How Do Happiness at Work and Perceived Organizational Support Affect Teachers’ Mental Health Through Job Satisfaction During the COVID-19 Pandemic?

Introduction

The COVID-19 pandemic has impacted the education sector, causing changes in teaching and learning practices worldwide, such as moving from onsite to online learning.^1,2 In Indonesia, the Ministry of National Education published the Letter of the Minister of Education and Culture of the Republic of Indonesia No. 4 of 2020, making this change mandatory. Such changes triggered by the COVID-19 pandemic have taken educators by surprise and posed challenges to both teachers and learners.³ Hence, many teachers continue to experience difficulties of increased workloads and uncertainty, thereby raising mental health (MH) concerns.^4–6

Teachers’ MH is a matter of widespread concern for education management worldwide, and poor MH is thought to be linked to high levels of stress and attrition.⁷ Working as a teacher is considered a mentally challenging job as teachers are obliged to manage numerous cultural, social, and community demands from multiple directions, resulting in a high number of daily emotional interactions.⁸ In addition, the teaching profession is considered to have a high level of burnout and stress.⁹ At the university level, lecturers face high levels of stress, which is thought to be associated with anxiety disorders, hypertension, headaches, psychosomatic disorders, and even heart disease.¹⁰

Braegunig et al asserted that teachers in Germany often experience health problems caused by stress.¹¹ Moreover, research conducted in China by Liu et al found that lecturers often experience pressure from balancing teaching work, research, and life demands, resulting in physical and mental suffering.¹⁰ In Indonesia, teachers’ MH challenges have received minimal research attention, hence the need for the present study.

According to Svalastog et al, a state of good health is a condition in which overall physical functions work well, there is an absence of pain, and a feeling of comfort arises due to harmonious organ functions.¹² This definition highlights the holistic implications of those factors that influence both physical health and MH. Specifically, MH is defined as the dynamic internal balance within an individual that enables harmony with the universal values of society.¹³ The World Health Organization defines MH as a state of wellbeing in which an individual realizes the desire to manage the problems of normal life and works productively to contribute to the community.¹⁴

One factor affecting teachers’ MH is happiness at work (HW), which is an aspect of subjective wellbeing in the form of short-term moods and emotions regarding one’s work experiences.¹⁵ HW can arise from positive experiences at work.¹⁶ Benevene et al found HW to have a positive effect on teachers’ overall health.¹⁶ When workers feel happy, they tend to be healthier and better able to cope with stress.¹⁷

Another factor affecting the MH of teachers is the perception of organizational support (POS). According to Eisenberger and Stinglhamber, POS is the belief that the organization in which a person works cares about the welfare of employees and values their contributions.¹⁸ POS is an aspect of organizational behavior theory that emphasizes the importance of valuing employees as organizational resources. Liu et al found the POS experienced by college teachers in China to positively affect both physical health and MH.¹⁰

Job satisfaction (JS) is also believed to affect MH. JS reflects a person’s affection or positive emotional perceptions toward work, coworker relationships, and the existing work environment, which arise on the basis of fulfilled expectations.^19,20 When this level of fulfillment is high, employees tend to feel satisfied. Research on high school teachers in Italy conducted by Capone and Petrillo found that teachers with high levels of JS tend to also have good MH.²¹ Furthermore, Chirico et al argued that low JS leads to poor MH.²²

Many MH studies have assessed factors of JS.^{16,19,21,23–25} Several studies have also sought to determine the role of POS on MH.^10,26,27 However, only a few have explored the effect of HW on MH in the education sector, particularly in Indonesia. Therefore, this research aims to fill the research gap by assessing the effects of HW on MH in the context of Indonesian teachers. Additionally, the effect of JS on MH has elicited contradictory results in previous studies, showing both positive and negative effects.^28,29 Hence, the mediating role of JS between HW and MH deserves further examination to address the current confusion and to examine the role of POS on teachers’ MH, with JS as a mediator in the context of the COVID-19 pandemic.

