Emotional Labor Mediates the Relationship between Clan Culture and Teacher Burnout: An Examination on Gender Difference

: Teacher burnout is a psychological syndrome affecting many teachers across the globe. Therefore, numerous studies have investigated antecedents of teacher burnout in order to provide recommendations to alleviate it. Although the studies pay attention to either the role of environmental factors, such as school culture, or individual factors, such as gender, in contributing to teacher burnout, they less frequently examine how teacher burnout is concurrently inﬂuenced by both factors. Thus, this study aims to understand the relationship between clan culture and burnout by examining the mediation effect of emotional labor and the moderating effect of gender. A sample of 467 primary and secondary schoolteachers from China participated in this study. The result demonstrated the following: (1) clan culture was negatively related to teacher burnout; (2) deep acting mediated the relationship between clan culture and teacher burnout, while surface acting did not; (3) the mediating effect of deep acting was only signiﬁcant in the female group of teachers, not the male group.


Introduction
Burnout, which refers to a syndrome of emotional exhaustion, depersonalization, and reduced personal accomplishment [1], has received much scholarly attention in educational research. A teacher with high burnout is more prone to experience illness, impatience, absenteeism, a lack of commitment, and poor mental health [2]. In addition to the deleterious effects of burnout on teachers themselves, teacher burnout has also been proven to be a strong predictor of student development outcomes, such as student autonomous motivation and exacerbated student behavior problems [3,4]. Given the significant effects of teacher burnout on both teachers and students, a considerable body of studies has observed the important antecedents of teacher burnout to identify effective strategies preventing teacher burnout and improving teacher wellbeing. The literature has investigated either the role of environmental factors, such as school culture [5] and policy factors [6], or the role of individual factors, such as gender [7] and thinking style [8], in contributing to teacher burnout. However, burnout should considered as being affected by both types of factors in the workplace setting [9]. Therefore, O' Brennan et al. [10] call for in-depth exploration into how both environmental factors and individual factors affect teacher burnout simultaneously.
To respond to the call, the present study aims to examine how teacher burnout is affected by clan culture as an environmental factor, and teacher emotional labor and gender as two individual factors. School culture has been considered as one important environmental factor affecting burnout [11]. For instance, the literature has shown that teachers are The emotional labor theory suggests that workers are inclined to be required to manage their feelings and displays in workplace if they need frequently engage in face-to-face or voice-to-voice interactions with clients [31]. Accordingly, teachers need perform emotional labor since they interact with students and parents every day [32,33]. By using emotional labor strategies, the emotional labor theory notes that teachers are capable of adjusting their psychological states in teaching [34]. According to the literature, surface acting and deep acting are two major emotional labor strategies [35]. The former is the strategy of faking unfelt emotions or hiding felt emotions for displaying required emotions, and the latter is the strategy of modifying felt emotions with cognitive techniques, such as distraction and self-persuasion, for displaying the required emotions [36]. In addition, some researchers like Ashforth and Humphrey [37] and Yin [36] suggest the expression of naturally felt emotions as another emotional labor strategy. Nevertheless, some other researchers like Brook [38] disagree, because genuine expression implies no actions or efforts for emotion management. Therefore, as Lu et al. [39] note that not every study on emotional labor takes the expression of naturally felt emotions into consideration.
According to Hochschild's [31] original theorization, performing emotional labor, whether it is surface or deep acting, will lead to emotional dissonance, i.e., the sense of discrepancy between felt and displayed emotion. She illustrates that chronic experience of emotional dissonance will make individuals feel depersonalized and dehumanized, resulting in psychological illness. Therefore, it has been assumed that emotional labor will cause burnout [40,41]. Nevertheless, studies increasingly demonstrate that the effects of surface and deep acting on burnout are different, i.e., the former has a positive effect while the later has a negative effect on burnout [21,42].
In addition to the effects of emotional labor strategies on burnout, the emotional labor theory proposes that emotional labor is a cultural phenomenon [43]. In social contexts, as Hochschild [44] shows, there are different emotional rules guiding how individuals feel and display emotions. Everyone is expected to feel and display appropriately according to the emotional rules, otherwise they will be socially regarded as deviant [45]. In school contexts, researchers have identified various emotional rules for teachers, including the following: avoid expressing extreme emotions, to love and to show enthusiasm for students, to be passionate and committed to teaching, to instrumentally make use of emotions to achieve instructional goals, and to hide negative emotions and maintain/demonstrate positive emotions [33,46,47]. In addition, as its emphasis is on teamwork, participation, and consensus, clan culture may require teachers to maintain supportive, trusting, and harmonious relationships [15]. In this sense, clan culture may also imply specific emotional  To fulfill the cultural requirement, teachers may manage their emotions  accordingly, by surface or deep acting. Thus, the emotional labor strategies of surface and deep acting may be associated with clan culture and be predictors of teacher burnout. In other words, they may constitute a mediation mechanism between clan culture and teacher burnout. Therefore, the following hypotheses are formulated: Hypotheses 2 (H2). Surface acting mediates the relationship between clan culture and teacher burnout.

