The relations between Principal support and work engagement and burnout: Testing the role of teachers’ emotions and educational level

Abstract

BACKGROUND:

One of the resources that may prevent burnout and foster the work engagement of teachers is Principal support. In this study we aimed to further explore the links between Principal support and work engagement and burnout, by testing the role of the teachers’ emotions and the educational levels of primary schools.

OBJECTIVES:

Testing 1) the mediating role of teachers’ emotions in explaining the relationship between perceived Principal support and work engagement and burnout, 2) differences in tested constructs, and the structural equivalence of the proposed relationship among constructs, between teachers from two educational levels of primary schools.

METHODS:

The study included subsamples of 868 class teachers (lower grades) and 1057 subject teachers (upper grades) employed at 104 primary schools in Croatia. Teachers filled out self-report scales measuring burnout, work engagement, perceived Principal support and the positive and negative emotions experienced in relation to students.

RESULTS:

Emotions partially mediate the relationship between perceived Principal support and work engagement and burnout among teachers. Although class teachers experience higher levels of positive emotions and work engagement compared to subject teachers, the results confirmed the structural equivalence of the hypothesized model across two education levels of primary school.

CONSLUSION:

The Principal support predicts work engagement and burnout in teachers both directly and indirectly via emotions.

Keywords

Primary school well-being work motivation exhaustion disengagement

1 Introduction

Bearing in mind the current socio-historical context of the teaching profession, which is marked by a great number of demands and challenges [1 –4], well as high levels of occupational stress and burnout in teachers [5 –8], research on teachers’ occupational well-being is very important from both theoretical and practical points of view. Today’s teachers cope with fast-changing world, whereby the most obvious influences on teachers’ work stem from rapid development of information and communication technologies and changes in families’ structure. Furthermore, the contextual features of teaching professions around the world include educational reforms with three interrelated policy technologies: the market orientation, managerialism and performativity [9]. Although these elements have different emphases in different countries, unique characteristics of educational reforms allude distancing from traditional public welfare tradition of education provision, i.e. aligning methods of work in schools, colleges and universities with those in private sector. Educational reforms are generally aimed at professionalism of teachers’ occupation. From teachers’ perspective, however, these changes are often reflected in less authority over teaching and students, reduced control over schools’ daily procedures, higher number of administrative duties and prescribed tasks, lower sense of autonomy, professional identity and status in the society, and higher level of accountability to the government, school managements, parents, students and whole society [2 , 10– 12]. All these changes have great potential to affect teachers’ occupational well-being. Bearing that in mind, in the present study we focused on analysing the relationships between several aspects of teachers’ occupational well-being, namely emotions experienced towards students, work engagement and burnout. In addition, we aimed at exploring the potential protective role of perceived Principal support as important factor in preserving these aspects of teachers’ occupational well-being.

1.1 Teaching profession in Croatian educational context

In the last several years, Croatia faced numerous economic difficulties that resulted in low investment in education. Economic downturns have considerably affected state schools and the role of teachers and Principal. In addition, due to political shifts, numerous non-systematic and incoherent educational reforms were applied through the past twenty years [13]. Hence, Croatia belongs to the group of countries with the lowest percentage of teachers who believe that their profession is viewed positively by society [14]. Recent investigations have shown that Croatian teachers at all education levels experienced high levels of stress at work [4, 15]; consequently, it is not surprisingly that near to 30% of them have reported readiness to leave teacher professions [16].

To date, Principal education/certification system is still not defined in Croatia, and employees on the position of Principal have unknown qualifications regarding capabilities and skills to cope with challenging and complex job demands, as well as to provide support to the members of their staff. However, they are responsible for school management, and are the most important for the development of a positive tone of school climate that provides the fulfilment of teachers’ needs for autonomy, competence and relatedness. In this way, Principal support could be crucial for reducing teachers’ stress and promoting their engagement. Teachers appear to be more satisfied in a collaborative school system (that includes reciprocal support among teachers and school principal) in which they play an active part in school managing and decision-making. In addition, various collaborative practices in schools appear to increase teachers’ perception of their profession as socially prestigious [14]. Considering mentioned idiosyncrasies of the contemporary Croatian educational context, we view Principal support as one of the most promising explanatory constructs with respect to teachers’ occupational well-being.

1.2 Defining teachers’ work engagement and burnout

Work engagement is defined as a positive, fulfilling, work-related state of mind that is characterized by three components: vigour (high levels of energy and mental resilience while working), dedication (being strongly involved in the work) and absorption (being fully focused and happily engrossed in the work) [17 –19]. Though some scholars consider work engagement as the opposite of burnout [20], empirical evidence suggests that burnout and work engagement are two distinct concepts that should be assessed independently [19]. Some work characteristics that positively correlate with work engagement are proposed as its resources, such as social support from co-workers and one’s superior, performance feedback, coaching, job autonomy, task variety and training facilities [17]. In accordance with this view, and in the line with the results of previous studies among teachers [e.g. 21, 22], the present study focuses on perceived support from the Principal (or head of school), which is considered as an important factor for developing and maintaining work engagement in teachers.

Burnout refers to “a syndrome of emotional exhaustion, depersonalization and reduced accomplishment which is a special risk for individuals who work with other people in some capacity” [23, p. 347]. It develops over time and is typically characterized by exhaustion (draining of mental energy), cynicism (negative attitude towards work), and reduced professional efficacy (belief that one is no longer effective in fulfilling job responsibilities) [24]. In the present study we have focused on two core dimensions of burnout, namely exhaustion and disengagement. Exhaustion is defined as a consequence of intense physical, affective and cognitive strain, while disengagement refers to distancing oneself from one’s work in general [25]. This kind of operationalization of exhaustion covers not only affective, but also physical and cognitive, aspects of exhaustion. Furthermore, the disengagement component is broader than Maslach’s depersonalization dimension of burnout, which refers primarily to relations directed towards customers [26]. The negative consequences of burnout among teachers are well documented [see review in 27]. On the other side, antecedents of teachers’ burnout are complex and multifaceted and, according to Chang [28], may be classified into various factors: individual (age, gender, personality, self-esteem, coping strategies, resilience etc.), organizational (work demands, class size, school SES, teacher participation in decision-making etc.) and transactional (e.g. perceived Principal support and perceived organizational leadership style). Based on the Chang’s propositions [28], it can be assumed that the lack of perceived Principal support leads to higher levels burnout, i.e. feelings of exhaustion and distancing oneself from the work.

