Emotional Intelligence and Teaching Satisfaction: The Mediating Role of Emotional Labor Strategies

The study examines the direct effect of four “emotional intelligence” attributes on teachers’ job satisfaction in Karachi’s private teaching institutions. The study also investigates the mediating effects of “emotional labor strategies” on teachers’ job satisfaction. We have used the questionnaire adopted from earlier studies. We distributed 550 questionnaires to respondents, of which we received 499 useable responses. The study has used Smart PLS version 3.3 for data analysis. Our results support only six hypotheses, including two direct and our indirect. This study has contributed to the body of knowledge in the following ways. First, it has measured the effects of the four attributes of emotional intelligence on job satisfaction. Second, most studies have examined the mediating effect of emotional labor strategies on emotional intelligence and other job satisfaction antecedents. Perhaps this is the first study that has examined the direct impact of the sub-factor of emotional intelligence on teachers’ job satisfaction. Additionally, it also looks at the mediating effect of emotional labor strategies on teachers’ satisfaction. There are several implications for managers. For example, the teaching institutes should provide counseling and training to teachers for enhancing their emotional intelligence. Emotional labor strategies help individuals control and monitor their emotions; therefore, educational institutions may also encourage their teachers to adopt these strategies.


Introduction
Educational institutes in many countries have not only adopted new technology but have also implemented various educational reforms. Despite all these measures, they still face certain challenges specifically related to teachers' satisfaction (Ignat & Clipa, 2012). These challenges are related to enhancing teachers' job-related performance, improving student attitudes towards learning, and balancing the workloads of teachers (Mérida-López, Extremera & Rey, 2017). Social and other job-related stress stimulate emotional stress and emotional exhaustion. However, teachers with strong emotional competencies can cope with stress (Li, Pérez-Díaz, Mao & Petrides, 2018). Many researchers have suggested a need to examine teachers' emotions and their effect on classroom learning, students' motivation, and teachers' job satisfaction (Ignat & Clipa, 2012). Nafukho (2009) argue that success in interpersonal relations and careers depends on how individuals learn to manage their emotions.
There is an abundance of studies on the effects of emotional intelligence (EI) on jobrelated antecedents. However, existing literature does not provide much evidence on the impact of emotional intelligence attributes on teachers' job satisfaction (JS). Perhaps no study is available that has examined the effects of antecedents of emotional intelligence on job satisfaction (JS). Given this gap, we have considered the impact of (OEA, SEA, ROE, UOE) on job satisfaction. Additionally, we have looked into the mediating effects of emotional labor attributes (i.e., " DSA, SA, and ENFE) on teachers' job satisfaction.

Literature Review Teacher's Job Satisfaction
Job satisfaction (JS) in general and teachers' job satisfaction, in particular, has been a problematic issue for decades. Its severity is more profound in developing countries where the compensation is lower than in other professions (Anastasiou, 2020). On the one hand, many teachers are constantly pursuing new employment opportunities due to various unfavorable conditions. On the other hand, the new generation prefers other professions over teaching (Eraldemir-Tuyan, 2019;Asforth et al., 1993). Eraldemir-Tuyan (2019) argues that teachers feel that modern society does not give due recognition to the teaching profession. Additionally, teachers' compensation has not increased significantly, while accountability, stress, and other job-related demands have increased considerably (Anastasiou, 2020;Wharton, 2009). Consequently, this disparity between job requirements and compensation of teachers has led to low job satisfaction. Teachers' motivation for joining the teaching profession is to provide intrinsic rewards and emotional benefits (Sahito & Vaisanen, 2020). Jones et al. (2002) argues that many teachers opted for the career because they feel that by imparting education to the future generation, they can contribute to society's development and progress. The research found that teachers who are changing their profession fall into two categories, which are beginners (that have worked up to five years) and veterans (that have worked more than 30 years in this profession) (Platsidou, 2010;Goleman, 1995).