Hypotheses 3 (H3).
Deep acting mediates the relationship between clan culture and teacher burnout.

Gender as a Moderator
The literature has illustrated that gender constitutes yet another important individual factor affecting emotional labor e.g., [48]. Fischer et al. [49] find that many societies share a similar pattern of gender role expectations, which expect males to be rational, assertive, aggressive, independent, masterful, etc., while females should be friendly, warm, interdependent, caring, emotionally expressive, etc. As influenced by the gender role expectations, males and females have different socialization experiences leading to gendered emotionality [50]. Thus, males tend to be more emotionally restrained, while females tend to be more emotionally expressive [51,52]. Similarly, the gender difference of emotional expressions also applies to the Chinese society. For example, compared to Chinese females, the Confucius cultural traditions tend to require males to control their emotions and put little emphasis on emotional expressions [53]. Therefore, the gender difference may shape how male and female teachers manage emotions in workplace differently. For example, numerous studies show that females prefer to use surface acting in the workplace than their male counterparts do e.g., [54,55]. In their study, Akın, Aydın, Erdogan, and Demirkasımoglu [23] observe that female teachers may generally use both surface and deep acting to manage emotions in classroom more often than male teachers do.
According to Timmers et al. [56], the gender difference in emotional labor is related to different motives for emotion management. They find that females tend to be motivated to manage their emotions for maintaining relationships and harmony, while men are inclined to be motivated to manage emotions for staying in control over themselves, situations, or others. Therefore, compared with male teachers, female teachers may look more favorably on school clan culture and more proactively manage their emotions to maintain relationships, harmonious teamwork, and achieve consensus in a school; in turn, they produce or reproduce school clan culture. In other words, school clan culture may affect teachers' emotional labor, but the effects may be moderated by gender. Therefore, the following hypotheses are formulated: Hypotheses 4 (H4). Gender moderates the relationship between clan culture and surface acting, indicating a gender difference for the foregoing relationship.

Hypotheses 5 (H5).
Gender moderates the relationship between clan culture and deep acting, indicating a gender difference for the foregoing relationship.
Based on the literature review, school clan culture may negatively affect teacher burnout, but its effects may be mediated by the emotional labor of teachers. Moreover, the relationship between school clan culture and emotional labor may also be moderated by teacher gender. The hypothetical relationships between these variables are illustrated in Figure 1, featuring the conceptual framework of the present study. relationship between school clan culture and emotional labor may also be moderated by teacher gender. The hypothetical relationships between these variables are illustrated in Figure 1, featuring the conceptual framework of the present study.

Participants
The study surveyed teachers who participated in professional development courses offered by a university locating in Beijing in 2018. After attending the courses, every teacher was invited to complete an online questionnaire, where informed consent was obtained from all participants in advance. The questionnaire was completed by 475 teachers. After deleting the 8 cases with missing data, this study ultimately included 467 participants. Among these participants, 107 were male (22.9%), while 360 were female (77.1%). The average age of the participants was 33 years old. The schools in which the participants worked varied in educational stage and educational quality. Specifically, 107 participants taught in kindergartens (22.9%), 203 in primary schools (43.5%), and 157 in secondary schools (33.6%). In terms of educational quality, defined by students' academic performance, 172 participants reported that they came from high-achieving schools (36.8%), 245 from ordinary schools (52.5%), and 50 from low-achieving schools (10.7%).

