Does Job Embeddedness Mediate the Effect of Resilience on Cabin Attendants’ Career Satisfaction and Creative Performance?

Abstract

An increasing body of research suggests job embeddedness (JE) as a motivational variable influencing employees’ attitudinal and behavioral outcomes such as quitting intentions and task performance. Personal resources have been reported to affect JE and these outcomes. However, little work has investigated the antecedents and consequences of JE among cabin attendants. There is also a dearth of empirical research regarding the mechanism linking resilience to cabin attendants’ affective and performance outcomes. Therefore, drawing on conservation of resources and JE theories, we propose a conceptual model that examines the interrelationships of resilience, JE, career satisfaction (CSAT), and creative performance (CPERF). Moreover, the model explores JE as a mediator of the impact of resilience on CSAT and CPERF. These linkages were tested via data collected from cabin attendants and their pursers. The findings from structural equation modeling reveal that resilience boosts cabin attendants’ JE, CSAT, and CPERF. As predicted, JE is a mediator between resilience and CSAT. Our paper culminates with implications for theory and practice as well as future research directions.

1. Introduction

Cabin attendants have frequent interactions with passengers, do emotional labor, and are among the most important actors in the service delivery process [1]. They ensure happiness among passengers and accomplish passenger satisfaction [2]. Cabin attendants also help the company maintain long-term relationships with passengers and achieve passenger loyalty. However, they suffer from the behaviors of aggressive passengers, boreout, work–family conflict, role stress, and burnout e.g., [1,2,3]. Such cabin attendants are expected to manage problems arising from stressors and strain. With this realization, airline companies should establish an environment in which cabin attendants can take advantage of their personal resources, display proclivity to stay with the employer, become satisfied with their career, and contribute to organizational performance via their novel ideas and feedback.

Personal resources are positive self-evaluations, represent employees’ sense of ability that enables them to control the work environment effectively, and play a critical role in their functioning in the workplace [4,5]. Resilience is one of these personal resources, e.g., [6,7]. Resilience is fundamental to understanding how employees adapt themselves to the work environment in the face of adversity [8]. Luthans [9] defines resilience as a “positive psychological capacity to rebound, to ‘bounce back’ from adversity, uncertainty, conflict, failure, or even positive change, progress and increased responsibility” (p. 702). Employees high on resilience can cope with difficulties associated with stressors and strain in the workplace [6]. Resilient employees can use their positive emotions to handle stressful situations [10]. However, workplace resilience or such a critical personal resource has been subjected to limited empirical inquiry [11].

Resilient employees are likely to exhibit higher job embeddedness (JE), which is a retention strategy [12]. JE is composed of three dimensions: “links”, “fit”, and “sacrifice.” Links are defined as “formal or informal connections between a person and institutions or other people” and fit refers to “an employee’s perceived compatibility or comfort with an organization and with his or her environment” [12], p. 1105. Sacrifice refers to “the perceived cost of material or psychological benefits that may be forfeited by leaving a job” [12], p. 1105. There are empirical studies that have demonstrated a positive association between resilience and job outcomes such as task performance, absenteeism, and job satisfaction [13,14].

1.1. Purpose

Against the above backdrop, our paper proposes a conceptual model that examines JE as a mediator of the impact of resilience on career satisfaction (CSAT) and creative performance (CPERF). Specifically, our paper tests: (a) the association between resilience and JE and the above-mentioned work-related consequences; (b) the impact of JE on CSAT and CPERF; and (c) JE as a mediator in these linkages. CPERF highlights cabin attendants’ helpful suggestions for service improvement and novel ideas for various passenger problems and complaints [15], while CSAT refers to “personal satisfaction with various aspects of career progress and success” [16], p. 283. Both CSAT and CPERF are among the crucial affective and performance consequences in frontline service jobs [17,18,19,20]. We use data collected from cabin attendants in Iran to gauge these direct and mediating effects.

1.2. Relevance and Significance of the Empirical Study

Our paper makes several important contributions to the literature. First, JE was first introduced in Mitchell et al.’s [12] study. It has been investigated in the fields of marketing, organizational behavior, air transport management, hospitality and tourism management, and human resource management, e.g., [15,21,22,23]. Studies illustrate that employees who are high on JE have proclivity to stay with the employer, display better task performance, and are service recovery performers at work, e.g., [24,25,26,27]. Since the turnover of cabin attendants is more than that of other professionals [28], examining the antecedents and outcomes of JE in an industry where cabin attendants display (in)voluntary turnover is important. However, evidence about the factors enhancing JE in the general service literature is still scarce [29,30,31]. Recognizing the gap in the literature, our paper uses resilience as a critical personal resource that can foster cabin attendants’ JE, CSAT, and CPERF. The importance of resilience is highlighted in Cooke et al.’s [7] study: “few firms see resilience as a set of skills and attitudinal qualities that can be developed proactively as part of strategic HRM to improve individual and organizational performance and well-being” (p. 1240).