1.3 Perceived Principal support, work engagement and burnout – the central role of teachers’ emotions

A Principal’s behaviour in interaction with teachers has a crucial role in ensuring environmental resources and autonomy for teachers, but also in creating the emotional tone of a school climate. According to Self-Determination Theory [29], supporting teachers’ feelings of competence, autonomy and relatedness facilitate their intrinsic motivation. Intrinsic motivation is seen through energized behaviour for its inherent satisfaction, rather than other consequences. In the model of work engagement proposed by Bakker and Demerouti [30], both intrinsic and extrinsic motivational roles of job resources – such as the Principal’s support – are important for work engagement. The extrinsic role of Principal support is thus instrumental for achieving educational goals in the classroom, while its intrinsic role fosters teachers’ growth, learning and development. Thus, we propose that Principal support may lead to work engagement and burnout, not only directly, but also through teachers’ positive and negative emotions, which result from the appraisal of either presence or lack of the support from the Principal. In other words, teachers’ emotions can be seen both as outcomes of perceived Principal support, as well as the antecedents of work engagement and burnout. While positive emotions are considered as a facilitating factor, negative emotions are viewed as a disturbing factor of work engagement in teachers. With respect to burnout, the opposite pattern of relationships can be expected. Expected links between emotions and work engagement and burnout are based on the results of previous studies which showed that frequent and diverse emotions (pleasant or positive emotions such as love and pride, and unpleasant or negative emotions such as anger and hopelessness) which teachers experience in the relation with students, parents, school members and system, have important influences on teachers, teaching and students [31 –33]. Regarding influences on teachers, it should be emphasized that emotions have an important role in forming teachers’ professional identity, commitment, effectiveness and well-being [34].

Since interactions with students are regarded as the most powerful in terms of evoking emotions [31], in this study we have focused on emotions which teachers experience in relation to their students. Furthermore, besides focusing on unpleasant / negative emotions exclusively, we also considered pleasant / positive emotions, since the relations of positive teachers’ emotions to burnout and engagement in teachers are generally less well explored than those of unpleasant or negative emotions. For example, Chang’s review [28] deals with relations between unpleasant emotions (only) and burnout, showing that repetitive and intense unpleasant emotions can be viewed as precursors of emotional exhaustion and burnout. In sum, we propose the model, according to which positive and negative emotions which teachers experience in relation to their students may mediate the relation between Principal support and work engagement and burnout.

1.4 Perceived Principal support, teachers’ emotions, work engagement and burnout – the role of educational level

The Croatian education system includes pre-school, primary, secondary and higher education. In this study we focused on teachers employed in primary schools. Basic compulsory primary education lasts eight years (grades from 1 to 8) and consists of two phases or levels. Teaching in the first four grades is based on a holistic approach to the curriculum and one class teacher teaches six extensive areas of knowledge. In the higher grades (grades from 5 to 8) curriculum is more differentiated and each subject is taught by subject teacher.

Knowledge about possible differences in constructs which are in the focus of this work, and more generally in work-related well-being, between teachers employed at different educational levels of primary schools is limited. On the other side, studies that compared levels of burnout, engagement and occupational stress between teachers from primary and secondary schools show that primary school teachers are more vulnerable group. For example, authors of study among teachers in Greece [35] based on the earlier findings hypothesized higher burnout and less engagement and performance in teachers who teach older children compared to the teachers who teach younger children. However, their data showed that primary school teachers in Greece experience more stress and higher levels of emotional exhaustion, as compared to the teachers of secondary education, which is explained as the possible effect of many reforms in Greek primary schools. Higher levels of stress in teachers of primary schools, compared to teachers of secondary schools, are also found in Hong Kong, which may be due to higher level of career instability in that education level [36]. Primary school teachers in Turkey also show higher levels of two burnout components (emotional exhaustion and depersonalization) in comparison to secondary school teachers [37]. The study performed on teachers in Macedonia [38] focused on differences between teachers from two levels of primary schools. The results showed that lower-grade primary school teachers in Macedonia experience some stressors (such as lack of authority and implementation of changes in system) as more severe than upper grade teachers. On the other side, data from studies on relatively small samples of teachers of primary schools in Croatia show that primary schools teachers in lower levels (class-teachers) are more satisfied with the job itself and with choice of occupation, and have more positive attitudes toward students compared to their colleagues in higher level, i.e. subject teachers in primary schools [39, 40], which could reflect also on the levels of work engagement and burnout. Data of another study on Croatian teachers [41) are in line with this and show that class teachers experience less burnout compared to subjects teachers. Obtained differences between Croatian teachers from two levels of primary schools are probably due to a number of factors, including differences in the nature of work and their previous education, attitudes toward occupation, the age of students, but also may be due to the differences in the levels of emotions experienced in the everyday work with students. Bearing in the mind the fact that work-related wellbeing of teachers is largely affected with specific national educational system, and the general lack of studies focused at the role of educational level of primary schools in the relation to the other studied constructs beside burnout (especially emotions experienced in the relation to students), we considered this issue as an additional important research objective.

1.5 Research objectives

The first objective relates to the examination of the mediating role of teachers’ positive and negative emotions in explaining the relationship between perceived Principal support and work engagement and burnout among teachers. We hypothesized that emotions would partially mediate this relationship i.e. that perceived Principal support would be related to work engagement and burnout both directly and indirectly via positive and negative emotions. Considering the direction of associations, it is expected that teachers who are more engaged in their work, perceive their Principal as more supportive and experience more positive, but less negative emotions in relation to their students. On the contrary, teachers who feel more “burned” out at work, perceive their Principal as less supportive and experience more negative, but less positive emotions. Finally, teachers who feel more supported by their Principal, experience more positive, and less negative, emotions towards their students.

Furthermore, we also aimed at testing the possible differences in Principal support, teachers’ emotions, work engagement and burnout between class teachers and subject teachers. Considering the results of previous studies performed in Croatian primary teachers [39 –41], we hypothesized that class teachers, in comparison to subject teachers, experience higher levels of positive and lower levels of negative emotions in relations to students, higher levels of work engagement and lower levels of burnout.

Our final objective is related to the testing of the structural equivalence of the proposed relations among Principal support, teachers’ emotions, work engagement and burnout across the two educational levels of primary school in order to prove conceptual robustness of the proposed model. Therefore, in this study we tested the structural equivalence of the hypothesized model across two educational levels – the lower and upper levels of primary education. We expected that the structural paths between variables in the model would be equivalent across samples of class teachers and subject teachers, i.e. the hypothesized model would demonstrate sufficient robustness across the two educational levels.

2 Method

2.1 Participants and procedure

Teachers employed at 104 primary schools in various towns in Croatia (N = 1925) volunteered to participate in the research. Considering the number of questionnaires sent to the members of target population, the response rate in this study was about 50%. In the identification of the sample of schools, we aimed to and succeeded in achieving maximum variations with regard of different areas of the country (twenty counties), as well as urban and rural parts, and school sizes. Schools’ sizes in the terms of number of pupils range from 26 to 1100. The study included teachers who work at the two levels of primary school:

868 teachers at the lower level: class teachers, who teach students from the 1st to the 4th year (students aged from 7 to 11 years old);

1057 teachers at the upper level: subject teachers, who teach different subjects to students from the 5th to the 8th year (students aged from 12 to 15). This sub-sample was heterogeneous regarding subjects, i.e. teachers of all subjects at the upper level (e.g. Croatian language and literature, foreign languages, mathematics, physics, biology, chemistry, geography, history, arts, physical culture etc.) were represented in the sample.

