Objective: To translate, adapt and validate emotional labour scale for Pakistani corporate employees.
Method: This study was conducted in locale of Rawalpindi and Islamabad from October 2014 to December 2015, and comprised customer service employees of commercial banks and telecommunication companies. It comprised of two independent parts. Part one had two steps. Step one involved translation and adaptation of the instrument. In the second step psychometric properties of the translated scale were established by administering it to customer services employees from commercial banks and the telecommunication sector. Data of the pilot study was analysed by using exploratory factor analysis to extract the initial factor of emotional labour. Part two comprised the main study. Commercial bank employees were included in the sample by using convenient sampling technique. SPSS 20 was used for data analysis.
Results: There were 145 participants in the first study and 495 in the second study . Exploratory factor analysis initially generated three-factor model of emotional labour which was further confirmed by confirmatory factor analysis suggesting that emotional labour had three distinct dimensions, i.e. surface acting, deep acting and genuine expressions of emotions.
Conclusion: The emotional labour scale was found to be a valid and reliable measure.
Keywords: Emotional labour, Surface acting, Deep acting, Genuine expression of emotions, Exploratory factor analysis, Confirmatory factor analysis. (JPMA 67: 233; 2017)
The term emotional labour was originally discussed by American sociologist Arlie Russell Hochschild as "the management of feelings to create a publicly observable facial and bodily display; emotional labour is sold for a wage and therefore has exchange value".1 Emotional labour is an area which has received considerable attention since its conception.2 Literature of emotional labour commonly supports the notion of emotional labour having mutually positive and negative effects for the employees and organisation proposed by Hochschild.1 Similarly, Morris and Feldman3 and Kim4 viewed emotional labour as having double-edged effects: it positively affects organisational outcomes but have negative influence on employees.
Literature of emotional labour reports the conflict findings about the outcomes of emotional labour. Precisely, literature suggests that emotional labour positively influence the service quality perceived by customers,5 by affecting customer loyalty, client satisfaction and financial outcomes.6 However, on the other hand, emotional labour has been negatively associated with the individual level outcomes. It is generally perceived that emotional labour has negative influence on job satisfaction,7 employees' well-being, emotional exhaustion,8,9 organisational commitment, intentions to leave,8,10 turnover11 and so forth. In general, it can be concluded from the available literature that emotional labour has less favourable outcomes on employees' health, well-being and work attitude. Some studies, however, reported vice versa. For instance, Wharton12 reported positive relationship between job satisfaction and emotional labour. Wharton12 explained these contradictory findings in a way that consequences of emotional labour are contingent on the strategies which an employee chooses to opt. Similarly, some other studies showed the inconsistent findings regarding the consequences of emotional labour, such as Grandey13 and Brotheridge and Lee,14 and have not reported any evidence for the relationship between emotional exhaustion and deep acting. In augmenting these inconsistent results of emotional labour consequences, several theorists15,16 recommended that researchers will be continuously producing fragmented work that would further bring more confusion about the area by debating about the outcomes of emotional labour without completely understanding dimensions and nature of the construct.
There could be much reason for these contradictory findings such as varying conceptualisations of emotional labour. A few researchers proposed the bi-dimensional perspective of emotional labour17,18 whereas others viewed it as multidimensional construct.17,18 Evidently, literature lacks the maximum support for consensus about the dimensions of emotional labour. The present research aimed to add to the literature by further clarifying and specifying the dimensions of emotional labour in Pakistani setting, specifically by addressing the validity and reliability of Diendroff et al’s19 three-dimensional scale of emotional labour.