Second, the general service literature presents studies about the mediating role of JE. However, to the best of our knowledge, this is the first study using JE as a mediator of the effect of resilience on crucial consequences such as CSAT and CPERF. Testing these interrelationships among cabin attendants is relevant and significant because they play a vital role in the provision of quality services to passengers and the achievement of passenger satisfaction and loyalty, e.g., [1,2,15]. Third, the findings of our paper provide useful implications for managerial practice about retention of cabin attendants who can enhance the service delivery process and contribute to organizational performance.

2. Theoretical Background, Hypotheses, and Research Model

2.1. Background

A comprehensive search made in the existing knowledge base reveals that the overwhelming majority of studies have examined the potential consequences of JE among service workers. Yet, what causes these employees to be embedded in their jobs has received little attention. More importantly, the relationship of personal resources to JE has been largely unexplored. For instance, Karatepe and Ngeche’s [32] work in Cameroon indicated that JE mediated the impact of employee engagement on task performance and quitting intentions. Akgunduz and Canli [33] indicated that perceived organizational support and employee advocacy enhanced JE in Turkish hotels. Afsar and Badir’s [34] work in China illustrated that JE moderated the positive effects of person–organization fit and organizational support on extra-role performance among hotel employees. An investigation of Iranian hotel employees showed that JE increased the negative effect of organizational justice on quitting intentions [35]. Karatepe [36] reported that work overload and conflicts in the work–family nexus reduced hotel employees’ JE via emotional exhaustion in Romania. Robinson et al.’s [37] research in Australia revealed that organizational sacrifice and community links stimulated organizational commitment, while they diminished quitting intentions. A study of hotel employees in Portugal illustrated that job satisfaction and JE acted as the mediators of the effects of different task characteristics on quitting intentions [29]. In their qualitative study conducted in Australia, Yam et al. [38] found that community attachment appeared to be an important factor triggering intent to stay with the company. Afsar et al.’s [39] study disclosed that JE mediated the impacts of high-performance work systems and trust in supervisor on quitting intentions among hotel employees in Thailand. In a study of hotel employees in Northern Cyprus, it was reported that organizational justice functioned as a mediator of the impact of favoritism on JE [40].

A study conducted with hotel employees in Malaysia illustrated that JE mediated the effect of felt obligation on proactive customer service performance, while the level of control over work hours activated JE [41]. Safavi and Karatepe’s [27] work in Iran documented that JE mediated the impact of job insecurity on quitting intentions and service recovery performance. In a study carried out in China, Lyu and Zhu [42] demonstrated that workplace ostracism eroded hotel employees’ JE, while intrinsic motivation enhanced it. They further demonstrated that JE positively influenced organizational commitment and diminished quitting intentions. Chen and Ayoun [43] disclosed that supervisor support and coworker socialization increased JE among restaurant employees. A recent study carried out in the hotel industry in South Korea reported that fit and sacrifice components of JE were positively linked to job satisfaction [44]. Yu et al. [45] also found that links and fits components of JE, coupled with work engagement, reduced hotel employees’ quitting intentions in South Korea.

Another recent study done in restaurants revealed that coworker support was negatively linked to quitting intentions through JE [30]. In addition, a study in Ghana indicated that job satisfaction and work engagement mediated the impacts of links, fit, and sacrifice as the components of JE on organizational commitment among hotel employees [24]. An investigation of Chinese hotel employees showed that JE mediated the effects of organizational inducements on work engagement [46]. A study conducted with small tourism entrepreneurs in China showed that place attachment functioned as a mediator of the impact of JE on external and internal corporate social responsibility [47]. Jolly and Self [48] found evidence in the restaurant industry that psychological diversity climate was positively associated with JE, while only sacrifice was negatively related to quitting intentions. A study carried out in restaurants in Iran documented that JE mediated the effects of benevolence, moral, and authoritarian leadership on creativity [26].

An empirical study done in Iran showed that JE mediated the impacts of high-performance work systems on cabin attendants’ creative and extra-role performances [15]. Shehawy et al.’s [49] research in the aviation industry in Egypt documented that supervisor support and employee advocacy enhanced JE, while JE increased organizational commitment and mitigated quitting intentions.

The extant literature also delineates studies concerning the critical role of resilience as a personal resource. For example, Wang et al. [11] demonstrated that resilience positively affected work engagement directly and indirectly through positive affect among information technology employees in China. A study carried out among employees in South Korea who experienced mergers and acquisitions indicated that resilience fostered relational psychological contract, while it reduced transactional psychological contract [50]. A study of Chinese public servers showed that resilience enhanced leader and team member exchanges and organizational commitment, while it mitigated burnout [51]. The findings of another study showed that resilience mediated the linkage between learning organization and work engagement among information technology professionals in India [52]. Furthermore, it was reported that resilience diminished quitting intentions and triggered work engagement among travel agency employees in Taiwan [53].

