The Buffering Effect of Spirituality at Work on the Mediated Relationship between Job Demands and Turnover Intention among Teachers

Abstract

The purpose of this study was to examine whether spirituality at work moderates the direct and indirect (through burnout) effects of quantitative and emotional job demands on turnover intention among teachers. The sample consisted of 952 Polish primary and secondary school teachers. Burnout mediated the relationship between both types of job demands and turnover intention. In the model with quantitative job demands as an independent variable, spirituality at work moderated the second stage path of the indirect effect, i.e., the relationship between burnout and turnover intention (b = −0.022; SE = 0.004; p < 0.001; β = −0.14). In the model with emotional job demands as an independent variable, spirituality at work moderated the first and second stage paths of the indirect effect, i.e., the relationship between emotional job demands and burnout (b = −0.001; SE = 0.001; p = 0.032; β = −0.05) and the relationship between burnout and turnover intention (b = −0.020; SE = 0.004; p < 0.001; β = −0.14). In both models, the indirect effect of job demands on turnover intention through burnout weakened as spirituality at work increased. The results of the study support the inclusion of spirituality at work as a subcategory of personal resources in studies using the job demands-resources model.

1. Introduction

Nowadays, teaching is regarded as a stressful and demanding occupation with high rates of exhaustion and turnover (Hakanen et al. 2006; Liu and Onwuegbuzie 2012; Lorente et al. 2008; Noor and Zainuddin 2011). This situation is determined, among other factors, by teachers’ multiple job requirements, duties and responsibilities, and the need to have multiple competencies and skills (Antoniou et al. 2013). A multitude and variety of duties can cause stress, negatively affecting work capacity in teaching, psychological well-being at work, and job satisfaction, often leading to burnout and turnover (Barkhuizen et al. 2014). The aim of the current study was threefold: (1) to examine the effect of qualitative and emotional job demands on turnover intention among teachers; (2) to test whether burnout mediates the relationship between job demands and turnover intention, and (3) to examine the role of spirituality at work as a moderator of the direct and indirect (through burnout) effects of job demands on turnover intention.

Job demands can be defined as job characteristics that require physical or mental effort and are therefore associated with certain physiological and psychological costs (Affrunti et al. 2018). Although job demands may be seen as challenges in work, which are required for personal growth and accomplishment, they may become stressors in situations that require high effort to sustain an expected performance level, inducing negative responses and sometimes even leading to energy depletion (Hakanen et al. 2006; Simbula 2010).

This study focuses on quantitative and emotional job demands. Quantitative job demands are related to the amount of work to be done in a limited time (Altaf and Awan 2011; Mazloum et al. 2008). In the case of teachers, quantitative job demands include a combination of preparation for lessons, testing, scoring, writing reports, administrative paperwork, and participation in professional development courses (Danielson 2013). Many teachers admit that they spend considerable time on unnecessary and time-consuming duties such as detailed and sometimes duplicated paperwork tasks, preparation for inspections, and continuously proving to principals and government that their work is fruitful (Ravalier and Walsh 2017). The rapid pace of changes in the educational system and countless demands from principals and supervisory authorities have also been listed as factors contributing to teacher stress (Skaalvik and Skaalvik 2018).

Apart from experiencing heavy workload and time pressure, teachers are also exposed to high emotional demands. These can be defined as aspects of work that require emotional effort due to the frequency, intensity, and variety of interpersonal interactions related to one’s job (Brotheridge and Lee 2002; De Jonge and Dormann 2003). Teaching includes long hours of face-to-face interaction with pupils (including those with behavioral problems and poor motivation or performance); conducting parent–teacher meetings; responding to high demands and pressure from parents; taking the floor during school board meetings; and frequent communication with the principal, other teachers, and administrative workers (Jensen 2021; Santavirta et al. 2011; Yin 2015). To fulfill these obligations, teachers are expected to learn how to regulate their own emotions in a proper way, i.e., showing some of them and suppressing others to influence the feelings and reactions of clients in the desired direction (Ogbonna and Harris 2004; Zapf 2002). Not surprisingly, such high emotional job demands and the restrictions imposed on teachers’ emotional reactions may lead to emotional exhaustion (Bodenheimer and Shuster 2020; Näring et al. 2006).

Differentiating between quantitative and emotional job demands in organizational studies is important, particularly in human service occupations, in which employees have to combat overload, time pressure, and emotional strain. Emotional job demands seem to be at least as important as quantitative ones when predicting burnout and the well-being of employees in human service occupations (Tuxford and Bradley 2015; Van Vegchel et al. 2004). Moreover, there is some evidence that different job demands and job resources may differ in the strength of their relationships with job outcomes (Skaalvik and Skaalvik 2018, 2020; Stelmokienė et al. 2019). Importantly, it has also been noted that the specific job resources might interact with specific job demands when predicting job outcomes (Bakker et al. 2005; Xanthopoulou et al. 2007a), so treating various job demands as indicators of a single latent construct may make it impossible to capture such differences in the moderating effects. Thus, in the current study, we distinguished between quantitative and emotional job demands.

It is also important to note that the current study was carried out at a very special time, i.e., during the second wave of the COVID-19 pandemic. In these extraordinary circumstances, teachers—as the front-line workers in the education system—have been exposed to a great amount of quantitative and emotional job demands related to the necessity to quickly implement the new teaching practices and methods and to adjust to remote teaching, at the same time dealing with organizational chaos, the problems with disobedience of some students and parents’ claims and high expectations, work–family conflict, parenting stress, and fear of COVID-19 infection (Carretero et al. 2021; Chmura-Rutkowska et al. 2020; Hong et al. 2021; Kim and Asbury 2020). This situation makes our study unique in terms of the opportunity to explore the moderating role of spirituality at work in the relationships between job demands, burnout, and turnover intention among teachers in such unprecedented circumstances.

1.1. Job Demands and Turnover Intention

In recent decades, teacher turnover has been identified as one of the major issues in educational settings (Hong 2010; Ingersoll 2001). For instance, in the U.S., 25% of teachers leave the teaching profession before their third year on the job, and almost 40% left the occupation within the first five years (Chang 2009; Guarino et al. 2006; Hakanen et al. 2006). Using data from the Survey of Health, Ageing, and Retirement in Europe, Federičová (2021) found an overall rate of 28% for teacher turnover across the nineteen European countries, with 25% noted in Poland. Moreover, many teachers quit their profession before they reach retirement age (Macdonald 1999). These numbers are alarming because the bulk of studies has shown that teacher turnover leads to serious consequences such as high monetary costs of recruiting and training new teachers, decrease in student learning and achievement, reduction of teamwork and diminishment of a school’s sense of community, and erosion of school climate (Hakanen et al. 2006; Ronfeldt et al. 2013; Watlington et al. 2010).

