different forms of social support as one construct. What may have been needed was the linkage of the appropriate type of social support to the appropriate stressors and correspondent strain. Research by Hochschild (1983) indicated that strong social support
21 among flight attendants enabled them to vent frustrations about passengers without violating role requirements and display rules. In this instance, the emotional support provided by coworkers helped to lessen the strain associated with the stressor of an obnoxious passenger. Other research (Baruch-Feldman, Brondolo, Ben-Dayan & Schwartz, 2002) has found that supervisor support was related to job satisfaction in a sample of traffic enforcement agents. Abraham (1998) suggests that support from supervisors and coworkers can be a moderator in the relationship between emotional labor and psychological well-being. She found that with low support there was a negative relationship between emotional dissonance and job satisfaction, however with high support there was a slight increase in job satisfaction when emotional dissonance was high. Thus, supervisor and coworker support, acting as moderators, prevent emotional dissonance from reducing job satisfaction (Zapf, 2002). Further research is needed to examine the moderating effect of supervisor and coworker support on the consequences of emotional labor. 9%8 % %%!&% %8 .%%! %%&%% %8 %!%%% .%!%%%& 9%8 % %"&%
22 %8 .% "%%&%% %8 "%% "%%% %& $ The current study investigates individual and organizational factors that affect the emotion regulation process and some potential consequences of emotional labor in a sample of customer service workers. Using GrandeyÂ’s (2000) model (Figure 1) as a guide I will examine a subset of the framework of emotional labor (see Figure 3). Individual factors such as emotional intelligence, self-monitoring, affectivity and gender will be measured to estimate their effect on the emotional labor process and its consequences. Job autonomy, coworker and supervisor support are the organizational factors that will be measured to determine their effect on the emotional labor process.
23 Method 4 A sample of 176 individuals from 10 different customer service organizations participated in this study. Participant organizations are identified by letter and number of participants contributed to this study in Table 1. Organizations A, B, and H are three branches of a large department store chain, and Organization E is one branch of a large supermarket chain. Organization F is a large financial services company, and Organization G is a large bank. Small groups of participants were obtained from Organization C, a hotel, and from Organization I, a small restaurant. Organizations J, K and L were obtained from a large metropolitan public university that provided the largest sample of participants. Consequently, participant jobs ranged from administrative assistants to customer service representatives, and encompass a wide range of customer service positions. Participants were required to engage in a significant amount of customer interaction as a part of their job, so this sample should be representative of customer service employees across a number of different organizations. The sample was 74 percent female and had an overall mean age of 35, with a range from 15 to 76. Average tenure for this sample was approximately five years and ranged from one month to about 40 years. Approximately 67 percent of the sample was White, 11 percent Black, 10 percent Hispanic, 6 percent Asian, and 6 percent classified themselves as Other.
24 $0:;.#<<9= "1 This scale is comprised of subscales that measure the six dimensions of emotional labor. The duration of customer interaction is assessed with a single free response question, which asks respondents to identify the actual duration of an average customer interaction. The remaining dimensions are measured with on a five-point Likert response scale (1 = 5 = % ). Participants are asked to answer items in response to the stem question, Â“On an average day at work, how often do you do each of the following when interacting with customers?Â” Higher scores on each of the subscales represent higher levels of the dimension being assessed. The subscale for the frequency dimension contains three items that address the frequency of the display of organizationally prescribed emotions. The intensity subscale consists of two items that assess how often the employee expresses strong or intense emotions. The variety subscale measures the variety of emotional expression on the job and contains three items. The three items in the deep acting subscale assess how much an employee has to modify feelings to comply with display rules. The surface acting dimension consists of three items that measure the extent to which the employee has to express emotions that are not felt. Confirmatory factor analyses provide support for the six factor measurement model because the values obtained for the goodness-of-fit index, the adjusted goodness-of-fit index, and the Tucker-Lewis-Index were all above 0.90, which is considered to be an acceptable level of fit (Brotheridge and Lee, 1998). Brotheridge and Lee (2002) report good combined coefficient alpha for the role
25 characteristics (frequency, intensity and variety) subscales ( = 0.71), as well as for the deep acting and surface acting subscales ( = 0.89, =.86). A free response question was added that asked respondents to identify the average number of customers that they served per day. ($$6> 0..(8;?.#<7<= ":1 This measure consists of three items that assess overall job satisfaction. A six-point Likert response scale is used where one corresponds to and six corresponds to A higher score indicates overall satisfaction with the job. Grandey (2003) reports a high alpha of 0.93 for this subscale. "0;(8.#<93= "1 These nine items comprise the emotional exhaustion subscale of the Maslach Burnout Inventory. The measure assesses how often respondents report feeling the symptoms of emotional exhaustion at work. A sample item is Â“I feel emotionally drained at work.Â” Higher scores on this measure suggest high levels of emotional exhaustion. Brotheridge and Grandey (2002) report high internal consistency reliability for this subscale ( = 0.91). +$0@;%.)AA)= "'1 This scale measures individual differences in the ability to identify and regulate emotions in the self and others. The scale consists of 16 items in a six-point Likert format where one corresponds to and six corresponds to High scores should correspond to high levels of emotional intelligence. The internal consistency reliability for this scale, as reported by Wong and Law (2002), is quite good ( = 0.94).
26 This measure of emotional intelligence demonstrates good convergence with two measures of emotional intelligence, the Trait Meta-Mood Scale (Salovey, Mayer, Goldman, Turvey & Palafi, 1995) and the EQ-I (Bar On, 1997) ( = .63). When correlated with the Big Five personality dimensions this scale demonstrated smaller correlations in comparison to the EQ-I, thus demonstrating its discriminant validity. Also in support of it discriminant validity, this measure also 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. (!@ !:$0(@$B?%8.".$ ; ?%.)AAA= "1 These 20 items represent the four subscales of the JAWS that fall along two dimensions, pleasurableness and arousal: high pleasurable-high arousal, high pleasurable-low arousal, low pleasurable-high arousal, and low pleasurablelow arousal. The scale measures an individualÂ’s emotional reactions to their job by requiring respondents to identify how often they had experienced specific emotions in the prior 30 days. A five-point Likert response format is used where one represents and five represents % High scores on this scale indicate high levels of affective well-being. Van Katwyk, et al., (2000) report an alpha of 0.95 for the 30-item version of the scale, which is comparable to the alpha of 0.93 obtained for the 20-item version of the JAWS used in this study.
27 45$ 045$@.8;. #<99= "1 This scale assesses an individualÂ’s trait of general affectivity with items that describe different feelings and emotions. The PANAS measures both positive and negative affectivity using a five-point Likert format that ranges from to Higher scores on positive or negative affectivity correspond to higher levels of positive and negative traits, respectively. Watson et al. (1988) report good internal consistency reliabilities for both the positive and negative affectivity scales ( = 0.88, = 0.87). ($('$0('$+8;'%. #<97= ",1 The job autonomy subscale of the JDS measures the level of autonomy in an individualÂ’s job. The subscale assesses autonomy with a seven-point Likert response format in which respondents are asked to respond to items that range from to High scores on this subscale suggest a high level of job autonomy. Fields (2002) reports alphas that range from 0.68 to 0.77 for the autonomy subscale. $!$0$;,.#<93= "1 This 18-item scale measures the level of self-monitoring in an individual. There are two response choices, true or false. High self-monitoring individuals should have high scores in the keyed direction, while low self-monitoring individuals answer in the alternative direction. Snyder and Gangestad (1986) report an acceptable level of internal consistency reliability for this scale ( = 0.70), where an alpha of 0.70 is the minimum considered acceptable (Nunnally & Bernstein, 1994).
