Blackbox between job crafting and job embeddedness of immigrant hotel employees: a serial mediation model

This study explores the influence of job crafting (JC) on migrant employees’ and its sensitivities on job embeddedness (JE) with the application of a serial mediation mechanism which takes into consideration the psychological capital and their work engagement as mediators. The data for this study was collected from immigrant employees in the hotel sector in North Cyprus. Both convenience and judgmental sampling techniques made up of 572 dyads were used. The findings of this study reveal that immigrant employees’ psychological capital and their work engagement has a mediating effect on the relationships between organizational crafting and the perceptions of employees’ job embeddedness. This study will serve as a substantial research evidence and tool for managers/owners in the hospitality industry for the purpose of retaining, engaging and embedding immigrant employees. The novelty of this study is the fact that it analyzes the work environment and conditions where organizational job crafting affects job embeddedness to determine the perceptions of immigrant employees using a serial mediation model. ARTICLE HISTORY Received 18 August 2018 Accepted 10 September 2019


Introduction
Recently, economic and political crises, information technology development, changing customers' needs, expectations and intensive competition have forced hospitality organizations' leaders to reexamine their strategic resources and their roles and functions in order to achieve organizational goals and objectives. As a labor intensive industry, hospitality's heavy reliance on human resources means that it must continually strive to remain competitive in a rapidly, changing environment (Sharpley & Forster, 2003;Zopiatis & Constanti, 2007). Due to the seasonality and other industry characteristics of hospitality, such as, low salaries and wages, anti-social job schedules, high stress, lengthy working hours drive local employees to other occupations (Arasli & Arici, 2019); therefore, local citizen's unwillingness to serve or undervalue the service jobs promote intervention of immigrant employees who would serve in majority to the survival or growth of the hospitality industry in tourism oriented countries (Janta, Ladkin, Brown, & Lugosi, 2011;Joppe, 2012). Hotel employees figure significantly within the total migrant worker community of European countries. For example, the rate is 8.6% in the UK,8.4% in Germany,14.7% in Spain,8.1% in Italy,and 4.8% in Turkey (Baum, 2012). In addition, the European Union's27 members, have 9.5 million people of which immigrant employees account for 6.2 percent of the EU's total population (Koehler, Laczko, Aghazarm, & Schad, 2010). The cost of turnover is high when subject to retention, selection, orientation, training, and replacement of new employees, which costs about half or a little over half percent of a full-time staff member's annual salary in the industry (Mitchell, Holtom, Lee, Sablynski, & Erez, 2001). Worse, these experienced employees may attract customers to potential competitors' businesses. Therefore, one of the most important problems of a hospitality setting is managing diverse workforces where migrant workers have been growing fast because of the aforementioned reasons in parallel to the fastdeveloping hospitality industry. Immigrant employees move to another country due to lack of employment and freedom, war, terrorism, and religious pressure in their native country. There are also positive attitudes behind their influx such as, finding a new job, a different culture, getting a better education, language skills, career opportunities and so forth. Like their local full-time counterparts, some immigrant employees were appointed by their employers not only because they are willing to work for low wages (Dench, Hurstfield, Hill, & Ackroyd, 2006;Wills et al., 2009), but also because they are perceived as having different attitudes, and other talents (Anderson, 2007). The vital factor was due to their boundary spanning roles, face-to-face and voice-to-voice relationships with guest who play a pivotal role in services delivery and products, as well as, service complaints and recovery to guests (Hartline & Ferrell, 1996;Gil, Berenguer, & Cervera, 2008;Arasli, Bahman Teimouri, Kiliç, & Aghaei, 2017).
Numerous employees working in the hotel organization can be migrant employees (de Castro, Fujishiro, Sweitzer, & Oliva, 2006), and thus the organizational success and development of the hotels are counting on keeping immigrant employees (Rozkwitalska & Basinska, 2015). Despite this, hospitality studies have paid insufficient attention to the immigrant employees' unique theoretical and managerial challenges (Chou, 2018). One of the biggest challenges that these employees face is the effective link and fit with the organization and the job, which have been acknowledged as two important dimensions of job embeddedness (Lang, Kern, & Zapf, 2016). Although the potential significance of immigrant employees remains clear, several scholars suggested that hotel organizations cannot keep these employees due to the turnover rate (Arasli & Arici, 2019;Tracey & Hinkin, 2008). The hospitality industry continuing quest to find ways to retain these employees remains a major managerial challenge (Karatepe, 2013). One factor for immigrant worker retention involves ensuring that workers feel linked to their work. One measure of this link is job embeddedness, which is a relatively growing concept in the current hospitality literature (Arasli et al., 2017;Karatepe & Ngeche, 2012). Job embeddedness consists of three important components, fit, link and sacrifice (Mitchell et al., 2001). Therefore, job embeddedness emerges as one of the main elements for promoting the retention of immigrant employees.
In fact, empirical literature investigating job embeddedness in the hospitality industry has been decidedly equivocal (Arasli et al., 2017;Karatepe, 2013), although job embeddedness is an important anti-withdrawal factor in HRM, and scholars have much to conduct exploration of job embeddedness (Wheeler, Harris, & Harvey, 2010). Furthermore, according to Mitchell et al. (2001), JE theory explained how organizational fit, links and sacrifice and community fit, links and sacrifice (off-the-job embeddedness) factors influence the retention of employees. For this study, we have directed our attention to JE. The reasons to examine JE are manifold. First, studies (Purba, Oostrom, Born, & van der Molen, 2016) have demonstrated that JE can reduce turnover. Second, trust in supervisors (TIS) (Goris, Vaught, & Pettit, 2003) and high performance work practices (HPWPs) (Combs, Liu, Hall, & Ketchen, 2006) are on-the-job aspects, and they are more likely related with JE within the organization. In addition, several scholars suggested that job embeddedness may be increased by such variables as: perceived supervisor support (Arasli & Arici, 2019), high-performance work practices, and work social support (Karatepe, 2013) However, job crafting as an organizational resource in the hospitality industry has yet to be investigated thoroughly. Employees have suggested that job crafting can improve the fit between their personal needs, abilities, and passions about the job through job crafting (Tims & Bakker, 2010).
Given this inadequate concern in the hospitality literature on JE of immigrant employees, this present study has begun to ask the job crafting-JE model so as to consider how (through what mediator variables) job crafting may lead to higher or lower JE. This shift in research focus has resulted in several significant insights. For example, scholars now know that relational and cognitive crafting results in a higher level of employees' fit when compared with the organization, while task crafting does not (Kim & Beehr, 2018) and job crafting increases flight attendants' work engagement and service recovery performance, decreases their intention to leave (Karatepe & Eslamlou, 2017).
Nevertheless, academic knowledge of the processes by which three dimensions of job crafting (i.e., relational, cognitive, and task crafting) promote or hinder immigrant employees' JE in the hospitality industry is far from complete, and several basic pieces of information are missing. For instance, higher level of immigrant employees JE is vital to retain these employees in the industry. While high turnover is one of the most important problems, these employees' embeddedness cannot be ensured without mediator effect as suggested by Whetten (1989). Whetten states that it is necessary for researchers to explain causal relationships as a phenomenon by determining mediators between antecedent and consequence variables. In order to better comprehend how job crafting can result in JE, we opened the black box by investigating the mediator effects of Psychological Capital (PsyCap) and work engagement (WE). In this study, we suggest that through providing PsyCap and WE of immigrant employees, job crafting may ensure these employees' job embeddedness.
Therefore, the main purpose of this study is to investigate the effect of job crafting (JC) on immigrant employees' perceptions of on the job embeddedness using a serial mediation mechanism which involved the employees' psychological capital and their work engagement as mediators. Furthermore, although several scholars suggest the importance of the mediators between these two variables (Whetten, 1989), an underexplored area among immigrant employees remain the precursors that poses a challenge which needs to be addressed. Thus, in this study, three (3) fundamental indicators of job crafting, PsyCap, and work engagement on JE were investigated. This also served as a key research question to be empirically answered.
Finally, prior research on the effects of job crafting on the job embeddedness has been widely conducted in Western countries (Elanain, 2009). To comprehend why employees, choose to quit or remain in an organization, most of these researches have recommended several models to be tested in non-Western contexts. Additionally, this research extends the knowledge of the factors triggering the relationship between job crafting's three dimensions and immigrant employees' job embeddedness by conducting this study in the small island hotel setting for examining the relationships mentioned above.