The principal component of all emotional labour theories is that the management and expression of emotions for a wage1 is the idea that organisations directly or indirectly communicate the desired emotional expressions, i.e. emotional display rules. An employee can achieve this desired emotional expressions either by surface acting or deep acting; the main emotional strategies suggested by Hochschild.1 Surface acting denotes to the expression of organisationally desired emotions without changing the inner feelings and state. However, the former refers to actually manipulation of inner feelings and felt emotions usually through reappraising the situation so that its emotional impact is lessened. It is derived from the past literature on emotional labour that Hochschild's1 description of emotional labour, i.e. deep acting and surface acting, was generally supported by many theorists.3,17,19 However, views about the nature and dimensionality of emotional labour remained contradictory. Hochschild identified two strategies of emotional labour i.e., deep and surface acting. Hochschild1 also acknowledged a third strategy known to be as passive deep acting by which spontaneously felt feelings are matched with display rules without much effort. However, there is a debate on whether to include this third technique in the conceptualisation of emotional labour because it does not involve the conscious effort and dissonance.20 Similarly, Grandey13 stated that both techniques, i.e. deep and surface acting, describe emotional labour as emotional regulation process. However, there are some other researchers who conceptualised emotional labour as multidimensional construct.3,14 For example, Morris and Feldman's3 conceptualisation of emotional labour included four dimensions: attentiveness to display rules, frequency of emotional display, variety of emotions to be expressed and emotional dissonance.
Similarly, Brotheridge and Lee14 and Brotheridge and Grandey17 updated emotional labour into two major kinds but with different labels: job-focused and employee-focused. Each category includes the surface and deep acting and other sub-factor, i.e. frequency, intensity and variety, and duration. Recently, Diefendorff et al.19 resisted that naturally felt or genuine emotions should be the part of the emotional labour construct; they presented multidimensional view of emotional labour, including surface acting, deep acting and naturally felt emotions as major dimensions of emotional labour.
As discussed earlier, existing literature holds diverse stand on the construct of 'emotional labour', for example two-factor model or three-factor model. The current study addresses the dimensionality of the construct "emotional labour" in local context through establishing the construct validation of three-dimensional view of emotional labour. Numerous researchers13,15 advocated that by debating the outcomes of emotional labour without completely understanding dimensions and nature of the construct, researchers will be continuously producing fragmented work that would further bring more confusion about the area. The present study was conducted to highlight the nature and dimensions of emotional labour by collecting the evidence from the Pakistani sample. Understanding the dimensionality's of emotional labour would allow future researchers to better differentiate emotional labour and clarify its relationships with various factors under different circumstances. For the purpose of establishing the construct validity of emotional labour in Pakistani Sample, this study tested model proposed by Ashforth and Humphrey21 conceptualisation of emotional labour having three components, i.e. surface acting, deep acting and genuine expression of emotions. The current study was planned to assess the validity of emotional labour construct conceptualised by Dienfendroff et al.19 within a non-western context, and to establish the psychometric properties of emotional labour scale.
Subjects and Methods
This study was conducted in Rawalpindi and Islamabad from October 2014 to December 2015, and comprised customer service employees of commercial banks and telecommunication companies Employees having direct dealing with customers, minimum 14 years of education and six-month experience at the present job were included.
Emotional labour scale19 was used to measure emotional labour. This scale measured the same typology as used in another study20 in order to differentiate three techniques' of emotional labour: deep acting, surface acting and genuine expression of emotions. The original questionnaire is comprised of 14 items anchored on five-point rating scale with 1 = "never" and 5 = "always". Surface acting has 7 items, deep acting has 4 items and a genuine expression has 3 items. There is no negative item in the questionnaire and high scores on each subscale are indicative of that emotional labour strategy.
The present research was executed in two independent studies. The first study was comprised of two independent phases. The purpose of the first phase was translation and adaptation of emotional labour scale. The second phase was aimed at establishing psychometrics and the initial validation of adapted version of emotional labour scale. The second study comprised of the main study with the aim to provide better evidence about the construct validity of Urdu version of emotional labour scale.
Translation and adaptation of emotional labour scale was completed in four steps: The first step was tryout of the original English version of the emotional labour scale. Initially, the original scale was administered to 20 bank employees. The main objective of the tryout was to identify the need to translate the original scale, to check the relevance of instruments and to assess the need of adding some new items. The employees were asked to identify any item which was difficult to understand and they were also briefed to add new information which might be relevant. On the basis of the tryout, it was assessed whether to translate and adapt the original scale. It was also reported to add local perspective on this construct.
