I. INTRODUCTION

Research has linked burnout to the quality of work done and the productivity among professionals, including teachers. Findings have also linked the high demand for high-quality teaching and good performance to emotional exhaustion, which is one of the three dimensions of burnout (Khan et al., 2014). Emotional exhaustion is a state of low energy and feelings of inadequacy experienced by an individual, setting the ground for eventual total burnout (Maslach, Schaufeli & Leiter, 2001; Muasya, 2016). The workload for teachers refers to all the duties officially allocated to teachers in school and which they are obligated to do. Shafie, Kadir, and Asimiran (2017) and Van-Droogenbroeck, Spruyt, and Vanroelen (2014) describe workload for teachers as the teaching and non-teaching responsibilities assigned to a teacher by the school and which the teacher must perform.

In Kenya, the teacher workload is not limited to actual classroom instruction. It comprises various curriculum activities that include; documentation of lesson plans, records of work, schemes of work, lesson notes, and students' records, preparation of the appropriate teaching and learning resources, actual classroom instruction and presentation, assessment and feedback given to learners on their academic progress (Cush, Pete, Ofafa & Otor, 2014; Teachers Service Commission, 2012, 2015). Other non-teaching responsibilities include: managing students' psychosocial welfare issues, administrative duties, and the supervision of co-curricular activities.

However, teaching workload is core and is done by all the teachers regardless of the other responsibilities (Teachers Service Commission, 2012, 2015).

Preparing for teaching duties and other activities demands that a lot of time and energy, emotionally, mentally, and physically, be dedicated to accomplishing them, thus subjecting the teachers to constant work overload and increased pressure to perform effectively. These tasks require many working hours and energy in preparation and, in turn, are likely to exert pressure on the person's emotional, mental, and physical strength. As a result, this pressure build-up might eventually lead to emotional exhaustion and loss of personal attachment to the work they could have previously enjoyed doing, resulting in burnout. It has been established that engaging in more than one responsibility is overloading (Afzal, Idrees, Fardous & Ambreen, 2019). In addition, Marina (2012), while expounding on the concept of workload, argued that more workload might occur when a person's work changes from a single item to multiple tasks. Furthermore, assigned workload demands that exceed the capacity to manage may lead to negative emotions because the individuals may feel stretched beyond their expectations and strength (Wang, Zheng, Hu & Zheng, 2014; Zheltoukhova, O'Dea & Bevan (2012).

Studies have established that workload is one of the noteworthy causes of burnout among teachers (Ali & Farooqi, 2014; El Helou, Nabhani, & Bahous, 2016). El Helou, Nabhani, and Bahous (2016) affirmed that workload was one of the major causes of emotional exhaustion. In the same strength, Ali & Farooqi (2014) conducted a study among the teaching staff at Gujranwala University to establish how workload affects work performance utilizing a sample size of 207 participants. The study findings revealed workload, when more than the individual's ability to manage, leads to one's inability to relax. As a result, the person suffers emotional stress, leading to job burnout. These results may have been a result of work overload. The outcome suggests that perceived or actual disparity between the assigned workload and the teachers' abilities to meet the work expectations may cause burnout. Moreover, in their study, Van Droogbroeck et al. (2014) affirmed that teaching and non-teaching workloads were strongly and negatively related to the three dimensions of burnout among teachers above 45. However, while the findings of these studies agree that workload relates to burnout positively, there is no clarification on how specific workload variables relate to burnout.