An overall evaluation of the JE-related studies suggests that JE enhances a number of various outcomes such as task performance, organizational commitment, and creative and extra-role performances, while it mitigates quitting intentions. Various different task characteristics, organizational support, employee advocacy, intrinsic motivation, supervisor support, and coworker socialization foster JE, while ostracism, work overload, and conflicts in the work–family interface erode JE.

An observation made based on the resilience-related studies suggests that resilience boosts work engagement, positive affect, organizational commitment, and relational psychological contract, while it alleviates quitting intentions, burnout, and transactional psychological contract.

Based on the empirical pieces given above, it seems that the relationship of JE to personal resources such as resilience has been largely unexplored. In addition, no empirical study has examined JE as a mediator of the impact of resilience on CSAT and CPERF among cabin attendants so far.

2.2. Hypotheses

2.2.1. Direct Effects

Conservation of resources (COR) theory can be used to develop the hypothesis regarding the effect of resilience on JE. Specifically, this theory advocates that individuals strive to obtain and protect their resources such as objects and energy [54]. When individuals acquire resources, they may invest them to acquire the new ones [55]. Resources (personal resources) tend to generate other resources (JE), resulting in resource caravans (resilience and JE together) [54,56]. Resilient cabin attendants are likely to know that they cannot move their links to other individuals in the company with them to the new work environment as a result of voluntary turnover or they may lose the perceived fit in the new work environment [55]. For instance, these cabin attendants cannot take their good connections with their colleagues to the new airline company or may not find a good fit between their values and career goals and the demands of their new jobs as well as the new airline company’s culture. This is not surprising because JE resources are restricted to the company and/or the position.

Although limited, there is evidence showing that personal resources enhance JE. For example, Sun et al. [57] found that psychological capital (i.e., self-efficacy, hope, resilience, and optimism) fostered nurses’ JE in the Chinese healthcare settings. Lev and Koslowsky [58] reported that conscientiousness was positively related to Israeli teachers’ JE. Singh’s [31] study in Trinidad illustrated that emotional stability, conscientiousness, and extraversion boosted service workers’ JE. Therefore, the following hypothesis is postulated:

Hypothesis 1.

Cabin attendants’ resilience relates positively to their JE.

As propounded by COR theory, the presence of sufficient personal resources contributes to employees’ well-being [59]. Within this theory, resilience is an important personal resource that can foster employees’ well-being [60]. As stated in Alarcon et al.’s [59] study, employees high on resilience “can thus be thought of as having a vast reservoir of resources from which to draw” (p. 822). Accordingly, we surmise that cabin attendants who can really adapt to challenging situations even when faced with adversity are likely to be satisfied with their career and contribute to the company through their CPERF.

In empirical terms, Avey et al. [10] reported a positive association between resilience and work-related performance. Karatepe and Karadas [61] found that hotel employees’ psychological capital in Romania boosted their CSAT. Lehoczky [62] reported a positive association between psychological capital and CSAT. On the other hand, Ngo et al.’s [63] research in China did not support the positive linkage between resilience and CSAT. Under the umbrella of COR theory and the above-mentioned findings, we derive the relevant hypotheses as follows:

Hypothesis 2.

Cabin attendants’ resilience relates positively to their CSAT.

Hypothesis 3.

Cabin attendants’ resilience relates positively to their CPERF.

JE is a retention strategy that enhances employees’ CSAT and fosters their CPERF. These associations can be developed based on JE theory [12]. Specifically, cabin attendants high on JE possess quality connections with their coworkers and pursers [15]. They feel there is a good fit between their values and knowledge and the requirements of the job as well as organizational values [12]. The ones with high JE know that they will lose various benefits if they exhibit voluntary turnover [32]. In a workplace where cabin attendants are aware of the critical role of links, fit, and sacrifice, they display higher CSAT, which is one of the indicators of subjective career success [64]. These employees also exhibit better CPERF, which enables them to share their new ideas for service improvement and present novel solutions for problems inherent in service encounters.

Dechawatanapaisal’s [65] study in Thailand indicated that JE enhanced accountants’ CSAT. Similarly, the results based on various samples used in Belgium highlighted a positive linkage between JE and CSAT [64]. Karatepe and Vatankhah [15] demonstrated that JE bolstered cabin attendants’ CPERF. Shah et al.’s [66] research conducted with academicians in Pakistan showed that JE was positively associated with CPERF. Accordingly, we hypothesize:

Hypothesis 4.

Cabin attendants’ JE relates positively to their CSAT.

Hypothesis 5.

Cabin attendants’ JE relates positively to their CPERF.

