Non-conformism as precursor for self-efficacy and well-being among schoolteachers in the Netherlands

In this study, a model is developed and tested to examine the influence of non-conformism and self-efficacy on affective well-being and burnout amongst 401 Dutch schoolteachers. The relationships between the level of non-conformism and self-efficacy and subsequently the level of affective well-being, and burnout are hypothesised. The results of this study showed that self-efficacy is positively related to affective well-being and strongly negatively related to burnout. The results of this study also showed non-conformism to be a significant predictor for self-efficacy. Because of the strategic importance of human capital in schools, it is important practitioners are provided with tangible suggestions, such as non-conformist behaviour, to improve self-efficacy and well-being.


Introduction
O ver recent years the pressure of work or the feeling of being under high pressure at work has developed very negatively in elementary education in the Netherlands (Traag 2018) leading to higher levels of burnout and consequently sickness absenteeism (Evers et al. 2005;Harmsen et al. 2018;Kodden and Hupkes 2019). Organisations, including the education sector, are faced with high costs as a result of this sickness absenteeism, declining productivity among schoolteachers, and a relatively high turnover rate with teachers leaving (Blomme and Kodden 2014;Consiglio et al. 2013;Kodden and Roelofs 2019;Pas et al. 2012). As a result, the Netherlands is facing a shortage in teachers due to the effects on reputation of the teaching job, spiralling towards more pressure at work and worsening of quality of education (Harmsen et al. 2018;Traag 2018). The role of teachers in students' success, the educational system, and society in the broadest sense is crucial (Keiler 2018;Vesely et al. 2013). Therefore, it is of the utmost importance to further research factors that are generally known to influence teachers' well-being at their jobs, such as self-efficacy (Consiglio et al. 2013;Skaalvik and Skaalvik 2014;Ventura et al. 2015;Kodden and van Ingen 2019), positive emotions (Bondarenko 2017;Schaufeli and Van Rhenen 2006), work engagement and burnout (Bakker and Demerouti 2008Kodden and Groenveld 2019). The well-being of schoolteachers is a broad concept, but in the work situation it can be associated with affects, such as emotions and moods arising from experiences during work (Bianchi et al. 2019;Hakanen et al. 2019;Holt-Lunstad et al. 2015). Looking at the impact of emotions that arise at work, and specifically the negative emotions that can lead to burnout in the longer term, it is interesting to study more closely the potential sources of these emotions which are related to affective well-being (Virgă et al. 2019). According to Demerouti (2014, 2017), the Job Demands-Resources (JD-R) theory offers a framework to explain and illustrate the mediating role of emotions (both positive and negative) in the relationships between demands and resources, and burnout and work engagement, respectively. The Job Demands-Resources (JD-R) theory focuses on both negative (burnout), as well as positive (work engagement) aspects of wellbeing by linking it to health impairment and motivational processes, respectively (Bakker and Demerouti 2014). The positive effects of self-efficacy are commonly explained by Social Cognitive Theory (SCT; Bandura 1997Bandura , 2000, which states that selfefficacious people have belief in their capabilities and are persistent to attain performance in the face of obstacles. Previous research has shown that self-efficacy has a negative effect on burnout and a positive effect on well-being (Brown 2012; Makara-Studzińska et al. 2019; Molero et al. 2018). The results of a number of other studies have shown that the level of selfefficacy has a positive effect on employee performance and work engagement, in which self-efficacy is considered a personal resource (e.g., Bakker and Demerouti 2008Kodden 2014;Ventura et al. 2015). Another theory that affects well-being is the Self Determination Theory (SDT; Gagnéand Deci 2005; Ryan and Deci 2000;Van den Broeck et al. 2016) in which autonomy is an important source of general psychological wellbeing (Gagnéand Deci 2005;Lock et al. 2018). Previous studies have shown that autonomy, as a job resource, is a driver for work engagement (e.g., Bakker and Demerouti 2008. Autonomy can be juxtaposed to control, in which behaviour is regulated by external forces, such as social pressures, group norms, regulations or compliance (Lock et al. 2018;Weinstein et al. 2012). Acting in accord with a group norm might be experienced as conformity and thus as a threat to autonomy rather than an expression of it (Gino 2016; Gray and Silbey 2014; Ryan and Deci 2000). People who are less susceptible to regulation, social pressures or group norms feel less obliged to conform and experience more autonomy (Lock et al. 2018;Weinstein et al. 2012). The pressure to conform can lead to lower work engagement, lower productivity and a lower level of innovation (Gino 2016;Grant 2016). According to Gino (2016), there should be a balance between conformity and non-conformity in employees, so that on the one hand the existing agreements on operational management are carried out, but on the other hand these agreements are not accepted as the status quo. In