Supervisors’ Leadership Styles’ Influence on Foreign Teachers’ Self-Efficacy in a Cross-Cultural Work Setting: A Moderated Mediation Analysis

Leadership style impacts on the manner and frequency of feedback transmission. However, communication challenges between superiors and subordinates originate from cultural differences, which undermine the usefulness of feedback. The study tested leadership style’s effect on self-efficacy through a moderated mediation approach, examined through the lens of the cross-cultural adaptation theory. Path analysis conducted on data from 281 foreign teachers in Chengdu, China, revealed that there is a positive effect of Chinese supervisors’ leadership styles on foreign teacher’s self-efficacy. Leadership style similarly has an influence on the nature of feedback. And the nature of feedback in turn mediates leadership style and self-efficacy. We establish in particular that transactional and transformational leadership styles, through the nature of feedback, influence self-efficacy of foreign teachers. Moreover, the association between the nature of feedback and self-efficacy is moderated by the perceived value of feedback. Employees’ perceptions are also found to be crucial in determining the value of feedback. It is thus imperative for supervisors and managers working with foreigners as subordinates to figure out when and how to provide valuable feedback. We conclude with suggested areas for further research.


Introduction
It has been established that individuals' assessment of their ability to accomplish a given task effectively predicts the level of determination in accomplishing that task (Bandura, 2010). In addition to understanding the nature of selfefficacy, the outcomes of self-efficacy are also critical. Consciously and unconsciously seeking information about the outcomes of one's self-efficacy is useful for interpreting and realigning oneself to fit the environment, and aids in the maintenance of positive behaviors. Feedback is the information people seek from others about themselves (Norcini, 2010) which then enables them to make adjustments in their actions.
Culture influences followers' behaviors toward feedback (Milliman et al., 2002). Working in other cultures or with people from other cultures requires an integrative communication pattern for appreciable comprehension and professional working relationship between the host and the foreigner. Cross-cultural adaptation (CCA) is a phenomenon that deals with how foreigners adapt to working conditions in their new environment (Kim, 2001). It also relates to how indigenes adjust to accommodate foreigners. In communication and social psychology, cross-cultural adaptation theory (CCAT) has been studied extensively, employing such concepts as acculturation, assimilation, and marginality to analyze adaptation experiences and psychological acculturation (Kim, 2001). To adapt successfully, one has to focus on obtaining new communication practices and leaving the old ones. Kim's assumption in the CCAT is that people adapt to problems from a given environment by maintaining a balance within the system through several communication styles. These modes of communication are found in feedback systems and influence the degree to which that feedback produces self-efficacy in followers. However, there is paucity of research addressing the extent to which cultural differences influence feedback and self-efficacy. It is this gap that our study contributes toward filling.
The broad framework that envelops the factors that enable or hinder adaptation to a host culture in cross-border migration is the CCAT. We apply this theory to a study of foreign teachers teaching English as a foreign language (TTEFL) in schools in Chengdu, the capital city of the Sichuan Province of China. Given that most studies on feedback hardly consider its perceived value to the receiver, we aim to provide insights into the moderation effect of perceived value on the relationship between leadership style and self-efficacy if analyzed through a cross-cultural lens. Also, we explore the existence of foreign employees under local managers in the conduct of this study, as exemplified by Gumah (2020a, 2020b).

CCAT
CCA is a multi-faceted process in which people's ability to function becomes useful in a culture other than their own. According to Kim (2001), CCA describes how settlers change from being strangers and the processes it entails. It is "the dynamic process by which individuals, upon relocating to an unfamiliar cultural environment, establish (or reestablish) and maintain a relatively stable, reciprocal, and functional relationship with the environment" (Kim, 2002, p. 260). Searle and Ward (1990) contend that it is the natural tendency to strive for internal equilibrium in times of challenging environmental conditions to promote the sociocultural adjustment and the psychological feeling of well-being. Motivational and behavioral cultural intelligence are relevant to cultural adaptation and used to support the feeling of selfefficacy (Bhaskar-Shrinivas et al., 2005). Therefore, intrinsic or extrinsic motivation, coupled with positive behavior, will increase CCA and, thereby, self-efficacy. Molinsky (2007, p. 623) calls it "cross-cultural code-switching," emphasizing a single interaction which he considers the foundation of unwavering adjustment to a new culture. In this regard, foreigners' reactions to situations influence their effectiveness and their performance in an organization (Bierwiaczonek & Waldzus, 2016).