Workload influences burnout in various ways (Afshar & Doosti, 2016; Anap, Iyer & Rao, 2017; Froese-Germain, 2014; Van Bogaert, Kowalski, Weeks, & Clarke, 2013. Afshar and Doosti (2016) argued that workload problems need to be controlled or addressed. If not, they might have severe consequences on one's personal, physical, psychological, social, and emotional health and intrapersonal and interpersonal relationships. On their part, Anap, Iyer, and Rao (2017) opined that excessive workload has harmful effects on an individual's health status and might result in negative emotions. Consequently, these negative emotions might lead to rising levels of burnout. Additionally, Froese-Germain (2014), in a study of teachers in Alberta, pointed out that workload impacted negatively on the wellbeing of various teachers, with $47\mathrm{\%}$ of the participants reporting regular experiences of workload-related depressed moods, a scenario that lowers job satisfaction. Van Bogaert et al. (2013) also associated workload with anxiety, stress, depression, and general burnout.

In another study conducted to examine the connection between burnout, and workload, the results indicated that a moderate relationship existed between workload, and burnout (Yilmaz, Altinkurt, Guner, & Sen, 2015). The study further revealed that, although burnout was present, it was within the acceptable range. In other words, a high workload causes emotional exhaustion. Still, if the individual can manage their work without being overwhelmed, emotional exhaustion is likely low and within the acceptable range. Yilmaz Altinkurt, Guner, and Sen (2015) study employed a cross-sectional descriptive design and obtained data from a sample of 242 participants drawn from a target population of 427 staff members in Kermanshah University of Medical Sciences in Iran. Data was collected using NASA-Task Load Index for workload and Maslach Burnout Inventory

(MBI-S) for burnout and analyzed using ANOVA and Pearson tests procedures.

Similarly, Afzal, Idrees, Fardous, and Ambreen (2019) reported that increased workload leads to more burnout. The study utilized a correlational study design and a sample of 200 teachers drawn from working teachers in public colleges of Pakistan who had been conveniently sampled to partake in the study. The research utilized the Maslach Burnout Inventory scale, questionnaires, and interviews to measure burnout and workload. However, Afzal Idrees, Fardous, and Ambreen (2019) limited their study to curriculum instruction, and there was a need to clarify the interaction of the variables in the development of burnout.

Khan et al. (2014) assessed burnout and professional growth among university academicians and used a sample of 160 academicians working in Pakistan Universities and the Oldenburg Burnout Inventory (OLBI) to evaluate burnout levels. The results indicated that mental workload increased burnout and negatively affected the teacher's job performance. Mental workload is associated with multitasking, and as a result, it causes mental demand, consequently leading to burnout, especially emotional exhaustion. Besides mental workload, other factors associated with emotional exhaustion are; time, pressure to complete the work demands, inadequate support from colleagues and supervisors, control, and autonomy. Moreover, high demand to produce high-quality teaching and good performance cause emotional exhaustion (Khan et al., 2014). Keser and Yilmaz (2014) also established that mental workload causes burnout as it puts pressure on the individual to balance the work expectations and other academic concerns.

Another study was conducted to assess the influence of workload on burnout among university academicians in Pakistan. The study was cross-sectional and utilized a quantitative research design and random sampling method to get a sample of 162 participants from public sector universities in the Malakand Division of Khyber Pakhtunkhwa, Pakistan. Data was gathered using a self-administered questionnaire which included workload and burnout items and analyzed using Hierarchical multiple regression analysis. The findings showed that workload positively correlated with emotional exhaustion (Khan, Rasil, Yasir, Khan, 2019). The study was based on the JD-R model. Furthermore, another study conducted to establish the implications of increased workload on burnout indicated that high workload was positively related to emotional exhaustion (Vesty, Sridharan, Northcott, & Dellaportas, 2016).

In another cross-sectional study conducted to establish burnout status using a sample of 490 Swedish school teachers and Maslach Burnout Inventory-General Survey to collect data, the study findings pointed to a negative relationship between workload and burnout.

Overall, teachers reported low emotional exhaustion associated with workload regardless of the high job demands (Arvidsson, Hakansson, Karlson, Bjork, & Persson, 2016). Another study was conducted among school teachers of Boven Digoel District, Papua. The research results indicated that workload harmed the teachers' emotional exhaustion (Werang, 2018). Further, the study related burnout to individual characteristics and the school climate.

