The moderating effect of perceived organizational support in the relationship between emotional labour and job attitudes: A study among health professionals.

AIM: The present study was set out to establish the link between emotional labour (surface and deep acting) and job attitudes (job satisfaction organizational commitment) by introducing perceived organizational support as a moderating variable.
DESIGN: The study made use of a cross-sectional design by sampling three hundred and forty-two (342) nurses and midwives from six health facilities in Ghana.
METHODS: The study employed a quantitative approach to examine the relationships between the study variables.
RESULTS: The study disclosed that whereas surface acting related negatively with job satisfaction but not with organizational commitment, deep acting did not relate significantly with both job satisfaction and organizational commitment. Perceived organizational support as a moderating variable showed a significant moderating effect between surface acting and job attitudes. However, Perceived organizational support moderated the relationship between deep acting and organizational commitment but not job satisfaction. The findings pose much contextual relevance to health professionals where emotional regulations are core responsibilities of health care.

INTRODUCTION

Organizational behaviour literature, in recent times, acknowledges the essence of emotional regulations in providing explanations to job‐related attitudes among employees, especially in the service sector (Choi, Tran, & Park, 2015; Dartey‐Baah & Mekpor, 2017). Among health professionals with particular reference to nurses and midwives, emotional regulation (ability of nurses and midwives to manage emotional situations of their profession) forms central part of healthcare delivery and has a direct impact on the quality of healthcare delivery (McQueen, 2004). Nurses and midwives are therefore required to enhance positive emotions while suppressing negative emotions (Hur, Han, Yoo, & Moon, 2015; McQueen, 2004). These occupational emotional requirements are conceptualized as emotional labour (Hochschild, 1983). In her seminal work on emotional labour, she argues that employees have to trade‐off their emotional expressions for a wage, and hence, engagement in emotional labour is demanding. Other recent studies (Dursun, Bayram & Aytaç, 2011; Grandey, Diefendorff, & Rupp, 2013; Mróz & Kaleta, 2016) have also confirmed emotional labour as a process of regulating emotional expressions in order to meet organizational goals. In relation to nursing, nurses are expected to show positive emotions including empathy towards patients/clients even when patients/clients show negative emotions such as anger.

In the process of engaging in emotional labour, two main strategies have been identified: surface acting and deep acting (Hennig‐Thurau, Groth, Paul, & Gremler, 2006; Hochschild, 1983). Whereas surface acting requires employees to alter their external emotional expressions to reflect organizational standards or expectations (acting in bad faith), deep acting requires an experience of internal emotions congruence with external emotional expressions and organizational standards (acting in good faith) (Hennig‐Thurau et al., 2006; Hochschild, 1983). Among health professionals such as nurses and midwives, when these health professionals encounter negative emotional situations such as angry and/or annoying patients, nurses may suppress the negative emotions of getting angry at the patients and rather show positive emotions such as trying to calm patients down. The suppression of negative emotions while trying to show positive emotions reflects surface acting. On the contrary, nurses may not necessarily suppress negative emotions but rather show genuine positive emotions even when faced with negative emotions such as an anger patient and this reflects deep acting.

Even though the seminal work of Hochschild identified that emotional labour demands both physical and psychological resources and, hence, has negative consequences on work‐related outcomes including employee satisfaction and well‐being (cited in Wang, Wang, & Hou, 2016), recent studies have reported inconsistent results on the consequences of emotional labour on job‐related attitudes (Bhave & Glomb, 2013; Ghalandar, Jogh, Imani & Nia, 2012). Per this study, job attitudes refer to the evaluations of health professionals (nurses and midwives) about their job which reflect their feelings towards the job, the beliefs about the job, and the attachment they have towards the job (Judge & Kammeyer‐Mueller, 2012; Judge, Weiss, Kammeyer‐mueller, & Hulin, 2017) which are measured by job satisfaction and organizational commitment. The underlining reason to the inconsistent link between these variables is mostly attributed to the diverse approaches employed in studying emotional labour (Grandey et al., 2013).