For effective cross-cultural interactions, cultural intelligence is vital among employers and employees. Cultural intelligence refers to an individual's use of reasoning and observation skills to understand unfamiliar gestures and situations and adopting suitable behavioral responses (Daft, 2008). Attar et al. (2019), in their study, specify a significant positive connection between the dimensions of cultural intelligence and leadership style (transactional and transformational leadership styles). Native supervisors and employees of foreign origin may lack cross-cultural skills or feel they lack it to positively produce the required behavior, which can be embarrassing (Molinsky, 2007). Communication within multicultural work settings can be useful when managers have cultural intelligence (Ochieng & Price, 2010). For foreigners to adapt to hostile environments, they sometimes require positivity, openness, and a strong personality (Kim, 2001). For Kim (2001), communication, environment, individual predisposition, and intercultural transformation are the facets and features that interact to impede or facilitate individuals' cultural adaptation.
Central to the adaptation process is the competence of the host and foreigner to communicate effectively. Communicating by the host culture's practices and norms indicates an active engagement in its social communication processes (Kim, 2001), which is vital to cultural adaptation. Multiple factors work simultaneously-from the level of personal disposition to the broader environmental conditions-to affect the communicative interface between the individual and the host environment, a process described by Deterding, (2011) as "linguistic code-switching." To adapt successfully, one should relegate old cultural predispositions to the background and acquire new cultural communication practices. To become conversant in the host communication system demands active involvement in the relational and the local community's mass communication processes (Kim, 2001). The perceived low confidence of the host in his or her communication abilities prevents him or her from interacting effectively with the foreigner. In this case, the low confidence of Chinese supervisors in speaking English to their foreign employees could significantly reduce the frequency with which they interact with foreigners. This perceived low self-efficacy in speaking English reduces their ability to provide foreign teachers with relevant feedback on their job.
Various attributes of leadership are found to affect successful maneuvering and multicultural teams. This study follows the lead of Kim (2001) in her identification of communication as a salient part of CCA. In employing CCAT, we acknowledge that communication challenges exist between foreigners and supervisors. Foreigners have to adjust to the communication norms of the host country, while supervisors have to also "code-switch" for better collaboration in the workplace. We envisage that the successful management of communication difficulties will influence CCA, thereby enhancing the self-efficacy of followers.

Leadership Style and Self-Efficacy
The centrality of leaders in influencing the self-efficacy of followers is undisputed (Damanik & Aldridge, 2017). Research shows that this is possible through spoken and unspoken means (Gagné & Deci, 2005). Fackler and Malmberg (2016) think that the leader's approach also exerts an enormous effect on the behaviors and perceptions of followers, including their self-efficacy. There is a saturation of theories and models of leadership style in the leadership literature that we do not intend to repeat. Current theories focus mainly on transactional and transformational leadership styles as the dominant approaches (Antonakis et al., 2003), which we exploit for our study.