2.2.2. Mediating Effects

Hypotheses 1–5 implicitly suggest JE as a mediator of the impact of resilience on CSAT and CPERF. The resilience → JE → CSAT and CPERF linkages can be developed in light of JE theory. Specifically, when cabin attendants are resilient to setbacks in the workplace and handle difficulties despite the disappointing outcomes, they are enmeshed in their jobs. This is due to the fact that they feel they fit well with the demands of the job and the company, possess quality relations with their colleagues and pursers within the company, and are aware of the potential loss of benefits when they leave the company cf. [12,36,57]. Such employees in turn are satisfied with different aspects of their career and contribute to the company via their novel ideas and fresh perspectives for service improvement. As Mitchell et al. [12] have stated, JE is a key construct linking on-the-job factors or resources to employees’ attitudinal and performance consequences. This is also supported by Holtom and Inderrieden [67], who showed that JE is a key mediating mechanism relating individual attitudes to employee outcomes.

There is little evidence showing that personal resources are linked to employee outcomes through JE. For example, Singh’s [31] research in different service contexts (e.g., hospitality and financial services) has showed that emotional stability, conscientiousness, extraversion, and networking boost JE, which in turn diminishes quitting intentions. Evidence has revealed that JE mediates the effects of personal resources such as psychological capital on nurses’ job performance [57]. Hence, we hypothesize:

Hypothesis 6.

JE mediates the impact of resilience of CSAT.

Hypothesis 7.

JE mediates the impact of resilience on CPERF.

2.3. Research Model

Figure 1 delineates the hypothesized linkages in the conceptual model. Cabin attendants with high resilience are more embedded in the job, have higher satisfaction with their career, and display CPERF at elevated levels. Cabin attendants who possess formal and informal connections with people in and outside the company, fit well into their jobs, and know that they will sacrifice valued things if they quit are more satisfied with their career and exhibit high levels of CPERF. The information given above implicitly demonstrates that JE mediates the impact of resilience on CSAT and CPERF. Gender and marital status are used as controls in our paper. This is because of the fact that these variables may be significantly associated with the study constructs and act as confounding variables, e.g., [7,15,34]. This raises the need for determining whether the significance of the findings remains the same with or without the control variables.

3. Method

3.1. Sample and Data Collection

Our study used data gathered from a judgmental sample of cabin attendants to assess the previously mentioned hypothesized associations. There are at least two criteria for choosing cabin attendants in the present study. First, cabin attendants spend so much time with passengers and are beset with work overload and passenger verbal aggression, and emotional dissonance [2]. Second, they are expected to manage various passenger demands and problems and respond to them in light of the company’s standards [15]. This makes resilience a viable personal resource in frontline service jobs.

After contacting the Iran Civil Aviation Organization at the time of the study, the researcher learnt that 13 private airline companies had international and/or domestic flights. However, only three airline companies agreed to take part in the study after the researcher had contacted these companies via a letter. In short, the participants were cabin attendants of three private airline companies having domestic and/or international flights in Iran.

The cabin attendant and purser surveys were prepared in light of the back-translation method. The surveys were originally prepared in English and then translated into Persian via the back-translation technique. Each survey at Time 1, Time 2, and Time 3 was assessed with a pilot sample of 5 cabin attendants, while the purser survey was tested with a pilot sample of 5 pursers. No amendments in the surveys were needed based on these pilot tests.

Data were gathered from cabin attendants with a time lag of two weeks in three waves and their pursers. Collecting data in different waves enables the researcher to separate the predictor variables from the criterion variables [68]. This is consonant with the work of Karatepe and Vatankhah [15]. Each participant completed the Time 1 (resilience and items about the participants’ profile), Time 2 (JE), and Time 3 (CSAT) surveys in the course of their briefing time in the central building of their company. The pursers rated cabin attendants’ CPERF via the purser survey. Identification codes were utilized to match the surveys.

The Shapiro–Wilk test was employed to check the normality of the data. The results illustrated that p-values were >0.05, presenting evidence of the normality of the data [69]. The participants’ profile is given in

Table 1. Participants’ profile (n = 121).

	No. of Respondents	%
Age
18–27	36	29.8
28–37	68	56.2
38–47	15	12.4
48–57	2	1.6
Gender
Male	48	39.7
Female	73	60.3
Education
Secondary and high school	3	2.4
Two-year college degree	18	14.9
Four-year college degree	70	57.9
Graduate degree	30	24.8
Organizational tenure
Less than 1 year	10	8.3
1–5	48	39.7
6–10	42	34.7
11–15	18	14.9
16–20	1	0.8
Longer than 20 years	2	1.6
Marital status
Single or divorced	83	68.6
Married	38	31.4

.