Nevertheless, extant literatures are mostly pointing to negative associations between surface acting and job attitudes (Bhave & Glomb, 2013; Wang et al., 2016). The underlining reason to the negative association is because engaging in surface acting creates an inner tension which tends to have negative consequences on work‐related attitudes (Grandey, 2003). Conversely, since engaging in deep acting implies showing genuine positive emotions, the consequences tend to be positive on job attitudes (Grandey, 2003; Wong & Law, 2002) even though context has been identified to play an important role in some discrepancies (Wang et al., 2016).

In African and, especially, the Sub‐Saharan region, empirical studies on the link between emotional labour and job attitudes are almost non‐existent. Available studies on the subject matter in this context are at the nascent stage with much focus on qualitative approach to understanding emotional labour and its consequences on a myriad of issues including job attitudes (Davel, 2014; Msiska, Smith, & Fawcett, 2014). The present study sought to examine the relationship between emotional labour strategies (surface acting and deep acting) and job attitudes (organizational commitment and job satisfaction) by introducing perceived organizational support as a moderating variable. The uniqueness of this paper lies in the theoretical argument used in providing explanation to the missing link between the study variables, the sample used in the study (health professionals) and the setting/context for the study (Africa typically, Ghana). The researchers also introduce another variable (perceived organizational support) to provide further information on the possible relationship between emotional labour and job attitudes.

THEORETICAL FRAMEWORK

Job demand‐resources model

The relationship between emotional labour and job attitudes as well as the moderating role of perceived organizational support is explained using the job demand‐resource model. The model postulates that working environment is associated with both physical and psychological demands which tends to deplete the resources of employees. The extent of the work demands gives birth to negative outcomes including, but not limited to, negative employee well‐being, turnover intentions, less satisfaction and less commitment (Bakker, 2011; Bakker & Demerouti, 2007). The theory also acknowledges the existence of job resources including personal, organizational and social resources upon which employees tend to fall on in the course of handling the demanding nature of their job. The availability and the richness of these job resources in the presence of job demands give birth to positive job attitudes (Bakker & Demerouti, 2007, 2014).

REVIEW OF RELATED STUDIES

Emotional labour and job attitudes

Among nurses, Yang and Chang (2008)'s study revealed that surface acting had no significant association with job satisfaction but related significantly with organizational commitment. Conversely, deep acting had a significant relation with job satisfaction but no significant relationship with organizational commitment. Another study conducted by Bogdan, Mariean, Avram and Stan (2010) disclosed a strong positive association between deep acting and surface acting on the one hand, and all the dimensions of job burnout on the other hand suggesting emotional labour gives birth to negative work attitudes and outcomes. Consistent with these findings, Cheung and Cheung (2013) in their study among teachers pointed out that while emotional dissonance (synonymous to surface acting) related positively with all dimensions of job burnout, there was a negative relationship between emotional dissonance and organizational citizenship behaviour. Cossette and Hess (2015) also conducted a similar empirical by examining the linkage between three strategies of emotional labour (naturally felt emotions, emotional suppression and reappraisal) and work‐related attitudes (commitment, turnover intentions and satisfaction), using teachers as their sample. They discovered that emotional suppression is negatively associated with affective commitment and job satisfaction while a combination of the three emotional regulation strategies generated positive consequences on job satisfaction and affective commitment. It is therefore postulated that:

Surface acting will negatively predict job attitudes (job satisfaction and organizational commitment).

Deep acting will positively predict job attitudes.

Perceived organizational support as a moderator

Extant studies and theoretical arguments point to the significant role of organizational resources such as organizational support serving as protective factors against the negative consequences of work demands (Bakker & Demerouti, 2007; Van Droogenbroeck, Spruyt, & Vanroelen, 2014). However, few empirical studies on emotional labour, perceived organizational support and job attitudes in the same study are available (JJohnson, Kraft, & Papay, 2012; Skaalvik & Skaalvik, 2007; Yang, Ge, Hu, Chi, &Wang, 2009). The few available studies argue that perceived organizational support (if positive) weakens the negative consequences of surface acting on job attitudes (Skaalvik & Skaalvik, 2007; Yang et al., 2009). Conversely, perceived organizational support (if positive) strengthens the association between deep acting and job attitudes (Skaalvik & Skaalvik, 2007; Yang et al., 2009).

It is also postulated that “the relationship between emotional labour (surface acting and deep acting) and job attitudes (organizational commitment and job satisfaction) will be moderated by perceived organizational support.”