this study non-conformity is defined as 'behavior that deviates from organizational norms, others' actions, or common expectations, to the benefit of the organization' (Gino 2016, p. 4). The non-acceptance of the status quo, or non-conformity, is advantageous for the organisation in having a positive influence on the work engagement of employees (Gino 2016). This study, therefore, aims to investigate the relationships between non-conformism, self-efficacy, burnout and well-being. Although some research has been done on non-conformism (e.g., Bellezza et al. 2014;Gino 2016;Ventura et al. 2015), to the best of our knowledge, the consequences of non-conformism as perceived by teachers have not yet been researched in relation to teachers' self-efficacy, burnout, and affective well-being. This study contributes to the literature by empirically researching teachers' non-conformism at the individual level and examining the impact of non-conformism on self-efficacy, burnout, and affective well-being amongst teachers. The research aims to enhance scholarly insights into the consequences of non-conformism and important antecedents of selfefficacy, affective well-being and burnout, as focused on in the existing literatures (i.e., JD-R theory, SCT, and SDT). In this study we test a hypothesised mediation model in which nonconformism is directly positively associated with self-efficacy, and the relationship between non-conformism and burnout and affective well-being, is fully mediated by self-efficacy. We do this by conducting a quantitative cross-sectional study among 401 teachers in the Netherlands using SPSS for analysis. In the next section we build on insights from JD-R theory, Social Cognitive Theory (SCT), and SDT to develop the hypotheses.

Theory and hypotheses
The job demands-resources theory. According to Demerouti (2014, 2017), with the Job Demands-Resources (JD-R) theory we are able to comprehend, explain and make predictions about well-being (e.g., work engagement, motivation, burnout). Regardless of work environment or job characteristics, the theory can be applied using two categories: job demands and job resources. JD-R links job demands and resources to two independent processes, a health impairment process (leading to exhaustion and consequently burnout) and a motivational process (leading to work engagement).
Job demands refer to 'those physical, psychological, social, or organizational aspects of the job that require sustained physical and/or psychological (i.e., cognitive or emotional) effort and are therefore associated with certain physiological and/or psychological costs' (Demerouti et al. 2001, p. 501). Examples of job demands among teachers are workload, disruptive pupil behaviours, and poor physical work environment (Bakker and Demerouti 2008. Job resources refer to 'those physical, psychological, social, or organizational aspects of the job that may reduce job demands and the associated physiological and psychological costs, are functional in achieving work goals, and stimulate personal growth, learning, and development' (Demerouti et al. 2001, p. 501). Examples of job resources are social support, feedback and autonomy (Bakker and Demerouti 2017). Job resources are not only necessary to deal with job ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-020-00551-6 demands and to get things done, but they are also important in their own right. Conversely, a lack job resources may invoke higher levels of burnout and thus negatively affects teachers' wellbeing. In sum, the JD-R theory proposes that high job demands and a lack of job resources form the basis for burnout and thus leads to reduced well-being.
Burnout. Burnout is defined as 'a prolonged response to chronic emotional and interpersonal stressors on the job, and is defined by the three dimensions of exhaustion, cynicism, and inefficacy' (Maslach et al. 2001, p. 397). Emotional exhaustion is seen as the stress component characterised by chronic fatigue, loss of energy, lack of emotional resources, and the feeling of being overstretched. Cynicism or depersonalisation refers to the indifference to or detachment of the work environment and its constituents, and is characterised by loss of interest in work, irritability, and negative or inappropriate, distant attitudes towards other people in the work environment. Inefficacy or lack of professional efficacy refers to a disbelief in one's competence and ability to achieve and accomplish tasks and goals in the job and the organisation. Previous research showed that emotional exhaustion is directly related to burnout and depersonalisation is indirectly related to burnout through emotional exhaustion. The relationship between burnout and the lack of professional efficacy is less strong and more complex and it appears that the lack of professional efficacy develops more independently than the other two factors (Bakker and Demerouti 2008Maslach et al. 2001). A limited supply of job resources is related most strongly with depersonalisation, and high job demands are most closely related to emotional exhaustion. Since burnout is primarily predicted by the presence of excessive job demands, it follows that when job demands are too high, emotional exhaustion is the most important variable causing burnout (Bakker and Demerouti 2008Maslach et al. 2001). Therefore, in this study emotional exhaustion reflects burnout in measurement.