3.2. Measurement

Well-established scale items were utilized to operationalize the study variables. Resilience was operationalized via six items from Luthans et al. [70]. A number of studies assessed employees’ resilience utilizing the same items [71,72]. The participants responded to the items on a 6-point scale (“6 = strong agree”, “1 = strongly disagree”). To assess JE, we utilized seven items from Crossley et al. [73]. This scale captures both on-the-JE and off-the-JE. Many studies tapped this scale to gauge JE, e.g., [22,74]. Using the recommendation given by Crossley et al. [73], the following instruction was given before the JE items were asked in the survey: “After considering both work-related (such as relationships, fit with job, benefits) and nonwork-related factors (such as neighbors, hobbies, community perks), please rate your agreement with the statements below.” The participants answered the JE items via a 5-point scale (“5 = strongly agree”, “1 = strongly disagree”). CSAT was assessed with five items from Greenhaus et al. [75]. Recent empirical studies also used this scale to assess CSAT [76,77]. The CSAT scale was measured utilizing a 5-point scale, anchored by “5 = strongly agree” and “1 = strongly disagree.” CPERF was operationalized with six items adapted from Wang and Netemeyer [78]. The same items were also used in different papers [15,20]. Items were scored on a 5-point scale, anchored by “almost always” and “never.”

3.3. Common Method Variance

We utilized several procedural remedies to minimize the threat of common method variance [36,68]. First, the information in the cover page was as follows: “There are no right or wrong answers in this questionnaire. Any sort of information collected during our research will be kept confidential. Participation is voluntary but encouraged. Management of your company fully endorses participation.” In addition, the participants were given assurance of the use of data for academic purposes. Second, anonymity was guaranteed. Third, each cabin attendant sealed the Time 1 survey in an envelope and put it in a designated box. This was repeated in all waves and for the purser surveys.

Fourth, Podsakoff et al. [68] have recommended that if the researchers are unable to collect multiple sources of data, they can rely on time-lagged data. We used both of them due to the rigor of the method. That is, data were gathered from cabin attendants utilizing a time lag of two weeks between each wave and their pursers for the assessment of CPERF. In short, the simultaneous utilization of time-lagged and multiple sources of data makes the research rigorous [79].

3.4. Data Analysis

Covariance matrix was tapped in LISREL (Linear Structural Equations) 8.30, which is regarded as the flagship of structural equation modeling (SEM) [80,81]. To test our analytical model, we followed the two-step modeling approach [82]. First, all the measures were subjected to confirmatory factor analysis (CFA) at the same time to provide evidence of validity and reliability [83]. That is, items for resilience, JE, CSAT, and CPERF were subjected to CFA simultaneously. Second, we gauged the hypothesized linkages through SEM, which explains “the pattern of a series of inter-related dependence relationships simultaneously between a set of latent (unobserved) constructs, each measured by one or more manifest (observed) variables” [84]. The maximum likelihood estimation method was performed.

The Sobel test was performed to test the mediating impacts. Baron and Kenny’s [85] four conditions for the mediation analysis were followed. The hypothesized model was compared with the fully mediated model via the Δχ² test to ferret out which model had a better fit to the data. We used a one-tailed test for the test of the hypothesized linkages (t > 1.65, p < 0.05 and t > 2.33, p < 0.01) [36]. Consistent with the works of, for example, Karatepe [36] and Aboramadan et al. [86], we tapped the following fit statistics: “χ²/df, comparative fit index (CFI), parsimony normed fit index (PNFI), standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA).”

We calculated a priori sample size requirement to detect the effect via SEM. The minimum sample size for detection of the effect was 51 (“anticipated effect size” = 0.5 large effect size; “desired statistical power level” = 0.95; “number of latent variables” = 4 (resilience, JE, CSAT, and CPERF); “number of observed variables” = 26 (six items each for resilience and CPERF, seven items for JE, five items for CSAT, gender, marital status; and “probability level” = 0.05). The sample size in our study (n = 121) was greater than the above-mentioned requirement [87].

The overall research process, which includes the survey design, common method variance check, sampling technique, pilot study, data collection, normality test, test of the measurement model, a priori sample size requirement for SEM, and assessment of the hypothesized model, is summarized in Figure 2.

4. Results

4.1. Measurement Model

The findings from CFA (

Table 2. Confirmatory factor analysis results.