The second step was the translation and adaptation of instruments. The scales were translated into Urdu language and some new items were also developed to measure the construct of emotional labour. Forward translation method was used to translate the original scales. In the first step, translation was carried out with the help of four bilinguals having Master of Studies (MS) degree in Psychology who were also proficient in Urdu and English language and had experience of test development and adaptation. These four bilinguals were instructed to translate the scales according to the criteria prescribed by Brislin,22 i.e. (1) to maximise the content similarity, (2) maintain relatively simple language level, and (3) translating the text without substitution or elimination of any item. Each of the four bilingual experts translated the questionnaires independently. Simultaneously, five new items were also developed to include the local perspective of emotional labour. These included 2 items for surface acting, 2 item for deep acting and 2 items for genuine expressions of emotions with the help of literature review and interviews with the bank employees during the tryout session. After the completion of initial translation, a committee approach was conducted to select the best translated items in Urdu and critically review the newly developed items. The committee comprised of four psychologists including researcher. This committee was independent of the previous four bilingual experts. They had not been exposed to the questionnaire before. Each item was discussed at length and some items were rephrased to make them more comprehendible. The primary emphasis was on checking whether these translated items conveyed the same meaning as that of the original scales, and to select the best-translated items and critically review the newly developed items. Finally, the Urdu translated items were arranged in the same order as given in the original language.
The third step was tryout of the Urdu version of emotional labour scale. The translated version of emotional labour scale comprised of 17 items, including 14 original items and 3 newly developed items. The Urdu version of the emotional labour scale was administered to 15 bank employees. The basic objective of the tryout was to identify any comprehension problem, and to check the cultural relevance of the items. The result of the tryout suggested that new Urdu version of emotional labour scale was suitable to apply on a Pakistani sample.
The fourth step was establishment of psychometrics properties and initial validation. The final form of Urdu version of emotional labour scale was administered on the participants.
SPSS 20 was used for data analysis. Factor analysis, item total correlation and reliability of the emotional labour scale were conducted to test the goodness of measure for the emotional labour construct.
To examine the factor structure of emotional labour scale, exploratory factor analysis with varimax rotation was used. The Kaiser-Mayer-Olkin (KMO) value was first calculated to check the adequacy of the sample size. The calculated KMO value was 0.72, which was more than the suggested minimum value of (0.5)23 and (0.6).24 Furthermore, the value of Bartlett's test was observed to be highly significant, i.e. (c2= 568.12), (p=0.00) which indicated a good correlation between individual items for running exploratory factor analysis (Table-1).
Initially, an analysis was run to obtain the number of component through the eigenvalues. There were three factors having eigenvalue greater than 1 as suggested by Kasier's criteria of 1 and these three factors explained 56.21% variance after the rotation.
Finally, only those items were retained that had loading of 0.4. This value of 0.4 was determined according to the Stevens' criteria for critical values against which factor loadings in a specified sample can be compared.
The second study was aimed at providing further evidence about the validity of Urdu versions of emotional labour scale via confirmatory factor analysis. A sample of commercials bank employees was undertaken in the second study. Employees having at least 14 years of education with minimum of 6 months of job experience were included.
Confirmatory factor analysis (CFA) was used to confirm the initial factor structure of emotional labour scale using AMOS (Analysis of Moment Structures) 20.
Of the 180 employees in the first study, 145(80.56%) valid cases were considered for analysis.
Factors on the basis of varimax method of factor analysis were in line as guided by the theory. These factors were further tested for the reliability analysis and item total correlation (Table-1).
All items of the emotional labour scales showed significant positive correlation with their subscales which indicated that the sub-scales were internally consistent (Table-2).
The reliability coefficients for subscales of emotional labour ranged from 0.75 to 0.85 which suggested that the scale was internally consistent. This was further supported by item total correlations of the sub-scale of emotional labour scale (Table-3).