Transformational leadership, according to Burns (1978), "occurs when one or more persons engage with others in such a way that leaders and followers raise one another to higher levels of motivation and morality" (p. 2). Studies have identified that transformational leaders motivate follows, ensure liberty, equality, and justice, which combine to appeal to, energize, inspire, and mobilize followers for action, and that, transformational leaders are mostly measured using four attributes, namely, personalized consideration, intellectual stimulation, idealized influence, and inspirational motivation (Burns, 2003). First, inspirational motivation signifies motivating and inspiring others by showing commitment to collective goals, demonstrating enthusiasm, and communicating high expectations (Leithwood & Jantzi, 2006). By motivating and inspiring followers, transformational leaders get followers to move along with them. Besides, the idealized influence attribute of transformational leaders is an attribute that distinguishes such leadership from controlling and autocratic leaders. Transformational leaders, through idealized influence, avoid using power for private gain, they are concerned about the needs of others, set challenging goals for others and demonstrate high moral standards (Walumbwa & Lawler, 2003). In addition, leaders who practice a transformational style of leadership are intellectual stimulators. They stimulate the intellectual capabilities of followers by encouraging innovation and creativity. Finally, individualized considerations represent the special treatment given to each individual and the provision of mentorship for the development of followers (Avolio & Bass, 2004). It also includes personalized support, advice, and attention to followers. Through these key attributes, recent studies have associated transformational leaders to organizational and work-related behavior outcomes (Azim et al., 2019;Eliyana et al., 2019). Transformational leadership is mostly preferred to transactional leadership, although in specific situations, transactional leadership is ideal.
Those who adopt the transactional leadership approach exhibit two main characteristics; contingent rewards and punishments (Camps & Torres, 2011), and active/passive management by exception (Penno, 2020). Contingent rewards imply the provision of material or psychological rewards conditioned upon the fulfillment of agreed contracts. Leaders who adopt this style see it as a contractual obligation based on the fulfillment of set goals. It is characterized by objective setting, monitoring, and the control of outcomes (Antonakis et al., 2003;Avolio & Yammarino, 2013). Active management by exception denotes the leader's active observance in ensuring that standards are met (Hackman & Johnson, 2009). They set goals, articulate explicit agreements regarding expectations, reward systems, and the provision of positive feedback to keep everybody focused (Vera & Crossan, 2004). The passive approach involves intervening and initiating corrective action to improve outcomes when subordinates do not meet acceptable performance levels (Hackman & Johnson, 2009). The transactional leader implements rules and regulations, stamps their authority, describes and emphasizes the achievement of defined goals, and guides employees to achieve these predefined targets (Yukl, 2013). Transactional leaders are also inclined to be command-driven and action-oriented. They are prepared to work within existing systems and negotiate to attain the goals of the organization. The followers' skills and experiences are eventually exploited to the maximum when people who perform are rewarded, and under-performers punished (Udoh & Agu, 2012).

Research Hypotheses
Types of leadership: Leaders influence and immensely contribute to follower's self-efficacy (Damanik & Aldridge, 2017) through feedback (Gagné & Deci, 2005). The leader's approach impacts followers' perceptions of themselves, their behavior, and their self-efficacy (Fackler & Malmberg, 2016). Current studies on teachers' self-efficacy have established the role of transformational (Damanik & Aldridge, 2017;Gkolia et al., 2018) and transactional leadership (Hoxha & Hyseni-Duraku, 2017; Jacobsen & Andersen, 2017) on teachers' self-efficacy. However, the study of feedback from local supervisors to foreign employees' and its impact on self-efficacy in cross-cultural contexts has received insufficient consideration. To contribute to the academic discourse in narrowing the above gap, we, hypothesize that: Hypothesis 1a: A Chinese supervisor's transformational leadership positively affects a foreign teacher's self-efficacy. Hypothesis 1b: A Chinese supervisor's transactional leadership positively affects a foreign teacher's self-efficacy. Nature of feedback: Feedback is information given to persons at the workplace based on the quantity or quality of their past performance that signifies their accomplishments. It offers an avenue for employees to improve their performance, understand the expectations of the workplace and identify reward opportunities (Bear et al., 2017). Feedback comes in various forms and dimensions. Negative and positive feedback is referred to as feedback valence (Lechermeier & Fassnacht, 2018), which we refer to as the nature of feedback. Negative feedback involves receiving a more direct indication of dissatisfaction from colleagues or supervisors for unimpressive performance, while positive feedback is a report of impressive performance (Herold & Greller, 1977). Lechermeier and Fassnacht (2018), in a review of studies on feedback, found that most studies on feedback valence supported the hypothesis that constructive feedback results in useful personal reactions than negative feedback. In other studies, positive feedback motivates and encourages desirable performance (Anderson et al., 2016), while negative feedback contributes to employees' self-regulation (O'Malley & Gregory, 2011). If adequately managed, a mixture of negative and positive feedback provides a reliable perspective of the individual's performance at work.