Turnover intention can be defined as the conscious and deliberate cognitive process of thinking, planning, and desiring to leave a current job (Hee and Ling 2011; Sousa-Poza and Henneberger 2004). It is significantly related to the actual turnover and is regarded as its proximal precursor (Griffeth et al. 2000; Kosi et al. 2015). Moreover, similarly to actual turnover, turnover intention has multiple negative effects on job outcomes, e.g., it is related to absenteeism from work, work withdrawal, lateness, reduced work performance, motivation and achievements, and a decrease in commitment within the organization (Chughtai 2013; Quick and Nelson 2011; Scanlan and Still 2019). Studying turnover intention has some advantages over studying actual turnover since its measurement is easier and more feasible and makes it possible to detect early symptoms of the problem, allowing remedial action to be taken before it develops into actual turnover (Lambert and Hogan 2009).

Quantitative and emotional job demands have been widely recognized—both in quantitative and qualitative studies—to be one of the primary factors increasing turnover intention among teachers, eventually leading to turnover and attrition in this group (Borman and Dowling 2008; Clandinin et al. 2015; Struyven and Vanthournout 2014). In line with the results of these studies, we expect that:

Hypothesis 1 (H1).

Quantitative job demands (H1a)/Emotional job demands (H1b) have a direct positive effect on turnover intention.

1.2. Burnout as a Mediator between Job Demands and Turnover Intention

According to one of the assumptions of the Job Demands-Resources (JD-R) model, when certain job demands are high, exerting high pressure on the routine working life of an employee, certain job resources are limited, short-term stress cannot be alleviated in time, and job burnout may develop (Bakker and Demerouti 2007; Demerouti et al. 2001; Maslach et al. 1996). In this study, we adopted the conceptualization of burnout proposed by Demerouti et al. (2001), according to which it has two core components: exhaustion and disengagement from work. Exhaustion is understood as “a consequence of intensive physical, affective and cognitive strain, that is, a long-term consequence of prolonged exposure to certain job demands” (Demerouti et al. 2010, p. 210). Disengagement is related to distancing oneself from one’s work and experiencing a negative attitude toward work in general, work object, and work content (Demerouti et al. 2001). It involves a devaluation, mechanical work performance, and negative emotions toward the work tasks.

It has been estimated that 5–20% of teachers in the U.S. manifested symptoms of burnout during their professional carrier (Farber 1991). In Finland, teachers have been found to experience stress and burnout more frequently than members of other professions (12% vs. 8%; Kauppinen et al. 2010). A study by Hallsten et al. (2002) demonstrated that 9.6% of Swedish teachers suffer from burnout syndrome. In Poland, Tucholska (2003) noted full-blown burnout in 20.7% of teachers. Lower estimates were made in a study by Sekułowicz (2002), in which 6.5% of Polish teachers from ordinary schools and 14% from special schools were suffering from burnout.

Many studies have shown that burnout is not only very costly for the education system but also has detrimental effects on teachers’ well-being and school functioning (Schonfeld 2001). It negatively affects mental and physical health (Maslach 2017; Salvagioni et al. 2017), and decreases job satisfaction, job involvement, and organizational commitment (Salvagioni et al. 2017; Swider and Zimmerman 2010). Moreover, teacher burnout affects students’ motivation and achievements, classroom climate, teacher–student relationships, and induces impaired organizational behaviors such as absenteeism and presenteeism (Maslach and Leiter 1999; Salvagioni et al. 2017; Shen et al. 2015; Swider and Zimmerman 2010).

Finally, burnout may result in the teacher’s decision to change the school or leave the profession (Chang 2009; Michniuk 2021; Skaalvik and Skaalvik 2020). There is substantial evidence that burnout is positively related to turnover intention (Allen and Mueller 2013; Chen and Yu 2014). This relationship occurs among various occupations and settings, and accounts for other relevant organizational variables (Schaufeli and Bakker 2004). This is consistent with Hobfoll’s (2001) conservation of resources (COR) theory in that the experience of burnout, as a result of depleted resources, can create the willingness to elude the exhausting experience, e.g., by quitting the job.

In several studies, burnout has been tested as a mediator between job demands and individual and organizational outcomes (Schaufeli and Bakker 2004). For example, a three-year cross-lagged study carried out among a large sample of Finnish dentists (n = 2555) showed that job demands predicted burnout over time, which, in turn, predicted future depression (Hakanen et al. 2008). Similarly, in a study among 316 Polish teachers, Baka (2015) noted the indirect relationship between job demands (i.e., interpersonal conflict at work, organizational constraints, and quantitative workload) and depression and physical symptoms through burnout. Sheikh et al. (2019) found that burnout mediated the relationship between job demands and work engagement among 271 faculty members working in different universities in Pakistan. In a study by Skaalvik and Skaalvik (2014), conducted among 2569 Norwegian teachers in elementary and middle schools, time pressure and discipline problems predicted emotional exhaustion, which, in turn, predicted increased motivation to leave and decreased job satisfaction.

Taking into account the fact that job demands impact the development of burnout (Azharudeen and Arulrajah 2018; Bakker and Demerouti 2007; Näring et al. 2006; Van Droogenbroeck et al. 2014), and that the burnout predicts actual turnover and turnover intention (Azharudeen and Arulrajah 2018; Swider and Zimmerman 2010; Van der Heijden et al. 2019), in the current study, we expected that:

Hypothesis 2 (H2).

Burnout mediates the relationship between quantitative job demands (H2a)/emotional job demands (H2b) and turnover intention.

1.3. Spirituality at Work as a Moderating Resource

The JD-R model states that job resources such as task autonomy may moderate the relationship between job stress and psychological outcomes (Demerouti et al. 2001). For instance, they can offset the impact of job demands on strain and burnout and hinder the development of exhaustion (Bakker et al. 2005). The JD-R model has undergone revision (Bakker and Demerouti 2007), resulting in the addition of internal resources (called “personal resources”) to it. Personal resources such as self-efficacy or resiliency involve aspects of self that are generally associated with resiliency and refer to the self-evaluations of one’s ability to manage effectively and influence their environment (Hobfoll et al. 2003; Xanthopoulou et al. 2007b). The extended JD-R model has been successfully implemented in education studies (see, e.g., Bakker et al. 2007; De Carlo et al. 2019; Lorente et al. 2008), suggesting that personal resources provide teachers with strategies with which to better manage demanding situations.

Bickerton et al. (2014) proposed extending the JD-R model by considering spiritual resources as a subcategory of personal resources, which comprises personal beliefs, practices, and experiences related to the sacred. In several studies, it has been supported that spirituality may provide additional resources for dealing with demanding aspects of the work environment, leading to higher work engagement and job performance, promoting a sense of meaning and wholeness, and encouraging personal growth (Bickerton et al. 2014; Cash and Gray 2000; Elmes and Smith 2001).