Theoretical Framework and Hypotheses Development

Relationship Between HW, JS, and MH

MH is an important construct in the teaching profession, especially during the COVID-19 pandemic. Sudden changes in the teaching practice have occurred, such as moving from onsite to online learning, and it is thought that teachers’ MH may suffer.³⁰ Almeida asserted that MH is a state of being free from psychological disturbances, as well as an individual’s capability to adapt to change, overcome crises, relate to other people, and enjoy life.³¹ Howard et al contended that MH disorders can be caused by various circumstances, such as work stress, which is often encountered by teachers.³² Kidger et al found low wellbeing experienced by teachers to be the result of stressful workloads, dissatisfaction with coworkers’ performance, and the inability to convey feelings and concerns to others, which can elicit moderate to high symptoms of depression.³³

MH is believed to be influenced by HW, which is the cognitive and affective process that appears to be related to the impact of the organizational characteristics of the workplace.¹⁶ HW is represented by employees’ moods and short-term emotions experienced based on relationships with colleagues, managers, leaders, and customers.¹⁵ HW is also shaped by perceptions of fair treatment, appreciation, and positive relationships.^34–36 During the COVID-19 pandemic, the pressure and stress caused by the dramatic work–life changes have had negative impacts on HW.³⁷ Previous research found that high HW positively affects individual health, including the reduction of stress and burnout.^15–17,38 By contrast, low HW results in withdrawn behaviors and mental suffering (HW→MH).³⁹

JS is believed to have a mediating relationship between HW and MH. JS reflects an employee’s affective emotions regarding personal accomplishments based on job expectations.^23,40 The fulfillment of personal and organizational goals in the workplace is a key JS factor. As an endogenous variable, JS is believed to be influenced by HW. However, the COVID-19 pandemic directly changed the workplace dynamics affecting JS.⁴¹ Benevene et al studied teachers at junior and middle schools, finding HW to have a significant positive effect on teachers’ JS (HW→JS).⁴²

Nandinloyi et al found a relationship between JS as an exogenous variable and MH, demonstrating higher JS to be associated with better MH, particularly in terms of reducing depression and building social relationships with others.⁴³ Relationship with colleagues also affects the balance of one’s psychological wellbeing.⁴⁴ In Japan and India, it was found that higher levels of teachers’ JS were related to better MH (JS→MH).^28,42

Based on the findings from previous research that HW affects JS,^15–17,39 and that JS affects MH,²⁸ suggest that these three variables are correlated. This study seeks to examine the positive effects of HW on MH through the mediation of JS. This mediation relationship has not been well-studied, except by Benevene et al.⁴² However, the direction of this study is novel in the context of educational management in Indonesia (HW→JS→MH).

Based on the association of the propositions derived above, the initial hypotheses of this study are as follows:

H1: HW positively affects teachers’ MH during the COVID-19 pandemic.

H2: HW positively affects teachers’ JS during the COVID-19 pandemic.

H3: JS positively affects teachers’ MH during the COVID-19 pandemic.

H4: HW positively affects teachers’ MH through JS during the COVID-19 pandemic.

Relationship Between POS, JS, and MH

The existence of the COVID-19 pandemic poses challenges and pressures for teachers in their work, and organizational support is required. Oubibi et al emphasized the importance of POS when dealing with educational crises (eg, COVID-19), in which high adaptation to changes is required. Notably, such changes clearly impact the emotional demands, pressures, and burnout symptoms of teachers.⁴⁵ Eisenberger and Stinglhamber defined POS as employees’ perceptions of organizations valuing their contributions and caring about their welfare.¹⁸ Another definition was proposed by Chatzittofis et al, as it pertains to employees’ perceptions of the extent to which an organization expresses concern with maintaining employees’ physical, psychological, and mental wellbeing.²⁷ According to Malik and Neem, employees’ POS is a vital aspect of work culture for organizational environments that require employees to perform more than their job descriptions.⁴⁶

POS is believed to influence employees’ MH. Chatzittofis et al found low POS to be associated with higher rates of depression, stress, and traumatic symptoms, indicating that the greater the POS, the better the employee MH.²⁷ This seems to have been particularly true during COVID-19.²⁷ Jin and Tang found higher POS experienced by employees at work to lower stress levels, further implying the correlation of POS to MH (POS→MH).⁴⁷

POS is also believed to affect JS.⁴⁸ Collegial relationships factor into high JS. Danielsson and Bodin found employees who share workspaces with others to have beneficial interactions, which positively affects JS.⁴⁹ Higher POS experienced by employees tends to result in higher JS, especially during COVID-19.^50–53 In the context of the teaching profession, Bogler and Nir investigated the experiences of elementary school teachers in Israel, finding that the greater the POS, the greater the JS (POS→JS).⁵⁴

The relationships between POS and MH^27,47 and POS and JS^52,54 suggest that all three variables are correlated. JS is believed to have a mediating role in the relationship between POS and MH. Indeed, a previous study found POS, JS, and MH to be correlated (POS→JS→MH).⁵⁵

Based on the association of the propositions derived above, the additional hypotheses of this study are as follows:

H5: POS positively affects teachers’ MH during the COVID-19 pandemic.