In line with the theoretical argument and the empirical studies reviewed, the present study generated a conceptual model which depicts the relationship between the study variables.

METHODOLOGY

Research design

Cross‐sectional design was used to collect data at one point in time (Creswell, 2014). The cross‐sectional design was considered as an appropriate design because compared to a longitudinal design, a cross‐sectional design enables researchers to collect a pool of data within limited time space and resources (Creswell, 2014).

Population, sample and sampling technique

Population for the study was nurses and midwives in Ghana which stand at forty‐two thousand (Ghana Health Service, 2015). With the aid of Krejcie and Morgan (1970) sample size determination formula, the sample for the study was estimated at 380. However, in the actual data collection, four hundred (400) questionnaires were distributed using quota and convenient sampling techniques out of which three hundred and forty‐two (342) were returned representing 85.5% response rate. The sample for the study was selected from six main health facilities within three different districts in the Greater Accra region of Ghana including three general hospitals, two polyclinics and one psychiatric hospital. The breakdown of the sample selected from each facility is presented as follows.

As indicated in Table 1, the size of workforce in each health facility served as the basis for estimating the sample selected from the health facility. That is, the population in all the health facilities is 720. Total workforce in the general hospitals is 370. As well, based on Krejcie and Morgan (1970) sample size determination of 380, the estimated sample to be selected from the general hospitals is generated. The same applies to the polyclinics and psychiatric hospitals.

Table 1

A summary of the sample size from the health facilities

Health facility	Quota sampling computation	Sample estimate selected	Actual sample retrieved
General hospitals	370/720 × 380	195	174
Polyclinics	200/720 × 380	106	103
Psychiatric hospital	150/720 × 380	79	65
Total		380	342

Instrumentation

Measures for the study included standardized instruments in assessing emotional labour, perceived organizational support, job satisfaction and organizational commitment. Whereas the 14‐item instrument on emotional labour by Brotheridge and Lee's (2003) was adopted to assess emotional labour (Cronbach's alpha between 0.71–0.90), perceived organizational support (POS) was assessed using the 8‐item POS scale developed by Eisenberger, Cummings, Armeli and Lynch (1997). Cronbach's alpha for POS reported to be 0.93 (Worley, 2006). A 7‐item overall job satisfaction scale (Cronbach's alpha = 0.82) developed by Taylor and Bowers (1974) was also used to assess respondents' level of job satisfaction, and organizational commitment (Cronbach's alpha = 0.93) was assessed with the aid of a 9‐item scale developed by Mowday, Porter, and Steers (1982). While emotional labour and job satisfaction were measured on a 5‐point Likert scale, POS and organizational commitment were on a 7‐point Likert scale as developed by the original authors.

Data collection procedure

The relevant ethical approvals to conduct the study were granted by the Ghana Health Service and University of Ghana Business School. The purpose of the study was explained to all respondents prior to administering the questionnaires, and other ethical principles including anonymity and confidentiality were also assured. In all, four hundred questionnaires were distributed (using quota and convenient sampling techniques) to these health professionals in the six health facilities used for the study. Questionnaires were left in the hands of the respondents to respond to them at their own time of convenience. After completion of the questionnaires, they were returned to the researchers who inquired about how long it took respondents to complete the questionnaires. Responding to the questionnaires took between 10–15 min, and most questionnaires were retrieved on the same day of administering. Data collection from the various health facilities lasted for six weeks after which three hundred forty‐two (342) questionnaires were returned out of the four hundred administered questionnaires. Collected data were then entered in SPSS version 21 after which data were analysed.

Inclusion and exclusion criteria

For the purpose of the study, only staff nurses and midwives who have been working with the health facilities for not <6 months and employed by the Ghana Health Service were included. This group of nurses and midwives were chosen for the study because it was assumed that job attitudes (job satisfaction and organizational commitment) are built over time and more affiliated to permanent workers. In line with this, student nurses, rotation nurses and nurses with less 6 months working experience with the health facilities were exempted from the study.

RESULTS

Statistical tests used in testing the relationship between the study variables were the Pearson r test and the hierarchical multiple regression. Preliminary analysis with the use of descriptive statistics such as means, standard deviation (SD), skewness and kurtosis was conducted to test assumptions underlining the use of regression. Details of the result from the preliminary statistical test are presented in Table 2 below.