Theoretical background
Job-Demand Resource (JD-R) theory and Conservation of Resources (COR) theory have underpinned the hypothesized relationships which link job and personal resources to employee job embeddedness through a serial mediation effect. According to JD-R theory, job-related features could be classified as job demands and resources in an organization . Job demands are defined as "those physical, social, or organizational aspects of the job that require sustained physical or mental effort and are therefore associated with certain physiological and psychological costs … ". Job resources are defined as "those physical, psychological, social or organizational aspects of the job that may do any of the following: a. be functional in achieving work goals; b. reduce job demands at the associated physiological and psychological costs; c. stimulate personal growth and development" (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001, p. 501).
Drawing on the JD-R theory, previous research revealed that job crafting results in enhanced job resources (Tims, Bakker, & Derks, 2013). Environmental stressors such as workload, emotional demands, and time pressure can be deemed as job demands, whereas organizational resources, such as job control, possibilities development, and social support could be listed as job resources (Demerouti et al., 2001). Job demands and resources are negatively associated, because job demands like workload, and emotionally demanding interactions with guests can hinder the mobilization of job resources (Bakker, Demerouti, De Boer, & Schaufeli, 2003;Bakker, Demerouti, & Euwema, 2005;Demerouti et al., 2001). Similarly, high job resources such as coworker support can mitigate job demands. Several scholars found that job demands result in negative employee outcomes, such as burnout and emotional exhaustion, whereas job resources are considered as the predictor of positive employee outcomes, such as work engagement (Bakker & Demerouti, 2017;Cheng & Yi, 2018). For example, Schaufeli and Bakker (2004) found that job demands positively affect burnout which leads to employee turnover intention. Hakanen, Schaufeli, and Ahola (2008) also found that job resources result in employee engagement which in turn leads to organizational commitment.
Recent approaches have described job crafting by drawing on JD-R theory , where followers could craft their jobs by using job resources and by defying job demands. Job demands and resources together could engender followers' advancement and development and encourage them to obtain further achievements (Karatepe & Eslamlou, 2017). Through utilizing job resources and challenging job demands, employees can display proactive behaviours and so they are able to start making changes to the borders of their jobs (Bakker and Demerouti, 2017). To illustrate, a follower could enhance social resources by asking a co-worker for a recommendation, enhance structural resources through endeavouring to get new job-related knowledge, increase challenging demands by undertaking extra jobs and responsibilities, and mitigate hindering demands by avoiding any interactions with undesirable guests . Based on the discussion, past research contended that individual employees who craft their jobs are more motivated to show better job performance (Karatepe & Eslamlou, 2017). Such employees are not only more engaged in their tasks, but the favourable experience of enhancing job resources and managing challenging demands might also increase their personal resources (Vogt, Hakanen, Brauchli, Jenny, & Bauer, 2016). Drawing on the JD-R theory, we conclude that job crafting in the form of enhancing job resources and challenging job demands may result in positive employee work outcomes (Tims et al., 2013). Current findings have provided empirical support for the supposition (Cheng & Yi, 2018;Karatepe & Eslamlou, 2017;Kim, Im, Qu, & NamKoong, 2018).
On the other hand, COR theory suggests that people try to acquire, maintain and foster valuable resources such as health, happiness, peace, and self-protection. Such resources are needed when dealing with physical and mental pressures and seeking to generate resource processions (Hobfoll, 1989;Karatepe & Eslamlou, 2017). The COR theory also suggests that people have to invest in resources so as to preserve against losing resources, being recused from losing resources and getting resources (Hobfoll, 1989). Kohn and Schooler (1982) claimed that job characteristics in a work-setting could influence personal resources since people generally use their experiences from the work in off-work circumstances. One personal resource of broad interest is PsyCap (Vogt et al., 2016). According to this theory, such resources are crucial since they create other valuable resources and underpin the development of resource processions which might result unfavourable job-related outcomes (Xanthopoulou, Bakker, Demerouti & Schaufeli, 2007). This is plausible, since work engagement originated from the aggregation of personal resources (Halbesleben & Wheeler, 2008;Karatepe & Eslamlou, 2017). Based on this theoretical framework, we conclude that PsyCap as a personal resource may lead to more engaged employees in the workplace which leads to higher levels of employee job embeddedness.
Drawing on the theoretical frameworks, the researchers believe that job crafting engenders hotel immigrant employees' PsyCap and job embeddedness. The researchers also argue that PsyCap results in highly engaged immigrant employees in the hotel industry, and highly engaged employees are more likely to embed in their jobs. As can be observed in the research model (Figure 1), PsyCap and work engagement are intervening variables that link job crafting to employee job embeddedness via a serial mediation effect. The hypothesized relationships are elaborated on in the following.