28 $ $ 0,.4;@.#<
29 4 Permission was obtained from the supervisors of all participants, or the Human Resources department at the organization. Some surveys were mailed directly to participants who returned them via intra-campus mail with a preaddressed envelope. Online survey completion took place at only one of the participant organizations, in which an email was sent to potential participants that explained the purpose of the survey and invited them to participate online. However, the majority of the surveys were made available at the organization, usually in a break room, where participants completed and returned them anonymously to a provided box to be retrieved at a predetermined time.
30 Results Scale scores were obtained for emotional labor, the subscales of deep acting and surface acting, and the proposed moderator variables of emotional intelligence, selfmonitoring, positive affectivity, negative affectivity, autonomy, coworker support and supervisor support. Scale scores were also calculated for the consequences of job satisfaction, affective well-being, and emotional exhaustion. Individual participantsÂ’ mean responses were substituted for each scale where there were one or two missing responses. If there were more than two missing responses for a particular scale that participant did not receive a score for that scale. The means, standard deviations, observed and possible ranges, as well as the CronbachÂ’s alphas for each scale variable are presented in Table 2. The majority of scales demonstrated good internal consistency reliability, where an alpha of 0.70 is the minimum considered acceptable (Nunnally & Bernstein, 1994). The one exception was the self-monitoring scale whose alpha ( = 0.64) was just below the accepted level. Table 3 displays the zero order correlations among all study variables. Gender was significantly correlated with three other variables of interest. Gender was negatively correlated with deep acting and coworker support, such that females engaged in significantly more deep acting and received more coworker support. Gender was also positively correlated with negative affectivity that is, males scored significantly higher in negative affectivity. Tenure displayed a significant positive relationship with
31 job satisfaction and affective well-being. Age was significantly related to a number of study variables. Age was negatively related to number of customers, the amount of emotional labor performed, surface acting, and emotional exhaustion. In addition, age showed positive relationships with affective well-being, self-monitoring, coworker support, and tenure. As ethnicity is a categorical variable, one-way ANOVAs were used to examine its relationship with other study variables. As shown in Table 4, ethnicity was significantly related to emotional labor, coworker, supervisor, and social support. Participants who classified themselves as Other engaged in significantly more emotional labor ( = 50.22) than all other groups including White participants who engaged in the least amount of emotional labor ( = 42.54). Black participants received the least amount of coworker, supervisor, and social support ( = 11.00, = 26.06, = 37.06), while Asian participants received the highest level of coworker support ( =13.67), and participants who classified themselves as Other received the highest level of supervisor ( = 36.00) and social support ( = 48.11). Hypothesis 1a predicted that surface acting is positively correlated with emotional exhaustion, and examination of the zero order correlation provided support for this hypothesis ( = .28, < .05). Hypothesis 1b proposed that deep acting is negatively correlated with emotional exhaustion, and was supported by examination of the zero order correlation ( = -.17, < .05). Hypothesis 2a examined whether surface acting is negatively correlated with job satisfaction, and Hypothesis 2b examined whether deep acting is positively correlated with job satisfaction. However, neither of these hypotheses
32 was supported as the zero order correlations were non-significant.Further examinations of the zero order correlations showed that surface ac ting is negatively related to affective well-being ( = -.24, < .05), and deep acting is positively related to affective well-being ( = .24, < .05). To test Hypothesis 3a moderated regression analyses were conducted in which emotional exhaustion was regressed on deep acting, the proposed moderator, gender and the interaction between deep acting and gender. As shown in Table 5, neither R2, nor any of the b weights were significant, therefore Hypothesis 3a was not supported. However, a one-way ANOVA of the relationship between gender and deep acting showed a significant difference between genders ( = 12.20, < 0.05), with females ( = 9.86) engaging in significantly more deep acting than males ( = 8.13). For hypothesis 3b, emotional exhaustion was regressed on surface acting, gender and the interaction between surface acting and gender. Both the R2 and the b weight for the interaction were significant ( = -1.32, < .05) thus supporting the role of gender as a moderator in the relationship between surface acting and emotional exhaustion. Thus, for women, as surface acting increases so will emotional exhaustion, and for men, there is little or no increase in emotional exhaustion as surface acting increases. Additional regression analyses indicate that gender also serves as a moderator in the relationship between surface acting and affective well-being ( = 2.14, < .05). Therefore, at high levels of surface acting, females experience lower levels of affective well-being than males at the same level of surface acting, and the converse occurs at low levels of surface acting.
33 Hypotheses 4a and 4b proposed that emotional intelligence moderates the relationship between emotional labor and affective well-being, and emotional exhaustion respectively. Hypothesis 4a was supported with a significant b weight for the interaction between emotional intelligence and emotional labor ( = .03, < .05), which indicates the presence of moderator effects. Consequently, for individuals low in emotional intelligence levels of affective well-being will d ecrease as the performance of emotional labor increases. Conversely, for individuals high in emotional intelligence, an increase in the level of emotional labor will contribute to a slight increase in affective well-being. Figure 2 illustrates the moderating effect of emotional intelligence on the relationship between emotional labor and affective well-being. Hypothesis 4b was also supported with a significant b weight for the interaction between emotional intelligence and emotional labor ( = -.03, p = 0.06). Therefore, individuals high in emotional intelligence actually experience a decrease in emotional exhaustion as the performance of emotional labor increases, while indi viduals low in emotional intelligence experience higher levels of emotional exhaustion with increased emotional labor. Emotional intelligence also played a moderating role in the relationship between emotional labor and job satisfaction with a significant b weight for the interaction ( = .01, < .05). Hence, individuals high in emotional intelligence will experience an increase in job satisfaction as the level of emotional labor perform ed increases. However, for individuals low in emotional intelligence, job satisfaction will be inversely related to the amount of emotional labor performed.