Jomuad et al. (2021) examined the association between burnout, workload, and job performance among educators using the Role Overload Questionnaire, Burnout Questionnaire, and Review Form (IPCRF) to collect data from 57 elementary teachers sampled using purposive and convenient sampling techniques. Results indicated that high workload contributed to high emotional exhaustion among teachers. Further, the study recommended addressing workload allocation to reduce burnout among teachers. According to Jomuad et al. (2021), more workload and its time-consuming nature, and a lack of mental and personal care, expose the teacher to burnout vulnerability. McTiernan and McDonald (2014) reported similar findings where they observed that burnout was positively linked to workload. Heavy workload and high demands from the administrators have also been linked to burnout. Subon and Sigie (2016) conducted a study using a sample of 50 primary and secondary school teachers of Samarahan District in Malaysia to assess the influence of workload on burnout. The study utilized a questionnaire to collect data and employed descriptive and inferential statistics to analyze the data. The study's results specified that teachers had a moderate level of emotional fatigue. Erat, Kitapci, and Gomez (2017) conducted a study to establish the influence of organizational loads on emotional commitment, stress, and turnover intention. The study collected data from 1043 participants using questionnaires. The results specified that workload and responsibility load was negatively correlated with emotional exhaustion. Similar findings were reported in other studies where researchers observed that work overload contributed to the emotional exhaustion of the workers (Arvidsson, Håkansson, Karlson, Björk, & Persson, 2016; Imo, 2017; Liu & Lo, 2018; Turtulla, 2017).

The relationship between workload and emotional exhaustion is inconsistent. Some scholars have reported positive relationships, whereas others have found negative relationships, with others reporting no relationship. Also, in most of these studies, the workload has been conceptualized as a single variable. Additionally, most studies have used samples drawn from other professions, such as health workers. Further, no research has ever been done in Kenya on the link between workload and emotional fatigue. The few studies on burnout have related burnout to poor performance and reduction of work commitment

(Baraza, 2017; Wang'eri & Okello, 2014). Therefore, there was a need for another study to explain the relationship between other workloads and emotional exhaustion among high school teachers. This study explains the link between emotional exhaustion and workload among high school teachers.

II. METHODOLOGY

This study was descriptive and had a random sample of 367 participants drawn from a target population of 4,447 teachers from public secondary schools in Kiambu County, Kenya. Sample determination was calculated using the formula provided by Yamane, 1973.

Where $n=$ sample size, $N=$ Total population size and $e=$ level of precision.

The researcher created strata samples proportional to the population from each stratum. The sample was chosen randomly within each stratum to ensure that individuals in a particular sub-group had equal chances of being included in the study sample. The purpose was to ensure that the sample included all the sub-groups and avoided bias. Teacher workload was measured using a workload questionnaire developed by the researcher. It contained twenty-four structured items measured on a 5- point Likert scale. The construction of the teacher workload Likert scale was based on the guidelines stipulated in the Teachers' code of regulation of 2012 and the information from the workload literature. The questionnaire was multidimensional, measuring the main four categories of teacher workload as conceptualized. They included eight (8) items to measure teaching load, six (6) items to measure management of students' psychosocial issues, four (4) items to measure the administrative workload, and six (6) items to measure co-curriculum activities supervision.

Emotional exhaustion was measured using the Maslach Burnout Inventory Scale (MBI-S) adapted and modified to fit the current study. The researcher obtained permission to use the research instrument. The MBI-S has 22 items measured on a 7-Likert Scale. Emotional fatigue was evaluated using items 1, 2, 3, 6, 8, 13, 14, 16, and 20 on the Maslach Burnout Inventory – Scale (MBI-S). The study used descriptive and inferential statistical techniques to analyze the data. Descriptive statistics were presented using the frequency tables, while inferential statistical analyzes employed multinomial logistic regression analysis procedures with STATA version 14.