Our study was focused on one of the spiritual resources, which is “spirituality at work” (also called “spirit at work”), as conceptualized by Kinjerski and Skrypnek (2004, 2006) and Kinjerski (2013). Such terms as “workplace spirituality,” “spirituality in the workplace,” “spirituality in organizations,” and “organizational spirituality” can also be found in the literature, generally capturing similar concepts (Karakas 2010; Kinjerski and Skrypnek 2004). Kinjerski and Skrypnek (2006, pp. 21–22) defined spirituality at work as a distinct experience, which comprises the following four components: (a) engaging work: a profound feeling of well-being, a belief that a person is engaged in meaningful work that has a higher purpose; an awareness of the consistency (alignment) between the individual’s core beliefs and the values of their organization; and a deep sense of authenticity; (b) spiritual connection: a sense of connection to something larger than self (e.g., a Higher Power or a God-within presence and/or a deep connection to humankind or nature); (c) sense of community: a sense of connectedness to others (feelings of love, genuine caring, and intimacy within the team) and common purpose; and (d) mystical or unitive experience: a positive state of energy or vitality; a feeling of perfection and transcendence; and a profound experience of bliss, joy, and ecstasy (see also Kinjerski 2013). This conceptualization is consistent with other definitions of spirituality at work, within which it has been seen as a multifaceted construct, associated with seeking meaningful and purposeful work, experiencing a deep connection or relationships with workmates, based on integrity, solidarity, and the common goal, and harmony or unity between one’s fundamental beliefs and their organizations’ values (Ashmos and Duchon 2000; Giacalone and Jurkiewicz 2003; Milliman et al. 2003). However, it is noticeable that Kinjerski and Skrypnek’s (2006) definition more clearly describes the nature of the individual experience of spirituality at work than other conceptualizations.

A large amount of data supports the beneficial role of spirituality at work in individual and organizational outcomes, including increased employee well-being, job satisfaction, employee job performance and organizational performance, work engagement, and organizational commitment (Fry 2003; Karakas 2010; Kinjerski and Skrypnek 2004; Kolodinsky et al. 2008; Kumar and Kumar 2014; Rego and Pina e Cunha 2008). Regarding employee burnout and turnover intention, only a few studies have considered the effect of spirituality at work on these outcomes. For example, Galea (2014) found that spirituality was related to lower burnout among nurses in difficult situations, even after adjusting for personality and well-being. In a study of 315 Italian employees (Dal Corso et al. 2020), workplace spirituality had a negative direct effect on burnout and mediated the relationship between supervisor integrity on employee burnout. Hong (2012) noted the effect of perceived workplace spirituality on intention to leave in a sample of community mental health center professionals. In another study, workplace spirituality was related to turnover intention among Jordanian academics, and this relationship was partially mediated through loneliness at work (Ghadi 2017). Aboobaker et al. (2019) showed that two dimensions of work spirituality—meaning at work and spiritual community—were directly and indirectly (via well-being) related to intention to stay among 523 teachers working in technical educational institutions in India.

Consistent with the JD-R model (Demerouti et al. 2001) and the moderating effects attributed to spiritual resources (Pargament 1997; Park et al. 1990), several studies have supported the buffering role of spirituality in the relationship between job demands and psychological, emotional, and health- and job-related outcomes (Zou and Dahling 2017). In the longitudinal study by Kim and Seidlitz (2002), spirituality has been found to moderate the effect of stress on negative affect and physical symptoms, controlling for the use of various coping strategies. Kumar and Kumar (2014) noted that workplace spirituality buffered the adverse effect of workplace stress on health among managers. Scherer et al. (2016) demonstrated that among 355 volunteers from various nonprofit organizations, the relationship between burnout and intention to quit was moderated by individual spirituality, such that the relationship was weaker when the spirituality was higher. The results of the aforementioned studies are generally consistent with the COR theory (Hobfoll 2001), which states that persons with better access to resources may be less negatively affected by resource depletion caused by severe stressful situations such as overload or emotional strain, thanks to the substitution for the lost or absorbed resources with additional resources.

Although the literature on the role of spirituality in organizations has proliferated in the past few decades (Dal Corso et al. 2020), there is still a shortage of studies exploring spirituality at work as a potential buffer between job demands and negative job outcomes such as burnout and turnover intention. Therefore, within the framework of the JD-R and COR theories and based on the results of previous studies on the benefits of spirituality, we expand our mediational model by including spirituality at work as a potential moderator of the relationships between job demands, burnout, and turnover intention. Conceptual models of moderated mediations (see Figure 1) were prepared based on the following hypothesis:

Hypothesis 3 (H3).

Spirituality at work moderates the relationships between job demands, burnout, and turnover intention. More specifically, spirituality at work attenuates the direct and indirect effects (via burnout) of quantitative job demand (H3a)/emotional job demands (H3b) on turnover intention.

2. Materials and Methods

2.1. Participants

Table 1. Sociodemographic characteristics of the sample.

	n	%
Gender
Female	834	87.6
Male	118	12.4
Age (M ± SD)	46.54	9.59
Religion
Religious	747	78.5
Non-religious	51	5.3
Agnostic	142	14.9
N/A	12	1.3
Relationship status a
Formal or informal relationship	751	78.9
Single	106	11.1
Divorced	58	6.1
Separated	10	1.1
Widow/widower	20	6.1
N/A	1	0.1
Job seniority (M ± SD)	21.12	11.29
School
Primary	816	85.7
Secondary	136	14.3
Type of school
Public	913	95.9
Private	33	3.5
N/A	6	0.6
Number of pupils in a school
<50	9	1.0
50–200	240	25.2
200–500	436	45.8
500–1000	243	25.5
>1000	16	1.7
N/A	8	0.8
Location of school
Village	276	29.0
Town with up to 20,000 inhabitants	170	17.8
Town with 20,000–100,000 inhabitants	213	22.4
Town with more than 100,000 inhabitants	290	30.5
N/A	3	0.3

Note: ^a For the relationship status, multiple responses were allowed. M = mean, SD = standard deviation. n = 952.

presents the sociodemographic characteristics of the participants. The sample consisted of 952 teachers: 834 women (87.6%) and 118 men (12.4%) from all (i.e., 16) provinces (voivodeships) in Poland. The average age of the participants was 46.54 years (SD = 9.59; min. = 23, max. = 72). The majority of participants (78.5%) declared themselves religious, mostly Roman Catholic (87.7% of religious participants). Other religious denominations were: Protestants (1.2%), Buddhists (0.3%), Jehovah’s Witnesses (0.3%), Seventh-day Adventists (0.1%), and Old Catholics (0.1%). Among religious participants, 10.3% did not identify with any particular religious denomination. Nearly four out of five participants (n = 751; 78.9%) were in formal or informal relationships. A similar percentage of participants (n = 750; 78.8%) had at least one child.