Clan Culture
Clan culture was measured by the clan culture subscale of the Organizational Culture Assessment Instrument (OCAI) developed by Cameron and Quinn [16]. It contained 6 items rated on a 5-point Likert-type scale from 1 (completely disagree) to 5 (completely agree). An example of the sample items: "Our school is a very personal place. It is like an extended family. People seem to share a lot of themselves." In this study, the Cronbach's alpha for clan culture was 0.909.

Surface Acting
Surface acting was measured by the surface acting subscale of the Teacher Emotiona Labor Strategy Scale (TELSS) developed by Yin [36]. It contained 6 items rated on a 5point Likert-type scale ranging from 1 (completely disagree) to 5 (completely agree). One example sample item: "The emotions I show to students or parents are different from what I really feel in my heart." In this study, the Cronbach's alpha for surface acting was 0.904

Deep Acting
Deep acting were measured by the deep acting subscale of the TELSS [36]. It contained 4 items rated on a 5-point Likert-type scale ranging from 1 (completely disagree)

Participants
The study surveyed teachers who participated in professional development courses offered by a university locating in Beijing in 2018. After attending the courses, every teacher was invited to complete an online questionnaire, where informed consent was obtained from all participants in advance. The questionnaire was completed by 475 teachers. After deleting the 8 cases with missing data, this study ultimately included 467 participants. Among these participants, 107 were male (22.9%), while 360 were female (77.1%). The average age of the participants was 33 years old. The schools in which the participants worked varied in educational stage and educational quality. Specifically, 107 participants taught in kindergartens (22.9%), 203 in primary schools (43.5%), and 157 in secondary schools (33.6%). In terms of educational quality, defined by students' academic performance, 172 participants reported that they came from high-achieving schools (36.8%), 245 from ordinary schools (52.5%), and 50 from low-achieving schools (10.7%).

Clan Culture
Clan culture was measured by the clan culture subscale of the Organizational Culture Assessment Instrument (OCAI) developed by Cameron and Quinn [16]. It contained 6 items rated on a 5-point Likert-type scale from 1 (completely disagree) to 5 (completely agree). An example of the sample items: "Our school is a very personal place. It is like an extended family. People seem to share a lot of themselves". In this study, the Cronbach's alpha for clan culture was 0.909.

Surface Acting
Surface acting was measured by the surface acting subscale of the Teacher Emotional Labor Strategy Scale (TELSS) developed by Yin [36]. It contained 6 items rated on a 5-point Likert-type scale ranging from 1 (completely disagree) to 5 (completely agree). One example sample item: "The emotions I show to students or parents are different from what I really feel in my heart". In this study, the Cronbach's alpha for surface acting was 0.904.

Deep Acting
Deep acting were measured by the deep acting subscale of the TELSS [36]. It contained 4 items rated on a 5-point Likert-type scale ranging from 1 (completely disagree) to 5 (completely agree). An example sample item: "I try to really feel the emotions that I have to show to students or parents". In this study, the Cronbach's alpha for deep acting was 0.847.

Teacher Burnout
Teacher burnout was measured by Li and Wang's [57] Chinese version of the Maslach Burnout Inventory-Educators Survey (MBI-ES) that was developed from Maslach, Jackson, and Leite's [58] MBI-ES. The Chinese version of the MBI-ES consists of 15 items measuring 3 dimensions of teacher burnout: emotional exhaustion (5 items), depersonalization (4 items), and reduced personal accomplishment (6 items). Each item was rated on a 7-point Likert-type scale ranging from 1 (completely disagree) to 7 (completely agree). Sample items for each dimension are as follows: "I feel emotionally drained from my work" (emotional exhaustion), "I doubt the significance of my work" (depersonalization), and "I have accomplished many worthwhile things in this job" (reduced personal accomplishment). The Cronbach's alpha of this scale was 0.909 as a whole.

Demographic Variables
The teachers' demographic information such as gender, age, teachers' educational attainments, professional level, administrative position, and schools' features (such as educational stage and educational quality) were also collected. Gender was treated as a moderator variable and the rest of demographic variables were controlled because they may influence teacher emotional labor and burnout [10,42].