Constructs and Items	M	SD	Standardized Loading	t-Value
Resilience (AVE = 0.46; CR = 0.80; α = 0.77)
When I have a setback at work, I have trouble recovering from moving on (-) *	-	-	-	-
I usually manage difficulties one way or another at work	4.48	1.05	0.71	8.45
Copyrighted item	4.69	1.26	0.77	9.50
Copyrighted item	3.89	1.27	0.45	4.92
Copyrighted item	4.54	1.35	0.58	6.40
Copyrighted item	4.66	1.16	0.80	10.20
Job embeddedness (AVE = 0.65; CR = 0.93; α = 0.92)
I feel attached to this airline company	3.85	0.95	0.80	10.44
It would be difficult for me to leave this airline company	3.85	1.03	0.88	12.29
I am too caught up in this airline company to leave	3.66	1.17	0.90	12.51
I feel tied to this airline company	3.61	1.17	0.88	12.35
I simply could not leave the airline company that I work for	3.72	1.20	0.86	11.83
It would be easy for me to leave this airline company (-)	3.53	1.29	0.52	5.93
I am tightly connected to this organization	3.41	1.24	0.72	9.08
Career satisfaction (AVE = 0.54; CR = 0.85; α = 0.83)
I am satisfied with the success I have achieved in my career	3.95	0.92	0.77	9.58
I am satisfied with the progress I have made toward meeting my overall career goals	3.80	0.89	0.81	10.32
I am satisfied with the progress I have made toward meeting my goals for income	3.35	1.15	0.55	6.28
I am satisfied with the progress I have made toward meeting my goals for advancement	3.79	0.97	0.82	10.21
I am satisfied with the progress I have made toward meeting my goals for the development of new skills	3.82	0.99	0.70	8.56
Creative performance (AVE = 0.74; CR = 0.95; α = 0.94)
This flight attendant carries out his/her routine tasks in ways that are resourceful	3.80	1.08	0.82	11.15
This flight attendant comes up with new ideas for satisfying passenger needs	3.75	0.97	0.88	12.40
This flight attendant generates and evaluates multiple alternatives for novel passenger problems	3.50	1.03	0.86	11.76
This flight attendant has fresh perspectives on old problems	3.52	1.13	0.82	11.00
This flight attendant improvises methods for solving a problem when an answer is not apparent	3.36	1.18	0.82	10.92
This flight attendant generates creative ideas for service delivery	3.71	1.08	0.96	14.40

Notes: All loadings are significant at the 0.01 level. AVE = Average variance extracted; CR = Composite reliability; α = Coefficient alpha. M = Mean; SD = Standard deviation. * Dropped during confirmatory factor analysis. (-) denotes reverse-coded items.

) resulted in deletion of one nonsignificant loading from the resilience measure. Deletion of this item resulted in the following fit statistics: χ² = 329.26, df = 217, χ²/df = 1.52, CFI = 0.94, PNF = 0.72, SRMR = 0.072, and RMSEA = 0.066. Though the χ²/df was not significant, the values of both SRMR and RMSEA were less than the cutoff level of 0.08, while the value of CFI was greater than the 0.90 threshold, while the value of PNFI was greater than 0.50. Overall, the four-factor measurement model fit the data well [36,88].

Most of the loadings were greater than 0.70. The average variance extracted (AVE) by resilience, JE, CSAT, and CPERF was 0.46, 0.65, 0.54, and 74, respectively. Composite reliability for resilience, JE, CSAT, and CPERF was 0.80, 0.93, 0.85, and 0.95, respectively. Although the AVE by resilience was slightly lower than 0.50, its loadings were significant and its composite reliability was 0.80 (>0.60) [89]. In addition, the AVE is quite a conservative test, and the AVE by a variable can be less than 0.50 though its composite reliability score is deemed acceptable [90]. Based on model fit statistics, significant loadings, and composite reliabilities greater than 0.70, convergent validity was verified [82,83].

The correlation matrix for latent variables was presented in

Table 3. Correlation matrix for latent variables.

Variables	RES	JE	CSAT	CPERF
Resilience (RES)	0.68
Job embeddedness (JE)	0.20	0.81
Career satisfaction (CSAT)	0.33	0.53	0.74
Creative performance (CPERF)	0.54	0.14	0.36	0.86

Notes: The √ of the average variance extracted by each latent construct was shown on the diagonal.

. The √ of the AVE was given on the diagonal in Table 3. Discriminant validity was verified since the √ of each AVE exceeded the respective correlation between variables shown in Table 3 [64]. In addition to composite reliabilities mentioned above, coefficient alpha for each observed construct was greater than 0.70 (Table 2). Overall, these findings indicated that the measures were reliable [20,89]. The mean and standard deviation for each observed item are given in Table 2. The means, standard deviations, and correlations of observed constructs are given in

Table 4. Descriptive statistics and correlations of observed variables.

Constructs	Mean	SD	Gender	MS	RES	JE	CSAT	CPERF
Gender	0.60	0.49
Marital status (MS)	0.31	0.47	−0.106	-
Resilience (RES)	4.45	0.88	0.043	0.038	-
Job embeddedness (JE)	3.66	0.94	−0.004	−0.020	0.243 **	-
Career satisfaction (CSAT)	3.74	0.77	−0.039	0.056	0.312 **	0.496 **	-
Creative performance (CPERF)	3.61	0.94	0.165 *	−0.024	0.490 **	0.156 *	0.370 **	-

Notes: SD = Standard deviation. Gender was coded as a binary variable (0 = male and 1 = female). Marital status was also coded as a binary variable (0 = single or divorced and 1 = married). * p < 0.05 and ** p < 0.01 (one-tailed test).

.

4.2. Structural Model Test Results

To assess JE as a mediator of the effect of resilience on CSAT and CPERF, we utilized the following conditions for the mediation test based on the guidelines given by Baron and Kenny [85] and the work of Chen et al. [91]: (1) the predictor variable should be significantly related to the mediator; (2) the predictor variable should have a significant association with the criterion variables; (3) the mediator should be significantly associated with the criterion variables; and (4) the predictor variable should have no significant association with the criterion variables when the mediator is controlled (fully mediated model) or the predictor variable should be significantly linked to the criterion variables when the mediator is controlled (partially mediated model). To gauge the last condition, the partially mediated model was compared with the fully mediated model [91].