Table 2

A summary of the descriptive statistics for the study variables

	N	Minimum	Maximum	Mean	SD	Skewness	Kurtosis
Deep acting	342	3.00	15.00	9.51	2.81	−0.26	−0.44
Surface acting	342	3.00	15.00	7.26	2.78	0.44	−0.13
POS	342	6.00	48.00	29.57	9.18	0.02	−0.40
Job satisfaction	342	11.00	35.00	25.20	4.61	−0.35	0.21
Organizational commitment	342	15.00	63.00	44.40	11.67	−0.40	−0.62

Abbreviation: POS: perceived organizational commitment.

As shown in Table 2, skewness and kurtosis values for each of the study variables were between −2–2 as recommended by George and Mallery (2010) indicating that the data variables were normally distributed.

As well, analysis of the demographic characteristics of the respondents was conducted and the result is presented in Table 3 below.

Table 3

A summary of the demographic characteristics of the research respondents

Variable	Frequency (N)	Percentage (%)
Gender
Male	57	16.7
Female	285	83.3
Agea
20–30 years	185	55.2
31–40 years	125	37.3
41–50 years	14	4.2
51–60 years	11	3.3
Marital statusa
Single	146	43.6
Married	188	56.1
Divorced	1	0.3
Education
SSCE/O' level/A' level	12	3.6
Diploma	191	55.8
Certificate	20	5.8
Bachelor's degree	83	24.3
Master's degree	16	4.7
Others	20	5.8
Cadres of nursinga
General nurse	109	33.3
Midwifery	71	21.7
Public health nurse	21	6.4
Community nurse	17	5.2
Psychiatric nurse	60	16.8
Nurse assistant	43	13.1
Others	12	3.4
Tenure of worka
1–5 years	249	75
6–10 years	59	17.7
11–15 years	15	4.5
16 years and above	9	2.7
Religiona
Christian	301	91.5
Islam	28	8.5

Missing data.

Table 3 indicates a summary of the demographic characteristics of the research respondents for the study which depicts the gender, age, marital status, educational background, religion, tenure of work and cadres of nurses used. Table 3 shows that majority of the respondents (83.3%) are females indicating that nursing and midwifery profession is dominated by females. Table 3 also indicates that more than 50% of the respondents were between 20–30 years indicating the youth nature of the respondents. Similarly, 56% of the respondents were married, 55.8% were diploma holders, majority of the respondents (33.3%) were general nurses, 75% of the respondents had worked with the organization between 1–5 years, and 91.5% were Christians.

A further statistical test was conducted testing the relationships between the study variables as presented in the correlation matrix (see Table 3). From the correlation matrix presented in Table 3, the highest relationship was witnessed between job satisfaction and organizational commitment [r = 0.62, p < 0.01] whereas the least relationship was observed between gender and perceived organizational support [r = −0.013, p > 0.05]. The table also reveals that while perceived organizational support related positively with both job satisfaction and organizational commitment [r = 0.523; 0.46; p < 0.05, respectively], surface acting related negatively with job satisfaction and organizational commitment [r = −0.22; −0.18; p < 0.05, respectively]. However, deep acting did not relate significantly with job satisfaction [r = 0.077, p > 0.05] but related significantly with organizational commitment [r = 0.12, p < 0.05]. Among the demographic characteristics, education negatively related with job satisfaction [r = −0.13, p < 0.05] and organizational commitment [r = −0.15, p < 0.05] and cadres of nursing also related significantly with job satisfaction [r = 0.093, p < 0.05]. Hence, cadres of nursing and education were treated as control variables in the hierarchical multiple regression analysis.