The class teachers were, on average, 42.2 years old (SD = 10.1) and had 17.5 years of teaching experience (SD = 10.8). The subject teachers were, on average, 41.2 (SD = 10.3) years old, and had 14.4 years of teaching experience (SD = 10.8). Females dominated both subsamples, as in the teacher population (821 of the class teachers and 862 of the subject teachers being female).

The study was conducted in autumn 2015 with the informed consent of the school Principal and teachers. Envelopes with questionnaires and cover letter were sent via postal service to schools and distributed to teachers. The teachers were informed about the confidentiality of the information they were sharing, as well as the importance of their contribution to the study. Participation in the study was voluntary and anonymous. Contact persons from schools gathered the completed questionnaires within a period of two weeks and sent them back to the researcher.

2.2 Measures

The questionnaire consisted of self-reported measures of perceived Principal support, teachers’ emotions, burnout, and work engagement. All scales had satisfactory levels of reliability as can be seen in the Table 1.

Table 1
Descriptive statistics and Pearson correlation coefficients between examined variables

Gend. Exper. Supp. Joy Pride Love Anger Fatig. Hopel. Vigor Dedic. Abs. Disen. Exh.

Gend. – 0.06^* – 0.02 0.13^ 0.13^ 0.14^** 0.05^* 0.09^ 0.07^ 0.00 – 0.01 0.00 – 0.03 0.12^

Exper. – – 0.08^ – 0.05^* 0.05^* 0.15^* 0.07^** 0.06^* – 0.03 0.06^* 0.10^ 0.10^ 0.01 0.06^

Supp. – 0.18^ 0.15^ 0.16^ – 0.13^ – 0.14^ – 0.12^ 0.25^ 0.24^ 0.21^ – 0.29^ – 0.22^

Joy – 0.63^ 0.44^ – 0.11^ – 0.03 – 0.09^ 0.34^ 0.33^ 0.32^ 0.63^ – 0.27^

Pride – 0.62^ – 0.05^* 0.02 – 0.06^* 0.43^ 0.38^ 0.38^ – 0.33^ – 0.15^

Love – – 0.19^ – 0.15^ – 0.22^ 0.51^ 0.47^ 0.48^ – 0.44^ – 0.23^

Anger – 64^ 73^ – 37^ – 0.40^ – 32^ 0.47^ 0.53^

Fatig. – 0.61^ – 0.36^ – 0.38^ – 0.32^ 0.51^ 0.74^

Hopel. – – 0.40^ – 0.41^ – 0.35^ 0.48^ 0.51^

Vigor – 0.90^ 0.86^ 0.43^ – 0.51^

Dedic. – 0.85^ 0.38^ – 0.53^

Abs. – 0.38^ – 0.46^

Disen. – 0.66^**

Exh. –

M – 15.76 3.74 4.75 4.46 4.17 2.32 2.88 2.63 4.69 4.78 4.59 1.99 2.22

SD – 10.72 0.65 0.37 0.51 0.61 0.71 0.86 0.71 0.91 0.83 0.84 0.46 0.50

α – – 0.78 0.85 0.86 0.87 0.78 0.90 0.86 0.90 0.82 0.81 0.75 0.83

	Gend.	Exper.	Supp.	Joy	Pride	Love	Anger	Fatig.	Hopel.	Vigor	Dedic.	Abs.	Disen.	Exh.
Gend.	–	0.06^*	– 0.02	0.13^**	0.13^**	0.14^**	0.05^*	0.09^**	0.07^**	0.00	– 0.01	0.00	– 0.03	0.12^**
Exper.		–	– 0.08^**	– 0.05^*	0.05^*	0.15^*	0.07^**	0.06^*	– 0.03	0.06^*	0.10^**	0.10^**	0.01	0.06^**
Supp.			–	0.18^**	0.15^**	0.16^**	– 0.13^**	– 0.14^**	– 0.12^**	0.25^**	0.24^**	0.21^**	– 0.29^**	– 0.22^**
Joy				–	0.63^**	0.44^**	– 0.11^**	– 0.03	– 0.09^**	0.34^**	0.33^**	0.32^**	0.63^**	– 0.27^**
Pride					–	0.62^**	– 0.05^*	0.02	– 0.06^*	0.43^**	0.38^**	0.38^**	– 0.33^**	– 0.15^**
Love						–	– 0.19^**	– 0.15^**	– 0.22^**	0.51^**	0.47^**	0.48^**	– 0.44^**	– 0.23^**
Anger							–	64^**	73^**	– 37^**	– 0.40^**	– 32^**	0.47^**	0.53^**
Fatig.								–	0.61^**	– 0.36^**	– 0.38^**	– 0.32^**	0.51^**	0.74^**
Hopel.									–	– 0.40^**	– 0.41^**	– 0.35^**	0.48^**	0.51^**
Vigor										–	0.90^**	0.86^**	0.43^**	– 0.51^**
Dedic.											–	0.85^**	0.38^**	– 0.53^**
Abs.												–	0.38^**	– 0.46^**
Disen.													–	0.66^**
Exh.	–
M	–	15.76	3.74	4.75	4.46	4.17	2.32	2.88	2.63	4.69	4.78	4.59	1.99	2.22
SD	–	10.72	0.65	0.37	0.51	0.61	0.71	0.86	0.71	0.91	0.83	0.84	0.46	0.50
α	–	–	0.78	0.85	0.86	0.87	0.78	0.90	0.86	0.90	0.82	0.81	0.75	0.83

Note: Gend. = Gender: 1 = male, 2 = female; Exper. = Working experience (years); Supp. = Perceived Principal support; Positive emotions: Joy, Pride, Love; Negative emotions: Anger, Fatig. = Fatigue, Hopel = Hopelessness; Work engagement: Vigor, Dedic. = Dedication, Abs. = Absorption; Burnout: Disen. = Disengagement, Exh. = Exhaustion; ^*p < 0.05, ^**p < 0.01.

Perceived Principal support was measured by the Croatian version of the Supervisor trust and support scale. The scale consists of 7 items (e.g. “My supervisor is willing to listen to my problems”). Participants were asked to respond to the items as they related to their Principal by using a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). High scores on the scale refer to high levels of perceived Principal support.

Teachers’ emotions were measured by the Teacher Emotions Questionnaire – TEQ [43]. The TEQ consists of 29 items distributed across 6 scales (joy, pride, love, anger, fatigue and hopelessness) each measuring a different emotion that teachers typically experience in relation to their students. Teachers respond to the items by indicating their level of agreement on a 5-point scale from 1 (strongly disagree) to 5 (strongly agree). A high level on each subscale refers to a high level of experiencing a certain emotion. The joy scale consists of 5 items (e.g. I am joyful when the class atmosphere is positive”), the pride scale consists of 6 items (e.g. I feel like a winner when my students succeed”), the love scale consists of 6 items (e.g. I feel warmth when I just think about my students), the anger scale consists of 5 items (e.g. Some students make me so angry that my face goes red), the fatigue scale consists of 7 items (e.g. At the end of my working day, I just want to rest) and the hopelessness scale consists of 6 items (e.g. I feel I cannot do anything more to correct the behaviour of some students).