Data Analysis
The Statistical Package for Social Science (SPSS 24.0) and PROCESS 3.0 were used to analyze the data. The variance inflation factor (VIF) of clan culture, surface acting, and deep acting were smaller than 2, indicating that collinearity was not a problem in this study [59]. Moreover, Harman's single-factor test suggested that a total of 5 characteristic roots were bigger than 1. The maximum factor variance interpretation rate was 30.91% (not exceeding the threshold of 40%), indicating that common-method biases were insignificant [60].
After descriptive statistics and Pearson correlation analysis, three steps were performed to further to test our hypotheses. First, the effect of clan culture on teacher burnout was estimated first after controlling demographic variables. Second, Model 4 of PROCESS was performed to investigate whether surface acting and deep acting mediated the effect of clan culture on burnout by using a bootstrapping estimation technique with 5000 samples to obtain the confidence intervals (CIs) [61]. Last, Model 7 of PROCESS was used to further examine the moderation effect of gender on the foregoing indirect relationships.

Preliminary Analysis
Descriptive statistics and Pearson correlation among the variables were calculated in Table 1. The result indicated that teacher burnout was correlated with clan culture negatively (r = −0.419, *** p < 0.001) and surface acting positively (r = 0.509, *** p < 0.001). Meanwhile, teacher burnout was insignificantly associated with deep acting (p > 0.05), while the insignificant correlation between them could still refer to significant indirect effects methodologically [61]. Moreover, clan culture was negatively associated with surface acting (r = −0.151, ** p < 0.01) and positively related to deep acting (r = 0.105, * p < 0.05).

The Main Effect of Clan Culture on Teacher Burnout
Taking demographic information as control variables, clan culture as an independent variable, and teacher burnout as the dependent variable, a linear regression analysis was computed. The results showed that the overall model fit was adequate (F = 7.429, p < 0.001, R 2 = 0.220). The impact of clan culture on teacher burnout was significant (β = −0.480, t = −9.490, p < 0.001), while none of the demographic variables had any significant effect (p > 0.05). Thus, H1 was preliminarily supported.

The Moderation Effect of Gender
According to the Model 7 of PROCESS (see Table 3), the interaction of clan culture and gender was negatively related to deep acting (β = −0.361, SE = 0.114, p < 0.01), indicating that gender moderated the relationship between clan culture and deep acting. Thus, the difference between two regression coefficients derived from two separate samples (the male and female groups, namely) was tested. In the female group, β was 0.268 (t = 4.269, p < 0.001, 95% CI = [0.144, 0.391]), while in the male group, β was −0.094 (t = −0.952, p > 0.05, 95% CI = [−0.288, 0.010]). The Fisher's Z score was 3.093, which was higher than 1.96, indicating a significant difference between the 2 samples.  Moreover, the moderated mediation effects in two groups were examined. According to the results (see Table 4), the mediating effect of deep acting was moderated by gender, meaning that there was a significant difference for the indirect effect from clan culture to teacher burnout via deep acting in the male and female groups. Specially, in the female group, deep acting mediated the relationship between clan culture and teacher burnout (β = −0.044, 95% CI = [−0.079, −0.016]). By contrast, the indirect effect was not significant in the male group (β = 0.016, 95% CI = [−0.023, 0.061]). Therefore, H5 was supported while H4 was not.