Job crafting and job embeddedness
Wrzesniewski and Dutton (2001) described job crafting as the duration in which employees engaged actively to alter the borders of their jobs. This can be due to a change by employees as a reaction to an alteration in job objectives (Griffin, Neal, & Parker, 2007). The three classifications of job crafting as described by Wrzesniewski and Dutton (2001) are task crafting, cognitive crafting, and relational crafting. Furthermore, Bardi (2011) described task as a peculiar activity or work. For example, the task of receptionist or front desk personnel in a hotel is specifically to deal with guests, receives payment from guests as well as guest check-in/check-out. Studies have shown that task crafting involves altering the successive, physical cases or behaviours in an organization which includes giving up or keeping tasks, restructuring tasks and regulating the impacts of time spent on the tasks (Wrzesniewski & Dutton, 2001).
Drawing on the JD-R theory, this present research claims that JE may be one of the possible consequences of job crafting. Job crafting gives another meaning to work because the shaping of cognitive tasks and relationships in the organization would allow employees to conduct their work differently resulting in reshaping the goal of their jobs (Wrzesniewski & Dutton, 2001). When immigrant hotel employees have the opportunity to craft their jobs by using job resources and challenging job demands, they can improve the fit between their personal needs, abilities, and passions about the job (Tims & Bakker, 2010). Hence, when crafting their job, such employees may demand more responsibilities from their supervisors, may sacrifice their priorities for completing jobs, may develop close links with colleagues by requesting their advice and aid, and may accept job demands when they sense these demands as a way for advancement and improvement Vogt et al., 2016), and may enhance their perceptions of fit, links, and sacrifice. That is, JE can be a potential consequence of enhancing job resources and challenging demands through job crafting. In empirical terms, several studies have to date investigated the effects of job crafting on several employee outcomes, such as work engagement Karatepe & Eslamlou, 2017;Tims et al., 2013;Vogt et al., 2016), job satisfaction , and burnout (Cheng & Yi, 2018;Demerouti et al., 2001) and job performance (Tims, Bakker, & Derks, 2015). Furthermore, drawing on the JD-R theory, several scholars have revealed the significant relationship between the amount of job resources and the level of work engagement (Xanthopoulou et al., 2007;Hakanen, Bakker, & Demerouti, 2005;Mauno, Kinnunen, & Ruokolainen, 2007). At this point, Karatepe and Eslamlou (2017) in their empirical study on job crafting, called for more investigation on the relationship between job crafting and JE which consists of links, fit, and sacrifice, through data collected from hotel employees. However, to date, no study has attempted to empirically examine the relationship between job crafting and JE in a hospitality setting. Considering this suggestion, the researchers aim to first test the relationship in order to expand on existing knowledge and understanding of job crafting and its consequences. Therefore, the following hypothesis is proposed: H1a: Immigrant hotel employees' task crafting is positively related to their JE.
H1b: Immigrant hotel employees' relational crafting is positively related to their JE.
H1c: Immigrant hotel employees' cognitive crafting is positively related to their JE.

Job crafting and psychological Capital
Drawing on the JD-R theory, previous research evidence suggests that job to off-job situations and experiences are mostly generalized by individuals, thus affecting their personal resources (Kohn & Schooler, 1982). This implies that employees may not only influence their work engagement as described earlier when crafting their jobs, but may also boost their personal resource, increase positive experience, job resources, and challenging demands when crafting their own jobs (Avey, Reichard, Luthans, & Mhatre, 2011;Luthans, Avolio, Avey, &Norman, 2007). PsyCap has been referred to as the positive psychological state of development of an individual characterized by several factors including: self-efficacy, optimism, and hope as well as resilience . PsyCap has been associated with employee attitudes, such as satisfaction and organizational commitment (Luthans, Norman, Avolio, & Avey, 2008), and with several measures of performance Avey et al., 2011). For example, an employee will feel highly efficacious and less hesitant about future proactive behaviour if he has successfully broken out of his or her daily routine and has expanded aspects of his or her current job. This implies that employees exert control over their environment by crafting their job, an experience which is central to building personal resources such as self-efficacy (Bandura & Wessels, 1997) and optimism (Seligman, 2006). Also, the ability to set goals and working hard to achieve them are critical steps in the job crafting process, and the development of personal resource of hope (Snyder, 2000). This notion has been substantiated by research evidence, in different work scenarios and has been shown to improve coping with difficult work situations, developing increased responsibility and performance, as well as building future resilience (Masten, 2001). In line with this notion, this present study hypothesizes that the activity of crafting job resources and challenging demands may boost the personal resources of hope, efficacy, resilience, and optimism. Hence, the following hypothesis is posited: H2a: Immigrant hotel employees' task crafting is positively related to their PsyCap.
H2b: Immigrant hotel employees' relational crafting is positively related to their PsyCap.
H2c: Immigrant hotel employees' cognitive crafting is positively related to their PsyCap.