In the second study, there were 495 participants. Factor loadings of confirmatory factor analysis for emotional labour scale and its three subscales ranged between 0.43 and 0.80. Based on the initial criteria only those items were retained having item loadings >0.40 (Table-4).
The initial CFA model results were unacceptable (c2=364 (df=116; p<0.05, comparative fit index [CFI]=0.901; root mean square error of approximation [RMSEA]=0.066). However, all items had loadings greater than 0.4 in the model. The factor loadings of the newly developed items ranged from 0.77 to 0.51 (item no 16, 0.51; item no 15, 0.54; and item no 17, 0.77). In refined model, all items were retained in the model. However, the model was refined with the help of modification indices. Some of the co-error variance was allowed which resulted in significantly improved fit of data (c2=277(df=112); p<.05; RMSEA=0.055; CFI=0.935; goodness of fit index [GFI]=0.939). Refined CFA model suggested that factor loadings of all 17 items ranged from 0.43 to 0.77 (Table-5).
Table-6 depicts the some of the descriptives, reliabilites and correlations among sub-scales of emotional labor. It indicates that all of the three sub-scales achieved satisfactory alpha level which ranged from .73 to 85 which is indicating that scales in internally consistent. Results are also indicative of the fact that sub dimensions of emotional labour are distinct but related which is proven through the values of correlation.
The main objective of the study was to adapt and validate the emotional labour scale in Urdu language for Pakistani sample. The primary aim of the present study was to establish the dimensionality of emotional labour construct in non-western culture like Pakistan. For the purpose, emotional labour scale developed by Diendroff et al.19 was selected. The present research was completed in two independent studies. The purpose of the study 1 was to translate, adapt and evaluate the factor structure of emotional labour scale. The scale was translated into Urdu language by four bilingual experts using the forward translation method. Meanwhile, certain new items were also developed to include the indigenous perspective on emotional labour scale. For item generation, interviews with bank employees were conducted during the pre-try out session. Finally, four new items were developed to capture the three dimensions of emotional labour, i.e. surface acting, deep acting and genuine expressions of emotions. The new Urdu version of emotional labour scale was ready to pilot test. It was administered on the sample (n=145) comprised of employees from commercial banks and the telecommunication sector. The results supported the three factor model of emotional labour in Pakistani settings. The results also provided initial evidence for the construct validity and internal consistency of emotional labour scale. Results of exploratory factor analysis supported a three-factor structure of emotional labour, suggesting deep acting, surface acting, and expression of naturally felt emotions are distinct constructs. These findings are consistent with ideas of Ashforth and Humphrey20 and Diefndroff and Grosserand25 which stated that emotional labour has three distinct types, i.e. surface acting, deep acting and expressions of naturally felt emotions. Furthermore, the scale achieved satisfactory Cronbach's alpha ranging from 0.75 to 0.85 which suggests that the scale is internally consistent. However, to provide further validity evidence, this adapted scale was administered on the sample of (N=500) commercial bank employees in study 2. The main objective of the study 2 was to confirm the factor structure of emotional labour scale by using CFA by using AMOS. The result of the study 2 was consistent with the results of study 1. The results of CFA supported the factor structure of emotional labour as extracted by exploratory factor analysis in study 1.
These finding are important because they support the idea that naturally felt emotions are one of the dimensions of emotional labour which is against the notion that it should not be included in dimensions of emotional labour as it does not involve any conscious effort or any felt dissonance.
Therefore, there is much work needed to be done to offer more in-depth insights into the dimensionality of the construct emotional labour. Some of the possible limitations of this study are to be considered for future research. First, the sample for present research was drawn from the banking sector in Pakistan, which could limit the generalisability of the findings to the other sectors. It would be valuable to replicate the present study using different sample for diverse occupations, societies and cultures. Secondly, the data in the study was obtained from the same single source, which could lead to common methods variance. Future research should adopt multi-method approach to collect data in order to reduce the common method variance.