34 Hypotheses 5a, 5b, and 5c proposed that self-monitoring is positively related to emotional labor and surface acting, and negatively related to deep acting. Significant zero order correlations for the relationships between self-monitoring and emotional labor ( = -.19, p < .05) and self-monitoring and surface acting ( = -.25, p < .05) provided support for Hypotheses 5a and 5b. However, Hypothesis 5c was not supported due to a non-significant correlation between self-monitoring and deep acting. Hypotheses 5d and 5e posited that self-monitoring moderates the relationship between surface acting and job satisfaction, and surface acting and emotional exhaustion respectively. Neither of these hypotheses was supported with non-significant b weights for both interactions. Examination of Hypotheses 6a and 6b did not provide support for the role of positive affectivity and negative affectivity as moderators in the relationship between emotional labor and emotional exhaustion, as the b weights for both interactions were non-significant. However, additional regression analyses indicate that positive affectivity moderates the relationship between surface acting and emotional exhaustion, ( = -.07, = .08). Thus, individuals with low levels of positive affectivity will experience higher levels of emotional exhaustion when engaging in high levels of surface acting. On the contrary, individuals high in positive af fectivity will not experience much change in emotional exhaustion when engaging in high or low levels of surface acting. Hypotheses 7a and 7b predicted that autonomy moderates the relationships between emotional labor and affective well-being (see Figure 4), and between emotional labor and emotional exhaustion (see Figure 5). Both hypotheses were supported with significant b weights for the interactions between autonomy and emotional labor ( =
35 .07, < .05), ( = -.05, < .05). These results are interesting as they indicate that individuals with high autonomy experience a much greater increase in affective wellbeing as the level of emotional labor performed increases than the individuals with low autonomy. With regard to Hypothesis 7b individuals with little job autonomy experience significantly more emotional exhaustion as the level of emotional labor increases, and individuals with high job autonomy experience a slight decline in emotional exhaustion as the performance of emotional labor increases. Further regression analysis determined that autonomy also moderates the relationship between emotional labor and job satisfaction with a significant b weight for the interaction between autonomy and emotional labor ( = .01, < .10). As a result, high autonomy individuals experience a slight increase in job satisfaction, and individuals with low autonomy will experience a marked decrease in job satisfaction as the performance of emotional labor increases. Additional regression results demonstrate that autonomy also acts as a moderator in the relationship between surface acting and all three outcome variables, and the relationship between deep acting and affective well-being and job satisfaction. As surface acting increases, individuals with low autonomy experience a sharp increase in emotional exhaustion and a sharp decrease in both affective well-being and job satisfaction, however with the increase in surface acting, high autonomy individuals experience a slight increase in emotional exhaustion, affective well-being and job satisfaction. As deep acting increases, low autonomy individuals experience a marked decline in both affective well-being and job satisfaction, while individuals with high levels of autonomy experience marked increases in affective well-being.
36 Hypotheses 8a and 8b posited that coworker and supervisor support moderate the relationships between emotional labor and emotional exhaustion and between emotional labor and affective well-being. While there were significant zero order correlations between the proposed moderators and the dependent variables, the b weights for the interactions were not significant; hence these hypotheses were not supported. However, regression analyses of the role of coworker support as a moderator in the relationship between emotional labor and job satisfaction did find a significant b weight for the interaction between coworker support and emotional labor ( = -.03, < .05). Thus, individuals with high coworker support experience greater job satisfaction which declines slightly as the performance of emotional labor increases. On the contrary, individuals with low coworker support experience lower job satisfaction that increases as emotional labor increases. A significant b weight for the interaction between supervisor support and surface acting ( = .07, < .05) indicates that supervisor support moderates the relationship between surface acting and affective well-being. That is, individuals with high levels of supervisor support will have a level of affective well-being that does not change much, or increases with the performance of emotional labor, and individuals with low supervisor support will experience a sharp decrease in affective well-being as the performance of emotional labor increases. Further analysis showed that social support also moderated the relationship between surface acting and affective well-being in a similar fashion.
37 Discussion The purpose of this study was to investigate the relationship between the emotional labor process and the long-term consequences of emotional exhaustion, job satisfaction and affective well-being at work. In addition, this study examines some individual and organizational factors that moderate the relationships between emotional labor and its potential consequences. In essence, this study aims to test a subset of GrandeyÂ’s (2000) emotion regulation framework in an organizational context. A number of interesting findings emerged from this study, as well as support for several of the proposed hypotheses. Among the more notable findings is the role of autonomy as a moderator in relationships between emotional labor, and the three outcome variables emotional exhaustion, affective well-being, and job satisfaction. That is, the more autonomy a service worker has, the less deleterious the effects of performing emotional labor. In fact, autonomy was significantly related to emotional exhaustion, affective well-being, and job satisfaction ( = -.21, = .29, = .27, < .05; respectively). This finding suggests that providing autonomy to customer service employees would alleviate potential negative effects that may stem from their performance of emotional labor. The significant relationship between autonomy and deep acting ( = .17, < .05) implies that when employees with high levels of autonomy are confronted with a situation that requires emotional labor they will be mo re likely to engage in deep acting in order to adhere to organizational display rules. The choice of deep acting over surface
38 acting is beneficial to the employee because while both forms of emotional labor require effort, surface acting is more taxing because it entails the effortful suppression of emotions as well as the production of the appropriate emotion (Richards & Gross, 2000). In all of the cases where autonomy acts as a moderator, the real difference is seen at the higher levels of emotional labor, deep acting or surface acting. That is, at the lower levels there is not much difference between the high and low autonomy individuals, but as the level of emotional labor, deep or surface acting increases, so too does the difference in the outcome variables for the high and low autonomy individuals. Therefore it is more crucial for organizations to provide service employees with autonomy in situations in which they have to engage in high levels of emotional labor, or deep or surface acting. These findings agree with WhartonÂ’s (1993) findings that autonomy served to alleviate the negative effects of jobs with high levels of emotional labor. Emotional labor displayed almost no relationship with emotional exhaustion, however the two mechanisms of emotional labor, surface acting and deep acting, displayed positive and negative relationships with emotional exhaustion. This is an important finding in that it supports WhartonÂ’s (1993) contention that it is not emotional labor itself that results in emotional exhaustion, but other facets of the job, or in this case, how the emotional labor is performed. The positive relation between surface acting and emotional exhaustion suggests that simply Â“faking itÂ” is detrimental to the employee. On the other hand, the negative relation between deep acting and emotional exhaustion indicates that the employee faced with a conflict between a felt emotion and the
39 organizationally mandated emotion would be better served to attempt to actually feel the required emotion, as less emotional exhaustion is likely with this method. Given this finding, organizations should provide training in techniques of deep acting such as cognitive reappraisal or attentional deployment, to their service employees in order to reduce emotional exhaustion. Research by Gross (1998a) found that while deep acting via reappraisal reduced expressed emotions as well as self-reported emotions, there was no clear effect on the physiological arousal caused by discrepant emotions. This suggests an interesting avenue for future research on the deep acting mechanism to determine whether it can have an effect on physiological arousal. Job satisfaction exhibited almost non-existent relationships with surface acting and deep acting. These relations were in the proposed direction with surface acting negatively related, and deep acting positively related, however the non-significance of these relations may be explained by the complexity of the job satisfaction construct. The scale used in this study was a three-item measure of overall job satisfaction (Cammann, et al., 1979), so it may not be fine-grained enough to determine the extent to which surface and deep acting may impact a particular facet of job satisfaction such as the nature of the work. Future research should take a facet approach in examining how surface acting and deep acting can affect job satisfaction. The lack of support for gender as a moderator in the relationship between emotional exhaustion and deep acting is unexpected as one-way ANOVA results revealed that females did engage in significantly more deep acting than males. Furthermore, gender did not moderate the relationship between deep acting and the other two outcome