Psychological capital and work engagement
The competence of employees' positivity is acknowledged by PsyCap. It accounts for one's psychological development positively in terms of resilience, self-efficacy, hope, and optimism (Vogt et al., 2016). It is asserted that individuals holding high-efficacy are prone to interrogate themselves as they have their own aims and investigate tough tasks. In addition, the ones being hopeful come up with realistic issues, however their goals are difficult, and their expectations include self-control, self-determination, energy, and individual perceptions. For the optimists, they are open to change and utilize the chances or focus on the opportunities for the future, and also employees having high resiliency attempts to keep positive attributions and go back while struggling with problems .
PsyCap has got a great deal of attention from scholars working on hospitality due to the positive supporting role it has on behaviours. According to Jung & Yoon (2015), workers with a substantial rate of PsyCap reported more job satisfaction when compared to their co-workers. Similarly, studies have affirmed the role of PsyCap on Romanian frontline employees' jobs in terms of life satisfaction and career (Karatepe & Karadas, 2015), and also in Korea (Paek, Schuckert, Kim, & Lee, 2015) by analyzing how PsyCap affected job engagement and the morale of the employees. However, Paek et al., (2015) emphasized that the process of developing employees' engagement remains blurred, hence the need for more studies on the relationship between work engagement and PsyCap. Considering the gap in existing understanding, this present research proposes that immigrant hotel employees who have more PsyCap are more engaged in their job.
For the relationship of PsyCap and work engagement, the COR theory ensures a theoretical framework. Abundant personal resources in a work setting could result in improvement of work engagement (Halbesleben & Wheeler, 2008). Several scholars suggested that the work-setting where there are abundant resources may engender followers' favourable emotions which are associated with personal resources. Development of favourable emotions and feelings in the organization may lead to higher levels of employee motivation (Walumbwa, Peterson, Avolio, & Hartnell, 2010).Individuals giving high points to PsyCap would be perceived to be more energetic and dedicated and more deeply involved in the job. Thus, the following hypothesis is formulated: H3. Immigrant hotel employees' ppsychological capital is positively related to their work engagement.

Work engagement and job embeddedness
A study by Halbesleben and Wheeler (2008) revealed that several changes including job resources and demand might change depending on certain factors surrounding job condition such as shocks and radical events, thus, making engagement unstable and flexible over time. Based on the JD-R theory, previous research suggested that engaged workers who possess favourable emotions, may generate or develop their job resources and challenging demands and transform engagement into their colleagues in the work-setting (Bakker & Demerouti, 2017). Accordingly, favourable employee outcomes like high performance and low intention to quit are listed as the consequences of work engagement (Bakker et al., 2003). Halbesleben and Wheeler (2008) reiterated that employees would likely be more engaged with their job and develop more embedded characteristics when they have ample resources in their organization. Empirical researches examining the relationship between work engagement and job embeddedness point to the fact that work engagement could indeed enhance job embeddedness. For example, Halbesleben and Wheeler (2008) investigated US employees' job outcomes and revealed work engagement to be significantly associated with job embeddedness. Congruent with this, Karatepe and Ngeche (2012) found that work engagement results in increased job embeddedness. On the basis of the theoretical and empirical knowledge given above, we formulate the following hypothesis: H4. Immigrant hotel employees' work engagement has a significant positive effect on their job embeddedness.

Serial mediation
The JD-R theory suggests that crafting job resources and challenging job demands result in employee work engagement which may induce positive employee job outcomes (Hakanen et al., 2008;Karatepe & Eslamlou, 2017). Hotel employees may increase their work engagement and PsyCap through job crafting attitudes (Vogt et al., 2016) and exert higher levels of job embeddedness.
Though limited, several scholars found empirical evidence regarding the mediator role of work engagement. For instance, Hakanen et al. (2008) found the mediation effect of work engagement on the relationship between job resources and job demands as well as organizational commitment. Karatepe and Eslamlou (2017) studied flight attendants job crafting and found that job crafting has a significant effect on flight attendants' service recovery performance and their intention to quit through the mediation effect of work engagement. Van Wingerden, Derks, and Bakker (2017) also showed the mediator role of work engagement on the effect of job crafting on job performance.
Further, congruent with the JD-R theory, there has been empirical evidence about the mediator effect of PsyCap between job resources and desirable employee outcomes. Specifically, Tims et al. (2015) verifies the positive impacts of job crafting on engagement through increasing psychological capital. These findings suggest indirect empirical justification for evaluating the association of job crafting with positive employee outcomes, such as job embeddedness. However, an empirical evidence for the serial mediation process of work engagement and PsyCap is still scarce.
To fill the abovementioned gap in recent knowledge, this research follows the guidelines produced by Whetten (1989).It is important for scholars to indicate causal associations in a phenomenon by focusing on mediators on the effect of an independent variable on a dependent variable. When one performs this line of idea, it may be possible to plan a causal relation between three dimensions of job crafting and an employee's job embeddedness. Thus, a serial mediation model suggested by Hayes (2018) can be performed. To illustrate, one may claim that job crafting may enhance employee's PsyCap. The PsyCap can also significantly influence employee's engagement at work, and the employee engaged in the job may be more embedded in his/ her job. In other words, in the hospitality work setting, where three dimensions of job crafting are commonly accepted (e.g., , hotel employees feel more PsyCap and they are more engaged in their work which in turn results in a higher level of employee job embeddedness, particularly in hotels where job crafting is important and popular (Chen, Yen, & Tsai, 2014). Thus, this serial mediation model proposes that job crafting has an indirect effect on job embeddedness via serial mediation effects of PsyCap and work engagement.
A serial mediation is of crucial significance in discovering the potential influence of the causation from three dimensions of job crafting to employee job embeddedness. This is particularly relevant since it may be denotative of the untested and underlying causal chain which is still unexplored. This will clarify if the mechanisms of mediation from the three dimensions of job crafting to employee job embeddedness are significant for immigrant hotel employees. Thus, this study posited the following hypotheses to expand our understanding of the serial mediation effect: H5a: Immigrant hotel employee's psychological capital mediates the relationship between task crafting and job embeddedness.
H5b: Immigrant hotel employee's psychological capital mediates the relationship between relational crafting and job embeddedness.
H5c: Immigrant hotel employee's psychological capital mediates the relationship between cognitive crafting and job embeddedness.
H6a: Immigrant hotel employee's work engagement mediates the relationship between task crafting and job embeddedness.
H6b: Immigrant hotel employee's work engagement mediates the relationship between relational crafting and job embeddedness.
H6c: Immigrant hotel employee's work engagement mediates the relationship between cognitive crafting and job embeddedness H7a: Immigrant hotel employee's task crafting is positively related to their job embeddedness through the chain of the immigrant hotel employee's psychological capital and work engagement.
H7b: Immigrant hotel employee's relational crafting is positively related to their job embeddedness through the chain of the immigrant hotel employee's psychological capital and work engagement.
H7c: Immigrant hotel employee's cognitive crafting is positively related to their job embeddedness through the chain of the immigrant hotel employee's psychological capital and work engagement.
The research model, including all hypotheses, is demonstrated in Figure 1.