III. RESULTS

The study's primary purpose was to establish the relationship between workload and emotional fatigue among teachers. The teacher workload had eight categories, namely; "1" teaching, management of students' psychosocial issues, administrative duties, and supervision of co-curricular activities; "2" teaching, management of students' psychosocial issues and administrative duties; "3" teaching, management of students' psychosocial issues, and supervision of co-curricular activities; "4" teaching, administrative duties, and supervision of co-curricular activities; "5" teaching

and management of students' psychosocial issues; "6" teaching and administrative duties; "7" teaching and supervision of co-curricular activities, and "8" teaching only. In the analysis, the workload category "1" was used as the reference through which interpretations for the other seven workload categories were made. The descriptive statistics on the distribution of workload responses by workload category are presented in Table 1.

Table 1 shows the distribution of responses by workload categories. Workload had eight categories. The majority of the respondents, 84 (24.42%) were in category '5' teaching and management of students' psychosocial issues) followed by 67 (19.48%) who were in category "6" (teaching and co-curricular activities). The minority of the respondents, 18 (5.23%), were in category "4" (teaching, administrative and co-curricular activities). Categories "1" (teaching, administrative, management of students' psychosocial issues and co-curricular), "2" (teaching, management of students' psychosocial issues and administrative), "3" (teaching, management of students' psychosocial issues and co-curricular), "7" (teaching and co-curricular) and "8" (teaching only) had 33 (9.59%), 21 (6.10%), 26 (7.56%) and 42 (12.21%) respectively. This distribution was expected because teachers engage in different responsibilities apart from teaching.

a) Distribution of Emotional Exhaustion within the Workload Categories

The study used descriptive statistical analysis to establish the link between workload and emotional fatigue. The researchers did a descriptive analysis based on the workload categories and emotional exhaustion. The distribution of emotional exhaustion within the workload categories is presented in Table 2.

Table 2 presents the distribution of the independent variable in each of the aspects of the dependent variable. The data reveals that; a majority of the respondents with emotional exhaustion, $25.7\mathrm{\%}$, were in workload category "5" (teaching and management) while the minority, $5.7\mathrm{\%}$, were in categories "2" (teaching, management of students' psychosocial issues and administrative duties) and "4" (teaching, administrative, and co-curricular activities) respectively. Workload categories "1", "3", "6", "7", and "8" had $13.3\mathrm{\%}$, $7.6\mathrm{\%}$, $15.2\mathrm{\%}$, $17.1\mathrm{\%}$, and $10.9\mathrm{\%}$, respectively.

b) Examination of the Relationship between Workload and Emotional Exhaustion

The study examined how emotional exhaustion and workload relate. The multinomial logistic regression analysis results of emotional exhaustion against workload are presented in Table 3.

Table 3 shows the analyzed results of the multinomial logistic regression done on emotional exhaustion. There were 331 observations. Based on the log-likelihood ratio (LR) test, the model containing the complete set of predictors represents a significant improvement in fit relative to a null model (LR ${\chi }^2(48)=190.74$, $p=.0000$ ). At least one population slope is nonzero. The model had a Pseudo R2 of.3273, which meant the proportionate improvement in fit relative to a null model was $32.73\mathrm{\%}$. The workload had eight categories identified as Categories "1", "2", "3", "4", "5", "6", "7", and "8". The study used workload category "1" (teaching, management of students' psychosocial issues, administrative duties, and the supervision of co-curricular activities) as a reference in the interpretation of the other seven workload categories w. From the table, results indicate that individuals in workload category "4" were more emotionally exhausted compared to those in workload category "1", ( $\beta =-.2331$, $t=0.23$, $p=0.814$ ). Those individuals in the workload categories "2", "3", "5", "6", "7" and "8" were found to be less emotionally exhausted than individuals in the workload category "1", ( $\beta =-1.6455$, $t=-1.99$, $p={0.046}^{\ast }$ ); ( $\beta =-.4597$, $t=-0.66$, $p=0.511$ ); ( $\beta =-1.9217$, $t=-2.77$, $p={0.006}^{\ast }$ ); ( $\beta =-1.7574$, $t=-2.01$, $p={0.045}^{\ast }$ ); ( $\beta =-1.7016$, $t=-2.30$, $p={0.021}^{\ast }$ ); ( $\beta =-.6208$, $t=-0.84$, $p=0.402$ respectively. Of the six workload categories with less emotional exhaustion, categories "3" and "8" had a moderate but negative relationship with emotional exhaustion.