40 variables, affective well-being and job satisfaction. These results indicate that there is a difference in the performance of deep acting but they do not clarify what effect this difference has on either gender. In contrast, gender did serve as a moderator in the relationships between surface acting and two of the outcome variables, emotional exhaustion and affective well-being. For females, the positive relationship between surface acting and emotional exhaustion provides support for the notion that they have been socialized to deal with the interpersonal demands of emotion management and consequently rely more on deep acting in order to adhere to display rules. As such, heavy reliance on surface acting leads to more emotional dissonance, which results in increased emotional exhaustion and decreased affective well-being. Results indicate that emotional intelligence serves as a moderator in the relationship between emotional labor and all three outcome variables, affective wellbeing, emotional exhaustion and job satisfaction. Therefore, individuals with high levels of emotional intelligence are be tter able to regulate their emotions to meet organizational display rules, which results in lower emotional exhaustion, and higher levels of affective well-being and job satisfaction.In each of these three findings, individuals with high emotional intelligence actually fared better as the level of emotional labor increased, while individuals low in emotional intelligence experienced more negative outcomes with the increase in emotional labor. This finding has interesting implications for the selection of customer service employees. That is, employers may want to assess the emotional intelligence of potential employees as it app ears to contribute to employee well-being in the service context. Interestingly, in the case of emotional exhaustion and job
41 satisfaction, when emotional labor is low the individuals high in emotional intelligence experience slightly more negative outcomes, that is more emotional exhaustion and less job satisfaction. This implies that low emotional intelligence individuals may not consistently recognize the need for emotion regulation at the lower levels of emotional labor, and so engage in less emotional labor which for them results in less emotional exhaustion and slightly higher job satisfaction. A high self-monitorÂ’s proficiency at monitoring and control of expressive behavior may help to explain why self-monitoring did not serve as a moderator in any of the relationships examined in this study. Having to engage in high levels of surface acting may be interpreted as a challenge to an individualÂ’s ability to self-monitor and hence result in lower levels of job satisfaction. That is, a high self-monitor is constantly monitoring and modifying their expressive behavior to match the social situation, and the extra psychic burden imposed by surface acting in order to meet display rules may be emotionally taxing. Little support was found for positive affectivity as a moderator in the relationships between emotional labor, deep acting, surface acting, and the three outcome variables. The one exception was the relationship between surface acting and emotional exhaustion. This study only examined service jobs that required the expression of positive emotions, accordingly individuals high in positive affectivity did not experience much change in emotional exhaustion with regard to the level of surface acting. A high level of positive affectivity makes it less demanding for an individual to engage in surface acting because positive emotions are often experienced, and consequently are easier to express even
42 when not experienced. It would be interesting to investigate whether this finding would translate to negative affectivity serving as a moderator in the relationship between surface acting and emotional exhaustion. Coworker support served as a moderator only in the relationship between emotional labor and one outcome variable, job satisfaction, and the form of the relationship is contrary to what would be expected based on the other hypotheses. To clarify, in this relationship individuals with low coworker support experienced increased job satisfaction, and individuals with high coworker support had a slight decrease in job satisfaction as emotional labor increased. This finding seems to support BeehrÂ’s (1995) notion of reverse buffering, in which social support may actually facilitate a positive relationship between stressors and strains. Coworker support is implicitly reciprocal such that individuals who indicate high levels of coworker support may also provide their coworkers with support, therefore as increases in emotional labor make the situation more emotionally demanding, having to provide support to colleagues may result in the slight decline observed in job satisfaction. The positive relationship between job satisfaction and emotional labor for those low in coworker support is a little more difficult to explain. Perhaps individuals who are unencumbered by the need to support colleagues during emotionally demanding times are able to attend to their task more appropriately, which could result in an increased sense of personal accomplishment which would contribute to increased job satisfaction. Supervisor support and the combined measure of social support served as moderators in the relationship between affective well-being and surface acting. In both
43 cases, the interaction was in the expected direction such that those with low support experienced a decline in affective well-being, and those with high levels of support experienced a slight increase in well-being as surface acting increased. This suggests that supervisors of service employees should provide support especially when the employees experience increased levels of surface acting. It should be noted that the combined measure of social support is more closely related to supervisor than coworker support because it is a combination of both scales and the supervisor support scale contained nine items as compared to the four items in the coworker support scale. The separation of coworker and supervisor support in this study allowed a closer examination of the role of social support as a moderator. It seems as though the social support suggested by Cohen and Wills (1985) to facilitate individual coping with job stressors in the service context is more likely to be supervisor support. On the contrary, coworker support appears to act as a double-edged sword because while beneficial at times, it can also be emotionally taxing. The cross-sectional nature and the use of only self-report measures do serve as limitations of this study as there is the concern that self-report methodology will lead to the overstatement of relationships between variables. It must be noted that the use of self-report measures may be appropriate in this instance because this study seeks to assess individual behaviors, attitudes and perceived outcomes. Nevertheless, future research in this area could improve on the methods used here by including some physiological measures of distress, perhaps during the performance of emotional labor.
44 To address the concern about the cross-sectional nature of this study, future research should adopt a longitudinal design where the levels of emotional labor and the outcome variables are assessed at various points in time so that inferences can be made about the causal nature of these relationships. The organizations that contributed to this sample were drawn from a number of different industries, two from hospitality, two from financial services, four from retail, one from technology, and one from the public sector, consequently the diversity of sources for this sample should contribute to the generalizability of these findings. In addition, organizational display rules will differ across organizations, so collecting data from a number of organizations facilitates the examination of emotional labor processes across varied organizational contexts. Overall, this study provides a valuable contribution to the literature on emotions in the workplace, and in particular it serves to clarify the how the process of emotional labor affects the service employee. Of the variables examined, emotional intelligence and autonomy seem to hold the most promise for future avenues of research. Emotional intelligence is relatively new construct that has a clear impact on the relationship between emotional labor and the outcome variables assessed. This study lends further empirical support to the use of emotional intelligence as a predictive tool in the service context. The strength with which autonomy alleviated the negative outcomes and contributed to the positive outcomes marks it as an essential variable in the service context worthy of further consideration.
45 In conclusion, this study provides useful information to organizations in the service industry, as well as to researchers because the negative consequences associated with performance of emotional labor can have immense personal and organizational costs. Understanding the emotional labor process and how it can result in negative consequences for employees is the first step in attempting to ameliorate the sometimes negative aspects of service work and reduce the related personal and organizational costs.
46 References Abraham, R. (1998). Emotional dissonance in organizations: Antecedents, consequences, and moderators. ,.$.,4 .#)/. 229-246. Adelmann, P. K. (1995). Emotional labor as a potential source of job stress. In S. L. Sauter & L. R. Murphy (Eds.), 68* (pp. 371381). Washington, DC: American Psychological Association. Arvey, R. D., Bouchard, T. J., Segal, N. L., & Abraham, L. M (1989). Job satisfaction: Environmental and genetic components. ( 4.7/. 187192. Ashforth, B. E., & Humphrey, R. H. (1993). Emotional labor in service roles: The influence of identity. C%.#9. 88-115. Baruch-Feldman, C., Brondolo, E., Ben-Dayan, D., & Schwartz, J. (2002). Sources of social support and burnout, job satisfaction and productivity. ( 6 4.70#1 84-93. Beehr, T. A. (1995). 4%8 New York: Routledge. Brotheridge, C. M. & Grandey, A.A. (2002). Emotional labor and burnout: Comparing two perspectives of Â“people workÂ”. (B:.3A 17-39. Brotheridge, C. M., & Lee, R. T. (2003). Development and validation of the Emotional Labour Scale. (6 64.730-1 365-379. Brotheridge, C. M., & Lee, R. T. (2002). Testing a conservation of resources model of the dynamics of emotional labor. (6 4. 70#1 57-67. Cammann, C., Fichman, M., Jenkins, D., & Klesh, J. (1979). 6> Unpublished manuscript, University of Michigan, Ann Arbor. Caplan, R. D., Cobb, S., French, J. R. P., Van Harrison, R., & Pinneau, S. R. (1980). ( %8 Ann Arbor: University of Michigan, Institute of Social Research.