Emotional Intelligence (EI)
EI has a close association with job satisfaction, organizational performance, and job success (Serrat, 2017). EI helps individuals manage job requirements and stress due to which they are more successful than others (Mattingly & Kraiger, 2019). Similarly, Miao, Humphrey & Qian (2017) also stress that individuals with high EI often are more successful at their jobs, as they are well equipped to use emotional knowledge to resolve personal and job-related issues (Miao, Humphrey & Qian, 2017). Serrat (2017) argues that besides the IQ level, EI is a critical precursor to JS and job success. EI has stemmed from social intelligence, which, according to Matttingly and Kragier (2019), enables individuals to manage others wisely and maintain sustainable human relations. The two facets of social intelligence are intrapersonal and interpersonal (Miao, Humphrey & Qian, 2017). Interpersonal knowledge enables individuals to interact with others effectively. As a result, such individuals earn the respect and cooperation of others (Mayer, Caruso & Salovey, 2016). Intrapersonal intelligence helps individuals judge their ability rationally. It also helps resolve personal, social, and job-related problems (Petrides et al. 2016). EI includes some important facets of both IQ and social intelligence.

Conceptual Framework
We have developed a new model in Figure 1 and have discussed the theoretical justification for the proposed hypotheses after the conceptual framework.

Emotional Intelligence (EI) and Job Satisfaction (JS)
Past studies have inconsistent results on the association between EI and teachers' JS. For example, Anari (2012) and Wong et al. (2010) found a positive association between teachers' EI and JS, while Platsidou (2010) found an insignificant association between teacher's satisfaction and EI. Goleman (1998) used the EI theory for understanding the association between EI and JS in several business domains. They concluded that the use of EI is not consistent in all industries but varies from one business sector to another. Also, emotionally intelligent individuals are more successful at work and society (Li, Pérez-Díaz, Mao & Petrides, 2018).
Emotional intelligence comprises of "Self Emotional Appraisal (SEA), Other Emotional Appraisal (OEA), Regulations of Emotions (ROE), and Use of Emotions (UOE)" (Mayer et al., 1990). SEA helps individuals to understand, appraise, and express their sentiments naturally. Consequently, these qualities enable individuals to improve interpersonal relationships (Miao, Humphrey & Qian, 2016;Wong & Law, 2002). OEA allows individuals to assess the emotions of others effectively. Therefore, they are more considerate and empathic to others (Toprak & Savaş, 2020;Wong & Law, 2002). ROE is a control mechanism of feelings and emotions. Individuals with this ability are capable of monitoring their emotions and sentiments. Additionally, such individuals can recover rapidly from emotionally stressed situations (Wong & Law, 2002). UOE helps individuals to use their feelings for enhancing job and non-job related performance. Singh & Kumar (2016) suggest that SEA helps individuals to appraise and control their emotions. At the same time, OEA enables individuals to judge the sentiments of friends and colleagues rationally. Therefore, such individuals are more satisfied with their jobs (Wen, Huang & Hou, 2019). Wen, Huang & Hou (2019) argue that emotionally intelligent teachers have full command of DSA and quickly adapt their sentiments to meet students' expectations.
Thus, due to superior emotion control mechanisms, emotionally intelligent teachers create an environment in a class where students feel comfortable and participate in the learning process. Consequently, this leads to students' achievements and teachers' satisfaction (Latif, Majoka & Khan, 2017). Teachers with high ROE have more control over their emotions due to which they promote positive emotions and sentiments in a class. Additionally, such teachers protect students from experiencing negative emotions, such as anger and fear. As a result, students remain focused on their studies and achieve better grades. Toprak & Savaş (2020) argue that teachers with a high level of ROE do not adopt emotional suppression strategies such as SA and DSA. Instead, they assume "cognitive appraisal, " which many researchers believe is an efficient approach for expressing the emotions expected by others. UOE helps teachers to respond to students with controlled emotions that promote an interactive environment in a class. As a result, both students and their teachers benefit. That is, teachers, benefit from a higher satisfaction level, and students benefit through better academic achievements (Miao Humphrey & Qian, 2016). Ho and Au (2006) and Weiss (2002) suggest that students' academic achievements stimulate teachers' pleasant emotions, which results in positive job satisfaction. A teacher's satisfaction level has a direct association with the fondness of job. It also motivates teachers to create an environment of social interactions, discussions, and debates (Hirschfeld, 2000;Yin et al., 2013). As previously discussed, we did not find a single study that has examined the impact of sub-factors of emotional intelligence on job satisfaction. Given this gap, we have proposed the following hypotheses: H1a: Other emotional appraisals (OEA) and teachers' satisfaction are positively associated.
H1b: Regulation of emotions (ROE) and teachers' satisfaction are positively associated. H1c: Self emotion appraisal (SEA) and teachers' satisfaction are positively associated.
H1d: Use of emotion (UOE) and teachers' satisfaction are positively associated.