IV. DISCUSSION

The study addressed the objective by hypothesizing that "there is a relationship between

The null hypothesis was that workload has no connection with emotional fatigue. The documented workload literature presents contradicting findings regarding the link between workload and teacher emotional exhaustion. Some of the studies indicate that workload relates to emotional exhaustion positively (Akca & Kucukoglu, 2020; Baraza, 2017; Chirimi, 2016; El Helou, Nabhani, & Bahous, 2016; Afzal, Idrees, Fardous & Ambreen, 2019; Khan et al., 2019; Shaheen & Mahmood, 2020; Van Bogaert et al., 2013; Vesty et al., 2016; Yilmaz et al., 2015; Zheltoukhova, O'Dea & Bevan, 2012). Other studies have reported a negative association between emotional exhaustion and workload (Arvidsson et al., 2016; Werang, 2018). These studies, however, have argued that the relationship is moderate (Turtulla, 2017; Ziaei, Yarmohammadi, Moradi & Khadnan, 2015). These inconsistencies may be due to the different perceptions that various teachers may have towards their profession. The current study conceptualized workload as a multidimensional construct comprising teaching and non-teaching workloads.

The quantitative findings of this study indicate that workload is related to emotional exhaustion both negatively and positively. Teacher workload categories "2", "3", "5", "6", "7" and "8" were more negatively related to emotional exhaustion than those in the workload category "1" (teaching, management of students' social issues, administrative duties and supervision of co-curricular activities). These results indicate that an increase in the workload unit in these workload categories leads to decreased teachers' emotional exhaustion. These findings reveal that most teachers were less likely to be emotionally exhausted due to workload. However, workload category "4", with teachers involved in (teaching, administrative duties, and the supervision of co-curricular activities) indicated a more robust and positive relationship with emotional exhaustion than the workload category 1. The findings unveiled statistically significant relationships between workload categories "2", "5", "6" and "7". These results are supported by the qualitative data as the majority of teachers in the discussion groups believed that although they had many activities to do on top of their teaching load, they still were energized to work and mentor young talents. For many teachers, emotional exhaustion resulting from workload was within manageable levels. However, during the discussion, the few teachers who expressed being exhausted with their work cited work overload, due to understaffing, multitasking, pressure, starting the day early and working late, paperwork demand, being out of school many times, filling and signing documents, supervision and several other activities outside the school time table.

The negative findings align with previous studies such as Arvidsson et al. (2016) and Werang (2018), who reported a negative association between emotional fatigue and workload. Elsewhere, Yilmaz et al. (2015) had reported a moderate negative relationship between workload and emotional exhaustion. The positive results of the current study also concur with the findings of other studies where workload had a positive relationship with emotional burnout (Van Droogbroeck et al., 2014; El Helou, Nabhani, & Bahous, 2016). These findings are consistent with Ali and Farooqi (2014), who found that when work is more than the individual's ability to manage, the individual cannot relax. As a result, the individual suffers emotional stress, leading to exhaustion. Similarly, Shafie, Kadir, and Asmiran (2017) reported that teachers who spent many hours doing administrative work and teaching were more emotionally exhausted. Moreover, Erat, Kitapci, and Gomez (2017) and Liu and Lo (2018) also argued that workload and responsibility load led to more emotional exhaustion and stress. Additionally, the positive results are similar to other findings where lesson plans preparation, marking and processing feedback, and managerial, mentorship, security, and social work were associated with high emotional exhaustion (Afzal, Idrees, Fardous & Ambreen, 2019). Elsewhere, high emotional exhaustion was due to increased workload, paperwork, multitasking, and time pressure (Acka & Kucukoglu, 2020; Keser & Yilmaz, 2014; Khan et al., 2014; Vesty et al., 2016).