The role of positive and negative emotions in teacher affective well-being. This study researches well-being in relation to the work environment, specifically amongst teachers in an educational setting. The past decades job stress an increase occurred in attention to the role of emotions as indicators of strain and well-being at work (e.g., Evers et al., 2015;Harmsen et al. 2018). Most job stress research has been limited to a small number of affective or affect-related strains, such as job dissatisfaction or work anxiety, and consequently insufficient attention has been paid to other positive and negative emotional states (Schaufeli and Van Rhenen 2006;Traag 2018). The affective aspect of wellbeing can be divided into emotions, moods and temperaments, which are interconnected but which differ in duration, focus and intensity (Schaufeli and Van Rhenen 2006). According to Schaufeli and Van Rhenen (2006) emotions are short-lived, intensive affective experiences that are focused on a specific object (e.g., fear), moods are affective experiences that are longlasting, less intense and more general, i.e., not focused on specific objects (e.g., sombreness), and temperaments are lifelong, stable affective response preferences that are generalised over situations (e.g., negative affectivity). It can be assumed that short-lived positive or negative experiences or emotions ultimately turn into positive or negative moods. Examples are feelings of enthusiasm and cheerfulness, which can result in a state of work engagement, and negative feelings of tiredness and frustration, which can lead to a state of burnout (e.g., Bakker and Demerouti 2008. In this study affective wellbeing is divided into positive emotions and negative emotions and burnout encompassed the dimension emotional exhaustion.
Following the above reasoning, this leads to the following two hypotheses: Hypothesis 1: Positive emotions are negatively related to emotional exhaustion.
Hypothesis 2: Negative emotions are positively related to emotional exhaustion.
Self-efficacy as personal resource. Self-efficacy is a key concept Social Cognitive Theory (SCT; Bandura 1997Bandura , 2000. Self-efficacy refers to the belief in one's capabilities and competences to master the job and to accomplish attainments (Bandura 1997(Bandura , 2000. In a work environment, self-efficacy differs from the dimension of professional efficacy, which is part of the burnout concept. These beliefs affects decisions on action, how much energy is put in activity, and positively influence people's perseverance in facing difficulties. Self-efficacy beliefs are context-specific since they are related to individual's capabilities and activities which are needed to be performed in a certain work environment, and these selfefficacy beliefs can be considered an important antecedent of organisational behaviour and job performance (Federici and Skaalvik 2012;Kodden 2014;Ventura et al. 2015). Personal resources are defined as 'aspects of the self that generally linked to resiliency' (Hobfoll et al. 2003, p. 632). These aspects refer to individuals' sense of their ability to control and impact upon their environment successfully, especially during challenging circumstances (Bandura 1997(Bandura , 2000Consiglio et al. 2013;Hobfoll et al. 2003;Pas et al. 2012;Skaalvik and Skaalvik 2014). From this point of view, self-efficacy can be considered a personal resource, and as such can be considered a personal resource in JD-R theory (e.g., Xanthopoulou et al. 2007). By combining the JD-R Theory (Bakker and Demerouti 2014), with SCT (Bandura 1997(Bandura , 2000 we can understand the burnout and engagement processes, from a perspective in which an individual beliefs in his or her ability to impact work and actively shapes the work environment. According to SCT, self-efficacy beliefs influence not only affect (e.g., emotions) and behaviour (e.g., Pajares 1997; Skaalvik and Skaalvik 2014), but also the perception of contextual and situational opportunities and threats (Bandura 1997(Bandura , 2000. Previous research showed that self-efficacy is positively related to work engagement and well-being (Bakker and Demerouti 2008;Kodden 2014;Stajkovic et al. 2018) and negatively related to burnout, and specifically emotional exhaustion (Brown 2012; Federici and Skaalvik 2012;Skaalvik 2010, 2014). The construct of self-efficacy suggests a protective effect when facing challenges or difficulties. An optimistic belief in one's competence to cope with daily adversities enhances the motivation to engage in constructive ways of dealing. Thus, self-efficacy, as a key construct in SCT, and considered a personal resource in JD-R theory, is likely to influence affective well-being, specifically positive and negative emotions, and burnout, specifically emotional exhaustion. This leads to the following hypotheses: Hypothesis 3: Self-efficacy is positively related to positive emotions.