Two dimensions of burnout – exhaustion and disengagement – were measured with the 16-item Croatian version of the Oldenburg Burnout Inventory – OLBI [44]. Both subscales have eight items (four positively-worded and four negatively-worded items). Examples of items: “There are days when I feel tired before I arrive at work” (for the exhaustion dimension) and “It happens more and more often that I talk about my work in a negative way” (disengagement dimension). Participants were asked to respond to the items by using a scale ranging from 1 (strongly agree) to 4 (strongly disagree). Responses were recoded so that high scores on both subscales refer to high levels of the measured dimension of burnout.

Work engagement was measured with the Croatian version of the Utrecht Work Engagement Scale – UWES [17, 18]. The UWES is a 17-item scale, consisting of three dimensions: vigour (6 items, e.g. “At my work, I feel bursting with energy”), dedication (5 items, e.g. “My job inspires me”), and absorption (6 items, e.g. “Time flies when I’m working”). The items are scored on a seven-point scale ranging from 0 (“never”) to 6 (“always”). High scores on all three work-engagement subscales refer to high levels of the measured dimension of work engagement.

3 Results

Descriptive statistics for all analysed variables as well as correlation between them are shown in Table 1. Even though most of the correlation coefficients between teachers’ gender and variables under study reached the threshold of statistical significance, it can be seen that these associations were modest at best and are probably more a result of great statistical power of this research, i.e. large number of participants than of substantial reasons. Nonetheless, it can be stated that female teachers in general report about slightly higher levels of emotions which is in line with typical gender differences that can be found in research on emotion. This similarly refers to the variable of working experience. All the correlations that were statistically significant were quite small in magnitude pointing to the conclusion that teachers report about similar levels of emotions, work engagement, burnout and Principal support across different levels of working experience. Other correlations reported in Table 1 were in line with the hypotheses. Teachers, who perceive higher levels of Principal support, also report about higher levels of joy, pride, love, vigor, dedication and absorption, as well as about lower levels of anger, fatigue, hopelessness, disengagement and exhaustion. Furthermore, teachers who experience more joy, pride and love towards their students, and less anger, fatigue and hopelessness, report about higher levels of vigor, dedication and absorption. The opposite pattern of correlations is found for dimensions of burnout. Due to weak association of gender and working experience with other variables under study, it was decided not to include them in further analysis.

3.1 Testing differences in Principal support, emotions, work engagement and burnout between teachers from two educational levels

Independent-samples t-tests were conducted to compare levels of perceived Principal support, positive and negative emotions, and measures of work engagement and burnout between two subsamples of teachers (class teachers and subject teachers).

Since we compared teachers from two educational levels by using multiple t-tests between (12 dependant measures), a Bonferroni-adjusted significance level of 0.0041 was calculated to account for the increased possibility of type-I error. Therefore, only correlations noted as ^***p < 0.001 should be considered as significant (see Table 2). According to the obtained results, teachers who teach in lower grades (class teachers) experience higher levels of all positive emotions (joy, pride, and love) in comparison to subject teachers who teach in higher grades of primary schools. Furthermore, class teachers reported higher levels of all three aspects of work engagement (vigor, dedication, and absorption), and lower level of one component of burnout (disengagement), compared to subject teachers.

Table 2
Results of t-tests and descriptive statistics for examined variables by educational level

M_class SD_class M_subject SD_subject t(df)

Perceived 3.74 0.64 3.74 0.66 – 0.19 (1899)

Principal support

Joy 4.78 0.35 4.72 0.38 3.60^* (1895)

Pride 4.54 0.46 4.39 0.53 6.65^* (1878)

Love 4.40 0.52 3.97 0.62 16.29^*** (1879)

Anger 2.82 0.67 2.36 0.75 – 2.25 (1880)^*

Fatigue 2.85 0.84 2.90 0.87 – 1.22 (1879)

Hopelessness 2.57 0.67 2.67 0.74 – 3.21 (1879)^

Vigor 4.83 0.85 4.57 0.94 6.39^* (1886)

Dedication 4.87 0.77 4.70 0.86 4.44^* (1882)

Absorption 4.73 0.80 4.48 0.86 6.49^* (1872)

Disengagement 1.92 0.45 2.04 0.46 – 5.76^*** (1849)

Exhaustion 2.21 0.49 2.23 0.50 – 1.10 (1858)

	M_class	SD_class	M_subject	SD_subject	t(df)
Perceived	3.74	0.64	3.74	0.66	– 0.19 (1899)
Principal support
Joy	4.78	0.35	4.72	0.38	3.60^*** (1895)
Pride	4.54	0.46	4.39	0.53	6.65^*** (1878)
Love	4.40	0.52	3.97	0.62	16.29^*** (1879)
Anger	2.82	0.67	2.36	0.75	– 2.25 (1880)^*
Fatigue	2.85	0.84	2.90	0.87	– 1.22 (1879)
Hopelessness	2.57	0.67	2.67	0.74	– 3.21 (1879)^**
Vigor	4.83	0.85	4.57	0.94	6.39^*** (1886)
Dedication	4.87	0.77	4.70	0.86	4.44^*** (1882)
Absorption	4.73	0.80	4.48	0.86	6.49^*** (1872)
Disengagement	1.92	0.45	2.04	0.46	– 5.76^*** (1849)
Exhaustion	2.21	0.49	2.23	0.50	– 1.10 (1858)

Note: class = class teachers (lower grades); subject = subject teachers (upper grades). ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, Bonferroni-adjusted significance level: 0.0041.

3.2 Testing hypothesized mediation model

In order to answer the research objectives, structural equation modelling was applied. All analyses were conducted using Mplus 6.0 [45], and the maximum-likelihood estimation method. Full-information maximum-likelihood procedures [46] were employed to compensate for the missing data. The quality of model fit was assessed by chi-square (χ²) test, comparative fit index (CFI), Tucker-Lewis index (TLI), root-mean-square error of approximation (RMSEA) and standardized root-mean residual (SRMR). Values of CFI and TLI close to or above 0.95, RMSEA below 0.08 and SRMR below 0.05 are indicative of a good model fit [47].