Research context
The North Cyprus economy is dependent on tourism and education (Roudi, Arasli, & Akadiri, 2018). By 2019, in North Cyprus there are 5 four-star and 22 five-star hotels. The hotel's casinos alone provide $600 million annually to the national economy. This means 20-25% of the North Cyprus budget.
By 2018, 56,653 immigrant workers from 126 countries have been working in the private sector of which 11.5 percent are employed in the hotel industry in North Cyprus (Department of Labor, 2018). In addition, hotels, restaurants, travel agencies, such as the inclusion of side sectors, total number of employees are 20 thousand which makes 32 percent of workforce in the hotel industry. (Department of Labor, 2018). A fluctuated economic situation in North Cyprus in the last few decades has resulted in labor migration. Such migration has generally been supported by the government of North Cyprus to ease the pressure on labor markets, to be able to serve better foreign customer needs and expectations, reduce employment costs and accelerate development especially in the tourism industry.

Sample and procedure
Data were collected from a sample of full-time immigrant employees (e.g. receptionists, chefs, sale-people, waiters, housekeepers, security and others) in four-and fivestar hotels in North Cyprus between August 2018 and March 2019. The statistics show that four and five star hotels are mostly chain hotels attracting the majority of international tourist and investments in the accommodation sector in North Cyprus (Ministry of Tourism Statistics, 2018). These hotels also provide job opportunities to immigrant employees from Turkey, Russia, Ukraine, Iran, Nigeria, Cameroon, Turkmenistan, Tajikistan and so forth. The statistics show that approximately 5500 immigrant employees were employed in the hotel industry in North Cyprus at the time of the research (Ministry of Tourism Statistics, 2018). There are important reasons for choosing this category of employees for this research.
First of all, immigrant employees' mind-set plays a crucial role in delivering service, coping with the complaint processes or filling logistic needs and expectations of hotel guests like their other counterparts. Secondly, immigrant employees (IEs) represent the hotel to customers and other social shareholders and increase the reputation of the accommodation organization (Bettencourt & Brown, 2003).Thirdly, IEs in accommodation, restaurants, casino and bars have close interactions with customers and need to have a civil relationship with guest behaviours as frequently as other employees (Daunt & Harris, 2011).However, research about these employees is sparse.
Based on the information received from the Ministry of Tourism Statistics (2018), just before the fieldwork, there were 22 five-star and 5 four-star hotels in North Cyprus. 5 hotels (100 percent) from four-star and 13 of them (60 percent) from fivestar hotels have been chosen based on judgmental sampling. These employees were deemed representative since they were chosen from the most immigrant employee populated chain hotels according to the statistics gathered from the tourism ministry on North Cyprus. The human resource (HR) department of each hotel was contacted using a formal letter which included information about the aim of the study and requested permission to collect data. The management of all four-star hotels and 13 five-star hotels agreed to help with this research. Since top management of the hotels did not allow the researcher communicate directly with the immigrant employees, the questionnaires were distributed by one of the authors who was accompanied by an assigned employee from the HR department. An assurance of confidentiality was provided to the participants on the first page of the questionnaires. In addition, after filling out the self-administered questionnaires, the employees placed the questionnaires into envelopes and placed them into the cartoon box located in the front office of the hotels.
The survey instruments were distributed to 615 immigrant hotel employees. After excluding 16 questionnaires that had reckless answers (mainly the same responses, like 1 for all items) and 27 instruments with incomplete responses, the sample consisted of surveys from 572 frontline employees with a response rate of 93.01 percent.
To form a valid and practical questionnaire to investigate the above-mentioned hypothesized relationships, the authors developed the questionnaire in English and then utilized two independent professional bilingual translators to translate it into Turkish. A back-translation was then completed by another academician, fluent in both languages as recommended by McGorry, (2000) to make sure that all item contents were cross-linguistically comparable and created in the identical context. The surveys were pretested with a pilot sample of 20 frontline employees to ensure the clarity of each question. The pilot study indicated that the wording, measurement scales, and sequence of questions were fine. Confidentiality and anonymity were ensured, and these were communicated to the participants before they agreed to participate in the study.
Harman's one-factor analysis as a statistical remedy was also conducted to mitigate the common method bias, since all data concerning the proposed model were selfreported. Of the six factors identified, the principal factor explained 21.28 per cent of the variance. Because one factor did not explain more than 50 per cent of the variance, common method bias has not been a problem in the dataset (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003).
The participant employees' profiles were as followed, 296 (75.9%) of the respondents were between 18-35 years, and this age group represented the majority of the employees. As far as gender was concerned, 225 (39.3%) of the employees were female and 347 (60.7%) were male. In regards to educational achievement, 56 (9.8%) had a primary school diploma; 220 (38.5%) had secondary and high school diploma; and 240 (42%) had an associate degree from a vocational school and this represented the majority of respondents; the rest had a bachelor degree (9.8%). In regard to organizational tenure, those respondents who worked in the organizations for less than six months were 96 (16.8%). Meanwhile, those who worked from six months to 1 year were 93 (16.3%), 1 to 3 years were 171 (29.9%), this represented the majority of the respondents, and those between 3 to 5 years were 118 (20.6%), 5-10 years were 65 (11.4%), and finally those who worked more than 10 years were 29 (5.1%). In terms of departments, 16.6% of the respondents worked in the front office; 39.5% worked in the food & beverage department that is the restaurant and kitchen; 6.5% worked in housekeeping; and the rest worked in other departments, such as sales, accounting, security, etc (Table 1).