Hypothesis 4: Self-efficacy is negatively related to negative emotions.
Hypothesis 5: Self-efficacy is negatively related to emotional exhaustion.
Non-conformism as a determinant of self-efficacy, burnout and affective well-being. The concept of non-conformity literally means any behaviour that is not conformity. According to Nail et al. (2000), non-conformity consist of two subtypes: independence and anti-conformity. Independence is defined as 'behavior or belief that results when the influence target gives zero weight to the norms, positions, or standards of another or others' (Nail et al. 2000, p. 456). Anti-conformity is defined as 'behavior or belief that is not consistent with the norms, positions, or standards of another or others based on one or more motives of the influence target' (Nail et al. 2000, p. 457). Non-conformity can be perceived as admirable behaviour that reflects high levels of autonomy and control (Bellezza et al. 2014). From an etymological view autonomy is indicated by autos (self) and nomos (rule or law). Thus autonomous individuals tend to act independently and behave according to their own rules. According to Self Determination Theory (SDT), autonomy and self-perceived competence are fundamental universal psychological needs that are important for motivation and psychological well-being (Gagnéand Deci 2005;Lock et al. 2018;Ryan and Deci 2000). Previous studies have shown that autonomy, as a job resource, is a driver for work engagement (e.g., Bakker and Demerouti 2008. The SDT makes a distinction between autonomous motivation and controlled motivation. Autonomous motivation arises from the individual's own interests. Controlled motivation comes from the need or pressure to carry out an action. Both forms of autonomy occur voluntarily (Gagnéand Deci 2005). According to the SDT, there are factors that influence people's activation or passivity and responsibility or laziness. In a social environment, people can be activated due to their basic need for self-development and their curiosity. There are also factors that encourage passivity and negativity, such as excessive control, suboptimal challenges and a lack of social cohesion. Autonomy can be juxtaposed to control, in which behaviour is regulated by external forces, such as social pressures, group norms, regulations or compliance (Lock et al. 2018;Weinstein et al. 2012). Acting in accord with a group norm might be experienced as conformity and thus as a threat to autonomy rather than an expression of it (Ryan and Deci 2000), which can lead to people taking the initiative less or taking responsibility less, and over time it can even lead to stress and psychological conditions (Ryan and Deci 2000). In this study conformity is defined as 'involving characteristic willingness to identify with others and emulate them, to give in to others to avoid conflict, and generally, to be a follower rather than a leader in terms of ideas, values, and behaviors' (Mehrabian and Stefl 1995, p. 257). The degree of conformity results from social interactions and is related to individual differences. It has been shown that resistance to conformity arises when the individual experiences being forced to conform, which restricts that individual's own freedom of choice (Brehm 1966). This motivates a person to react in order to overturn this restriction on freedom of choice (Brehm 1966). The pressure to conform might thus lead to lower work engagement, lower productivity and a lower level of innovation (Gino 2016;Grant 2016). This leads to the following hypothesis: Hypothesis 6: Non-conformism is positively related to selfefficacy.