The current findings fit the Multidimensional Theory of burnout (Maslach, 1998; Maslach, Schaufeli & Leiter, 2001) and Job Demand-Resource theory (Demerouti, 2014) proposals. Maslach Schaufeli and Leiter (2001) propose that burnout generally results from a perceived discrepancy between work demands and the individual's coping ability. Thus, whenever the workload is manageable, emotional exhaustion is reduced. The current results show that most workload categories had less emotional exhaustion than those that combined all the workload attributes. Also, during the interviews, several teachers revealed that it was not the amount of frustrating workload but spending time on activities that were perceived not to add value to their work. Low burnout among teachers is associated with the motivation to mentor and transform the students' lives and see students excel academically and non-academically. Therefore, as long as the outcome of the work adds value to the students, the teachers are energized to work and are less frustrated.

According to Muasya (2015), burnout is manageable as long as the person is not exposed to stressful situations. However, as suggested by Maslach, Schaufeli, and Leiter (2001), Leiter and Maslach (2016), and Maslach and Leiter (2016), few teachers had high burnout, which was associated with inadequate time, doing a lot of paperwork, poor results, ability to balance teaching time and other activities, students' indiscipline, and multitasking. Furthermore, administrative duties require that one spends many hours doing work in the office, attending to many meetings in a week, both formal and informal, pursuing work deadlines, a lot of paperwork and supervision to do, all of which are time demanding as well as physically, mentally and emotionally draining. In addition, during the co-curricular activities (music and drama, athletics and sports), teachers spend many hours and energy supervising and accompanying the students for inter-school competitions. At times, engaging in these activities can cause teachers to get drained and miss out on other activities that make them feel unaccomplished, leading to burnout.

Leiter, Bakker, and Maslach (2014) maintain that people get emotionally exhausted when faced with demanding work situations. Administrative tasks and supervision of co-curricular activities are time-demanding. If teachers have difficulties keeping deadlines, sit for many hours in the offices compiling reports and organizing time for outdoor activities with the students, they may experience pressure (Ali & Farooqi, 2014). Accordingly, the situation may cause intra-personal conflicts leading to emotional exhaustion. The findings are supported by the qualitative data where the participants lamented that they spend a lot of time coaching and taking students out at the expense of their class and personal time. Hence, teachers' emotional exhaustion is attributed to the pressure created by various workloads (classroom teaching, management of students' social issues, administrative and co-curricular activities. Teachers who have been exposed to high work demands for a long time and lack the abilities to cope may experience burnout due to workload, whereas those who perceive no discrepancy in their work are more likely to manage with moderate burnout (Lesener, Gusy, & Wolter, 2019; Maslach & Leiter, 2016; Maslach & Leiter, 2017). However, the current guidance and counselling program in schools emphasizes the needs of pupils and students (Wango, 2015).

V. CONCLUSION

The current study explored the connection between workload and emotional exhaustion among high school teachers. They found that several teachers had moderate emotional exhaustion. The study concluded that workload as a multidimensional concept is related to teachers' emotional exhaustion, with some categories having negative relationships and others having positive relationships. Further, there were statistically significant negative relationships between workload categories "2", "5", "6" "7" and emotional exhaustion. Workload category "4" was positively related to emotional exhaustion. Emotional exhaustion can affect teachers' physical, mental, emotional, and social wellbeing. Therefore, the research recommends that school counselling departments be made comprehensive enough to accommodate workplace counselling to assist teachers in dealing with emotional exhaustion, thus acquiring overall mental health, which would enhance their performance.