47 Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. 4:.<9. 310-357. Cordes, C. L. & Dougherty, T. W. (1993). A review and integration of research on job burnout. C%.#90/1. 621-656. Davis-Blake, A., & Pfeffer, J. (1989). Just a mirage: The search for dispositional effects in organizational research. C%.#/. 385-400. Diener, E., Larsen, R., & Emmons, R. (1984). Person x situation interactions: Choice of situations and congruence response models. (4$ 4./7. 580-592. Ekman, P. (1973). Cross culture studies of facial expressions. In P. Ekman (Ed.) '% % 169-222. New York: Academic Press. Fields, D. L. (2002). Taking the measure of work: A guide to validated scales for organizational research and diagnosis. Thousand Oaks, CA: Sage. Gardner, W. L., & Martinko, M. J. (1988). Impression management in organizations. (.#/. 321-338. Goleman, D. (1995). New York: Bantam Books. Grandey, A. A. (2000). Emotion regulation in the workplace: A new way to conceptualize emotional labor. (6 4.20#1. 95-110. Grandey, A. A. (2003). When Â“the show must go onÂ”: Surface acting and deep acting as determinants of emotional exhaustion and peer-rated service delivery. (./30#1 86-96. Greenhaus, J. H., Parasuraman, A., & Wormley, W. M. (1990). Effects of race on organizational experiences, job performance evaluations, and career outcomes. (.--0#1 64-86. Gross, J. (1989). Emotion expression in cancer onset and progression. $$ .)9. 1239-1248. Gross, J. (1998a). Antecedentand response-focused emotion regulation: Divergent consequences for experience, expression, and physiology. (4 $4.7/0#1 224-237.
48 Gross, J. (1998b). The emerging field of emotion regulation: An integrative review. C%,4.)0-1. 271-299. Grove, S. J., & Fisk, R. P. (1989). Impression management in services marketing: A dramaturgical perspective. In R. A. Giacalone & P. Rosenfeld (Eds.), + 427-438. Hillsdale, NJ: Erlbaum. Idaszak, J.R., & Drasgow, F. (1987). A revision of the Job Diagnostic Survey: Elimination of a measurement artifact. ( 4.7) 69-74. Hochschild, A. R. (1979). Emotion work, feeling rules, and social structure. ($.92. 551-575. Hochschild, A. R. (1983). Berkeley: University of California Press. House, J. S. (1981). @8 London: Addison-Wesley. Jackson, S. E., Schwab, R. L., & Schuler, R. S. (1986). Toward an understanding of the burnout phenomenon. ( 4.7#. 630-640. Kahn. R. L. (1964). Role conflict and ambiguity in organizations. 4 .<. 8-13. King, L. A. & Emmons, R. A. (1990). Conflict over emotional expression: Psychological and physical correlates. (4$4.29. 864877. Kruml, S., & Geddes, D. (1998, August). D%8 Paper presented at the First Conference of Emotions in Organizational Life, San Diego, CA. Kruml, S., & Geddes, D. (2000). Catching fire without burning out: Is there an ideal way to perform emotional labor? In N. M. Ashkanasy, C. E. J. Haertel, & W. J. Zerbe. (Eds.) %8 C. (pp. 177-188). Westport, CT: Quorum Books. Maslach, C. (1982). : Englewood Cliffs, NJ: Prentice Hall. Maslach, C., & Jackson, S. (1986). (2nd ed.). Palo Alto, CA: Consulting Psychologists Press. Maslach, C., Jackson, S. & Leiter, M. (1996). (3rded.). Palo Alto, CA: Consulting Psychologists Press.
49 Mayer, J. D., & Salovey, P. (1997). What is emotional intelligence? In P. Salovey & D. J. Sluyter (Eds.), (pp. 3-31). New York: Basic Books. Middleton, D. R. (1989). Emotional style: The cultural ordering of emotions. #70)1. 187-201. Morris, J. A., & Feldman, D. C. (1996). The dimensions, antecedents, and consequences of emotional labor. C%.)#0/1. 986-1010. Morris, J. A., & Feldman, D. C. (1997). Managing emotions in the workplace. ( +.<0-1 257-274. Nunnally, J. C. & Bernstein, I. H. (1994). 4 New York, NY: McGraw-Hill. Parasuraman, A., Zeithaml, V. A., & Berry, L.L. (1985). A conceptual model of service quality and its implications for future research. (8./<0/1 4150. Rafaeli, A. (1989). When clerks meet customers: A test of variables related to emotional expressions on the job. ( 4.7/0-1. 385-393. Rafaeli, A., & Sutton, R. I. (1987). Expression of emotion as part of the work role. C%.#). 23-37. Rafaeli, A., & Sutton, R. I. (1989). The expression of emotion in organizational life. C6:.##. 1-42. Repeti, R. L., & Cosmas, K. A. (1991). The quality of the social environment at work and job satisfaction. ( $4.)#0#A1 840-854. Richards, J. M., & Gross, J. J. (2000).Emotion regulation and memory: The cognitive costs of keeping oneÂ’s cool. (4$4.7< 410-424. Riggio, R. E., & Friedman, H. S. (1982). The interrelationship of self-monitoring factors, personality traits, and nonverbal social skills. (5:.7. 33-45. Salovey, P., Bedell, B. T., Detweiler, J. B., & Mayer, J. D. (2000).Current directions in emotional intelligence research. In M. Lewis & J. Haviland-Jones (Eds.), 8 (pp.504-520). New York, NY: Guilford Press.
50 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.), . 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. C.//. 583-602. Schneider, B., & Bowen, D. E. (1985). Employee and customer perceptions of service in banks: Replication and extension. ( 4.7A0-1. 423433. Sheldon, K. M., Ryan, R. M., Rawsthorne, L. J. & Ilardi, B. (1997).Trait self and true self: Cross-role variation in the big-f ive personality traits and its relations with psychological authenticity and subjective well-being. (4 $4.7-031. 1380-1393. Smith, T. W. (1992). Hostility and health: Current status of a psychosomatic hypothesis. 4.##. 130-150. Snyder, M. (1974). Self-monitoring of expressive behavior. (4 $4.-A. 526-537. Snyder, M., & Gangestad, S. (1986). On the nature of self-monitoring: Matters of assessment, matters of validity. (4$4. 2#0#1 125-139. Staw, B. M., Bell, N. E., & Clausen, J. A. (1986). The dispositional approach to job attitudes: A lifetime longitudinal test. $>.-#. 5677. Sutton, R. I. (1991). Maintaining norms about expressed emotions: The case of bill collectors. $>.-3. 245-268. Sutton, R. I., & Rafaeli, A. (1988). Untangling the relationship between displayed emotions and organizational sales: The case of convenience stores. (.-#. 461-487. Van Katwyk, P. T., Fox, S., Spector, P. E., & Kelloway, E. K. (2000). Using the Jobrelated Affective Well-being Scale (JAWS) to investigate affective responses to work stressors. (6 4 2 219-230.