H6: POS positively affects teachers’ JS during the COVID-19 pandemic.

H7: POS positively affects teachers’ MH through JS during the COVID-19 pandemic.

Materials and Methods

Research Design

This research was conducted using a quantitative approach that applied a cross-sectional study method with path analysis measured by partial least squares structural equation modeling (PLS-SEM). This method was chosen because it suits the data used, which otherwise does not accommodate classical assumption tests.⁵⁶ PLS-SEM is also suitable for testing for unobservable latent variables using indicators from the theory of each construct.⁵⁷ In this study, MH was an endogenous variable, HW and POS were exogenous variables, and JS was assessed as a mediating variable. The research proposal, which included an introduction, theories for hypothesis development, and methods, was started in June 2021. It was then submitted for review to our institution’s Internal Review Board (IRB) on October 7, 2021 and was approved on November 19, 2021. Research data were acquired via an online questionnaire over 10 days following ethical clearance from our IRB.

Participants

The participants in this research were kindergarten–high school-level teachers in the Special Capital Region of the Jakarta province, Indonesia. The baseline population of 80,416 participants was provided by the Special Capital Region of Jakarta province based on academic year 2020 information.⁵⁸ Referring to Krejcie and Morgan’s sample calculation theory, the number of samples required for a thorough representation of this population is 384.⁵⁹ In this study, the number of responses was 490 after eliminating invalid entries. Because this study was conducted during the COVID-19 pandemic, convenience sampling was used as the data collection technique, wherein the research data were based on respondents’ completion of the questionnaire. The questionnaire was distributed online using a Google form. The questionnaire link was distributed through teacher representatives at schools in the Special Capital Region of Jakarta province to individual teachers. The research data were then analyzed using a bootstrapping technique for 5000 observations to ensure that the sample was truly representative of the population under study.⁵⁶

Measures

All construct measurements of this study were adapted from previous theories and research, which were translated into Indonesian to make it easier for participants to understand and respond to the online questionnaire. The questionnaire is presented in Supplementary Table S1. The questionnaire used through the PLS-SEM statistical validation process refers to Hair et al with the rule of thumb 0.7 for each loading factor.⁵⁶

MH was measured using six items in the questionnaire using a Likert scale of 1–5, referencing Kodraliu et al and Ware et al, integrating the four indicators of (1) good social relations, (2) good emotional roles, (3) good vitality, and (4) good mood.^60,61 Examples of items in the questionnaire include “I tend to be calm in dealing with problems at work” and “I am able to use my time for various activities during the COVID-19 pandemic.” The Cronbach’s alpha of this variable was 0.883.

HW was measured using six items in the questionnaire using a Likert scale of 1–5, referencing Fisher and Lutterbie and Pryce–Jones, including the three dimensions of (1) happiness toward work, (2) contextual job features, and (3) the entire organization.^35,62 Examples of items in the questionnaire include “I can convey ideas comfortably to coworkers” and “I am enthusiastic about carrying out home teaching during the COVID-19 pandemic.” The Cronbach’s alpha of this variable was 0.871.

POS was measured using seven items in the questionnaire using a Likert scale of 1–5, referring to Eisenberger and Stinglhamber and Eisenberger et al, assessing the five indicators of (1) the organization values employee contributions, (2) the organization pays attention to employee complaints, (3) the organization cares about employee welfare, (4) the organization shows great concern for employees, and (5) the organization appreciates employees’ extra efforts.^18,63 Examples of items in the questionnaire include “The school cares about my complaints during the COVID-19 pandemic” and “The school provides adequate work facilities to work from home during the pandemic.” The Cronbach’s alpha of this variable was 0.927.

JS was measured using five items in the questionnaire using a Likert scale of 1–5, referring to Dolbier et al and Tsai, including the five indicators of (1) satisfaction with compensation, (2) satisfaction with the promotion, (3) satisfaction with supervisor’s supervision, (4) satisfaction with coworkers, and (5) satisfaction with assigned work.^64,65 Examples of items in the questionnaire include “I am happy with the allowances provided by the school during the COVID-19 pandemic” and “I am happy with the way the leaders at my school provide feedback on my performance.” The Cronbach’s alpha of this variable was 0.914.