The average job seniority as a teacher was 21.12 years (SD = 11.29; min. = 1; max. = 52). The majority of participants were primary school teachers (85.7%) who worked in public schools (95.9%), in schools located in towns with more than 20,000 inhabitants (52.9%), and in schools with 50–500 pupils (71.0%). The participants were teaching various subjects such as languages (e.g., Polish, English, German, and French), natural sciences (e.g., chemistry, biology, physics, mathematics, and information technology), social sciences (geography, history, and social studies), the arts (music and fine arts), religion, physical education, early childhood education, pedagogy, and psychology.

2.2. Measures

The factorial structure of all measures used in the study was presented in Supplementary Material.

2.2.1. Quantitative and Emotional Job Demands

Quantitative and emotional job demands were assessed using the items from the Copenhagen Psychosocial Questionnaire (COPSOQ; Kristensen et al. 2005; Polish adaptation by Baka 2019). Quantitative job demands were measured with three items (e.g., “Do you have to work very fast?”), scored on a 5-point Likert scale ranging from 1 = “never/hardly ever” to 5 = “always.” Emotional job demands were assessed using two items (e.g., “Is your work emotionally demanding?”). The response options ranged from 1 = “to a very small extent” to 5 = “to a very large extent.” The total score for each scale was calculated by averaging the responses to the relevant items. A higher score indicates a higher level of quantitative/emotional job demands. In the present study, Cronbach’s alpha coefficient was 0.77 for quantitative job demands and 0.76 for emotional job demands.

2.2.2. Spirituality at Work

To measure spirituality at work, we used the Spirit at Work Scale (SAWS; Kinjerski and Skrypnek 2006) in the Polish version prepared by Charzyńska and Gardzioła (2015). The SAWS is made up of 18 items, comprised of 4 subscales: “engaging work” (7 items, e.g., “I am fulfilling my calling through my work.”), “sense of community” (3 items, e.g., “I experience a real sense of trust and personal connection with my coworkers.”), “spiritual connection” (3 items, e.g., “I experience a connection with a greater source that has a positive effect on my work”), and “mystical experience” (5 items, e.g., “At moments, I experience complete joy and ecstasy at work.”). Items are assessed using a 5-point Likert scale ranging from 1 (“completely untrue”) to 6 (“completely true”). The score for each subscale is calculated by summing up the responses to relevant questions. The total score for spirituality at work is calculated by summing up the scores for the three subscales. Higher scores indicate higher levels of spirituality at work. In the current study, we used only the total score for the SAWS. Cronbach’s alpha reliability for the SAWS was 0.91.

2.2.3. Burnout

To measure burnout, we used the Polish adaptation (Baka and Basinska 2016) of the Oldenburg Burnout Inventory (OLBI; Demerouti et al. 2003, 2010). The OLBI is made up of 16 items, which constitutes two subscales: “exhaustion” (8 items, e.g., “After working, I have enough energy for my leisure activities”) and “disengagement” (8 items, e.g., “It happens more and more often that I talk about my work in a negative way”). Both subscales include four positively worded items and four negatively worded items. Items are scored using a 4-point Likert scale, ranging from 1 (“Agree”) to 4 (“Disagree”). A higher level of the subscale score indicates a higher level of exhaustion/disengagement. Based on the results of the examination of the OLBI’s factorial validity (see Supplementary Material), in this study, we calculated the total score for the OLBI by averaging responses to solely negatively worded items. Cronbach’s alpha coefficient for this modified measure was 0.87.

2.2.4. Turnover Intention

The turnover intention was assessed using a single question prepared by Spector et al. (1988): “How often have you seriously considered quitting your current job over the past 6 months?”. The item was scored using a 6-point Likert scale (from 1 = “Never” to 6 = “Extremely often”). A single item for measuring turnover intention has been commonly used in studies in the field of occupational psychology and teacher education, and its reliability and validity have been repeatedly confirmed (Hong 2012; Sprung et al. 2012; Troesch and Bauer 2020).

2.3. Procedure

The research was carried out in accordance with the Declaration of Helsinki, and its ethical guidelines were approved by the Research Ethics Committee at the University of Silesia in Katowice, Poland. The study was conducted online between January and March 2021. The research team members, assisted by the I.T. department of the University of Silesia in Katowice, sent 9072 e-mails to primary and secondary schools from all over Poland (i.e., from all 16 voivodeships). Among these, 30 e-mail addresses were invalid and thus did not reach the recipients. The e-mails were addressed to the headteachers and included an introductory note describing the purpose of the study, the procedure of the study, the anonymous and voluntary nature of the study, and the rights of the participants, followed by the link to the online study (LimeSurvey platform). The headteachers were asked to disseminate the information about the study among the teachers from their schools and encourage them to participate in the study. Online informed consent was obtained from all the participants. The average time of completing the questionnaires was 17 min (SD = 9).

2.4. Statistical Analysis

The data analysis began with the inspection of missing data. Persons who only opened the questionnaires but did not respond to any questions were excluded from further analyses. To examine the missingness mechanism for the rest of the data, we performed the Little’s (1988) Missing Completely at Random test. After imputing the data with the Expectation-Maximization (EM) algorithm, we calculated the means, standard deviations, and bivariate correlations between the study variables.

The mediation and moderated mediation models were tested using the PROCESS macro for SPSS, version 3.5 (Hayes 2013). To examine the significance of the indirect effects, we used the bootstrapping method, which is widely regarded as the optimal solution (Shrout and Bolger 2002). We computed k = 10,000 bootstrapping subsamples and used a 95% confidence interval (CI 95%). The indirect effect is deemed significant if the particular confidence interval does not contain zero.

First, we tested two mediational models using Model 4 (Hayes 2013) to examine whether the relationship between quantitative/emotional job demands and turnover intention was mediated by burnout. Model 4a included quantitative job demands as the independent variable (X), burnout as the mediator (M), and turnover intention as the dependent variable (Y). Model 4b was identical with Model 4a, except for the independent variable, which, in Model b, was emotional job demands. We tested two separate models to avoid the problem of multicollinearity due to a fairly strong correlation between two types of job demands (i.e., r = 0.56; see also Van Vegchel et al. 2004).

Next, we used Model 59 (Hayes 2013) to examine two moderated mediation models, in which spirituality at work was predicted to moderate the mediational relationships between quantitative (Model 59a)/emotional job demands (Model 59b) and turnover intention through burnout. All variables that defined the product term were mean-centered. Based on the previous studies, all models were adjusted for gender and job seniority (Bauer et al. 2006; Borman and Dowling 2008; Guarino et al. 2006).

Effect sizes for indirect effects and moderation were calculated as completely standardized indirect effects (ab_CS) and the effect-size metric f², respectively (Aiken and West 1991; Preacher and Kelley 2011). All the calculations were computed using IBM SPSS version 26.0 (IBM Corp. 2019).