The mediating role of self-efficacy. According to JD-R theory Demerouti 2014, 2017), personal resources may function either as a moderator or as a mediator in the way people comprehend the work environment, formulate it, and react to it (Bandura 2000;Brown 2012;Molero et al. 2018;Xanthopoulou et al. 2007). Previous studies have shown that the relationship between autonomy, as a job resource, and emotional exhaustion, mediated by self-efficacy as a personal resource (Makara-Studzińska et al. 2019) showed that a resourceful work environment activates employees' 'psychological capital' (i.e., hope, optimism, efficacy, and resilience). In relation to JD-R theory, job resources, such as a control or autonomy over one's decisions and actions, will invoke a sense of importance to individuals, leading to feel efficacious, optimistic about their future, and, consequently, engaged in their work Demerouti 2014, 2017;Xanthopoulou et al. 2007). The degree of external social pressure and the requirement to conform appear to have a strong influence on the degree of self-efficacy that is experienced. When schoolteachers do not have to conform much, they can introduce their own standards and methods into the lessons and align these to their pupils as they see fit. This denotes a logical connection between the experience of self-efficacy and non-conformity. When an individual has to conform to a high degree, that person will primarily experience low self-efficacy. Research on the degree of self-efficacy in schoolteachers has shown that the degree of self-efficacy has a negative effect on emotional exhaustion, which is the most important dimension of burnout (Skaalvik and Skaalvik 2014). Schoolteachers who work in an environment with a high degree of self-efficacy will see teaching as an opportunity (Skaalvik and Skaalvik 2014). It is expected that a high degree of non-conformity and therefore a high degree of self-efficacy will lead to more feelings of well-being and positive emotions, and less burnout. In sum, by influencing selfefficacy, more non-conformist behaviour can bring about indirectly, more positive and fewer negative emotions (and thus greater affective well-being), and reduced burnout, specifically reduced emotional exhaustion. Following this reasoning leads to the following hypotheses: Hypothesis 7: Self-efficacy fully mediates the positive relationship between non-conformism and positive emotions.
Hypothesis 8: Self-efficacy fully mediates the negative relationship between non-conformism and negative emotions.
Hypothesis 9: Self-efficacy fully mediates the positive relationship between non-conformism and emotional exhaustion.

Method
Sample and procedure. Our sample consisted of 401 schoolteachers. The participants in this study were informed upfront and agreed to participate voluntarily. Participants consented with the anonymous processing and secure storage of their data ensuring confidentiality and anonymity. The data have been collected using an online questionnaire, distributed via mail, containing an anonymous link, and by means of convenience and snowball sampling. A total of 337 (84%) of all respondents were women. According to Bontekoning (2011), the age groups were as follows: generation Y (19 to 33 years) 24.70%, pragmatic generation (34 to 48 years) 42.90%, generation X (49 to 63 years) 29.90%, and the protest generation (>63 years) 2.50%. In this sample, 1.00% of the respondents had secondary vocational education as their highest educational qualification; 85.53% had Higher Professional Education (HBO); and 12.21% had a university education.

Measures
Burnout. In this study burnout was measured via emotional exhaustion. We used the eight-item Utrechtse Burnout Schaal-Leraren (UBOS-L) for emotional exhaustion (EE), which is a specific questionnaire for schoolteachers (Breeman et al. 2015;Schaufeli and Van Dierendonck 2000). Teachers could respond to a 7-point Likert-type scale ranging from (score 0) never to (score 6) always. The scale for measuring emotional exhaustion (EE) concerns the feeling of being completely 'finished' or 'empty' and is based on work-related tensions, mainly in the contact with students (Schaufeli and Van Dierendonck 2000). Cronbach's alpha was found to be 0.92, which exceeds the threshold of 0.70 (Hair et al. 2010).

ARTICLE
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-020-00551-6 Affective well-being. The Job-related Affective Well-being Scale (JAWS) was used for the concept of work-related affective wellbeing (Van Katwyk et al. 2000). Further research on the various possibilities for studying affective well-being has shown that the shortened JAWS with 2 factors of 6 positive and 6 negative emotions is capable of measuring affective well-being (Schaufeli and Van Rhenen 2006;Van den Heuvel et al. 2015). Teachers could respond to a Dutch 5-point Likert type scale ranging from (score 1) never to (score 5). The Cronbach's alpha coefficient for the dimension with positive emotions was 0.90 and for negative emotions it was 0.87, both exceed the threshold of 0.70 (Hair et al. 2010). The Cronbach's alpha coefficient for the variable Affective Well-being, based on the total eigenvalue, was 0.95, which exceeds the threshold of 0.70 (Hair et al. 2010).