Employees obtain feedback by informally monitoring the environment through direct inquiry and formal feedback mechanisms (Sully De Luque & Sommer, 2000). Herold and Greller (1977) distinguish between hierarchical and nonhierarchical forms of feedback, depending on the nature of the feedback deliverers. The feedback that is offered by supervisors is hierarchical; peer-or colleague-initiated feedback is nonhierarchical. Feedback can be delivered in a controlling and informational manner (Ryan, 1982). The controlling style of feedback emphasizes the desires and demands of superiors to elicit specific behaviors from recipients. Feedback that is informational improves the sense of internal causality of the recipient (Zhou, 1998). Factors such as consistency, distinctiveness, and consensus determine a supervisor's behavior in giving feedback on an employee's performance (Larson, 1986). These factors are used to give feedback and for understanding why specific types of feedback are given. These factors are also largely influenced by the context within which feedback is given.
In communicating performance feedback, supervisors are critical, and they support subordinates to accomplish their objectives by clarifying goals while lessening ambiguity (Geijsel et al., 2009). Balyer and Özcan (2012) have established that teachers sometimes positively rate the communication skills of their headmasters. When communication is inadequate, self-image is threatened, which tends to reduce feedback effectiveness (Ten Cate, 2013). Positive language, however, improves feedback reception, and task performance (Nguyen et al., 2017). This supports the view that supervisors who adequately communicate expectations assist in improving performance. The communication competence of supervisors working with employees from various cultural backgrounds will be critical in the delivery of feedback. We, therefore, hypothesize that:  (Scheeler et al., 2004). Feedback frequency is information that allows persons to know how good or bad a job is done (Herold & Greller, 1977). Feedback frequency also reflects the regularity in the giving of feedback within a specified period (C. F. Lam et al., 2011). There is the perception that the more frequent the feedback, the more the possibility of self-correction and seeking help for deficiencies (Sato & Loewen, 2018). The contrary view is that more is not always better (C. F. Lam et al., 2011). A study has found that the best results are achieved when professionals received infrequent (monthly) but detailed feedback (Casas-Arce et al., 2017), corroborating Lam's assertion. Holderness et al. (2020) found a nonlinear but mixed connection between relative performance information frequency and performance. In other instances, while initially performance increases with frequent feedback, performance subsequently decreases. These varied findings on the effect of feedback frequency motivate our hypothesis that: Feedback is an intrinsic motivator (Fong et al., 2019) that influences self-efficacy (Abbas & North, 2018). Studies related to feedback have proved that practice conditions that induce positive feelings related to performance outcomes can increase perceptions of competence and self-efficacy (Mouratidis et al., 2008). Therefore, the nature or frequency of feedback can individually or collectively influence selfefficacy. Supervisors' feedback (frequency and nature) play a significant role in enhancing or diminishing the selfefficacy of employees. For instance, people who received good feedback when learning motor tasks demonstrated high levels of intrinsic motivation and self-efficacy (Abbas & North, 2018). The attitude of giving feedback by supervisors characterized by its frequency and its nature play a significant role in enhancing or diminishing the self-efficacy of employees. We, therefore, posit that when an employee does not perceive feedback to be valuable in enhancing self-efficacy and performance, feedback in its nature, source, and frequency will make a limited contribution to self-efficacy. The hypothesis thus follows: Hypothesis 4a: Frequent feedback positively affects foreign teachers' self-efficacy in a cross-cultural teaching environment. Hypothesis 4b: The nature of feedback positively affects foreign teachers' self-efficacy in a cross-cultural teaching environment. Hypothesis 5a: Feedback mediates the effect of transformational leadership and self-efficacy in a cross-cultural teaching environment. Hypothesis 5b: Feedback mediates the effect of transactional leadership and self-efficacy in a cross-cultural teaching environment.