Measures
Nine items for job crafting were drawn from Slemp and Vella-Brodrick's (2013) study. Job crafting consisted of three (3) categories, namely, task crafting, cognitive crafting, and relational crafting. The researchers measured responses on the five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). The illustrative items were "change the scope or types of tasks that I complete at work," and "make an effort to get to know people well at work." The coefficient alpha for the task crafting scale was 0.80, for the relational crafting scale it was 0.82, and for the cognitive crafting scale, it was 0.88.
PsyCap was measured using the five-point Likert scale (1 ¼ strongly disagree to 5 ¼ strongly agree) composted by Luthans, Youssef, and Avolio (2007) consisting of 24 items with four dimensions, efficacy (e.g., I feel confident contributing to discussions about the company's strategy), hope (e.g., If I found myself in a jam at work, I could think of many ways to get out of it), resilience (e.g., I usually manage difficulties one way or another at work), and optimism (e.g., I always look on the bright side of things regarding my job), each including six items. The reliability coefficient for this measurement scale was 0.84.
A nine-item scale adapted from the shortened version of the Utrecht Work Engagement Scale (Schaufeli, Bakker, & Salanova, 2006) measured work engagement. Hotel employees were asked their opinion of each statement utilizing the seven-point rating scale. Sample items were "At my work, I feel bursting with energy", and "I am enthusiastic about my job." The alpha coefficient demonstrated high internal consistency (a ¼ 0.92).
Job embeddedness was measured by utilizing the 7-item scale developed by Crossley, Bennett, Jex, and Burnfield (2007). The participants were asked to express how much they agreed or disagreed with seven statements using the five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Sample items were "I feel attached to this hotel", and "I feel tied to this hotel." The alpha coefficient for this scale was 0.81.

Data analysis
In order to detect the convergent and discriminant validities of the factors, confirmatory factor analysis (CFA) was used by following the recommendations of Anderson and Gerbing (1988). Pearson correlation analysis was also conducted to identify the relationships among the independent variable, the mediating variables, and the dependent variable. The study hypotheses were tested through performing a serial mediation analysis with the help of an SPSS macro developed by Hayes & Scharkow (2013) by testing model (6) of the process by introducing three dimensions of job crafting (i.e., task crafting, relational crafting, and cognitive crafting) as predictor variables, PsyCap and WE as mediator variables, and JE as the outcome variable. Mediation was also tested with the bootstrapping method recommended by Hayes & Preacher (2014). This involved a 95 percent confidence interval (CI) with 1,000 bootstrap samples. Hayes and Preacher presented SAS and SPSS syntax for an alternative "bootstrap" test of the indirect effect that is almost always more powerful than Sobel's test used in Baron and Kenny's (1986) mediation analysis (Zhao, Lynch, & Chen, 2010). Therefore, Hayes's approach is more powerful by using bootstrapping in mediation analysis because the Sobel test is low in power compared to a bootstrap test popularized by Preacher and Hayes (2004).

Measurement model
CFA demonstrated that factor loadings were significant, ranging from 0.61 to 0.91 (p < .05). In addition, all constructs showed the acceptable composite construct reliabilities (CCR) ranging between 0.81 and 0.94 (see Table 2). The average variance extracted (AVE) scores of constructs were between 0.58 and 0.74, demonstrating sufficient convergent validity. Analyses demonstrated that the hypothesized model yielded an acceptable fit to the data (v 2 ¼ 2524.03; df ¼ 975; p < .01; comparative fit index (CFI) ¼ .94; goodness-of-fit index (GFI) ¼ .84; Tucker-Lewis index (TLI) ¼ .95; root-mean-square error of approximation (RMSEA) ¼ .058; and standardized root-mean-square residual (SRMR) ¼ .044), which recommended that these variables should be conceived as distinct constructs. Overall, the ratio of the AVE in each construct was greater than the square of the correlation coefficient between variables, ensuring discriminant validity (Fornell & Larcker, 1981). Table 3 depicts the mean scores, standard deviations, and correlations for the study constructs. The correlation table showed that there was a statistically significant and positive correlation between task crafting and PsyCap (r ¼ .287, p < .01), relational crafting and PsyCap (r ¼ .240, p < .01), and cognitive crafting and PsyCap (r ¼ .269, p < .01). The correlation between PsyCap and WE was significant and positive (r ¼ .401, p < .01). There was also a significant positive correlation between WE and JE (r ¼ .297, p < .01). These findings provided preliminary support for the hypothesized relationships.