Self-efficacy. We used the 10 item Teacher Self-Efficacy Scale (Holzberger et al. 2013;Schwarzer et al. 1999). This self-rating questionnaire was created to assess teachers' self-efficacy in four major teaching areas: performance; skill development; social interaction, and coping with stress. Teachers could respond to a 4point Likert type scale from (score 1) completely false to (score 4) completely true. Cronbach's alpha was found to be 0.81, which exceeds the threshold of 0.70 (Hair et al. 2010).

Non-conformism.
To measure the construct non-conformity, we used the reversed 11 item conformity scale (Alquist et al. 2013;Mehrabian and Stefl 1995). Teachers were asked to respond to questions such as 'I am more independent than conforming in my ways' and 'I don't give in to others easily'. Teachers could respond to a 9-point Likert scale from (score 0) strongly disagree to (score 9) strongly agree. Cronbach's alpha was found to be 0.91, which exceeds the threshold of 0.70 (Hair et al. 2010).
Data analysis. The data were analysed with Pearson correlations and regression analyses in SPSS version 25. Mediation analyses were performed using PROCESS v3.1 for SPSS.

Results
Correlation. Table 1 displays descriptive statistics, correlations, and coefficient alphas for all factors extracted from the survey data. To provide a general overview of the relationships between the constructs, their correlations have been calculated and presented in Table 1, Cronbach's alphas are between brackets.
Regression analyses. This section shows the results of regression analyses in order to evaluate the effects of the control variables.
Direct effect of non-conformity on self-efficacy. The data in Table 2 shows the direct effect of non-conformity on self-efficacy (path A). The explained variance (R 2 = 0.10) is significant and non-conformity appears to be a significant predictor (B = 0.12, p < 0.001) of selfefficacy. Hypothesis 6 can also be confirmed with this data.
Mediating effect of self-efficacy on the target variables. This section describes the effect that a schoolteacher's self-efficacy has on the target variables positive emotions, negative emotions and emotional exhaustion. It also evaluates which type of mediation can be derived from the results. In the case of partial mediation, there is a significant mediating effect between the predictive variable and the target variable, but a significant direct effect still remains between the predictive and the target variable. In the case of full mediation, the direct effect between the predictive and the target variable is no longer significant if the mediator is added to the model. A significant mediating effect exists if it is observed that the value between the Boot LLCI (Lower Level Confidence Interval) and Boot ULCI (Upper Level Confidence Interval) do not pass through the 0 value (Field 2013). There is also an indirect effect when, prior to studying the mediating effect, it is shown that there is a significant effect between the predictive and the target variable.
Direct and indirect effect on positive emotions. The direct effect of self-efficacy on positive emotions presented in Table 3 shows significant explained variance (R 2 = 0.19, p < .001). Self-efficacy is shown to be a significant predictor (B = 0.71, p < .001) for positive emotions. As a result, hypothesis 3 can be confirmed.
Self-efficacy has a significant indirect effect on the relationship between non-conformity and positive emotions, because the LLCI (0.01) and the ULCI (0.18) do not pass through the 0 value. However, there is no mediating effect, because there is no  HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-020-00551-6 ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2020) 7:56 | https://doi.org/10.1057/s41599-020-00551-6 significant relationship between non-conformism and positive emotions. Therefore, there is no support for hypothesis 7.
Direct and indirect effect on negative emotions. The direct effect of self-efficacy on negative emotions (path B) shown in Table 4 has significant explained variance (R 2 = 0.10) and self-efficacy is shown to be a significant negative predictor of negative emotions (B = −0.49, p < 0.001), thus support is found for hypothesis 4. Self-efficacy shows an indirect negative effect on the relationship between non-conformity and negative emotions, because the LLCI (−0.10) and the ULCI (−0.03) do not pass through the 0 value. There is no mediation because there is no significant relationship between non-conformity and negative emotions. The research data therefore indicates that hypothesis 8 cannot be confirmed.