To test the validity of the hypothesized mediation model, analysis was conducted on the whole sample of teachers in two steps [48]. First, through confirmatory factor analysis [CFA) the validity of the measurement models (e.g. the relations of the observed variables with their respective latent variables) was simultaneously tested. The sum of the scores on the joy, pride and love scales served as indicator of the latent variable of positive emotions, the sum of the scores on the anger, fatigue and hopelessness scales served as indicator of the latent variable of negative emotions, the sum of the scores on the exhaustion and disengagement scales was the indicator of the latent variable of burnout, and the sum of the scores on vigour, dedication and absorption was the indicator of the latent variable of work engagement. For the variable of Principal support, three parcels of items were created to serve as indicators [49]. Overall, the CFA yield the following goodness-of-fit statistics: χ² = 1198.002, df = 67, p < 0.01, CFI = 0.937, TLI = 0.915, RMSEA = 0.094 (CI.089–.098), SRMR = 0.052. Since most of the fit indices were below the acceptable threshold, it was decided, based on the proposed modification indices, to introduce the residual correlation between joy and pride, as well as the correlation between fatigue and exhaustion. We considered such an approach as justifiable due to the fact that joy and pride are emotions of the same valence that may share their antecedents and components [50]. Also, fatigue, as a negative emotion experienced in the classroom, and exhaustion, as a key component of burnout, can be considered as conceptually overlapping constructs (e.g. both of them imply feelings of weariness and tiredness). After introducing these two parameters in the model, its fit greatly improved: χ² = 540.273, df = 65, p < 0.01, CFI = 0.974, TLI = 0.963, RMSEA = 0.062 (CI.057–.067), SRMR = 0.040. The correlation between fatigue and exhaustion was positive and moderate (r = 0.598), as well as the correlation between joy and pride (r = 0.418).

In the second step, the hypothesized causal relations between the latent variables were assessed (i.e. full structural model). In order to test the hypothesis regarding the mediating role of emotions in explaining the relationship between Principal support and work engagement and burnout, two competing models, viz. full vs. partial mediation, were tested. According to the full mediation model, Principal support predicts work engagement and burnout only indirectly via emotions. In contrast, the partial mediation model supposes that there is also a direct link from Principal support to work engagement and burnout. The results of the model-fitting procedure for the full mediation model were as follows: χ² = 601.737, df = 67, p < 0.01, CFI = 0.970, TLI = 0.960, RMSEA = 0.064 (CI: 0.060– 0.069), SRMR = 0.046; and, for the partial mediation model: χ² = 540.273, df = 65, p < 0.01, CFI = 0.974, TLI = 0.963, RMSEA = 0.062 (CI.057–.067), SRMR = 0.040. As the obtained model fit indices suggest, both models fit the data well. Still, to further test the better-fitting model, χ² difference test was calculated: $χ_{diff}^{2}$ = 61.464, df = 2, p < 0.01. Moreover, the direct paths from Principal support to work engagement and burnout were statistically significant but modest in size (see Fig. 1). Based on these results, the partial mediation hypothesis can be confirmed.

Fig. 1

Final (partial mediation) model on the whole sample of participants (N = 1886). Notedie. All parameters shown in Figure are statistically significant at p < 0.001.

Finally, the bias-corrected bootstrapping method (n = 5000 bootstrap resamples) was employed to assess the indirect effects [51]. The significance of these indirect effects was determined using 95% confidence intervals – the indirect effect is considered significant if the CI of the unstandardized mean beta weight does not include zero. The unstandardized mean regression coefficients for indirect effect of Principal support on work engagement via positive emotions and negative emotions were b = 0.236 (95% CI: 0.166,.315) and b = 0.143 (95% CI: 0.101– 0.193) respectively. Furthermore, the unstandardized mean regression coefficients for indirect effect of Principal support on burnout via positive emotions and negative emotions were b = – 0.091 (95% CI: – 0.123, – 0.064) and b = – 0.153 (95% CI: – 0.203, – 0.109) respectively. Thus, it can be concluded that both of these indirect effects are statistically significant at p < 0.05.

To conclude, it was confirmed that emotions partially mediate the relationship between perceived Principal support and work engagement and burnout among teachers. In other words, Principal support predicts work engagement and burnout both directly and indirectly via emotions. Those teachers who perceive their school Principal as more supportive also experience more positive and less negative emotions in relation to their students, as well as more work engagement and less burnout. Furthermore, teachers who experience more positive emotions and less negative emotions are more engaged in their work. On the opposite, teachers with higher levels of burnout usually experience less positive and more negative emotions towards their students.

3.3 Testing structural equivalence of hypothesized model across two educational levels

In order to answer the third research objective, i.e. to test the equivalence of the causal structure of our hypothesized model across the two educational levels, multi-group SEM was conducted. In general, equivalence testing of structural relations (e.g. path coefficients, latent correlations etc.) among latent variables in the model follows the confirmation of measurement invariance to make sure that the measurement instruments operate the same across different samples [52, 53]. Thus, to demonstrate the invariance of the measurement part of our model, we tested three competing models: 1) configural invariance model, i.e. same number of factors and same patterns of free and fixed factor loadings across groups of class teachers and subject teachers, which also served as baseline model for subsequent testing of increasingly restrictive nested models), 2) metric invariance, i.e. invariance of factor loadings across groups, and 3) scalar invariance, i.e. invariance of factor loadings and indicator intercepts across groups. The nested models were evaluated on the basis and ΔCFI criteria, and not based on the χ² difference test, since χ² values tend to be high when sample size is large (as in this study). According to ΔCFI criteria, the more constrained model is suggested to fit significantly worse than the less constrained one when the difference between the CFI values of the two models exceeds 0.01 [54]. Configural invariance model served as a baseline for all subsequent model comparisons. The results of measurement invariance testing are shown in Table 3.

Table 3
Model fit indices for the multi-group SEM analysis across educational levels

Model χ ² df CFI ΔCFI TLI SRMR RMSEA (90% CI)

CFA/SEM Configural 582.604^* 130 0.975 – 0.965 0.042 0.060 (0.055, 0.065)

CFA Metric 617.011^* 144 0.974 0.001 0.967 0.075 0.058 (0.054, 0.063)

CFA Scalar 954.903^* 158 0.956 0.019 0.949 0.119 0.072 (0.068, 0.077)

CFA Partial Scalar 723.742^* 157 0.968 0.007 0.963 0.061 0.061 (0.057, 0.066)

SEM Structural 728.874^* 164 0.968 0.007 0.965 0.075 0.060 (0.055, 0.064)

Model	χ ²	df	CFI	ΔCFI	TLI	SRMR	RMSEA (90% CI)
CFA/SEM Configural	582.604^*	130	0.975	–	0.965	0.042	0.060 (0.055, 0.065)
CFA Metric	617.011^*	144	0.974	0.001	0.967	0.075	0.058 (0.054, 0.063)
CFA Scalar	954.903^*	158	0.956	0.019	0.949	0.119	0.072 (0.068, 0.077)
CFA Partial Scalar	723.742^*	157	0.968	0.007	0.963	0.061	0.061 (0.057, 0.066)
SEM Structural	728.874^*	164	0.968	0.007	0.965	0.075	0.060 (0.055, 0.064)

Note: ^*p < 0.001.

From the results presented in Table 3 it can be concluded that both configural and metric invariance are achieved. However, ΔCFI value, when comparing the scalar invariance model to configural invariance model, was above the recommended threshold indicating the difference in indicator intercepts across two subgroups of teachers. Based on the size of the proposed modification indices, it was decided to release the intercept of love scale since teachers from different educational levels obviously reported about different level of love towards their students, i.e. class teachers experience more love and affection in relation to their students.