To test the precise relationships between emotional labour (deep acting and surface acting), perceived organizational support and job satisfaction, education and cadres of nursing were treated as controlled variables as seen in step 1. The two variables collectively explained 3% of the changes on job satisfaction [R 2 = 0.03, p < 0.05]. However, only education was the significant predictor of job satisfaction [β = −0.13, p < 0.05]. In step 2, Table 4 reveals that deep acting, surface acting and perceived organizational support collectively explained 27% of the changes on job satisfaction [R 2 change = 0.27, p < 0.01]. Whereas surface acting [β = −0.11, p < 0.05] and perceived organizational support (POS) [β = 0.49, p < 0.05] related significantly with job satisfaction, deep acting did not predict job satisfaction [β = 0.05, p > 0.05]. In step 3, the moderating role of POS was tested. The analysis shows that whereas POS significantly moderated the relationship between surface acting and job satisfaction [β = −0.13, p < 0.05], POS did not moderate the relationship between deep acting and job satisfaction [β = 0.08, p > 0.05] (Tables 5 and 6).

Table 4

A summary of the correlation matrix between the study variables

	1	2	3	4	5	6	7	8	9
1. Job satisfaction	1.000
2. Gender	0.05	1.000
3. Age	0.02	0.12**	1.000
4. Education	−0.13**	−0.15**	0.04	1.000
5. Cadres of nursing	0.09*	−0.09*	−0.13**	0.01	1.000
6. Deep acting	0.077	0.11*	0.10*	0.03	0.05	1.000
7. Surface acting	−0.22**	−0.10*	−0.18**	0.05	−0.00	0.08	1.000
8. Perceived organizational support	0.52**	−0.013	0.13**	−0.08	0.04	0.08	−0.23**	1.000
9. Organizational commitment	0.62**	0.08	0.08	−0.15	−0.02	0.12	−0.18**	0.46**	1.00

Abbreviation: POS: perceived organizational commitment.

*p < 0.05. **p < 0.01.

Table 5

A summary of the hierarchical multiple regression showing the predictors of job satisfaction

	β	t	p	R 2 change	F change
Step 1				0.03*	2.13
Education	−0.13*	−2.16	0.03
Cadres of nursing	0.10	1.74	0.08
Step 2				0.27**	37.50
Deep acting	0.05	0.96	0.34
Surface acting	−0.11	−2.19	0.03
POS	0.49	9.55	0.00
Step 3				0.03**	4.90
Deep × POS	0.08	1.60	0.11
Surface × POS	−0.13	−2.59	0.01

Abbreviation: POS: perceived organizational commitment.

*p < 0.05. **p < 0.01.

Table 6

A summary of the hierarchical multiple regression showing the predictors of organizational commitment

	β	t	p	R 2 change	F change
Step 1				0.02	2.22
Education	−0.14	−2.39	0.02
Cadres of nursing	−0.00	−0.07	0.95
Step 2				0.22**	26.36
Deep acting	0.09	1.72	0.09
Surface acting	−0.08	−1.51	0.13
POS	0.43	7.96	0.00
Step 3				0.02**	8.56
Deep × POS	0.10	1.97	0.05
Surface × POS	−0.19	−3.52	0.00

Abbreviation: POS: perceived organizational commitment.

p < 0.01.

To testing the predictors of organizational commitment, education and cadres of nursing were treated as control variables. In step 1, it is revealed that education and cadres of nursing collectively explained 2% of the changes in organizational commitment [R 2 = 0.02, p > 0.05] but only education independently predicted organizational commitment [β = −0.14, p < 0.05]. In step 2, deep acting, surface acting and perceived organizational support (POS) jointly contributed 22% of the changes in organizational commitment [R 2 change = 0.22, p < 0.01]. However, POS was the only significant predictor of organizational commitment [β = 0.43, p < 0.01]. In testing the moderating effect of POS (step 3), the result demonstrates that POS moderated the relationship between deep acting and organizational commitment [β = 0.10, p = 0.05] as well as surface acting and organizational commitment [β = −0.19, p < 0.01].

DISCUSSION

The significant negative relationship between surface acting and job satisfaction as revealed from this study is attributed to the fact that engaging in surface acting (acting in bad faith) creates an inner tension in employees which emanates from employees only altering their outer feeling but not necessarily changing their inner feelings. This inner tension is what tend to have negative bearing on the level of job satisfaction among the nurses and midwives (Grandey, 2003). This finding is not far from the reach of empirical studies; it corroborates extant literature (see Cossette & Hess, 2015; Wang et al., 2016). Thus, when nurses and midwives suppress negative emotions in order to show positive emotions, the inconsistency between the suppressed emotions and the emotions they show tends to have negative consequences on their commitment to the organization and their satisfaction to the job. Hence, nurses and midwives need to be equipped to less engage in surface acting since this has been found to have negative consequences on their job attitudes.