Direct and indirect effect on emotional exhaustion. The direct effect of self-efficacy on emotional exhaustion in Table 5 has a significant explained variance (R 2 = 0.09, p < 0.001)) and selfefficacy is a significant negative predictor (B = −0.77, p < 0.001) of emotional exhaustion, thus support is found for hypothesis 5. The data from Table 3 shows that self-efficacy has a significant indirect effect on the relationship between non-conformity and emotional exhaustion, because the values of the LLCI (−0.15) and the ULCI (−0.04) do not pass through the 0 value. However, there is no significant relationship between the non-conformism and emotional exhaustion. Therefore, there is no support for hypothesis 9.

Discussion and conclusion
This study aimed to enhance scholarly insights into the consequences of non-conformism and into important antecedents of self-efficacy, affective well-being and burnout, as focused on in the existing literatures (i.e., JD-R theory, SCT, and SDT; Bakker and Demerouti 2014;Bandura 1997Bandura , 2000Gagnéand Deci 2005;Ryan and Deci 2000). In this study we tested a hypothesised mediation model in which non-conformism is directly positively associated with self-efficacy, and the relationship between nonconformism and burnout and affective well-being, is fully mediated by self-efficacy. Below, the results of our study are summarised and reflected upon and its implications for theory and management practice are discussed.
Main findings and theoretical implications. First, in line with expectations and JD-R theory, we found that positive emotions are significantly negatively related to emotional exhaustion and the negative emotions are significantly positively related to emotional exhaustion Demerouti 2014, 2017).
Second, also in line with expectations and SCT and JD-R theory, we found that self-efficacy is significantly positively strongly related to positive emotions (B = 0.71, p < 0.001), and significantly negatively moderately related to negative emotions (B = −0.49, p < 0.001) and significantly negatively strongly related to emotional exhaustion (B = −0.77, p < 0.001), respectively (Bakker and Demerouti 2014;Bandura 1997Bandura , 2000Xanthopoulou et al. 2007). Thus, this study shows similarities with previous studies (e.g., Bakker and Demerouti 2017;Bandura 2000) and found additional empirical support for the theoretical relations derived from SCT and JD-R theory between self-efficacy and affective well-being and burnout (Bakker and Demerouti 2014;Bandura 1997). Third, we found that non-conformism was directly significantly positively related to self-efficacy. Our results showed that non-conformity appears to be a significant predictor (B = 0.12, p < 0.001) of self-efficacy. In this study we gained insight into the underlying mechanism explaining the relationship between non-conformism and self-efficacy at the individual level, thereby adding to the literature on non-conformity. Nonconformity is likely to produce a feeling of authenticity (Gino et al. 2015), of being independent (Nail et al. 2000), of challenging or non-accepting the status quo (Gino 2016), thus nonconformity is likely to invoke experiencing higher levels of autonomy and control (Bellezza et al. 2014;Ryan and Deci 2000). This self-perceived autonomy and competence are the fundamental universal psychological needs that are important for motivation and psychological well-being (Gagnéand Deci 2005;    Ryan and Deci 2000), and are key constructs in SDT. This aligns with JD-R theory in which autonomy, as a job resource, is a driver for work engagement (e.g., Bakker and Demerouti 2008, and autonomy is known to positively affect self-efficacy (Gagnéand Deci 2005; Makara-Studzińska et al. 2019; Molero et al. 2018;Xanthopoulou et al. 2007). Nonconformism might be considered a personal resource in JD-R theory Demerouti 2014, 2017;Gino 2016), since it provides a sense of autonomy, and ability to control and impact an individual's environment successfully (Grant 2016;Hobfoll et al. 2003). Thus, non-conformism may directly and positively influence self-efficacy (Bandura 1997(Bandura , 2000, and possibly indirectly positively influence work engagement and negatively influence burnout Demerouti 2014, 2017). In consequence, one can cautiously conclude that non-conformity might be considered a predictor of feelings of greater well-being and less burnout. Fourth and finally, no support was found for self-efficacy mediating the relationship between non-conformism and positive emotions, negative emotions, and emotional exhaustion, respectively, due to non-significant relationships between non-conformism and the target variables. Considering the reasoning above, which states that non-conformism is likely to invoke experiencing higher levels of autonomy and might be considered a personal resource in JD-R theory, non-conformism is expected to significantly relate to affective well-being (i.e., positive and negative emotions) and burnout (i.e., emotional exhaustion) (e.g., Makara-Studzińska et al. 2019; Ryan and Deci 2019). A possible explanation might be that we measured the extent to what level teachers perceive themselves to be nonconformist, irrespective of how other people respond to the teacher's non-conformist behaviour and to what level teachers give weight to the norms, positions, or standards of another or others', i.e., the difference in non-conformity subtypes independence and anti-conformity (Nail et al. 2000). For example, if other people are happy with the positive effects of teacher's non-conformist behaviour, this will likely result in positive emotions and less burnout among teachers and vice versa. Further research including the perceptions of others and non-conformity subtypes (Nail et al. 2000) seems necessary to investigate the relationships between non-conformist behaviour and well-being and burnout.