The emotional labour scale was found to be a valid and reliable measure.
Conflict of Interest: None.
Source of Funding: None.
1. Hochschild A. The managed heart: Commercialization of human feeling. Berkeley: University of California Press; 1983.
2. Hoschchild A. Emotion work, feeling rules, and social structure. Ame J Sociol1979; 85: 551-75.
3. Morris JA, Feldman DC. The dimensions, antecedents, and consequences of emotional labor. Acad Manag Rev 1996; 21: 986-1010.
4. Kim HJ. Hotel service providers' emotional labor: The antecedents and effects on burnout. International J Hospitality Manag 2008; 27: 151-61.
5. Bowen DE, Schneider B. Service marketing and management: Implications for organizational behavior. Research in organizational behavior 1988; 10.
6. Heskett JL, Schlesinger LA, Sasser WE. The service profit chain: How leading companies link profit and growth to loyalty, satisfaction and value. New York: The Free Press; 1997.
7. Johnson HA, Spector PE. Service with a smile: Do emotional intelligence, gender, and autonomy moderate the emotional labor process? J Occup Health Psychol 2007; 12: 319-33.
8. Yang FH, Chang CC. Emotional labor, job satisfaction, and organizational commitment amongst clinical nurses: A questionnaire survey. Int J Nurs Stud 2008; 45: 879-87.
9. Ang MCH, Rostinah S, Japang MC, Nasah CRJ. Emotional labor, employee outcomes, support, and job autonomy: An empirical investigation of hotel workers in Sabah and FT Labuan. Psychology 2009; 133: 319-33
10. Abraham, R. The impact of emotional dissonance on organizational commitment and intention to turnover. J Psychol 1999; 133: 441-55.
11. Meier KJ, Mastracci SH, Wilson K. Gender and emotional labor in public organizations: An empirical examination of the link to performance. Public Administration Review 2006; 66: 899-909.
12. Wharton AS. The affective consequences of service work. Work and Occupations 1993; 20: 205-32.
13. Grandey AA. Emotion Regulation in the Workplace: A New Way to Conceptualize Emotional Labor. J Occup Health Psychol 2000; 5: 95-110.
14. Brotheridge CM, Lee RT. On the dimensionality of emotional labor: Development and validation of the emotional scale. Paper presented at the First Conference on Emotions in Organizational Life. San Diego: 1998.
15. Fisher CD, Ashkanasy NM. The emerging role of emotions in the work life: An introduction. J Organ Behav 2000; 66: 339-57.
16. Grandey AA. Emotion regulation in the workplace: A new way to conceptualize emotional labor. J Occup Health Psychol 2000; 21: 123-9.
17. Brotheridge CM, Grandey AA. Emotional labor and burnout: Comparing two perspectives of people work. J Vocational Behav 2002; 60: 17-39.
18. Kruml S, Geddes D. Catching fire without burning out: Is there an ideal way to perform emotional labor? In: Askanasy NM, Haertel CEJ, Zerbe WJ, eds. Emotions in the workplace: Research, theory, and practice. Westport, CT: Quorum Books; 2000, 177-88.
19. Diefendorff JM, Croyle MH, Grosserand RH. The dimensionality and antecedents of emotional labor strategies. J Vocational Behav 2005; 66: 339-57.
20. Mann S. Hiding what we feel, faking what we don't: understanding the role of your emotion at work. New York: Harper-Collins; 1999.
21. Ashforth BE, Humphrey RH. Emotional labor in service roles: The influence of identity. Acad Manag Rev 1993; 18: 88-115.
22. Brislin RW. Back-Translation for Cross-Cultural Research. J Cross Cultural Res 1970; 1: 185-216.
23. Diefendorff MJ, Grosserand RH. Understanding the emotional labor process: a control theory perspective. J Organ Behav 2003; 24: 945-59.
This journal is a member of and subscribes to the principles of the Committee on Publication Ethics.