Hypothesis testing
The first three hypotheses of this study proposed that three dimensions of job crafting (i.e., task crafting, relational crafting, and cognitive crafting) have a significant positive effect on employees' job embeddedness JE. To examine the hypotheses, a hierarchical regression analysis was performed. The results showed that there was a significant and positive relationship between task crafting and JE (b ¼ .30, t ¼ 7.65, p < .001, Table 4, Model 2); between relational crafting and JE (b ¼ .22, t ¼ 5.30, All loading values are significant at the 0.05 level. p < .001, Table 4, Model 3), and between cognitive crafting and JE (b ¼ .25, t ¼ 6.14, p < .001, Table 4, Model 4), supporting H1a, H1b, and H1c respectively. Further, this study investigated the effects of three dimensions of job crafting on hotel employees' JE through the mediating effects of PsyCap and WE. The sequence of our study model demonstrated the favour of serial mediation. Therefore, a serial mediation analysis was performed via SPSS macro developed by Hayes and Scharkow (2013). The regression results indicated that the effect of task crafting on PsyCap was positive and significant (b ¼ .19, t ¼ 7.14, p < .001, Table 5, Model 1). This finding supported Hypothesis 2a. The results also demonstrated that relational crafting has a significant and positive effect on PsyCap (b ¼ .15, t ¼ 5.91, p < .001, Table 6, Model 1). Thus, Hypothesis 2 b was supported. In addition, cognitive crafting significantly and positively influences PsyCap (b ¼ .17, t ¼ 6.67, p < .001, Table 7, Model 1), supporting Hypothesis 2c.
The analysis that examined the third hypothesis demonstrated that these two variables are positively associated, which means that the hotel immigrant employees, who possessed a higher level of PsyCap are more likely to engage in their work (b ¼ .40, t ¼ 10.44, p < .001, Table 4, Model 1). Thus, Hypothesis 3 was supported.
The result showed that they are significantly and positively associated, which means that the hotel immigrant employee, who possessed a higher level of WE tended to show a higher level of JE (b ¼ .30, t ¼ 7.41, p < .001, Table 4, Model 5). Thus, Hypothesis 4 was supported.  The hypotheses that involved the mediating effect were estimated using the serial mediation model, which is the method suggested by Hayes and Scharkow (2013). The result from the serial mediation analysis endorsed the mediating role of employee's PsyCap. That is, the finding displayed that the indirect effect of task crafting through PsyCap (b ¼ .217, Table 5) was significant as the lower and upper bounds of the 95 percent CI did not contain 0 [lower bound CI ¼ .087; upper bound CI ¼ .511]. Therefore, Hypothesis 5a was supported.
Furthermore, PsyCap has a mediator role in the relationship between relational crafting and JE (b ¼ .316, Table 6). The lower and upper bounds of the 95 percent CI did not comprise 0 [lower bound CI ¼ .113; upper bound CI ¼ 1.213]. Thus, Hypothesis 5 b was supported.
The effect of cognitive crafting on JE was also mediated by PsyCap (b ¼ .273, Table 7). The lower and upper bounds of the 95 percent CI did not include 0 [lower bound CI ¼ .102; upper bound CI ¼ .684], supporting Hypothesis 5c.
The findings also showed that WE mediated the indirect effects of three dimensions of job crafting on JE (for task crafting b ¼ .165,

Discussion
Based on the JD-R theory and COR theory, the purpose of this study was to investigate the effect of job crafting on immigrant hotel employees' job embeddedness through a serial mediation mechanism that employed the employees' PsyCap and their work engagement as mediators. PsyCap and work engagement, both crystallize the relationship between job crafting and immigrant employees' job embeddedness. In other words, job crafting as an organizational and PsyCap as a personal resource, affect employees' work engagement. Thus, this study advances knowledge by providing substantial empirical evidence regarding the relationships between the considered constructs with the data gathered from the immigrant employees in a small island setting, North Cyprus.
In line with the suggestion of JD-R theory (Bakker & Demerouti, 2017), our findings revealed task crafting, relational crafting, and cognitive crafting, as the three most important job design tools (Simons & Buitendach, 2013) that trigger immigrant hotel employees' job embeddedness and PsyCap. Karatepe and Eslamlou (2017) recommended that future researches should examine JE which consists of links, fit, and sacrifice as a potential outcome of job crafting in the hospitality work environment. This research responds to this call through demonstrating that job crafting positively and significantly affect JE in the hotel industry, while also elaborating on the association between job crafting and JE. In addition, congruent with the guideline of COR theory, PsyCap increases employee work engagement. Current knowledge concerning the relationship between PsyCap and work engagement is scarce (Kang & Busser, 2018;Karatepe & Karadas, 2015). This important finding expands our knowledge regarding the aforementioned relationship in the hospitality industry context. The finding recommends that immigrant hotel employees who have high PsyCap seem more energetic and engaged in their job in a hotel organization. The significant positive effect of PsyCap on WE show that personal resources engender hotel employees to be immersed in their job. In line with Xanthopoulou et al. (2007) study, our findings also show the importance of personal resources in hotel immigrant employees' work engagement. This result also backs up limited empirical investigations (Karatepe & Karadas, 2015).
The results of this present research also demonstrate that work engagement significantly affect hotel immigrant employees' job embeddedness which provide support to empirical findings of Karatepe and Ngeche (2012) and Halbesleben and Wheeler (2008). This finding recommends that engaged immigrant employees are more embedded in the hotel organization.
Our results also show that hotel immigrant employees' job crafting behaviours affect their job embeddedness via mediating roles of work engagement and PcyCap. Congruent with the JD-R theory (Bakker & Demerouti, 2017), workers may generate their resources and work engagement through job crafting. If crafting their jobs, hotel immigrant employees may gain support from their co-workers and supervisors and get empowerment and new duties (Vogt et al., 2016). This leads to employees' feeling themselves more energetic and dedicated in their work. These employees taking individual initiative for crafting their jobs are more likely to engage in their works and be more embedded in their jobs (e.g., Halbesleben & Wheeler, 2008).
Furthermore, our findings also recommend that PsyCap mediates the relationship of job crafting and job embeddedness. This finding is congruent with the COR theory (Hobfoll, 1989). Particularly, employees' job crafting behaviours through job resources engender their PsyCap. Such employees having job resources are optimistic concerning being successful at the moment and in the future  and have the potential to get rid of hardship . These hotel immigrant employees in turn demonstrate higher levels of job embeddedness.
The findings of this study also reveal a serial mediation model of PsyCap and work engagement in the relationship between job crafting and job embeddedness. In line with theories of JD-R and COR, the findings show that job crafting behaviours result in high PsyCap and work engagement respectively which in turn ultimately engender employee job embeddedness in the hotel industry.