Sample
Foreign English teachers in Chengdu, China, were the sample used for the study. From February to June 2018, an online survey was administered to 281 foreign teachers, constituting our total sample size for study. We employed a convenience sampling procedure to obtain participants from the target population based on accessibility and readiness (Creswell & Creswell, 2018). Only foreign teachers TEFL and supervised by a Chinese national participated. Foreigners were obtained using social media platforms of Wechat and QQ. Initially recruited participants were asked to pass on the questionnaire to friends and acquaintances who satisfied the condition for participation. No form of compensation was offered to participants for their involvement in the study.

Measurement of Variables
Avolio and Bass' (2004) 9-item multifactor leadership questionnaire (MLQ) was used to measure leadership style on a 5-point Likert-type scale from 1 (not at all) to 5 (always). To measure Transformational Leadership, statements were used to define the attributes of intellectual stimulation, idealized influence, individualized considerations, and inspirational motivation. We used statements such as re-examines critical assumptions to question whether they are appropriate, considers the moral and ethical consequences of decisions, helps others to develop their strengths, goes beyond self-interest for the good of the group. Similarly, Transactional Leadership was measured using responses to statements related to contingent reward (makes clear what one can expect to receive when performance goals are achieved) and passive management by exception (waits for things to go wrong before taking action).
The study adopted and modified the 12-item Teacher Sense of Efficacy Scale (TSES; Tschannen-Moran et al., 1998). Student engagement, classroom management, and instructional strategies are the three critical aspects of teaching measured by this scale. This scale has been used not only for general teachers but also for language teachers (see Eslami & Fatahi, 2008;Swanson, 2012). We used questions such as "How much can you do to motivate students who show low interest in learning English?"(student engagement), "How much can you do to make students follow classroom rules?" (classroom management), and "how much can you do to implement alternative instructional strategies in your English class?" (instructional strategies). Respondents showed their level of agreement with each statement by selecting nothing (1) to a great deal (5).
Feedback's frequency was measured based on how often feedback is given within a stipulated period. We used the question; how often do you receive feedback from your supervisor? The respondent is required to select a range from 1 (yearly) to 5 (weekly).
The Nature of Feedback shows whether the supervisors' feedback to the employee is positive or negative. Using responses from 1 (unsatisfactory) to 5 (exceptional), we asked the question; what feedback have you received on your recent performance? Negative feedback is when respondents are informed that their performance is "unsatisfactory" or "needs improvement." Responses such as "meets expectation," "satisfactory," and "exceptional" are regarded as positive feedback.
Perceived value of feedback is measured using the single question: is feedback valuable to your performance? This is similar to the question used by Liu and Gumah (2020a) in their study. An attribution scale is used in responding to this question ranging from 1 (not at all) to 4 (a great deal). Responses such as somehow, quite a bit, and a great deal are an indication of the positivity of feedback.

Procedure
We used path analysis through structural equation modeling to establish the relationships in this study as shown in Figure  1. Path analysis is a research analysis tool used to determine whether a multivariate set of nonexperimental data collected fits its causal model (Wuensch, 2015). We used SPSS 25 for correlation analysis and AMOS 21 for moderation and mediation analysis. Schreiber et al. (2006) provide a benchmark for overall model fit indices. Based on this benchmark, the data show an adequate fit for the hypothesized model. The data produced a satisfactory fit index for Tucker-Lewis Index (TLI) = .986, comparative fit index (CFI) = .992, and root mean square error of approximation (RMSEA) = .065.