Mediating Role of Surface Acting (SA)
All the facets of emotional labor, including SA, directly and indirectly, impact teachers' JS (Grandey et al., 2013). Individuals use SA based on their built-in capabilities and the requirement of situations. Qi, Ji, Zhang, Lu, Sluiter, and Deng (2017) argue that SA usage is not consistent in all domains and industries. It is generally high in businesses where personal and social interactions with employees are high (Winograd, 2005). On the other hand, it is low in sectors where social interaction with coworkers is minimal. SA is a phenomenon where an individual reacts to others' aggressive behavior, suppresses his/her natural emotions, and fakes a positive emotional expression (Winograd, 2003).
Thus, SA is sometimes important for maintaining a sustainable social interaction environment in an organization. Although individuals with high SA change their outer emotional feelings and expressions, their internal personal feelings remain intact. Continued SA may not only adversely affect individuals' wellbeing, but it may also negatively affects their attitude towards the job (Lee, Pekrun, Taxer, Schutz, Vogl & Xie, 2016). Many teachers, despite the aggressive behavior of management and students, display pleasant emotions. However, this does mean that these teachers are satisfied with the organizational environment (Asrar-ul-Haq, Anwar & Hassan, 2017;Hayes, 2003).
The literature suggests inconsistent results on the relationship between SA and EI. A few studies found they both are negatively associated, while others found insignificant links between SA and EI (Austin et al., 2008;Mikolajczak et al., 2007). These studies also concluded that individuals with high EI orientation have a low inclination towards SA and vice versa. Similarly, we found inconsistent results in the literature on the association between SA and JS. For example, some studies suggest that SA negatively stimulates JS (Beal, Trougakos, Weiss, and Green, 2006;Brotheridge & Lee, 2002;Grandey, 2003), while other studies stress that SA and JS have an insignificant association (Cheung et al. 2011;Hargreaves, 1998).
Given the inconsistent results, there is a need to incorporate a mediator that may bring more insight into EI components' relationships. Given this background, we have formulated the following hypotheses: H2c: Surface acting (SE) mediates the self-emotional appraisal (SEA) and job satisfaction (JS) relationship.

Mediating Effect of Deep Surface Acting (DSA)
On many occasions, teachers, despite having negative emotions, display a positive attitude to others. This behavior is known as deep surface acting (DSA). In DSA, individuals show different emotions, but the real sentiments do not change (Schirmer & Adolphs, 2017). EI levels vary from one individual to another. Individuals with high EI are better equipped to cope with the job induced stress. Therefore, they generally do not adopt DSA (Lee, Pekrun, Taxer, Schutz, Vogl, & Xie, 2016). Since DSA exhibits the emotional reaction that other people anticipate, many studies suggest a strong association between SA and EI (Xanthopoulou, Bakker, Oerlemans & Koszucka, 2018).
Past studies found inconsistent results on the association of DSA and EI. For example, Karim and Weisz (2011) and Liu et al. (2008) cited that emotionally intelligent teachers often resort to DSA. Therefore, they concluded that DSA and EI have a positive association. On the contrary, Mikolajczak et al. (2007) suggest a negative association between EI and DSA. Similarly, previous research has also examined the association between DSA and JS and found conflicting results. For example, Brotheridge and Lee (2002) and Grandey (2003) found a positive association between DSA and JS. Contrarily, others have found an insignificant association between deep DSA and JS (Cheung, Tang, & Tang, 2011;Mayer et al., 1990). Given the conflicting findings, we have formulated the following hypotheses:

Mediating effect of Expression of Naturally Felt Emotions (ENFE)
The expression of naturally felt emotions (ENFE) is the third kind of emotional labor (EL) (Mikolajczak et al., 2007). In this case, individuals express their true emotions, unlike SA. Past studies have found conflicting and heterogeneous EI outcomes. Austin et al. (2008) found a positive association between EI and EL, while Mikolajczak et al. (2007) concluded that these two variables have an insignificant association.
Teachers, due to emotional labor (EL) suppress their feeling and sentiments, which adversely affects their job-related outcomes (Austin et al., 2008;Karakucs, 2013). In contrast, a few studies suggest that when individuals express their true emotions, they are less stressed, due to which they develop positive attitudes towards personal and job-related outcomes (Serrat, 2017;Mattingly & Kraiger, 2019;Lee & Ok, 2012). Given the conflicting findings, we have formulated the following hypotheses: H4a: Expression of naturally felt emotions (ENFE) mediates the self-emotional appraisal (SEA) and job satisfaction (JS) relationship.

Methodology Population and Sample
The research population of the study comprises of faculty members working in private teaching institutions of Karachi. From this population, the authors collected data from five leading business schools. The authors personally visited the selected universities and distributed 550 questionnaires. Of this total, we received 499 complete and useable responses. The profile of the respondents is presented in Table 1.

Scales and Measures
The questionnaire we have used in the study has 38 items. Of this total, 4 questions are related to demographics, based on a nominal scale. As many as 34 items are based on a rating scale of 1 to 5. The summary of the questionnaire used in the study is presented in Table 2.

Data Analysis
We have used the Smart PLS software (version 3.3) for statistical analysis, considered useful for estimating complex models (Henseler et al., 2014). Partial least squares (PLS) is a technique that links latent and indicator variables. The questionnaire used in the study has three latent variables (with seven factors) and 34 indicator variables. The reliability analysis was based on Cronbach's Alpha values, which should be greater than 0.6 (Tabachnick & Fidell, 2007). Convergent validity was examined based on composite reliability and AVE (Refer to Table 3). We have used the Fornell & Larcker (1981) criterion, cross-loadings and the Heterotrait-Monotrait (HTMT) ratio for discriminant validity.

Descriptive Analysis
For descriptive analysis, we have analyzed convergent validity, reliability, crossloadings of items, and constructs in Table 3. The summary of the results suggests that the Cronbach's Alpha value is the highest for surface acting (SA) (α=0.854), and the lowest is for expression of naturally felt emotions (ENFE) (α=0.853). Thus, we have inferred that the constructs have internal consistency (Hair et al., 2014). All the items' factor loadings are as high as 0.853 and as low as 0.653 and are statistically significant. Additionally, "the AVE value is greater than 0.60, and composite reliability values are also greater than 0.70. " Thus, we have inferred that the data fulfills convergent validity requirements (Hair et al., 2014).

Discriminant Validity
We have ascertained the discriminant validity of the constructs based on two criteria, i.e. (1) on Fornell & Larcker (1981) and (2) cross-loading. These approaches have been discussed in the following sections:

Discriminant Validity using Fornell & Larcker (1981) Criteria
The first criteria we have used to assess discriminant validity is of Fornell & Larcker (1981). It compares the values of the square root of AVE with the Pearson correlation values. We have depicted a summary of the results in Table 4.  The results show that the highest Pearson correlation value (R=0.428) is for the pair SA and deep surface acting (DSA). The lowest Pearson correlation value (R=0.006) is for the pair surface acting (SA) and ENFE. The lowest value for AVE's square root is for UOE (0.719), and the highest value is for SEA (0.805). Since the square root of AVE is greater than the values of Pearson correlation, therefore the results fulfill the first criteria of discriminant validity (Brienam & Friedman, 1985).

Discriminant Validity Based on Cross Loadings
The second criteria we have used for examining the discriminant validity is loading and cross-loading. The summary of the results are shown in Table 5.