Limitations and future research. This study has several limitations. First, this study is cross-sectional in nature and definite causal inferences cannot be drawn (Schaufeli and Van Rhenen 2006). This involves the assumption that short-lived emotions lead over the long term to longer-lasting moods, and that in the case of negative emotions, these can ultimately result in burnout. Longitudinal research is needed to gain insights in possible causal relationships and the potential reverse causation between personal resources, such as self-efficacy, engagement and job demands to better identify the interaction between these concepts. Second, since this study focuses on individual perceptions, self-reported single-source data have been used. To ensure minimal effects of common method bias, participants were assured of total confidentiality and anonymity and stimulated to respond to the questions as frankly and intuitively as possible. This was done to help lessen any evaluation comprehension and make them less likely to give socially desirable answers and different types of Likert scales were used to help reduce the effects of one type responses. Third, since the focus in this study was on education, specifically schoolteachers, the results may not be generalisable to other sectors. In order to see whether the results are generalisable to other sectors, future research in this field should focus on other sectors. Fourth, no clear explanation could be given for the nonsignificant relationships between non-conformism and affective well-being, and burnout, respectively. Although, there seem to be a theoretical rationale combining JD-R theory, SCT, and SDT. Since this study was, to the best of our knowledge, one of the first to research non-conformism from an educational perspective at an individual level, future research is recommended to further examine theory (i.e., JD-R theory, SCT, and SDT) and do further empirical studies in this field. Fifth and finally, several limitations around the conceptualisation and measurement of job burnout have been recently highlighted (Maslach 2017;Bianchi et al. 2019), as well as missing the social support construct in the jobdemands resource model that might have a greater contribution to health and well-being outcomes in those involved in helping occupations (cf., Holt-Lunstad et al. 2015).
Practical implications. This study appears to support the significant importance of self-efficacy for a schoolteacher on affective well-being and reduced risk on burnout. In view of the annual increase in cases of burnout in the Netherlands, and the fact that research indicates that burnout is highest in the education sector, there is an urgent need to address this issue The degree of the schoolteacher's self-efficacy has been shown to have a large significant effect on reducing emotional exhaustion. The results of this study show that non-conformism has a significant positive effect on self-efficacy. Following Gino (2016), this practically implies that there should be a balance between conformity and non-conformity schoolteacher's work behaviour. Practical suggestions might be act in a non-conforming way towards workload (e.g., ignore certain administrative tasks that add no value to the learning of students) and towards the physical work environment (e.g., change learning locations, more field trips; Demerouti 2014, 2017). The non-acceptance of the status quo, or non-conformity, is advantageous for the students' learning and non-conformism is likely to have a positive influence on the work engagement of schoolteachers (Gino 2016;Grant 2016).

Conclusion
Burnout and emotional exhaustion are a serious threat to the educational sector (Consiglio et al. 2013;Demerouti 2014, 2017). The role of teachers in students' success, the educational system, and society as whole (Keiler 2018;Vesely et al. 2013), is of crucial importance. Teachers' well-being needs to be nurtured, by providing resources or by reducing the risk on burnout. The results of this study showed that self-efficacy affects emotions and reduces burnout of schoolteachers, and in turn, non-conformism positively influences self-efficacy. Consequently, non-conformity might be considered a predictor of feelings of greater well-being and less burnout in the educational sector.

Data availability
The datasets generated and analysed during this study are not publicly available due to the compromise of individual privacy of the participants but are available from the corresponding author on reasonable request.