Limitations of the Study
The study suffers two main methodological limitations, mainly; sampling error/bias and common method bias. Convenience sampling, which is widely used, suffers from sampling error/bias, which prevents the generalization of the findings and sometimes affects external validity. Since the total population of foreign teachers in Chengdu is not available, and efforts to obtain same from relevant authorities failed, this sampling technique was the most appropriate to use. We, therefore, concede that the study is a representation of the sampled population and not necessarily the general foreign teacher population in Chengdu. With access to the general foreign teacher population in Chengdu, it will be appropriate to employ a probability sampling technique in future studies.
Common method bias stems from a single source of data collection. We used both ex-ante and ex-post strategies (Podsakoff et al., 2012) in data collection and analysis to remedy this limitation. In this instance, we designed the questionnaire using different measuring scales and assuring respondents of their anonymity in the security of the data collected while making the questions precise, accurate, and unambiguous.

Descriptive Statistics
The respondents and their work-related profile are shown in Table 1. Males (61.9%) and persons between the ages of 25 to 34 (64.7%) dominate the population sampled. The majority of respondents had attained a Bachelor's degree (65.8%). Teachers in their current teaching job of more than 13 months but up to 59 months make up 45.2% of the sampled population. Foreign teachers from Africa make up 70% of the sampled population. The control variables used were educational level, the region of origin, place of work, and years on the job. This is because, these variables are noted in various studies as contributors to teachers' self-efficacy. For example, a study found that ethnicity played a role in counseling self-efficacy (S. Lam et al., 2013).
We followed the three-step procedure for establishing moderation (Baron & Kenny, 1986). We first conducted a correlation analysis to establish the relationship between the variables as shown in the correlation matrix in Table 2. The table shows a statistically significant correlation between the variables. It also presents the means and standard deviation of the variables.
We used bootstrapping in AMOS to establish the direct and indirect effects of the relationships as shown in Tables  3 and 4. We also used bootstrapping in AMOS as our main analysis technique. Hayes (2009) explains that the advantage with bootstrapping is that estimates are based on the indirect effect alone and make no suppositions about whether a sample is normally distributed, thereby avoiding this challenge.
We established the indirect and direct effects in the model before determining whether there is moderation. Through regression analysis, we determined how perceptions on the value of feedback moderate the link between the nature/ frequency of feedback and self-efficacy. According to Champoux and Peters (1987), the procedure in regression analysis used for testing moderation effect is generally recognized and statistically straightforward. In doing this, the mediators and moderators were centered and interacted to produce an interaction term. This approach solves the problem of nonessential multicollinearity (Fairchild & Mcquillin, 2017). There was a positive and a statistically significant (β = .041, p < .05) effect of the first interaction term (nature of feedback*perceived value) on self-efficacy, supporting Hypothesis 6a. Meanwhile, hypothesis 6b is not supported because the second interaction term (feedback frequency*perceived value of feedback) negatively affects self-efficacy (β = −.040, p > .1).

Discussion and Implication
In this study, we set out to establish how leadership style influences self-efficacy through feedback in a cross-cultural context. Based on the analyzed data, we demonstrate that the qualities of both transactional and transformational leaders as exhibited by Chinese supervisors, positively influence the self-efficacy of foreign teachers. Active management by exception is an attribute of transactional leaders that depicts their enthusiasm in guaranteeing that standards set are followed. Chinese supervisors also maintain existing norms and intervene when employees are incapable of achieving set performance standards (Hackman & Johnson, 2009). Indicating reward systems after setting goals, providing clear expectations, and giving constructive feedback are critical to transactional leaders (Vera & Crossan, 2004). Alternatively, transformational leadership induces inspiration, emotional strength, and commitment while employees' performance is defined by ability and diligence (Avolio & Bass, 2004). Our study strengthens the view that the school's principal's leadership style can be a substantial influential factor in teachers' self-efficacy (Damanik & Aldridge, 2017;Gkolia et al., 2018). We posited that the content of feedback, whether negative or positive (feedback nature) will impact an individual's self-efficacy. Teachers who stated that they received positive feedback experienced higher self-efficacy. Conversely, negative feedback decreases intrinsic motivation (Fong et al., 2019). Kim (2001) suggests that a strong personality and interactional ability, such as openness and positivity, are sometimes required for effective CCA. The ability to linguistically code-switch is advantageous to local supervisors who communicate with foreign teachers. This highlights the value of CCA strategies for enhancing intrinsic motivation. Praise as a type of positive feedback increases intrinsic motivation by enabling an internal locus of causality and perceived competence (Henderlong & Lepper, 2002). Also, positive feedback is known to have a motivational and positive affective effect that produces creativity than negative feedback (Zhou, 1998). These are effects that leadership can affect. Positive feedback through its potentials could, therefore, influence CCA.