51 Van Maanen, J., & Kunda, G. (1989). Â“Real feelingsÂ”: Emotional expression and organizational culture. In L. L. Cummings & B. M. Staw (Eds.), C .##. 43-103. Greenwich, CT: JAI Press. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. ( 4$4.2/. 1064-1070. Watson, D., & Tellegen, A. (1985). Toward a consensual structure of mood. 4:.<90)1 219-235. Weiss, H., & Cropanzano, R. (1996). Affective events theory: A theoretical discussion of the structure, causes, and consequences of affective experiences at work. C6:.#9. 1-74. Wharton, A. S. (1993). The affective consequences of service work. @8 6 .)A. 205-232. Wharton, A. S., & Erickson, R. J. (1993). Managing emotions on the job and at home: Understanding the consequences of multiple emotional roles. C%.#9. 457-486. Wong, C., & Law, K. S. (2002). The effects of leader and follower emotional intelligence on performance and attitude: An exploratory study. >.#-0-1 243-274. Zajonc, R. B. (1985). Emotion and facial efference: An ignored theory reclaimed. $.))9. 15-21. Zapf, D. (2002). Emotion work and psychological well-being. A review of the literature and some conceptual considerations. CC%.#). 237-268.
52 Table 1 Number of Participants from each Organization Organization Number of Participants A 7 B 11 C 6 D 24 E 14 F 20 G 11 H 21 I 5 J 11 K 27 L 19
53 Table 2 Descriptive Statistics and CronbachÂ’s Alpha for Study Variables Scale N Mean SD Alpha Range (Observed) Range (Possible) Emotional labor 174 43.14 7.93 0.84 14 Â– 65 14 -70 Deep acting 174 9.40 2.96 0.90 3 Â– 15 3 Â– 15 Surface acting 174 8.66 2.64 0.76 3 Â– 15 3 Â– 15 Job satisfaction 172 14.11 3.27 0.81 3 Â– 18 3 Â– 18 Emotional exhaustion 174 18.89 12.62 0.90 0 Â– 52 0 Â– 54 Emotional intelligence 175 80.05 7.47 0.86 58 Â– 96 16 Â– 96 Affective well-being 173 69.17 13.13 0.93 33 Â– 100 20 Â– 100 Autonomy 172 15.75 4.72 0.91 3 Â– 21 3 Â– 21 Self-monitoring 173 27.35 3.14 0.64 20 Â– 34 18 Â– 36 Coworker support 176 12.30 2.59 0.78 4 Â– 16 4 Â– 16 Supervisor support 174 31.21 9.50 0.94 9 Â– 45 9 Â– 45 Social support 174 43.52 10.49 0.81 16 Â– 61 13 Â– 61 Life satisfaction 174 23.19 6.88 0.87 5 Â– 35 5 Â– 35 Positive affectivity 172 36.38 6.50 0.87 16 Â– 50 10 Â– 50 Negative affectivity 172 17.01 5.72 0.85 10 Â– 40 10 Â– 50
54 Table 3 Intercorrelations Between Study Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 178 18 1. Number 2. Duration -.17* 3. EL .12 -.04 4. DA .01 .02 .68* 5. SA .07 .05 .54* .08 6. JS -.15 -.09 .05 .10 -.07 7. EE .10 .05 .04 -.17* .28* -.52* 8. EI -.11 .14 .06 .11 -.08 .12 -.22* 9. AW -.14 -.09 .03 .24* -.24* .69* -.74* .26* 10. AU -.22* -.06 .07 .17* -.13 .27* -.23* .18* .29* 11. SM -.13 -.16 -.16* -.05 -.25* .04 -.09 -.14 .15* -.03 12. CS -.12 -.11 -.08 .03 -.22* .32* -.43* .04 .44* .05 .19* 13. SS .02 -.18* -.01 .12 -.14 .43* -.40* -.03 .39* .24* .03 .27* 14. SU -.02 -.19* -.03 .12 -.18* .46* -.46* -.02 .46* .23* .08 .49* .97* 15. PA -.17* .14 .14 .22* -.06 .36* -.42* .51* .53* .24* -.04 .21* .12 .15* 16. NA .02 -.05 .16* -.10 .41* -.19* .52* -.34* -.50* -.26* -.23 -.26* -.13 -.17* -.32* 17. Gender -.08 .05 -.14 -.26* .08 -.02 .03 .03 -.06 .10 -.17 -.17* -.00 -.03 -.02 .17* 18. Tenure -.09 -.01 -.13 -.14 -.08 .17* -.12 -.01 .16* -.03 .12 .13 -.10 -.06 .12 -.09 .04 19. Age -.33* -.02 -.22* -.07 -.20* .14 -.19* -.01 .28* .03 .34* .31* -.07 .01 .07 -.15 -.03 .48* Note. EL = emotional labor, DA = deep acting, SA= surface acting, JS = job satisfaction, EE = emotional exhaustion, EI = emoti onal intelligence, AW = affective well-being, AU = autonomy, SM = self-monitoring, CS = coworker support, SS = superv isor support, SU = social sup port, LS = life satisfaction, PA = positive affectivity, NA = negative affectiv ity. Gender: 1 = female, 2 = male. < 0.05.
55 Table 4 One-way ANOVAs for Ethnicity Ethnicity Variables F value Asian Black Hispanic White Other Emotional labor 2.22* 42.78B 43.06B 43.53B 42.54B 50.22A Deep acting 1.22 8.78B 9.75AB 9.60AB 9.43AB 11.44ASurface acting 1.20 9.67A 8.50A 7.67A 8.31A 9.33AEmotional exhaus tion .25 13.56A 17.94A 17.33A 16.81A 15.67AAffective well-being 1.52 79.38A 68.18B 66.47B 70.40AB 71.11ABJob satisfaction 1.57 14.89AB 13.47AB 12.60B 14.39AB 15.44AEmotional intelligence 1.65 79.78A 82.29A 82.73A 78.69A 80.78AAutonomy .71 14.33A 16.31A 16.40A 15.47A 17.56ASelf-monitoring .92 25.89A 27.25A 27.67A 27.86A 27.56ACoworker support 2.53** 13.67A 11.00B 12.13AB 12.76AB 12.11ABSupervisor support 2.51** 34.33A 26.06B 30.07AB 32.33AB 36.00ASocial support 3.22** 48.00A 37.06B 42.20AB 45.11A 48.11ALife satisfaction 1.24 21.33A 21.65A 20.93A 23.71A 20.22APositive affectivity .80 38.63A 36.56A 36.40A 35.61A 38.67ANegative affectivity .39 18.63A 17.00A 16.73A 16.34A 16.00A < .10, ** < .05
56 Table 5 Results of Moderated Regression AnalysesEmotional Exhaustion Affective Well-Being Job Satisfaction Independent variable R2 b R2 b R2 b H3a.03.07**.02 Gender 3.35 -.07 -.10 Deep acting -.14 -7.55 -1.37 Gender Deep acting -.38 .88 .17 H3b.10**.10**.03 Gender 12.36* -20.08** -.62** Surface acting 2.92** -3.93** -3.65* Gender Surface acting -1.32** 2.14** .41** .02 .02.01 Gender 18.45* -20.17 -2.0 Emotional labor .58* -.51 -.04 Gender Emotional labor -.41** .44 .05 .09** .13***.02 Emotional intelligence -.08 -.03 .01 Deep acting 1.96 -3.23 .25 Emotional intelligence Deep acting -.03 .05 .00 .14*** .13***.02 Emotional intelligence .15 -.09 -.03 Surface acting 5.75 -6.10 -.81 Emotional intelligence Surface acting -.06 .06 .01 H4b.08**H4a.10**.04* Emotional intelligence .69 -.85 -.28 Emotional labor 2.16** -2.54** -.62* Emotional intelligence Emotional labor -.03* .03** .01** .04** .09**.02 Self-monitoring -.15 1.60 .17 Deep acting -.13 3.87 .46 Self-monitoring Deep acting -.02 -.10 -.01 H5e.07**.07**H5d.02 Self-monitoring -.29 .57 .40 Surface acting .51 -.56 1.05 Self-monitoring Surface acting .02 -.02 -.04 .01 .03.02 Self-monitoring -.20 1.34 .64 Emotional labor .07 .52 .39 Self-monitoring Emotional labor -.00 -.02 -.01 .19*** .31***.14*** 5 Positive affectivity -.45 .57 .04 Deep acting .64 -1.04 -.50 Positive affectivity Deep acting -.03 .05 .02 < .10, ** < .05, *** < .0001