In the final step of our analysis we tested two additional SEM models: configural model, i.e. factor loadings and intercepts constrained to be equal and structural paths between latent variables set free to vary across groups, and constrained model, i.e. factor loadings, intercepts (with the exception of the intercept of love scale) and structural paths constrained to be equal across groups. The results of this analysis are also shown in Table 3. According to both evaluation criteria, imposing constraints on structural paths between latent variables did not lead to loss of fit, indicating that the causal structure of the hypothesized model can be considered equivalent across the two groups of teachers. In other words, direction and strength of the associations between examined constructs were the same at both lower and upper educational levels.

4 Discussion

The study presented here tested a hypothetical model according to which teachers’ positive and negative emotions partially mediate the relationship between perceived Principal support and work engagement and burnout among teachers. The obtained results showed that Principal support predicts work engagement and burnout both directly and indirectly via emotions (Fig. 1).

As we expected, higher levels of perceived Principal support predicted higher levels of positive emotions and work engagement, but lower levels of negative emotions and burnout. These results are in line with previous research, according to which Principal support has been recognized as an important organizational characteristic that promotes well-being in teachers and prevents their burnout [21, 22]. Furthermore, higher levels of work engagement are predicted by higher levels of positive emotions, and this is in line with the previous observation that engaged workers, often experience positive emotions, such as happiness, joy, pride, challenge and enthusiasm [30]. Moreover, one of the basic distinctive items in the general definition of work engagement relates to a ‘positive’ emotional tone – ‘positive work-related state of mind’ [17 –19]. Therefore, links obtained between emotions of joy, pride and love on one hand, and the three aspects of work engagement (vigor, dedication and absorption), on the other, are theoretically meaningful. Higher work engagement was also predicted by lower levels of negative emotions, which is in line with the proposition that engaged employees are characterized by high positive affect and to a somewhat lesser degree by low negative affect [17, 18]. Furthermore, according to Kahn [55], to become engaged in an activity, individuals must have psychological availability that includes the emotional, cognitive, and physical resources. Negative affect may reduce psychological availability and thus diminish work engagement.

In addition, the results showed that teachers with higher levels of burnout usually experience more negative emotions and less positive emotions towards their students. A review of the literature on teacher burnout and emotions clearly shows that teachers’ repeated experience of unpleasant emotions leads to burnout, which is attributed mostly to habitual patterns in teachers’ judgments [28]. Experience of positive emotions, such as love, joy and pride is seen more in teachers with lower levels of exhaustion and disengagement, which is conceptually logical when we focus on the definitions of the burnout dimensions [25]. Therefore, our results are in accordance to empirical evidence suggesting that burnout is strongly associated with increased negative affectivity – or, conversely, decreased positive affectivity [56].

Next, our main hypothesis that perceived Principal support is associated with work engagement and burnout also indirectly i.e. via positive and negative emotions was also confirmed. Obviously, the perceived support from the Principal, with its intrinsic and extrinsic components, plays an important role in the experience of teachers’ emotions, which in turn contribute to their work engagement and burnout. Perceived support from the Principal implies a better overall emotional climate in the school, which also reflects on emotions in relation to the students. On the other hand, lack of support from the Principal leads to negative emotions, which disturbs work engagement and may lead to burnout.

Finally, the results of the present study have demonstrated the structural equivalence of the hypothesized model across the two education levels of primary school, which confirm the conceptual robustness of the proposed relationships among analysed constructs. However, although relations between tested constructs are shown to be equivalent across two education levels, the levels of positive emotions, disengagement and three components of work engagement differed significantly by educational level.

Several factors may be proposed as the possible sources of the obtained differences. First of all, class teachers are spending their working days with the same students over an extended period of time, therefore they are being in the situation that leads to greater intimacy and concern for their students what consequently may increase the level of their emotional experience of joy, love and pride. These positive emotions related to the students probably reflect also positively on the levels of work engagement and burnout [31 –33]. Further source is the nature of class teachers’ work itself (holistic approach to curriculum, whereby the same class teacher teaches all subjects to his/her students during first four grades), which is why they are better acquainted with the characteristics and the needs of the students. Therefore, in comparison to subject teachers who are more limited by time in their teaching [40], the nature of class teaching gives teachers greater flexibility, freedom and autonomy in the organisation of their work day, which positively reflects of their occupational well-being. Moreover, obtained differences between class teachers and subject teachers may be also partly based on the differences in the age of students, considering the fact that subject teachers teach students in puberty age which may be accompanied by lower academic motivation [40]. Finally, higher work engagement in class teachers compared to subject teachers may also be explained by the fact that they are in general more satisfied by the choice of their occupation [40]. Class teachers are educated and prepared for the profession of school teacher, which is not the case for all subject teachers [39, 41]. Therefore, obtained differences between class teachers and subject teachers may also be the reflection of the differences in their intrinsic motivation for the work in the primary school.

The present study has some limitations that should be mentioned. First of all, the data were collected at a single point in time, so it is not possible to draw conclusions about the causal relationships between the constructs tested. Furthermore, all the data are based on self-reports. Although we used reliable and validated measures for all the constructs tested, possible error coming from the common-variance method cannot be excluded. Moreover, forthcoming studies should include data from other sources, such as ratings of work commitment from colleagues or students, for example, and/or behavioural measures relating to work commitment, such as absenteeism and work performance. Finally, although a response rate of 50 % is more satisfactory than the average rate in psychological research, our sample, which is based on voluntary participation, may be biased, and that fact limits the generalizability of the results. Further research should assess the role of teachers’ emotions relating to other sources beside students, such as emotions relating to school staff, parents of students, and the education system. This is in line with the call for new research, in the field of teachers’ emotions, which supports the complexity of teachers’ emotions, i.e. considering the social, cultural and political factors which influence teachers’ emotions [43, 58].

Nonetheless, despite these limitations, this study presents an important step forward in understanding the role of teachers’ emotions experienced at workplace in understanding individual differences in their occupational well-being. Even though emotions are considered as integral components of teachers’ lives [31], they were rather neglected in previous research. However, thanks to the recent development of suitable measurement instruments aimed at assessing teachers’ affective experience [43], more research on antecedents and effects of teachers’ emotions is expected.

Finally, the possible practical implications of this study go in two directions. The first relates to Principal support as a crucial job resource [58], which plays an important role in promoting teachers’ occupational well-being. Different aspects of support, such as instrumental, informational and emotional support, as well as trust, help, and an appropriate attitude on the part of the Principal probably promote the well-being and work motivation of teachers. This issue is of crucial importance for achieving educational goals The importance of school management’s support in reducing time pressure, preventing burnout and attrition, and promoting well-being and motivation is documented in previous studies in teachers outside of Croatia [59, 60]. Thus, in the process of selection and training of Principal, special attention should be given to their skills of human-resource management. While Principal evidently need help with dealing with a great number of demands (administrative, financial, public etc.), the Principal is the first person in the school who must promote the concept of a healthy organization. According to Wilson et al. [61], a healthy organization may be defined as an organization in which work processes promote and maintain a state of complete physical, mental and social well-being among employees, which leads to higher levels of performance.