3. Results

3.1. Missing Data Analysis

The percentage of missing data was 0.53%. There were no variables with the missing rate exceeding 2.5%. The insignificant p-value for the Little’s MCAR test (χ²(1815) = 1910.35; p > 0.05) indicated the lack of relationship between the missingness of data and any values (observed or missing). Based on these results, to manage the missing data, we performed a single imputation by EM algorithm using the Missing Value Analysis Module in SPSS version 26.0 (IBM Corp. 2019).

3.2. Preliminary Analysis

Table 2. Means, standard deviations, and zero-order correlations between the study variables.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)
(1) Quantitative demands	1
(2) Emotional demands	0.56 ***	1
(3) Spirituality at work	−0.13 ***	−0.05	1
(4) Burnout	0.41 ***	0.37 ***	−0.44 ***	1
(5) Turnover Intention	0.29 ***	0.29 ***	−0.40 ***	0.54 ***	1
(6) Gender a	0.04	−0.09 **	−0.05	0.03	0.07 *	1
(7) Job seniority (years)	0.07 *	0.08 *	0.12 ***	−0.01	−0.04	0.00	1
M	9.16	11.27	54.91	2.31	2.09	87.6% a	21.12
SD	2.30	2.04	12.84	0.66	1.36	–	11.29
Range	3–15	3–15	13–78	1–4	1–6	–	1–52

Note: * p < 0.05, ** p < 0.01, *** p < 0.001. Gender was dummy-coded (0 = women, 1 = men). ^a The value presents the percentage of women in the sample. M = mean, SD = standard deviation. n = 952.

presents means, standard deviations, and zero-order correlations between the study variables. Nearly one-third of participants (31.1%) had at least sometimes considered quitting their job during the previous six months. Quantitative and emotional job demands correlated positively with burnout (p < 0.001) and turnover intention (p < 0.001). There was also a positive correlation between burnout and turnover intention (p < 0.001). Spirituality at work correlated negatively with quantitative job demands (p < 0.001) but not with emotional job demands (p = 0.14). Spirituality at work was also negatively related to burnout (p < 0.001) and turnover intention (p < 0.001). Being a man correlated negatively with emotional job demands (p = 0.008) and positively with turnover intention (p = 0.026). Longer job seniority was positively related to quantitative (p = 0.027) and emotional job demands (p = 0.021), and spirituality at work (p < 0.001).

3.3. Mediation Models

3.3.1. Model 4a: Quantitative Job Demands, Burnout, and Turnover Intention

The detailed results for the mediation models are presented in Table S1 in the Supplementary Materials. The direct effect of quantitative job demands on turnover intention was significant (b = 0.053; SE = 0.018; p = 0.003; β = 0.09). The point estimate of the indirect effect between quantitative job demands and turnover intention through burnout was 0.120 (SE = 0.012), and the 95% bootstrap confidence interval ranged from 0.097 to 0.144, which indicated that the indirect effect was significant. More specifically, quantitative job demands were positively related to burnout (b = 0.116, SE = 0.009, p < 0.001; β = 0.41) which, in turn, increased the turnover intention (b = 1.035, SE = 0.062, p < 0.001; β = 0.50). The effect size for the indirect effect was at least medium (Kenny 2021; Preacher and Kelley 2011), which was suggested by the value of ab_CS at 0.203 (CI 95% (0.167, 0.240)). The total effect of quantitative job demands on turnover intention was significant (b = 0.173; SE = 0.018; p < 0.001; β = 0.29).

3.3.2. Model 4b: Emotional Job Demands, Burnout, and Turnover Intention

For Model 4b (see Table S1), the direct effect of emotional job demands on turnover intention was significant (b = 0.079; SE = 0.020; p < 0.001; β = 0.12). We also noted a significant indirect effect of emotional job demands on turnover intention via burnout (b = 0.124; SE = 0.013; CI 95% (0.099, 0.150)). Specifically, emotional job demands were positively related to burnout (b = 0.122; SE = 0.010; p < 0.001; β = 0.38), which in turn increased turnover intention (b = 1.018; SE = 0.061; p < 0.001; β = 0.49). The effect size was at least medium (ab_CS = 0.186; CI 95% (0.151; 0.222); Kenny 2021; Preacher and Kelley 2011). The total effect of emotional job demands on turnover intention was significant (b = 0.204; SE = 0.021; p < 0.001; β = 0.31).

3.4. Moderated Mediation Models

3.4.1. Model 59a: Spirituality as a Moderator of the Relationships between Quantitative Job Demands, Burnout, and Turnover Intention

The results of testing the moderated mediation models are presented in

Table 3. Moderated mediation analyses when using spirituality at work as a moderator.

	Outcome: Burnout			Outcome: Turnover Intention
Predictors	b	SE	t	b	SE	t
Model 59a
Quantitative demands (quant. dem.)	0.101	0.008	13.04 ***	0.061	0.017	3.57 ***
Spirituality (spir.)	−0.020	0.001	−14.50 ***	−0.018	0.003	−5.58 ***
Interaction 1 (quan. dem. × spir.)	0.000	0.001	−0.70
Gender	−0.003	0.054	−0.06	0.169	0.108	1.56
Job seniority	0.001	0.002	0.33	−0.002	0.003	−0.61
Burnout (burn.)				0.840	0.066	12.75 ***
Interaction 2 (quant. dem. × spir.)				0.001	0.001	1.16
Interaction 3 (burn. × spir.)				−0.022	0.004	−5.02 ***
R	0.564			0.591
R2	0.318			0.350
F statistic	F(5, 946) = 88.13; p < 0.001			F(7, 944) = 72.54; p < 0.001
Model 59b
Emotional demands (emo. dem.)	0.114	0.009	13.06 ***	0.089	0.019	4.66 ***
Spirituality (spir.)	−0.021	0.001	−15.53 ***	−0.019	0.003	−6.05 ***
Interaction 1 (emo. dem. × spir.)	−0.001	0.001	−2.15 *
Gender	0.089	0.054	1.65	0.229	0.109	2.10 *
Job seniority	0.001	0.002	0.41	−0.002	0.003	−0.66
Burnout (burn.)				0.816	0.066	12.38 ***
Interaction 2 (emo. dem. × spir.)				0.001	0.001	0.85
Interaction 3 (burn. × spir.)				−0.020	0.004	−4.52 ***
R	0.567			0.596
R2	0.321			0.356
F statistic	F(5, 946) = 89.54; p < 0.001			F(7, 944) = 74.40; p < 0.001

Note: * p < 0.05, ** p < 0.01, *** p < 0.001. b = unstandardized coefficient; SE = standard error; t = t-test; n = 952.

and Figure 2. For Model 59a, spirituality moderated only the relationship between burnout and turnover intention (b = −0.022; SE = 0.004; p < 0.001; β = −0.14). This relationship was significant at all three levels of the moderator (i.e., one standard deviation below the mean, the mean, and one standard deviation above the mean) but decreased with the increase in the level of spirituality at work (see Table S2); for the consecutive levels of moderators, β was 0.54, 0.41, and 0.27, respectively.