57 Table 5 (Continued) Results of Moderated Regression AnalysesEmotional Exhaustion Affective Well-Being Job Satisfaction Independent variable R2 b R2 b R2 b .27***.33*** .14*** Positive affectivity -.10 .61 .28** Surface acting 3.86** -3.07* .37 Positive affectivity Surface acting -.07* .05 -.01 H6a.20***.28*** .14*** Positive affectivity -.33 .43 -.04 Emotional labor .54 -.64 -.20 Positive affectivity Emotional labor -.01 .02 .01 .28***.30*** .05** Negative affectivity .67* -.67 .04 Deep acting -1.12 1.77** .37 Negative affectivity Deep acting .04 -.05 -.02 .27***.25*** .04* Negative affectivity 1.19** -.99* -.17 Surface acting .95 -.12 -.10 Negative affectivity Surface acting -.03 -.01 .01 H6b.26***.26*** .04* Negative affectivity 1.08 -.98 -.09 Emotional labor -.03 -.27 -.04 Negative affectivity Emotional labor -.00 .01 .001 .09**.16*** .10** Autonomy .20 -.63 -.15 Deep acting .96 -1.68 -.54* Autonomy Deep acting -.09 .16** .04** .17***.20*** .11** Autonomy .77 -1.27** -.23 Surface acting 3.41** -4.71*** -.78** Autonomy Surface acting -.15** .23** .05** H7b.10**H7a.13*** .09** Autonomy 1.32 -1.93* -.27 Emotional labor .86** -1.06** -.17 Autonomy Emotional labor -.05** -.07** .01* .19***.25*** .12*** Coworker support -2.40** 3.77** .75** Deep acting -1.44 3.13** .58 Coworker support Deep acting .07 -.17 -.04 .20***.21*** .10** Coworker support -2.65** 3.02** .71** Surface acting .45 .45 .38 Coworker support Surface acting -.11 -.10 -.03 < .10, ** < .05, *** < .0001
58 Table 5 (Continued) Results of Moderated Regression AnalysesEmotional Exhaustion Affective Well-Being Job Satisfaction Independent variable R2 b R2 b R2 b H8b.17***H8a.20*** .12*** Coworker support -5.06** 5.70** 1.56** Emotional labor -.93 1.10* .36** Coworker support Emotional labor .07 -.08 -.03** .18***.20*** .19*** Supervisor support -.31 .50* .15** Deep acting .07 .77 .08 Supervisor support Deep acting -.02 .00 -.001 .21***.22*** .18*** Supervisor support -.46 -.06 .15* Surface acting .92 -2.94** -.02 Supervisor support Surface acting .00 .07** .00 H8b.17***H8a.17*** .19*** Supervisor support -.64 .51 .32* Emotional labor -.06 .02 .15 Supervisor support Emotional labor .00 .00 -.004 .23***.25*** .22*** Social support -.42* .62** .17** Deep acting -.13 1.17 .19 Social support Deep acting -.01 -.01 -0.003 .25***.26*** .21*** Social support -.45* .02 .14** Surface acting .89 -3.31** -.02 Social support Surface acting -.00 .06** .00 H8b0.21**H8a0.22*** .23*** Social support -.61 .48 .29** Emotional labor -.08 -.03 .18 Social support Emotional labor .00 .00 -.003 < .10, ** < .05, *** < .0001
59 Figure Captions #& Conceptual framework of emotion regulation performed in the work setting. From Â“Emotion Regulation in the WorkplaceÂ” by A. Grandey 2000, ( 6 4.2. p. 101. Copyright 2000 by the Educational Publishing Foundation. Reprinted with permission of the author. NA = negative affect; PA = positive affect. )& Graph of the proposed moderating effect of emotional intelligence on the relationship between emotional labor and affective well-being. -& A subset of the emotion regulation framework proposed by Grandey (2000). /& Graph of the moderating effect of autonomy on the relationship between emotional labor and affective well-being. 2& Graph of the moderating effect of autonomy on the relationship between emotional labor and emotional exhaustion.
60 Figure 1 Situational Cues Emotion Regulation Process Long-Term Consequences Interaction Expectations Frequency Duration Variety Display rules Emotional Labor Deep acting: Modify Feelings Attentional deployment Cognitive change Surface acting: Modify Expressions Response modulation Individual Well-Being Burnout Job satisfaction Organizational Well-Being Performance Withdrawal behavior Individual Factors Gender Emotional expressivity Emotional intelligence Affectivity (NA/PA) Organizational Factors Job autonomy Supervisor support Coworker support Emotional Events Positive events N e g ative events
61 Figure 2 High High EI Affective Well-Being Low Low EI Low High Emotional Labor
62 Figure 3 Emotion Regulation Process Long-Term Consequences Emotional Labor Deep acting: Modify Feelings Surface acting: Modify Expressions Individual Well-Being Job satisfaction Affective well-being Emotional exhaustion Individual Factors Gender Emotional intelligence Self-monitoring Affectivity (NA/PA) Organizational Factors Job autonomy Supervisor support Coworker support
63 Figure 4 Hypothesis 7a11.99 76.76 -5.11 123.92 -50 0 50 100 150 low ELhigh ELEmotional LaborAffective Well-being low AU high AU
64 Figure 5 Hypothesis 7b13.9 50.11 25.06 15.37 0 10 20 30 40 50 60 low ELhigh ELEmotional LaborEmotional Exhaustio n low AU high AU
66 Appendix A Emotional Labor Scale (Brotheridge & Lee, 1998) On a typical day I have ____ customers. A typical interaction I have with a customer takes about ____ minutes. ON AN AVERAGE DAY AT WORK, HOW FREQUENTLY DO YOU DO EACH OF THE FOLLOWING WHEN INTERACTING WITH CUSTOMERS? PLEASE CIRCLE THE ONE NUMBER FOR EACH QUESTION THAT COMES CLOSEST TO REFLECTING YOUR OPINION ABOUT IT. Never Rarely Sometimes Often Always 1 Interact with customers. 1 2 3 4 5 2 Adopt certain emotions as part of your job. 1 2 3 4 5 3 Express particular emot ions needed for your job. 1 2 3 4 5 4 + Express intense emotions. 1 2 3 4 5 5 Show some strong emotions. 1 2 3 4 5 6 B Display many different kinds of emotions. 1 2 3 4 5 7 Express many different emotions. 1 2 3 4 5 8 Display many different emotions when interacting with others 1 2 3 4 5 9 Make an effort to actually feel the emotions that I need to display to others. 1 2 3 4 5 10 Try to actually experience the emotions that I must show. 1 2 3 4 5 11 Really try to feel the emotions I have to show as part of my job. 1 2 3 4 5 12 $ Resist expressing my true feelings. 1 2 3 4 5 13 Pretend to have emotions that I donÂ’t really have. 1 2 3 4 5 14 Hide my true feelings about a situation 1 2 3 4 5
67 Appendix B Job Satisfaction from the Michigan Organizational Assessment Questionnaire (Cammann, Fichman, Jenkins & Klesh, 1979) PLEASE CIRCLE THE ONE NUMBER FOR EACH QUESTION THAT COMES CLOSEST TO REFLECTINGYOUR 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
68 Appendix C 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 have 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 on 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