Data Analyses

The data in this study were classified into two groups: respondent profile (nominal) and research questionnaire (interval). Respondent profile data describe the demographics of the respondents, including sex, educational background, teaching level, and teaching experience. This data group was analyzed descriptively by calculating the percentages. Research questionnaire data reflect respondents’ answers in the range [1–5] and were analyzed descriptively and inferentially.

Descriptive analysis for the questionnaire data was used to present the means and standard deviations of respondents’ answers. In addition to the cross-sectional method analysis requirement, Harman’s single factor test was conducted using the SPSS v.25 software to ensure no common method bias because respondents’ answers may be influenced by the format and context of the questions, resulting in over- or under-stated tendencies reported by the respondents.⁶⁶

PLS-SEM inner and outer inferential statistics models were used to validate the hypotheses using the Smart-PLS v.3.2 software.⁵⁶ The outer model test focuses on validity testing with the loading factor, average variance extracted (AVE), convergent and discriminant validity, and data reliability testing with Cronbach’s alpha and composite reliability.⁵⁶ The inner model test examines the multicollinearity of the data, predicts the coefficient determinant (R²), and tests the hypotheses via path analysis.⁵⁶

Results

The respondents in this research were predominantly female teachers with the highest educational background of a bachelor’s degree. Participants were from different school departments, with the largest number being of elementary school teachers. Most respondents had been teaching for more than 12 years at the time of questionnaire completion. The details of the respondents’ demographic data are presented in Table 1.

Table 1 Respondent Profile

Common Method Bias Test

The common method bias test was carried out using Harman’s single factor test (result 38.48%) to ensure that the cross-sectional research results were not biased by instrument artifacts.⁶⁶ This result meets the rule of thumb, which is less than 50%.⁶⁶ Hence, common method bias did not influence the results. The result of the common method bias test is shown in Table 2.

Table 2 Harman’s Single Factor Test Results

Outer Model

The outer model test measures the validity and reliability of the data. The descriptor of each construct was assessed as valid as it met the loading factor of >0.7 with an AVE of >0.5.⁵⁶ All constructs were confirmed to be reliable as their composite reliability and Cronbach’s alpha values exceeded the 0.7 requirement,⁵⁶ and all constructs were considered reliable with values above 0.7. Details are found in Table 3.

Table 3 Loading Factor, AVE, and Reliability Test Results

In the outer model, discriminatory validity testing was conducted to ensure that each construct of the study differed sufficiently.⁵⁶ To assess discriminant validity, heterotrait–monotrait ratio (HTMT) testing was carried out by calculating the mean of the indicator correlations of all constructs whose rule of thumb was less than 0.85.⁶⁷ On the basis of the results of the data analysis presented in Table 4, all constructs have values less than 0.85; hence, discriminant validity is confirmed.

Table 4 Descriptive Statistics and Discriminant Validity Test Result (HTMT) Ratios

Inner Model

Inner model testing is divided into three parts. The first includes multicollinearity testing, which aims to ensure that all exogenous variables lack high collinearity.⁵⁶ This test was conducted by examining the value of the variance inflation factor, which must be >5.00.⁵⁶ The results of the data analysis presented in Table 5 reveal that the exogenous variables did not exhibit multicollinearity.

Table 5 Multicollinearity Test Results

The second inner model applied determinant coefficient testing, which assesses the prediction accuracy of the research model.⁵⁶ The value of the determinant coefficient was ascertained from the effect of the exogenous variable as a whole and has a value in [0, 1]; the higher the value is, the better the research model is.⁵⁶ Hair et al provided R² values of 0.25, 0.5, and 0.75 as criteria, indicating weak, moderate, and substantial predictions, respectively.⁵⁶

The R² value in this study indicated that the research model predicted MH constructs well and JS constructs very well as 60.8% of JS findings were explained by the HW and POS variables, whereas the remaining 39.2% were explained by other variables not examined in this study. It was also concluded that 49.2% of the MH findings were explained by HW, POS, and JS, whereas the remaining 50.8% were explained by other variables not examined in this study. The values of the determinant coefficients for JS and MH are presented in Table 6.