The interaction term accounted for a large (f² = 0.027) amount of unique variance in turnover intention (Kenny 2018). To facilitate the interpretation of the moderating effect, we plotted the predicted turnover intention by burnout, separately for low, mean, and high levels of spirituality at work (see Figure 3).

The comparison of conditional indirect effects using pairwise contrasts showed that although the indirect effect of quantitative job demands on turnover intention was significant in all groups differing in levels of spirituality at work (i.e., low, medium, and high levels), the effect decreased (β = 0.20, 0.14, and 0.09, respectively; see also Table S3) as the level of spirituality at work increased. In other words, the relationship between quantitative job demands and turnover intention through burnout was stronger for those teachers who scored low in spirituality at work and weaker for those high in spirituality at work.

3.4.2. Model 59b: Spirituality as a Moderator of the Relationships between Emotional Job Demands, Burnout, and Turnover Intention

For Model 59b, spirituality moderated two paths (see Table 3): the relationship between emotional job demands and burnout (b = −0.001; SE = 0.001; p = 0.032; β = −0.05), and the relationship between burnout and turnover intention (b = −0.020; SE = 0.004; p < 0.001; β = −0.12). For both paths, the relationship between the variables decreased with the increase in the level of spirituality at work (i.e., low, medium, and large), despite being significant at all levels of the moderator (emotional job demands–burnout: β = 0.40, 0.35, and 0.30, respectively; burnout–turnover intention: β = 0.51, 0.39, and 0.27; see also Table S4). Figure 4 illustrates the moderating effect of spirituality at work on both paths. The effect size of the moderating effect was 0.005 for the relationship between emotional job demands and burnout, and 0.022 for that between burnout and turnover intention. The first effect is small, whereas the second may be regarded as quite large (Kenny 2018).

Similar to the results for Model 59a, the comparison of conditional indirect effects showed that with an increase in the level of spirituality at work, the conditional indirect effect decreased (β = 0.21, 0.14, and 0.08, respectively; see also Table S5). This means that teachers with higher levels of spirituality at work experienced a weaker relationship between emotional job demands and turnover intention through burnout than teachers with lower levels of spirituality at work.

4. Discussion

The main purpose of the current study was to examine the moderating role of spirituality at work in the direct and indirect relationships between job demands and turnover intention. The study was conducted among teachers because teaching is a demanding and stressful occupation and many teachers as human service professionals are susceptible to suffering from burnout and, in consequence, consider leaving their job (Hakanen et al. 2006; Lorente et al. 2008). Moreover, teachers are generally eager to find meaning and purpose in what they do at their workplace (Khasawneh 2011), which makes the exploration of spirituality at work highly relevant in this group.

As expected, we noted a direct effect of quantitative job demands/emotional job demands on turnover intention (H1a and H1b supported). This finding is consistent with the results of the previous studies, which noted that heavy workload, time pressure, and emotional strain stemming from the countless and sometimes tense and unpleasant relations with students, their parents, other teachers, and principals predicted turnover intention (Liu and Onwuegbuzie 2012; Räsänen et al. 2020).

Consistently with H2, burnout mediated the relationship between quantitative job demands (H2a)/emotional job demands (H2b) and turnover intention. This is in line with the results of the recent longitudinal study among 262 Norwegian high school teachers (Skaalvik and Skaalvik 2020), which showed the indirect effect of time pressure on motivation to quit through emotional exhaustion. In a similar vein, a study by Scherer et al. (2016) demonstrated the mediational role of burnout between poor person–organization fit and intention to quit among volunteers. The findings of the current study are also consistent with the considerable number of studies showing the positive relationship between job demands and burnout (Baka 2015; Stelmokienė et al. 2019; Tuxford and Bradley 2015) and burnout and turnover intention (Madigan and Kim 2021; Scherer et al. 2016). Workload, a fast pace of working, and emotional job demands may exceed teachers’ adaptation abilities, contributing to growing feelings of exhaustion and disengagement, which, in turn, are related to the increased consideration for leaving the current job. A recent metanalysis by Madigan and Kim (2021) showed that burnout and job satisfaction together explained 27% of the variance in teachers’ intention to quit; importantly, burnout symptoms accounted for the majority of this explained variance. Personal (e.g., health problems and lack of rest), familial (e.g., work-family conflict), and social (limited contacts with friends and colleagues) costs of continuing the current job, which the individual perceives as draining their resources, may be so severe that they begin to think about leaving their job (De Carlo et al. 2019; Demerouti et al. 2001). In this sense, increased turnover intention may be seen as a coping response to the threat and a way of protecting oneself from losing more resources (Baka 2015; Hakanen et al. 2006). Considering quitting the job may also bring some relief and comfort to teachers suffering from burnout, reassuring them that if their occupational situation worsens, they can leave their current job and find another, which would be less demanding and more rewarding.

To verify H3, we built and tested the moderated mediation models, in which spirituality at work served as a moderator of the direct and indirect relationships between job demands and turnover intention. The results of the analysis showed that the indirect relationship through burnout was moderated by spirituality at work for both quantitative and emotional job demands; more precisely, the indirect effect of job demands on turnover intention through burnout was weaker for teachers with higher levels of spirituality at work compared with teachers with lower levels. By contrast, moderation of a direct effect of either quantitative or emotional job demands on turnover intention was not noted. Thus, H3a and H3b were partially corroborated. These results support the role of spirituality at work as a spiritual resource of teachers (Dal Corso et al. 2020; Hong 2012) and are generally congruent with the assumptions of the JD-R and COR models.

Although in the current study the indirect effect of job demands on turnover intention was conditional on values of spirituality at work in both moderated mediation models, it should be noted that for the first stage of the moderated mediation, a significant interaction effect was noted between emotional job demands and spirituality at work, at the same time being insignificant for quantitative job demands and spirituality at work. When discussing this result, it should be highlighted that because factors causing stress are specific for each person (Kyriacou 2001), the key role in its development and progression may be attributed not to the objective presence and intensity of the potential stressors (stress per se) but the way they are interpreted (stress appraisal; Lazarus and Folkman 1984). In other words, beliefs about stressors often play a crucial role in the stress response. Therefore, those teachers who think that emotionally stressful situations at work can bring some benefits to them and their schools may feel less stress from their occupational duties and thus experience burnout symptoms less frequently than teachers who perceive such situations as only demanding and heavy (Kim et al. 2020). Having spiritual resources in the form of spirituality at work may allow the individual to interpret emotional job demands as less stressful, thus protecting teachers from the development of burnout (Kumar and Kumar 2014). In the face of emotional job demands, spirituality at work may foster effective coping and help balance job demands with the meaningfulness of the job-related relationships, in this way protecting a person from the depletion of resources and thus from developing burnout (Zou and Dahling 2017). By contrast, quantitative job demands refer to more objective work conditions—the time and resources the employee has to complete the task are limited, which objectively hinder the accomplishment of duties on time, being less dependent on subjective interpretation of the stressful situation. This result supports the need for distinguishing between quantitative and emotional job demands when studying their interplay with spirituality at work in job-related outcomes (Van Vegchel et al. 2004).