69 Appendix D Emotional Intelligence Scale (Wong & Law, 2002) PLEASE CIRCLE THE ONE NUMBER FOR EACH QUESTION THAT COMES CLOSEST TO REFLECTINGYOUR OPINION ABOUT IT.Strongly disagree Disagree Slightly disagree Slightly agree Agree Strongly agree 1 I have a good sense of why I ha ve 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 friendsÂ’ emotions from their behavior 1 2 3 4 5 6 6 I am a good observer of othersÂ’ emotions 1 2 3 4 5 6 7 I am sensitive to the feelings and emotions of others 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 difficulties 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
70 Appendix E Job-related Affective Well-Being Scale (Van Katwyk, Fox, Spector & Kelloway, 1999) Below are a number of statements that describe different emotions that a job can make a person feel. Please indicate the amount to which any part of your job (e.g., the work, coworkers, supervisor, clients, pay) has made you feel that emotion in the past 30 days. PLEASE CHECK ONE RESPONSE FOR EACH ITEM THAT BEST INDICATES HOW OFTEN YOU HAVE EXPERIENCED EACH EMOTION AT WORK OVER THE PAST 30 DAYS.Never Rarely Sometimes Often Always 1 My job made me feel at ease. 1 2 3 4 5 2 My job made me feel angry 1 2 3 4 5 3 My job made me feel anxious 1 2 3 4 5 4 My job made me feel bored 1 2 3 4 5 5 My job made me feel calm 1 2 3 4 5 6 My job made me feel content 1 2 3 4 5 7 My job made me feel depressed 1 2 3 4 5 8 My job made me feel disgusted 1 2 3 4 5 9 My job made me feel discouraged 1 2 3 4 5 10 My job made me feel energetic 1 2 3 4 5 11 My job made me feel excited 1 2 3 4 5 12 My job made me feel ecstatic 1 2 3 4 5 13 My job made me feel enthusiastic 1 2 3 4 5 14 My job made me feel frightened 1 2 3 4 5 15 My job made me feel furious 1 2 3 4 5 16 My job made me feel gloomy 1 2 3 4 5 17 My job made me feel fatigued 1 2 3 4 5 18 My job made me feel inspired 1 2 3 4 5 19 My job made me feel satisfied 1 2 3 4 5 20 My job made me feel relaxed 1 2 3 4 5
71 Appendix F PANAS (Watson, Clark & Tellegen, 1988) This scale consists of a number of words that describe 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
72 Appendix G Job autonomy (Job Diagnostic Survey: Idaszak & Drasgow, 1987) PLEASE CIRCLE THE ONE NUMBER FOR EACH ITEM THAT COMES CLOSEST TO REFLECTING YOUR OPINION ABOUT IT.Very inaccurate Mostly inaccurate Slightly inaccurate Uncertain Slightly accurate Mostly accurate Very accurate 1 I decide on my own how to go about doing the work. 1 2 3 4 5 6 7 2 The job gives me a chance to use my personal initiative or judgment in carrying out the work. 1 2 3 4 5 6 7 3 The job gives me considerable opportunity for independence and freedom in how I do the work. 1 2 3 4 5 6 7
73 Appendix H Self-Monitoring Scale (Snyder & Gangestad, 1986) PLEASE CIRCLE ONE RESPONSE FOR EACH ITEM THAT BEST INDICATES HOW YOU FEEL ON AVERAGE True False 1 I find it hard to imitate the behavior of other people. (F) (R) 1 2 2 At parties and social gatherings, I do not attempt to do or say things that others will like. (F) (R) 1 2 3 I can only argue for ideas which I already believe. (F) (R) 1 2 4 I can make impromptu speeches even on topics about which I have almost no information. (T) 1 2 5 I guess I put on a show to impress or entertain others. (T) 1 2 6 I would probably make a good actor. (T) 1 2 7 In a group of people I am rarely the center of attention. (F) (R) 1 2 8 In different situations and with different people, I often act like very different persons. (T) 1 2 9 I am not particularly good at making other people like me. (F) (R) 1 2 10 IÂ’m not always the person I appear to be. (T) 1 2 11 I would not change my opinions (or the way I do things) in order to please someone or win their favor. (F) (R) 1 2 12 I have considered being an entertainer. (T) 1 2 13 I have never been good at games like charades or improvisational acting. (F) (R) 1 2 14 I have trouble changing my behavior to suit different people and different situations. (F) (R) 1 2 15 At a party I let others keep the jokes and stories going. (F) (R) 1 2 16 I feel a bit awkward in company and do not show up quite as well as I should. (F) (R) 1 2 17 I can look anyone in the eye and tell a lie with a straight face (if for a right end). (T) 1 2 18 I may deceive people by being fr iendly when I really dislike them. (T) 1 2
74 Appendix I Supervisory Support (Greenhaus, Parasuraman & Wormley, 1990) PLEASE CHECK ONE RESPONSE FOR EACH ITEM THAT BEST INDICATES HOW YOU FEEL ON AVERAGE.Strongly agree Agree to some extent Uncertain Disagree to some extent Strongly disagree 1 My supervisor takes the time to learn about my career goals and aspirations 1 2 3 4 5 2 My supervisor cares about whether or not I achieve my goals 1 2 3 4 5 3 My supervisor keeps me informed about different career opportunities for me in the organization 1 2 3 4 5 4 My supervisor makes sure I get the credit when I accomplish something substantial on the job 1 2 3 4 5 5 My supervisor gives me helpful feedback about my performance 1 2 3 4 5 6 My supervisor gives me helpful advice about improving my performance when I need it 1 2 3 4 5 7 My supervisor supports my attempts to acquire additional training or education to further my career 1 2 3 4 5 8 My supervisor provides assignments that give me the opportunity to develop and strengthen new skills 1 2 3 4 5 9 My supervisor assigns me special projects that increase my visibility in the organization 1 2 3 4 5
75 Appendix J Coworker Support (Caplan, Cobb, French, Van Harrison & Pinneau, 1980) PLEASE CIRCLE THE ONE NUMBER FOR EACH QUESTION THAT COMES CLOSEST TO REFLECTING YOUR OPINION ABOUT IT.Not at all A little Somewhat Very much 1 How much do your coworkers go out of their way to things to make your work life easier for you? 1 2 3 4 2 How easy is it to talk with your coworkers? 1 2 3 4 3 How much can your coworkers be relied on when things get tough at work? 1 2 3 4 4 How much are your coworkers willing to listen to your personal problems? 1 2 3 4
76 Appendix K Demographic Information __________ Female __________ Male _____ Asian _____ Black _____ Hispanic _____ White _____ Other How long have you worked for this company (in months)? ____________ Age in years: ___________ Indicate your type of service job: _____________________________________