Table 6 Determinant Coefficient Test Results

The final inner model test assessed the research hypotheses by investigating the path coefficient value and its significance, which can be ascertained using two events: t-statistic value > 1.96 and p-value < 0.05.⁵⁶ The results showed that HW positively and significantly affected teachers’ MH (β = 0.659, p = 0.000). HW also had a positive effect on JS (β = 0.194, p = 0.000). Therefore, Hypotheses 1 and 2 were supported. Meanwhile, JS was found to affect teachers’ MH positively but not significantly (β = 0.060, p = 0.173). Therefore, Hypothesis 3 was not supported. Hypothesis 5 was also supported as POS was found to affect teachers’ MH positively but not significantly (β = −0.028, p = 0.304). POS was also found to positively affect JS (β = −0.687, p = 0.000). This path is therefore considered strong and significant. Therefore, Hypothesis 7 was supported. Table 7 presents the detailed results of the path analysis between variables.

Table 7 Hypotheses Test Results

Mediated Effect

Table 7 also presents the results of the indirect relationship coefficient path test. In the mediating relationship, JS did not successfully mediate the positive effect of HW on teachers’ MH (β = 0.012, p = 0.178). This result was smaller than the direct effect of HW on teachers’ MH and was not significant. Referring to Zhao et al and Hair et al, the first measurement of the mediating effect depends on the significance of the resulting mediation effect.^56,68 If the value is not significant, then a mediation relationship does not exist (fails). The subsequent mediating relationship involving JS was found to fail to mediate the positive effect of POS on teachers’ MH (β = 0.041, p = 0.175). This result was higher than the direct effect on Hypothesis 5, but it was not significant. Following the rules for calculating the variance accounted for (VAF) proposed by Hair et al, these two mediating relationships were not significant as they had a p-value > 0.05; thus, the VAF test was not carried out.⁵⁶ Therefore, Hypotheses 4 and 7 were not supported. Details of the results of the hypothesis testing are presented in Table 7.

The path analysis results of the research model are presented in Figure 1.

Figure 1 Research model with path coefficient to estimate the structural equation model.

Based on these results, two structural equations based on the two endogenous variables emerged from this study:

Discussion

Relationship Between HW, JS, and MH

The results of this research confirmed that HW has the greatest influence on teachers’ MH. The first hypothesis test showed that HW positively and significantly affected teachers’ MH, which can be generalized to the entire teacher population in the Special Capital Region of Jakarta province. Thus, improvements that evoke happiness in the workplace (school) will improve teachers’ MH. This deduction is supported by the results of Benevene et al, De Stasio et al, Graziotin et al and Salas-Vallina et al indicating that HW affects MH.^15–17,39

Based on the descriptive statistics from the questionnaire items, the teachers felt enthusiastic when conducting remote teaching activities during the pandemic. This suggests that teachers’ happiness is related to their work.³⁵ Additionally, teachers felt comfortable conveying ideas to colleagues, indicating that the teachers had a good work environment that allowed good social interactions among colleagues.⁶² The HW felt by teachers, therefore, impacts MH. The descriptive statistical data confirmed that teachers felt they had good physical conditioning, and they were enthusiastic about working during the pandemic. The teachers also agreed that they could think rationally regarding the problems encountered at work. These findings suggest that teachers’ emotions and moods affect their MH.⁶⁰

The second hypothesis was supported by the positive and significant effect of HW on JS. This reflects previous findings that HW positively influences teachers’ JS.⁴² Although the results of our study showed a positive effect, the value of this effect was small and weak. Thus, teachers’ JS is minimally influenced by HW. The descriptive statistics suggest that although teachers may feel happy at work, the level of JS tends to be low, as revealed by respondents’ average answers showing hesitancy regarding JS questionnaire items.

The third hypothesis was not supported as the results revealed a positive but insignificant effect. These results indicate that in the context of the research sample, JS positively affects teachers’ MH. However, these results cannot be generalized for all teachers in the Special Capital Region of Jakarta province. Based on the context of the sample data, the results are congruent with those of Nagai et al for teachers in Japan, who showed a positive influence of JS on teachers’ MH.²⁸ However, when generalized for the entire population of this study, the results support those of Kuwato and Hirano, who found JS to negatively affect teachers’ MH in Japan.²⁹

The results of Hypothesis 3 help fill the research gap on the relationship between JS and MH,^28,29 which impacts the mediating role of JS between HW and MH. The proposed mediating role in Hypothesis 4 was not supported as the test results revealed a positive influence of HW on teachers’ MH through JS, which was not significant. Therefore, these results contradicted those of Benevene et al, who found JS to successfully mediate the effect of HW on teachers’ MH.⁴² Although the effect was positive, it was not greater than the direct effect on Hypothesis 5; hence, the proposal of the mediating role of this study was unsuccessful as the participants expressed doubt regarding all JS questionnaire items. This indicates that teachers tended to be dissatisfied with their work during the COVID-19 pandemic based on the allowances provided by the schools and the ways in which leaders provided work feedback.