Regarding the second stage of the moderated mediation models, spirituality at work buffered the relationship between burnout and turnover intention for both tested models (i.e., 59a and 59b). This means that when experiencing burnout, teachers with higher levels of spirituality at work consider quitting their job less frequently, compared with teachers with lower levels of spirituality at work. When experiencing burnout, spiritual teachers may resort to their spiritual resources, which could be related to spiritual coping or meaning-making processes that help dissipate the burnout symptoms over time (Scherer et al. 2016). A deep sense of meaning and purpose in life provided by teaching may enhance teachers’ resilience and willingness to combat tough situations in the workplace and not resign from their job even despite adverse consequences of burnout (Van Wingerden and Poell 2019). Spirituality at work also helps create meaning and purpose in the work, promoting motivation and engagement (Fouché et al. 2017; Van Wingerden and Van der Stoep 2018), in this way potentially mitigating the relationship between burnout and turnover intention. Moreover, spirituality at work may diminish organizational cynicism (Shrestha and Jena 2021) and thus enhance trust in co-workers and perceived organizational support, both of which may function as job resources attenuating the effect of burnout on turnover intention.

Consistent with our results, in a study by Scherer et al. (2016) volunteers with higher individual spirituality were less likely to intend to quit, even if they experienced burnout symptoms, compared with volunteers with lower individual spirituality. The authors elegantly summarized this result with the phrase “Grin and bear it.” Analogously, teachers with high levels of spirituality at work may identify with their job very closely, treat teaching not merely as a formal profession but rather as a vocation or a calling, and highly appreciate transcendent experience through the work process (Aboobaker et al. 2019; Paloutzian et al. 2003). The experience of meaningful work may motivate teachers to go above and beyond the average requirements of their work. This explanation is consistent with the fact that people with high levels of spirituality demonstrate more perceived control than less spiritual peers, which may give them more strongly the belief that they can overcome work-related difficulties (Jackson and Bergeman 2011). Thanks to these beliefs, highly spiritual teachers may view problems as opportunities for personal growth rather than hindrances (Zinnbauer et al. 1999).

This line of reasoning is further supported by the fact that our sample consisted mostly of religious people, predominantly Roman Catholics. Catholics are encouraged to treat their work as a sacred duty, a way to serve God, and thus are expected to maintain patience, stoic perseverance, and hope (King et al. 2020), even if they encounter serious obstacles on their way (see Scherer et al. 2016). If this is the case, highly spiritual teachers may feel morally obliged to fulfill their duties as best as possible to please God and thus, when experiencing burnout, tend to consider quitting their job less frequently than their less spiritual peers.

4.1. Practical Implications

This study does not only contribute to the current knowledge on the role of spirituality at work in the educational context but also provides important implications for school policymakers and administrators. First of all, the results of the study revealed that both quantitative and emotional job demands are significant predictors of burnout and turnover. This means that increasing teachers’ quantitative and emotional job demands may result in long-term adverse consequences for both employees and schools. Having said that, quantitative and emotional job demands are often unavoidable, especially in such demanding settings as schools (Hakanen et al. 2006). Principals and other school administrators should try to protect the teachers from the effects of job demands, provide them with adequate job resources, and enhance the development of employees’ personal resources (Stelmokienė et al. 2019). Improvements can involve reducing the repetition of paperwork and taking care of a good fit between school tasks and teachers’ abilities, which would be possible thanks to recognizing each teacher’s potential (Ravalier and Walsh 2017). Some benefits can also be brought by offering teachers training in time planning.

Regarding emotional job demands, both principals and colleagues should provide teachers with more emotional support, especially in the case of frequent and intense interactions or difficult situations with students or their parents (Geiger and Pivovarova 2018; Skaalvik and Skaalvik 2018). Emotionally demanding tasks can be varied with nonemotionally demanding ones, giving some relief and distance from work (Van Vegchel et al. 2004). The perceived emotional burden may also be mitigated by offering the teachers emotion regulation training (Chang 2009; Pietarinen et al. 2013), emotional intelligence training (Schoeps et al. 2019), or stress management programs (Ansley et al. 2021; Taylor et al. 2021), which help reduce perceived stress and increase the ability to cope with negative emotions. Principals and school managers should also emphasize the importance of rest and relaxation and encourage teachers to engage in hobbies and pleasant activities in their spare time. Moreover, teachers may be encouraged to participate in daily spiritual practices such as meditation, yoga, or prayer, which have been considered emotion-focused strategies that help reduce and manage negative emotions in the workplace (Ancona and Mendelson 2014; Chirico et al. 2020; Giacalone and Jurkiewicz 2003). For instance, a quasi-experimental study by Sharp Donahoo et al. (2018) demonstrated that prayer and mindfulness effectively reduced levels of stress and compassion fatigue among teachers and professional staff working in special education. Mindfulness practices and job crafting have also been suggested to reduce teacher stress and burnout (Ancona and Mendelson 2014; Van Wingerden et al. 2017).

The current study supports the notion that spirituality at work buffers the impact of job demands on turnover intention through burnout. The results indicate the need to create a workplace environment that will satisfy teachers’ spiritual needs (Dal Corso et al. 2020) by helping them find meaning in their work and nurturing and nourishing the team spirit and the sense of community through positive relationships (Aboobaker et al. 2019). A large number of studies showed that making the workplace (including school environment; Janik and Rothmann 2015) meaningful and purposeful fosters employees’ job satisfaction and protects them from experiencing burnout and leaving the job. May et al. (2004) identified three factors that enhance meaningfulness at the workplace: work-role fit (the match between the employee’s values, beliefs, and behaviors and the work-role; Brief and Nord 1990), job enrichment (engaging in interesting and challenging tasks that facilitate personal growth; Chen et al. 2011), and co-worker relationships (communality, sense of belongingness, social support between the members of staff, and a common sense of purpose; Geiger and Pivovarova 2018). The sense of connectedness and attachment to the purpose is also facilitated by supportive supervisor–employee relationships (Janik and Rothmann 2015), in which the leader does not only focus on the instrumental part of work (job being carried out) but also on employees’ needs, experiences, and work-related problems (Dal Corso et al. 2020). The positive organizational climate in schools may also be shaped by spiritual leadership, which is defined as a value-based holistic approach to leadership promoting integrity, goodness, teamwork, knowing, wholeness, and interconnectedness (Afsar et al. 2016). A spiritual leader can create a positive workplace by emphasizing spiritual values, being authentic, responsive, and honest, sharing their own spiritual well-being with the staff members and encouraging them to express their spirituality through their work (Dal Corso et al. 2020; De Carlo et al. 2019). Increasing the principals’ awareness and knowledge concerning spiritual leadership and fostering their skills, competencies, and motivation required to embed spirituality into school environments may bring benefits in the form of teachers’ higher job satisfaction, lower burnout, and reduced intention to leave the job (Yang and Fry 2018).