The second implication is based on our data, which show that affective processes partially mediate links between organizational characteristics such as Principal support, on one hand, and work engagement and burnout, on the other. Since the teachers’ profession is recognized as prone to the experience of occupational stress and burnout [5– 8 , 22], which threatens the productivity and motivation of teachers and their retention in the profession, practical measure may be recommended (e.g. workshops for teachers whose primary goal relates to increasing the awareness of the importance of regulation of emotions). In promoting well-being and work engagement among teachers, special attention should be given to strategies aimed at maintaining or enhancing positive emotional experiences in teaching. Through experience of positive emotions, people become more resilient [62], which promotes positive outcomes at both individual and organisational levels. This is especially important among subject teachers, who according to our data, experience lower levels of positive emotions in work with students, and are generally less engaged in their work.

Finally, obtained data may be also used as an additional starting point for implementing planned changes and innovations in national educational system, including both teaching curriculum and education of teachers.

Conflict of interest

None to report.

Footnotes

Acknowledgments

This research was supported by Croatian Science Foundation (Grant No. 5065).

References

Gewirtz

, Mahony

, Hextall

, Cribb

. Changing teacher professionalism: International trends, challenges and ways forward. London: Routledge; 2009.

Fuller

, Goodwyn

, Francis-Brophy

. Advanced skills teachers: Professional identity and status. Teach Teach. 2013;19(4):463–74. doi: 10.1080/13540602.2013.770228

Hargreaves

, Cunningham

, Everton

, Hansen

, Hopper

, McIntyre

, et al.. The status of teachers and the teaching profession: Views from inside and outside the profession: Interim findings from the teacher status project. Research Report 755. London: DfES; 2006.

Slišković

, Burić

, Macuka

. The voice of Croatian elementary school teachers: Qualitative analysis of the teachers’ perspective on their profession. Teach Teach. 2016;23(5):518–31. doi: 10.1080/13540602.2016.1206521

Johnson

, Cooper

, Cartwright

, Donald

, Taylor

, Millet

. The experience of work-related stress across occupations. J Manage Psychol. 2005;20(2):178–87. doi: 10.1108/02683940510579803

Kyriacou

. Teacher stress: Directions for future re-search. Educ Rev. 2001;53(1):27–35. doi: 10.1080/00131910120033628

Kyriacou

. Teacher stress: From prevalence to resilience. In Langan-Fox

, Cooper

, editors. Handbook of Stress in the Occupations. Cheltenham: Edward Elgar Publishing Ltd; 2011, pp. 161–77.

Maslach

, Jackson

, Leiter

. Maslach burnout inventory manual, 3rd Ed. Palo Alto, CA: Consulting Psychologists Press; 1996.

Ball

. The teacher’s soul and the terrors of performativity. J Educ Pol. 2003;18(2):215–28. doi: 10.1080/0268093022000043065

10.

Beck

. Governement professionalism: Re-professiona-lising or deprofessionalising teachers in England? Brit J Educ Stud. 2008;56(2):119–43. doi: 10.1111/j.1467-8527.2008.00401.x

11.

Gür

. Deskilling of teachers: The Case of Turkey. Educ Sci – Theor Pract. 2014;14(3):887–904.

12.

Ingersoll

. Power, accountability, and the teacher quality problem. In: Kelly

, editor. Assessing teacher quality: Understanding teacher effects on instruction and achievement. New York: College Press; 2012, pp. 97–109.

13.

Vizek Vidović

. Cjeloživotno obrazovanje učitelja i nastavnika: Višestruke perspektive [in Croatian]. Zagreb: Institut za društvena istraživanja; 2005.

14.

European Commission/EACEA/Eurydice. The Teaching Profession in Europe: Practices, Perceptions, and Policies. Eurydice Report. Luxembourg: Publications Office of the European Union; 2015. Available from: http://ec.europa.eu/eurydice.

15.

Slišković

. Stres kod nastavnika u visokom obrazovanju [in Croatian]. Psihologijske Teme. 2011;20(1):67–90.

16.

Sorić

. Samoregulacija učenja. Možemo li naučiti učiti [in Croatian]. Jastrebarsko: Naklada Slap; 2014.

17.

Schaufeli

, Bakker

. UtrechtWork Engagement Scale [UWES): Preliminary manual. Department of Psychology, Utrecht University; 2003. Available from: www.wilmarschaufeli.nl/publications/Schaufeli/Test%20Manuals/Test_manual_UWES_English.pdf.

18.

Schaufeli

, Bakker

. Defining and measuring work engagement: Bringing clarity to the concept. In: Bakker

, Leiter

, editors.Work engagement:Ahandbook of essential theory and research. New York: Routledge; 2010, pp. 10–24.

19.

Schaufeli

, Taris

, Van Rhenen

. Workaholism, burnout and engagement: Three of a kind or three different kinds of employee well-being? Appl Psychol. 2008;57(2):173–203. doi: 10.1111/j.1464-0597.2007.00285.x

20.

Maslach

. Job burnout: New directions in research and intervention. Curr Dir Psychol Sci. 2003;12(5):189–92. doi: 10.1111/1467-8721.01258

21.

Brackett

, Palomera

, Mojsa-Kaja

, Reyes

, Salovey

. Emotion-regulation ability, burnout, and job satisfaction among British secondary-school teachers, Psychol Schools. 2010;47(4):406–17. doi: 10.1002/pits.20478

22.

Hakanen

, Bakker

, Schaufeli

. Burnout and work engagement among teachers. J School Psychol.. 2006;43(6):495–513.

23.

Leiter

M P

, Maslach

. The impact of interpersonal environment on burnout and organizational commitment. J Organizational Behav. 1988;9:297–308. doi: 10.1002/job.4030090402

24.

Maslach

, Schaufeli

, Leiter

. Job burnout. Ann Rev Psychol. 2001;52:397–422. doi: 10.1146/annurev.psych.52.1.397

25.

Demerouti

, Mostert

, Bakker

. Burnout and work engagement: A thorough investigation of the independency of both constructs. J Occup Health Psych. 2010;15(3):209–22. doi: 10.1037/a0019408

26.

Maslach

. Burnout: A multidimensional perspective. In: Schaufeli

, Maslach

, Marek

, editors. Professional burnout. Recent developments in theory and research. New York: Hemisphere; 1993, pp. 19–32.

27.

Ghanizadeh

, Jahedizadeh

. Teacher burnout: A review of sources and ramifications. Brit J Educ Soc Beh Sci. 2015;6(1):24–39. doi: 10.9734/BJESBS/2015/15162

28.

Chang

. An appraisal perspective of teacher burnout: Examining the emotional work of teachers. Educ Psychol Rev. 2009;21(3):193–218. doi: 10.1007/s10648-009-9106-y

29.

Ryan

, Deci

. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. Am Psychol. 2000;55(1):68–78. doi: 10.1037110003-066X.55.1.68

30.

Bakker

, Demerouti

. Towards a model of work engagement. Career Dev Int. 2008;13(3):209–23. doi: 10.1108/13620430810870476

31.