Relationship Between POS, JS, and MH

In this study, POS was found to positively affect teachers’ MH (0.083), which was not significant. Thus, Hypothesis 5 was not supported. The effect of POS on teachers’ MH was weak and did not represent the entire population. Therefore, the results do not support the findings of Chatzittofis et al and Jin and Tang, who found higher POS to be correlated with lower stress levels and superior MH.^27,47

The low value of the effect of Hypothesis 5 was caused by the teachers’ hesitation regarding the POS questionnaire items. Based on the results of the descriptive statistics, most teachers responded doubtfully regarding whether schools provided adequate work facilities for working from home during the pandemic. The teachers also answered doubtfully regarding whether their leaders appreciated their best efforts. These findings further suggest that teachers perceived a lack of support from their organizations during the COVID-19 pandemic. In terms of MH, the teachers indicated that they think rationally, are not often anxious, use their time well, and have good vitality, which implies that they have good MH.^60,61 Although they felt less supported by their organizations, they maintained good MH. Therefore, POS was not shown to have a major influence on teachers’ JS.

For Hypothesis 6, POS was found to positively and significantly affect teachers’ JS. The effect value was considerably high, indicating that POS is a strong predictor of teachers’ JS. This further implies that increases in teachers’ POS may have a high impact on JS. The influence value is also significant, indicating that it can be generalized to the entire teacher population. These results support the findings of Bilgin and Demirer, Maan et al and Zeng et al.^50–52

A notable finding from this research emerged from Hypothesis 7, revealing that POS has a positive and insignificant effect on teachers’ MH through JS. The value of the effect was greater than the direct effect of Hypothesis 5, but it was not significant; thus, the proposed mediating role of this study was unsuccessful. This finding is interesting as POS was shown to have a strong effect on teachers’ JS, whereas JS does not mediate the effect of POS on MH. This finding differs from those of Bravo–Yáñez and Jiménez-Figueroa, in which POS, JS, and MH were shown to have positive correlations.⁵⁵

Conclusion

This study applied an appropriate model for predicting teachers’ MH and JS, as confirmed by the value of the coefficient determinant (R²). HW was found to be a predictor of MH with the highest influence, whereas JS had the smallest influence. The results also showed that POS is a strong predictor of teachers’ JS.

This study successfully fills the research gap regarding the confusion among previous results about the effects of JS on teacher’s MH in Indonesia, which further impacts the mediating role of JS in the relationship of HW with MH. This study confirms that JS positively affects teachers’ MH while not being significant to the wider population, as reflected by the findings of the unsuccessful mediating role of JS in this research model.

Managerial Implications

The results of this research prompt several managerial implications for leaders of educational institutions in the Special Capital Region of Jakarta province. HW was found to be a strong predictor of teachers’ MH. Therefore, school leaders should attend to increasing opportunities to evoke HW in schools to improve teachers’ MH. This can be achieved by establishing a pleasant working climate, such as with teacher-appreciation gatherings. School leaders can also improve HW by assigning jobs according to teachers’ abilities and interests at appropriate workload levels.

The next managerial implication is related to JS. POS was found to be a strong predictor of JS; therefore, school leaders should increase the general support provided to teachers. The JS of the teachers participating in this study tended to be low. Therefore, school leaders should adequately attend to teachers’ complaints and concerns, provide support for remote learning (eg, internet quotas and training), assist teachers who are sick, particularly during the COVID-19 pandemic, and show meaningful appreciation for teachers’ contributions.

Limitation and Research Suggestions

This research is certainly not without limitations. The study was conducted during the COVID-19 pandemic, and the data collection method applied a convenience sampling technique that relied on the completion of an online questionnaire. As a result, the number of respondents per school may be disproportionate, including gender distribution. It is suggested that future researchers accommodate a proportional number of respondents to elicit more comprehensive results. Additionally, this study found that JS did not operate as a mediating variable. Therefore, future research could investigate other variables for this purpose regarding the effects of HW and POS on teachers’ MH. In terms of the instrument and research design, this study used a self-reported questionnaire in a cross-sectional study design; hence, it could not fully explore causal hypotheses.