When discussing the positive outcomes of implementing spirituality into the workplace, the risk of misuse or misappropriation of employee’s spirituality should not be omitted (Gotsis and Kortezi 2008; Lips-Wiersma et al. 2009). The awareness of many benefits brought by employee’s spirituality to the workplace in some cases may evoke the inclination to manipulative or instrumental use of their spiritual resources just to increase productivity, without caring about the employee’s well-being. In this context, the result of the study by Scherer et al. (2016) is worth mentioning since it showed that spirituality at work attenuates the relationship between burnout and turnover intention. This result, along with the result of the current study, suggests that highly spiritual teachers tend to stay in their jobs even if they feel burned out. Employers should not take advantage of this situation or belittle the problem by claiming that since an employee does not want to quit, they do not experience burnout, or its symptoms are not “serious enough” to require any actions. Instead, in such a situation, employers should try to diagnose the causes of burnout, change “the alterables” in schools and individual teachers, and help the teacher find job resources that can effectively reduce burnout and its adverse health consequences (Hakanen et al. 2006).

4.2. Strengths and Limitations of the Study

To the best of our knowledge, this is the first study to explore the moderating role of spirituality at work in the relationships between job demands, burnout, and turnover intention in educational settings. Our study demonstrated that the indirect effect of job demands on turnover intention through burnout decreased as the level of spirituality increased. The results of the study point out that spirituality at work acts as a protective factor, mitigating the negative impact of teachers’ job demands on their turnover intention. One of the strengths of our study lies in the size of the sample (n = 952) and the fact that the teachers who participated worked in schools all over Poland and were widely varied in terms of sociodemographics and school-related characteristics. This increases the ecological validity of the study. Another strength is that we used reliable and valid tools to measure the study variables, whose factorial structure we carefully examined before conducting further analyses. In addition, the study was carried out during the second wave of the COVID-19 pandemic, which allowed us to test the extended JD-R model in the extraordinary conditions in which teachers were forced to work.

Besides its strengths, this study also has some limitations, which should be discussed. The main limitation lies in the cross-sectional nature of the study, which precludes us from making unequivocal causal inferences regarding the relationships noted. It is, for example, possible that teachers who experience higher levels of burnout perceive their work as more demanding than teachers who experience lower levels of burnout. Although alternative directions of the relationships between the studied constructs are indeed possible, it should be highlighted that our models were grounded in theory, especially in the JD-R model and, to a lesser extent, on the COR theory, which supports the assumed directions of the relationships between the variables. Nevertheless, longitudinal and intervention studies are required to further explore the buffering effect of spirituality at work in the relationships between job demands, burnout, and turnover intention among teachers. In addition, future studies may consider applying multilevel analysis, in which school-level (e.g., school spiritual climate or spiritual leadership) and teacher-level (e.g., individual spirituality or spiritual coping) spiritual predictors would be included in the same model to comprehensively understand their interactions with job demands and burnout in explaining turnover intention.

Another limitation of the study stems from the fact that we cannot calculate the actual response rate due to the study design and recruiting teachers via convenience sampling. This limits the generalizability of the results of our study. It is likely that teachers who did not decide to participate in the study differed from the participants as regards the study variables. Future studies would benefit from using a probability-based sample of Polish teachers to increase the confidence in the findings. Moreover, our sample was over-represented by women, which constituted 87.6% of the sample. However, this proportion between women and men in the current study roughly mirrors the gender differences in the population of teachers in Poland (i.e., 82% of female teachers; Centre for Education Development 2013). The current sample predominantly consisted of religious teachers, mostly Roman Catholics, which could have affected the results since one of the crucial variables in the current study was spirituality at work, which is related to religiosity to a considerable extent (Oman 2013; Paloutzian and Park 2005). The high rate of religious teachers in the sample is well-justified because Poland is a highly religious country, with more than 90% of Poles describing themselves as religious (Centre for Public Opinion Research 2020). Therefore, our study requires replications in more secular and more religiously diverse countries.

In the current study, we used self-report questionnaires, which may affect the quality of the data to some extent. Future studies may benefit from combining data on job demands from teachers’ self-assessment and principals’ or other teachers’ assessments. In addition, the turnover intention was measured with a single question, which might have reduced the between-person differences in this variable and the strength of its relationships with other variables. However, a single item measuring turnover intention has been frequently used in previous organizational studies, and its good psychometric properties have been supported (Hong 2012; Troesch and Bauer 2020).

As mentioned earlier, a special time in which the current study was carried out (i.e., COVID-19 pandemic) is one of the strengths of this study. However, at the same time, it can be considered as its limitation. During the pandemic, due to sudden changes in the school functioning and the necessity to adjust to them and to reconcile work with family life, teachers might perceive higher levels of job demands than they tended to before the pandemic, which, in consequence, could also have affected the levels of the studied variables. Even if that happened, it probably did not affect the relationships between the variables to a great extent since the teaching profession is stressful and demanding not only in critical situations, and teacher burnout and turnover intention do not only occur in tough circumstances. Therefore, it can be expected that the buffering role of spirituality at work in educational settings will be supported in more ordinary circumstances as well, though perhaps with a smaller effect sizes. Our expectation is also based on the results of previous studies with similar scope (see, e.g., Kim and Seidlitz 2002; Kumar and Kumar 2014; Scherer et al. 2016) and the results of the recent study (Sokal et al. 2021), which generally supported the JD-R model in the teaching context during the COVID-19 pandemic and showed quite similar patterns of relationships between job demands, job resources, and burnout to those noted before the COVID-19 outbreak.

5. Conclusions

This study is innovative in terms of conceptualizing and testing theoretical models in which two kinds of job demands (quantitative and emotional) turned out to be related to the intention to quit through burnout, with the magnitude of the indirect relationships being modulated by the level of spirituality at work. The results of the study bring several important conclusions for theory and practice, especially for occupational psychology and teacher education literature. First, they indicate the need to reduce both qualitative and emotional job demands to prevent teacher burnout and turnover. Second, the study supports the usefulness of including spiritual resources as a subcategory of personal resources when testing the JD-R model in educational settings. Third, it suggests the importance of distinguishing between quantitative and emotional job demands when testing their interactions with spirituality at work and highlights the buffering role of spirituality at work in the relationship between emotional demands and burnout. Last but not least, the results of the study stress the usefulness of considering teachers’ spirituality at work to design an employee-friendly, meaningful, and supportive environment at schools and when planning interventions based on personal resources, which are aimed at reducing teachers’ burnout and turnover intention.