The study also established that Chinese supervisors who are transactional leaders impact feedback's frequency. The inference is that supervisors will be motivated to give frequent feedback to highlight reward and punishment for performance. Intuitively, teachers who work under transactional Chinese supervisors are predicted to be efficacious because of feedback's corrective effect (Hattie & Timperley, 2007). Frequent feedback could also be used for emphasis, especially when the supervisor is less confident in communicating in a foreign language. Although some studies on feedback frequency do not positively link feedback frequency with self-efficacy (Casas-Arce et al., 2017; C. F. Lam et al., 2011), our study establishes the contrary. In this study, frequent feedback could be weekly or monthly. As it has been found that frequent feedback positively affects performance and that performance gradually dwindles, it is plausible to argue that the initial effect of feedback frequency will be for purposes of CCA, but which effect then begins to dwindle thereafter.
The study establishes that feedback mediates the link between transformational leader's attributes and self-efficacy. This relationship infers that the combined effect of feedback's nature and feedback's frequency significantly enhances foreign teacher's self-efficacy under transformational leadership. We thus confirm that transformational leadership traits positively affect performance feedback (Lee et al., 2019) which may signify cross-cultural intelligence and communication.
Feedback's perceived value could be recognized as a personal attribute relevant for determining their self-efficacy. We found a moderation effect of the perceived value of feedback in the relationship between feedback's nature and selfefficacy, which supports the studies of Gumah (2020a, 2020b). It illustrates that depending on the nature of feedback, perceived value has a strong effect on the link between the nature of feedback and teacher's self-efficacy.

Conclusion
We set out to show how leadership style exercised through the medium of feedback impacts self-efficacy. This, we accomplished using the CCAT. A moderation effect was introduced as our innovation to determine how the value of perceived feedback affects the supervisor-employee relationship. To increase the uniqueness of our study, we used local supervisors who exercise supervisory roles over foreign employees in contrast to several studies that mainly focus on expatriate managers. The main findings are that supervisors' leadership style (transactional and transformational) positively affect foreign teachers' self-efficacy. Supervisors can kindle intrinsic motivation through their respective attributes and influence teachers' self-efficacy. Also, the nature of feedback is influenced by the supervisors' leadership style, which influences the foreign teachers' self-efficacy. This implies that the appropriate use of CCA strategies such as communication and cultural intelligence can enhance self-efficacy. In addition, the study reveals an important mediation role of feedback's nature on the relationship between transformational leadership and self-efficacy. A critical practical contribution that provides direction to managers and supervisors is the effect of the interaction between feedback frequency and a teacher's perception of the value of feedback on self-efficacy. The implication is that managers must be sufficiently equipped in providing feedback to make subordinates value it for it to be relevant for intrinsic motivation and consequently influence CCA.
The study makes a modest contribution to the theory of feedback by showing that the nature of feedback depends mostly on the communication ability of a supervisor; therefore, we add that cross-cultural communication plays a significant part in the leadership style and self-efficacy relationship. Our study as well demonstrates how selfefficacy could also be a significant determinant of CCA, an area worth considering for further studies. Besides, it will be worth studying the negative moderation effect of feedback's perceived value on the relationship between feedback frequency and self-efficacy.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.