The results in Table 4 show that the first three conditions were met. Specifically, resilience had a positive association with JE (r = 0.243, p < 0.01), CSAT (r = 0.312, p < 0.01), and CPERF (r = 0.490, p < 0.01). JE was also positively related to CSAT (r = 0.496, p < 0.01), and CPERF (r = 0.156, p < 0.05). For the last condition, the partially mediated model (χ² = 359.19, df = 256) was compared with the fully mediated model (χ² = 389.34, df = 258) using the chi-square difference test. The result was significant (Δχ² = 30.15, Δdf = 2, p < 0.01).

In the fully mediated model, resilience was found to be significantly and positively related to JE, while JE had a significant positive association with CSAT. However, the direct path from resilience to CSAT and CPERF was not freed. The findings in the fully mediated model did not improve fit statistics. For example, CFI was 0.92, while SRMR was 0.13.

The partially mediated model fit the data well: χ² = 359.19, df = 256, χ²/df = 1.40, CFI = 0.94, PNF = 0.71, SRMR = 0.077, and RMSEA = 0.058. The findings from SEM were presented in

Table 5. Structural model test results.

Hypotheses	Path Estimate	t-Value	Supported/Not Supported
H1 Resilience → Job embeddedness (β21)	0.20	1.97	Supported
H2 Resilience → Career satisfaction (β31)	0.26	2.85	Supported
H3 Resilience → Creative performance (β41)	0.54	5.19	Supported
H4 Job embeddedness → Career satisfaction (β32)	0.48	4.89	Supported
H5 Job embeddedness → Creative performance	0.04	0.47	Not supported
	z-Value
H6 Resilience → Job embeddedness → Career satisfaction	1.88		Supported (partial mediation)
H7 Resilience → Job embeddedness → Creative performance	-		Not supported
Gender → Creative performance (γ41)	0.16 (t-value = 2.01)

Notes: t-values: one-tailed test t > 1.65, p < 0.05; and t > 2.33, p < 0.01.

. Hypothesis 1 suggested that resilience enhances cabin attendants’ JE. This hypothesis received support since resilience positively influenced JE (β₂₁ = 0.20, t = 1.97). In short, resilience is a personal resource that increases cabin attendants’ JE.

Hypothesis 2 predicted that resilience fosters cabin attendants’ CSAT, while hypothesis 3 proposed that resilience activates their CPERF. The findings were in support of both hypotheses 2 and 3. This was due to the fact that resilience had a positive impact on CSAT (β₃₁ = 0.26, t = 2.85) and CPERF (β₄₁ = 0.54, t = 5.19). Resilient cabin attendants are satisfied with different aspects of their career and contribute to the airline company by making novel suggestions about improvement in service delivery.

Hypothesis 4 suggested that JE enhances cabin attendants’ CSAT, while hypothesis 5 proposed that JE fosters their CPERF. The findings showed that hypothesis 4 was supported because JE positively affected CSAT (β₃₂ = 0.48, t = 4.89). However, hypothesis 5 was not supported since JE did not significantly affect CPERF (β₄₂ = 0.04, t = 0.47). Cabin attendants who are embedded in their jobs exhibit satisfaction with their career in the present airline company.

The Sobel test results indicated that the indirect impact of resilience on CSAT through JE was significant (z = 1.88). Hence, hypothesis 6 was supported. Resilient cabin attendants are highly embedded in their jobs and therefore exhibit CSAT at high levels. Hypothesis 7 cannot be supported because JE did not exert a significant impact on CPERF.

In addition, female employees reported higher levels of CPERF. The findings accounted for 4% of the variance in JE, 35% in CSAT, and 33% in CPERF. When the control variables were excluded from the analysis, the results did not demonstrate any changes concerning the significance of the impacts tested.

5. Discussion

5.1. Key Findings

The overriding goals of our paper were to assess (1) the effect of resilience on JE, CSAT, and CPERF, (2) the impact of JE on CSAT and CPERF, and (3) JE as a mediator in these associations. Unlike most of the other studies, data gathered from cabin attendants in three waves and their pursers in Iran were utilized to evaluate the hypotheses. The implications of these results are discussed below.

First, the research findings suggest that resilience enhances cabin attendants’ JE. Resilience is a viable personal resource and a theoretically relevant variable. Consonant with COR theory, resilient employees accumulate their resources, resulting in resource caravans [56]. Resilient employees do not show willingness to leave the company because they are aware of the fact that they cannot take their quality connections with their coworkers in the organization and pursers to the new company and may sacrifice a number of benefits, which may not be offered to them in the new workplace; cf. [55]. With this realization, these employees accumulate their personal resources and the ones arising from JE to have new additional resource gains. In short, employees who are high on resilience fit well with the demands of the job in the aviation industry, develop and have quality connections with their coworkers and pursers, and know what benefits would be sacrificed easily as a result of quitting.