Nurses and midwives also tend to fall on their organizational resources (perceived organizational support) to enable them to manage the emotional requirements of their profession. This organizational resource (if positive) serves as a protective factor against the negative consequences of these emotional requirements. It is not surprising that the present study disclosed the significant moderating effect of POS on the relationship between surface acting on the one hand and job satisfaction and organizational commitment on the other hand. This finding is consistent with other empirical studies conducted by Hur, Won Moon and Jun (2013) as well as Hur et al. (2015). Conversely, whereas the present study found a significant moderating effect of POS on the relationship between deep acting and organizational commitment, POS did not moderate the relationship between deep acting and job satisfaction. This finding presupposes that employees' engagement in deep acting will have significant bearing on the level of employee commitment to their organization, and they fall on the level of their organization's support; when they see their organization to be more supporting, the more likely that acting in good faith will have consequences on their organizational commitment. This finding is also consistent with the empirical studies by Hur et al. (2013) as well as Han et al. (2015) who pointed to the moderating effect of POS in the relationship between deep acting and organizational commitment. The finding suggests that nurses and midwives seek support from their organization in the process of managing the emotional demands of the profession. As such, when nurses and midwives see their organizations to be more supporting, their emotional regulations will have positive consequences on their job attitudes particularly commitment to the organization.

The insignificant relationship between deep acting on the one hand, job satisfaction and organizational commitment on the other hand contradicts empirical studies (see Cossette & Hess, 2015; Yang & Chang, 2008), and other empirical studies have disclosed inconsistencies on the link between emotional labour and attitudes (see Bhave & Glomb, 2013; Grandey, 2000; Kim, 2008). The present study's finding adds to the inconsistencies in the extant literature on the association between emotional labour and job attitudes. These inconsistencies are mostly attributed to the approaches used in studying these variables but the present study revealed POS as a potential missing link.

LIMITATIONS OF THE STUDY

The findings from the current study, like many other studies, need to be explained within context due to limitations surrounding the research design employed. The study employed a cross‐sectional design which is correlational in nature, and therefore, establishing cause‐and‐effect relationships between the study variables is impossible. Caution should also be taken when generalizing the findings of the present study to health professionals since the present study made use of nurses and midwives which are just a fraction of the health professionals' category. Furtherance, the study employed a quota and convenient sampling techniques which affect the generalizability of the study. Nevertheless, this study is one of a kind in the Ghanaian and Sub‐Saharan African context and, therefore, opens the door for further empirical studies including longitudinal studies to be conducted in this context.

IMPLICATIONS

Findings from the study also have both theoretical and practical implications. To the job demand‐resources model, the significant moderating effect of POS in the relationship between emotional labour and job attitudes indicates the relevance of organizational resources in mitigating the negative consequences of emotional demands on job attitudes. That is to say, when employees perceive their organizations to be more supporting, the negative consequences of the emotional demands on their job attitudes are buffered. Hence, organizations, especially human resource managers, should be more willing to initiate and implement policies that elicit concerns for employee well‐being and valuing their contributions. Similarly, nursing education and practice should also pay attention to equipping students and professionals in terms of how best to handle the emotional demands of the profession as this has been found to have bearing on their job attitudes. Continuous Professional Education (CPEs) programmes that are organized generally for health professionals including nurses should have modules on how to handle emotional labour and regulate their emotions. Nurses and midwives overall emotional regulation skills have consequences on job attitudes, and this study has showed the impact and need for emotional regulation in professional health practice.

CONFLICT OF INTEREST

The authors declare no conflict of interest in publishing this study.

AUTHOR CONTRIBUTIONS

The first author was responsible for the design of the study and data collection. All other authors were involved in the analysis of the data as well as developing the manuscript. All authors read through and approved the final manuscript.

PATIENT CONSENT

The study did not involve patients but rather nurses and midwives. Respondents' content for participation was sought.

ETHICAL STATEMENTS

Ethical approval to conduct the study was sought from the University of Ghana Business School, Department of Organisation and Human Resource Management as well as the Ghana Health Service.