Sutton

, Wheatley

. Teachers’ emotions and teaching: A review of the literature and directions for future research. Educ Psychol Rev. 2003;15(4):327–58. doi: 10.1023/A:1026131715856

32.

Frenzel

. Teacher emotions. In: Linnenbrink-Garcia

, Pekrun

, editors. International Handbook of Emotions in Education. New York: Routledge; 2014, pp. 494–519.

33.

Frenzel

, Goetz

, Stephens

, Jacob

. Antecedents and effects of teachers’ emotional experiences: An integrated perspective and empirical test. In: Schutz

, Zembylas

, editors. Advances in Teacher Emotion Research: The Impact on Teachers’ Lives. New York: Springer; 2009, pp. 129–52.

34.

Day

, Qing

. Teachers’ emotions: Well-being and effectiveness. In: Schutz

, Zembylas

, editors. Advances in teacher emotion research: The impact on teachers’ lives. New York: Springer; 2009, pp. 15–32.

35.

Antoniou

A-S

, Ploumpi

, Ntalla

. Occupational stress and professional burnout in teachers of primary and secondary education: The role of coping strategies. Psychol. 2013;4(3):349–55. Available from: http://file.scirp.org/Html/29268.html

36.

Chan

AHS

, Chen

, Chong

EYL

. Work stress of teachers from primary and secondary schools in Hong Kong. Proceedings of the International Multiconference of Engineers and Computer Scientists, HongKong, 17-19 of March 2010. Available from: http://www.iaeng.org/publication/IMECS2010/IMECS2010_pp1903-1906.pdf

37.

Yavuz

. An investigation of burn-out levels of teachers working in elementary and secondary educational institutions and their attitudes to classroom management. Educ Res Rev. 2009;4(12):642–49. Available from: http://www.academicjournals.org/article/article1379621016_Yavuz.pdf

38.

Agai– Demjaha

, Minov

, Stoleski

, Zafirova

. Stress causing factors among teachers in elementary schools and their relationship with demographic and job characteristics. OA Mac J Med Sci. 2015;3(3):493–99. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4877846/

39.

Vidić

. Zadovoljstvo poslom učitelja u osnovnoj školi [in Croatian]. Napredak. 2009;150(1):7–20. Available from: https://hrcak.srce.hr/82813

40.

Koludrović

, Jukić

, Reić Ergovac

. Sagorijevanje na poslu kod učitelja razredne i predmetne nastave te srednj-oškolskih učitelja [in Croatian]. Život i škola. 2009;22(2):235–49. Available from: http://hrcak.srce.hr/file/73551

41.

Kuzijev

, Topolovčan

. Uloga odabira učiteljske profesije i samopoštovanja u sagorijevanju učitelja u osnovnoj školi [in Croatian]. Andragoški Glasnik. 2013;17(2):125–44. Available from: http://hrcak.srce.hr/116177

42.

Jones

, James

. Psychological climate: Dimensions and relationships of individual and aggregated work environment perceptions. Organ Behav Hum Perf. 1979;23:201–50. doi: 10.1016/0030-5073[79)90056-4

43.

Burić

, Slišković

, Macuka

. A mixed-method approach to assessment of teachers’ emotions in the workplace: Development and validation of Teacher Emotion Questionnaire. Educ Psychol. 2017;38(3):325–49. doi: 10.1080/01443410.2017.1382682

44.

Demerouti

, Bakker

. The Oldenburg Burnout Inventory: A good alternative to measure burnout and engagement. In: Halbesleben

, editor. Stress and burnout in health care. New York: NovaSciences, Hauppage; 2008, pp. 65–78.

45.

Muthen

, Muthen

. Mplus User’s Guide. Seventh Edition. Los Angeles, CA: Muthen & Muthen; 2014.

46.

Byrne

. Structural Equation Modeling with AMOS, EQS, and LISREL: Comparative approaches to testing for factory validity of a measuring instrument. Int J Test. 2001;1(1):55–86. doi: 10.1207/S15327574IJT0101_4

47.

, Bentler

. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Modeling. 1999;6(1):1–55. doi: 10.1080/10705519909540118

48.

Bollen

. Structural equations with latent variables. New York: Wiley-Interscience; 1989.

49.

Little

, Cunningham

, Shahar

, Widaman

. To parcel or not to parcel: Exploring the question, weighing the merits. Struct Equ Modeling. 2002;9(2):151–73. doi: 10.1207/S15328007SEM0902_1

50.

Pekrun

, Goetz

, Perry

, Kramer

, Hochstadt

, Molfenter

. Beyond test anxiety: Development and validation of the Test Emotions Questionnaire (TEQ). Anx Stress Copin. 2004;17:287–316. doi: 10.1080/10615800412331303847

51.

Preacher

, Hayes

. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Beh Res Methods. 2008;40(3):879–91. doi: 10.3758/BRM.40.3.879

52.

Byrne

. Structural Equation Modeling with Mplus. Basic Concepts, Applications, and Programming. New York: Routledge; 2012.

53.

Wang

, Wang

. Structural equation modeling: Applications using Mplus. Chichester, UK: Wiley; 2012.

54.

Cheung

, Rensvold

. Evaluating goodness-of-fit indexes for testing measurement invariance. Struct Equ Modeling. 2002;9(2):233–55. doi: 10.1207/S15328007SEM0902_5

55.

Kahn

. The psychological conditions of personal engagement and disengagement at work. Acad Manage J. 1990;33:692–724. doi: 10.2307/256287

56.

Brotheridge

, Grandey

. Emotional labor and burnout: Comparing two perspectives of “People Work”. J Vocat Behav. 2002;60(1):17–39. doi: 10.1006/jvbe.2001.1815

57.

Fried

, Mansfield

, Dobozy

. Teacher emotion research: Introducing a conceptual model to guide future research. Iss Educ Res. 2015;25(4):415–41. Available from: www.iier.org.au/iier25/fried.html.

58.

Torenbeek

, Peters

. Explaining attrition and decreased effectiveness of experienced teachers: A research synthesis. Work. 2017;57(3):397–407. doi: 10.3233/WOR-172575

59.

Nilsson

, Blomqvist

, Andersson

. Salutogenic resources in relation to teachers’ work-life balance. Work. 2017;56(4):591–602. doi: 10.3233/WOR-172528

60.

Hoefsmit

, Cleef

. If it isn’t finished at five, then I’ll continue until it is. A qualitative study of work pressure among employees in vocational education. Work. 2018;61(1):69–80. doi: 201810.3233/WOR-182780

61.

Wilson

, Dejoy

, Vandenberg

, Richardson

, McGrath

. Work characteristics and employee health and well-being: Test of a model of healthy work organization. J Occup Organ Psych. 2004;77(4):565–88. doi: 10.1348/0963179042596522

62.

Tugade

, Fredrickson

. Regulation of positive emotions: Emotion regulation strategies that promote resilience. Journal of Happiness Studies. 2007;8(3):311–33. doi: 10.1007/s10902-006-9015-4