Discussion
The literature implies that clan culture tends to provide a favorable school environment for teachers to search for competence, autonomy, and relatedness, which are the basic human psychological needs [26], because of its value of participation, collaboration, cohesiveness, and social harmony [27]. Thus, as the theory of basic psychological needs predicts, the clan culture tends to be negatively associated with teacher burnout, because satisfying basic psychological needs is conductive to psychological wellbeing and preventive of ill-being [25]. This prediction is supported by the present study. It shows a significant negative association between clan culture and teacher burnout and its regression analysis further suggests a negative prediction of clan culture to teacher burnout when demographic variables are controlled. Moreover, similar to other studies e.g., [21,42], the study finds that teacher burnout is negatively associated with deep acting but positively associated with surface acting. This means that managing emotions with the strategy of deep acting may alleviate burnout, while managing emotions through surface acting may aggravate burnout in teaching.
Nevertheless, further data analyses indicate a more complex mechanism between clan culture, emotional labor, and teacher burnout than what has been described above. First, as the emotional labor theory suggests [14,22], the findings imply that teachers may emotionally respond to school clan culture by emotional labor strategies to adjust and shape their psychological states. In particular, the study identifies that teachers are more likely to engage in deep acting instead of surface acting under school clan culture. A possible reason for this phenomenon is that school clan culture may create a friendly and harmonious environment in which the teachers tend to feel supported, trusted, and recognized [16]. Therefore, they may be willing to make efforts to manage emotions in an appropriate way for maintaining such an environment. Although surface acting requires less effort for emotion management than deep acting [31,36], they may not prefer to perform surface acting. This is because it may be easier for others to sense that they intentionally fake or hide emotions leading to an impression of inauthenticity, which may harm interpersonal relationships. To avoid this outcome, it is possible for them to use deep acting instead of surface acting. Therefore, surface acting may not play a mediating role to the relationship between school clan culture and teacher burnout, but deep acting may. Moreover, although this study is conducted in the Chinese clan-family cultural context, it is also worthwhile noting that school clan culture tends to be universal globally. For example, Berkemeyer et al. [62] observed that clan culture may be a more common school culture in practice compared with the market and/or hierarchical culture, indicating that the findings could also potentially apply to other countries and areas.
Second, the mechanism may become more complex when gender is taken into consideration. According to the findings, the mediating effect of deep acting is only significant in the female group of teachers but not the male group. This suggests that, compared with male teachers, female teachers are more inclined to spend resources and time on building and maintaining interpersonal relationships by deep acting, to respond to school clan culture and to adjust their psychological states. The findings are different from numerous previous studies, which illustrate that females prefer to use surface acting, in contrast to men [54,55]. To explain the finding, the disposition of female teachers needs to be taken into account. As Timmers, Fischer, and Manstead [56] suggest, compared with males, females tend to value harmonious social relationships and be motivated to manage such relationships. This disposition may make female teachers a better fit for the environment of school clan culture than male teachers. According to Kim [63] and Lam et al. [64], a person-environment fit is inclined to provoke positive attitudes of employees toward an organization, and encourage them to devote more conscious efforts for deep acting rather than faking their emotions by surface acting. Therefore, female teachers may be more willing to engage in deep acting than males in school clan culture. If this explanation holds, it implies that the existing research that investigated gendered emotional labor e.g., [55] may have overlooked the interactive effects of gender and environmental factors on emotional labor. In this sense, it is suggested that further studies pay attention to the embedded and interactive effects in order to improve our understandings of emotional labor, and hence extend the emotional labor theory [65].
To sum up, the present study aims to understand the relationship between clan culture and burnout by examining the mediation effect of emotional labor and the moderating effect of gender, simultaneously. The results reveal that clan culture was negatively related to teacher burnout. Moreover, deep acting mediates the relationship between clan culture and teacher burnout, while surface acting did not. Finally, the mediating effect of deep acting is only significant in the female group of teachers but not the male group, indicating that gender plays an important role in how individuals face the external environment and how they use emotional labor strategies. Accordingly, school administrators should build school clan culture for their schools by emphasizing teacher collaboration, delegating power, valuing teachers' strengths, facilitating teachers' abilities to gain competence and autonomy, and providing a safe environment for teaches to take risks. This is because clan culture can generally alleviate teacher burnout, supporting and sustaining teacher development. Moreover, school administrators should account for the moderating and mediating effects of emotional labor on the positive effects of clan culture. Thus, they should provide in-service training or workshops to facilitate teachers' development of appropriate strategies to manage their emotions. The training or workshops need addresses the gender difference, since our findings suggest male and female teachers may have different concerns about clan culture. To achieve that, they should examine the needs of the male and female teachers in their schools and design the training or workshops based on the results.
There are some limitations to the study. First, the study only surveyed 467 Chinese teachers through convenient sampling, meaning there is a limitation to the generalizability of the findings. The second limitation is that female participants outweigh males greatly in this study, constituting a percentage of 77.1%. The equal distribution of female and male teachers in sampling for future research would be a more consistent way to judge the effect of gender. Lastly, although we explored the combined effects of both clan culture and gender on emotional labor, further research can aid in understanding the mechanism of emotional labor by integrating other environmental and personal factors in future studies.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Ethics Committee of the Faculty of Education Beijing Normal University (protocol code BNU202109100022, approved by 28 September 2021).