Direct Effects
We have proposed four direct hypotheses which we tested through Smart PLS. The summary of the results is depicted in Table 6. of emotions has a positive effect on teachers satisfaction, and use of emotions has a positive impact on teacher satisfaction. "

Mediating Effects
We have proposed 12 mediating relationships. These are "the mediating effects of deep surface acting (DSA), surface acting (SA), and expression of naturally felt emotions (ENFE) on teachers' satisfaction. " A summary of the results is depicted in Table 7. The results suggest that of the 12 mediating relationships, six were accepted and the other six were rejected.

Discussion and Conclusion
This study examines the direct effects of emotional intelligence constructs (i.e., OEA, ROE, SEA, UOE) on TS. It also looks at the mediating effect of SA, DS, and ENFE on TS. Our results supported only six of the 12 hypotheses, including two direct and four mediating (Refer to Tables 7 and 8) The literature suggests that emotional intelligence elevates teachers' satisfaction level, enhancing their behavior and attitude towards work. Consequently, teachers feel happy, and their wellbeing improves significantly (Bar-On, 2010; Jones, et. al., 2002;Hochschild, 1983 emotions are often more satisfied. Mayer, Caruso & Salovey (2016) suggest that teachers with UOE can direct their emotional stress productively. Consequently, such teachers create an environment that motivates students towards learning and achievements (Mayer, Caruso & Salovey, 2016;Hamachek, 2000).
Teachers who can control their own emotions and appraise others' feelings are considered emotionally intelligent (Johnson & Spector, 2007;Mayer et al., 2004). Contrary to our results, the literature suggests that teachers with SEA and UOE have higher satisfaction levels towards their jobs (Joseph et all., 2010). DSA enables teachers to monitor and control their emotional feelings due to which they are more productive and conducive to the work environment (Yin et al., 2013;Hostani et al., 2011). Moreover, SA enhances the association between ENFE and teachers' JS. The literature suggests that teachers in higher educational institutions with EI can adopt different strategies to manage difficult situations. Also, teachers with low EI cannot develop positive psychological feelings due to which their satisfaction is low (Grandey, 2000).

Practical Implications
This study has implications for the management of higher education institutions. The results suggest that emotional intelligence is a critical asset. Teachers who can use emotional intelligence adequately are capable of making rational decisions in overstressed situations (Intrator, 2006;Jones et al., 2002). Emotional intelligence is a naturally gifted trait, but institutions, through counseling and training, can increase the personal intelligence level of their employees.
Thus, the management of universities should primarily focus on enhancing this capability through well developed and structured training programs. These training programs may help teachers improve their expertise and skills of comprehending, controlling, and monitoring their feelings. Moreover, these training programs would help teachers build emotional associations, refine their cognizance, and upgrade their regulation capability. In addition to that, universities should counsel the teacher on the importance of learning and utilizing emotional labor strategies (i.e., SA, DSA, and ENFE) favorably.

Limitations and Future Research
This study has some constraints and provides directions for future research-the sample for the study consist of permanent and adjunct faculty of private teaching institutes of Karachi. Permanent and adjunct faculty members' emotional intelligence and satisfaction level may not be the same. Future studies may explore the difference in the attitude of permanent and adjunct faculty towards job satisfaction. Since this study's scope was towards one city, i.e., Karachi, other researchers can extend the developed conceptual framework to other cities and industries. We have examined the indirect effect of emotional labor strategies (i.e,. SA, DSA, and ENFE). Future studies can examine the mediating effects of other antecedents of job satisfaction. The demographic and cultural aspects were beyond the scope of this study. However, future academicians may consider these aspects in their studies.

College of Management Sciences
Volume 15, Issue 2 December 2020 I work at developing the feelings inside of me that I need to show to students or their parents

Expression of Naturally Felt Emotions (ENFE)
The emotions I express to students or their parents are genuine The emotions I show students or their parents come naturally The emotions I show students or their parents match what I spontaneously feel Teacher Satisfaction Scale (TS) In most ways, being a teacher is close to my ideal My conditions of being a teacher are excellent. I am satisfied with being a teacher. So far I have gotten the important things I want to be a teacher If I could choose my career over, I would change almost nothing