Theoretical implications
Regarding the relationship between Job Crafting, Psychological Capital, Work Engagement and Job Embeddedness from the perspective of immigrant employees, this empirical study highlights specific implications which help hospitality researchers, managers and practitioners gather useful insights in the three, four and fivestar hotels.
Three kinds of crafting are task, interaction and cognition crafting. This is the first time they have been taken into consideration by a study from the perspective of immigrant employees on their psychological capital which was not tested in previous studies. The first empirical finding was in line with the previous researches which claim that organizational resources increase positive psychological behaviours (Luthans et al., 2008) or enhance individual's attitudes like work engagement and on the job embeddedness outcomes (Macky & Boxall, 2007).
The three crafting movements predicted and enhanced WE. These results show that when employees are supported by management in terms of task design, relationship flexibility and judgment or positive perception, the changes toward the value of their jobs may bring work engagement.
The joint effects of the psychological capital's components have been found influential on work engagement which is in line with few studies in frontline positions (Karatepe & Karadas, 2015). This is very important insights for the managers who should understand that self-efficacious, hopeful, optimistic and resilient immigrant employees will get engaged to their work who will come to their job with high energy, forget how the time passes while job going on or devote themselves to their job.
Another empirical finding is the joint effects of psychological capital components have a positive impact on work outcomes like job embeddedness (Nafei, 2015). The finding points out that well-engaged employees would be better embedded. Job embeddedness significantly presents itself through job crafting in the productivity, pro-activeness and innovative workforce of employees in an organization, which eventually leads to an increase in their general well-being (Schoberova, 2015). This empirical study can advance the current knowledge regarding immigrant employees' job crafting as how it may influence on the job embeddedness directly, and through which mechanism indirectly.

Practical implications
The managers in the hospitality industry of North Cyprus need to show maximum attention to the job crafting practices regarding immigrant employees. Such supportive proactive strategies when they are provided help immigrant employee's psychological capital and work engagement increase. In return, high levels of psychological capital may result in engagement and the employees become embedded which results in extra-role performance (Chaurasia & Shukla, 2014), service performance (Sun, Zhao, Yang, & Fan, 2012) creative performance (Sweetman, Luthans, Avey, & Luthans, 2011), and job satisfaction (Bitmiş & Ergeneli, 2013). The immigrant employees with high levels of optimism, hope, self-efficacy, and resilience should be preferred at the interview and promotion stage in the selection processes. We can indicate that in the hospitality industry in North Cyprus, the management of hospitality institutions may keep, retain, engage and embed such employees via offering job crafting opportunities to the employees.
The findings of this study show that job crafting within hospitality firms in North Cyprus does not only gives a signal to an immigrant employee about caring his/her value, but also it will enrich the job descriptions that are significant in hiring immigrant employees in North Cyprus hospitality industry. Thus, by using job crafting which enables extending the task, relationship and cognition boundaries of employees, hospitality managers in North Cyprus may strengthen their retention policies and so lessen the turnover rates. Turnover rate is a crucial problem in the hospitality industry of North Cyprus (Nadiri & Tanova, 2010) and reducing the gap between the employees expectations and industry's negative image such as low salary payment, anti-social working hours, high task responsibilities and stress, low quality employee food and shelter, limited training and development opportunities, long career schedule or other negative issues.
The joint implementation of three job crafting strategies assists immigrant employees to enhance their psychological capital and engagement to their organization which may lead to job embeddedness. Thus, hotel managers in the hospitality industry in North Cyprus with stiff domestic and international competition (Nadiri & Tanova, 2010) should be aware of immigrant employees' job crafting behaviours that will allow them to improve more embedded immigrant employees in their jobs through PsyCap and work engagement which may assist them in increment their competitiveness.

Limitations and future research suggestions
This study investigates the mediating roles of PsyCap and work engagement in the relationship between job crafting and employees' on the job embeddedness among immigrant hotel employees in North Cyprus. North Cyprus as a developing country has become an emerging new market for European tourists (Roudi et al., 2018) and today, its economy is highly dependent on the tourism industry (Nadiri & Tanova, 2010). Similar to others, this empirical study has its limitations. First, to reduce the effect of common method and single source biases, this study might have used different sources for data collection. Thus, for further generalizations, a larger sample size and multiple resources of data in other geographies is needed. It is suggested that other researchers conduct similar studies in different countries and other service sectors such as restaurants, airlines, and travel agencies, and/or in other industries (i.e. banking, hospital industry) to minimize the common method bias and to provide better evidence regarding the causality issue as well as the basis for conducting a crossnational study. Examining the study model through cross-national samples like Southern Cyprus, Greece, Italy, and Spain, would increase knowledge and understanding about the effects of job crafting in the hospitality industry in Mediterranean countries.
Second, other direction for future research includes a serial mediation analysis of job crafting and employees' on the job embeddedness link via a causal chain of PsyCap and work engagement from the other employees' perspective, such as seasonal employees and part-time employees. The potential distinct of further exploration outcomes enable considering the effects of job crafting with the findings of this present study and enlarging the academic understanding of the hypothesized relationships.
Lastly, it would also be useful to examine the consequences of three dimensions of job crafting on other employee-level outcomes, such as creativity, service, innovative behaviour, organizational citizenship behaviour, and extra-role performance in the hotel industry. Hence, this study recommended further investigation in this avenue.

Disclosure statement
No potential conflict of interest was reported by the authors.