Second, the present study empirically documents that cabin attendants high on resilience exhibit higher CSAT in the organization. This is consonant with COR theory that employees’ well-being is triggered by their personal resources [59]. Further, our study empirically shows that cabin attendants with high resilience exhibit higher CPERF. Resilient cabin attendants can adapt to challenging service encounters despite the presence of obstacles and bounce back from failures. Such employees are accustomed to these difficult situations, gain satisfaction from the work of serving passengers, and want to overcome difficulties associated with their personal and organizational goals. Under these circumstances, they display higher CSAT and CPERF.

Third, the research findings suggest that JE enhances cabin attendants’ CSAT. This is in line with JE theory [12] and recent evidence [65]. Cabin attendants are satisfied with different aspects of their career since they perceive that they have good links to other people and activities within the company and fit well with the organization. In addition, they are aware of the consequential things they would lose if they left the company.

Fourth, JE is a key variable linking resilience to CSAT. JE partly mediates the impact of resilience on CSAT. This lends support to the works of Afsar et al. [39] and Singh [31] and is congruent with JE theory [67]. On the other hand, the findings do not empirically support JE as a mediator of the influence of resilience on CPERF. This is because of the fact that JE does not significantly influence CPERF. This finding does not corroborate the other studies which have reported that JE is a predictor of performance-related variables, e.g., [15]. It seems that cabin attendants do not need to feel attached to the company or take advantage of links, fit, or benefits to make constructive suggestions for service improvement and share novel ideas for a better service delivery process. Their resiliency appears to be a key to their CPERF.

Whereas most of the empirical studies tested the consequences of JE, e.g., [27,31], our paper enhances current knowledge by reporting that resilience is an important personal resource triggering cabin attendants’ JE and CPERF and JE is a mediator of the impact of resilience on CSAT.

5.2. Implications for Practice

The results present several recommendations for managerial action. First, airline companies should conduct surveys with the current cabin attendants to ascertain the level of their resilience. In addition, specific training programs can be arranged to strengthen cabin attendants’ resilience. In these programs, mini case studies can be used to understand how resilient cabin attendants are and how their resilience can be enhanced. Since resilient cabin attendants contribute to the airline company via elevated levels of CPERF, retention of a resilient workforce is important to the accomplishment of sustainable competitive advantage in a competitive market environment where customer expectations are rising [7]. Second, management should pay utmost attention to the hiring process. That is, it is important to provide the candidates with detailed information about the requirements of the job, career opportunities within the company, and the presence of formal or informal connections with other individuals in the company. Otherwise, it would not be possible to find the right individuals whose knowledge, skills, abilities, and values fit with the requirements of the position.

Third, online workshops can be organized to encourage cabin attendants to share their novel ideas for service improvement and make suggestions in the passenger complaint-handling process. When cabin attendants high on JE perceive that management really takes into consideration their ideas and suggestions, they can help the company achieve passenger satisfaction and loyalty. Fourth, airline companies need to establish a relaxed work environment where cabin attendants do not feel that their career is not jeopardized when they voice their ideas about new solutions for old problems in the company. In any event, management should reward cabin attendants’ CPERF to achieve passenger satisfaction and loyalty. Lastly, management should establish an intranet or a private network, which can be used by cabin attendants and pursers. In this network, cabin attendants can share their ideas and provide significant feedback about how they can solve new passenger problems, increase the level of passenger satisfaction and loyalty, and contribute to the profitability of the company.

5.3. Limitations and Future Research

As with any paper, the present study has several limitations. First, we used resilience as an important personal resource affecting cabin attendants’ JE and job outcomes. Future research can incorporate other critical personal resources such as emotional intelligence and hardiness to ascertain their effects on JE, CSAT, and CPERF among cabin attendants. Second, the data utilized in our paper came from cabin attendants in Iran. Future research can focus on cross-national data (e.g., Iran, the United Kingdom, and China) to gauge the study linkages. By doing so, it would be possible to broaden the database.

Third, we utilized JE as a mediator of the impact of resilience on CSAT and CPERF due to the dearth of empirical evidence in the relevant literature. Other variables such as work engagement and work passion are also critical and can also mediate the influence of resilience on the aforesaid consequences [92,93]. With this recognition, future research can incorporate both work engagement and work passion as the mediators into the model and assess whether these mediators exceed JE in explaining the linkages between resilience and job outcomes.

Fourth, CPERF is one of the most important performance outcomes in frontline service jobs in the aviation industry [15]. Adaptive performance, which refers to individuals’ ability to adapt their behavior to different challenging service encounters [94], is a significant performance variable among cabin attendants. However, our search in the air transport management literature highlights a lack of evidence about adaptive performance. In future studies, addition of adaptive performance to the proposed model would enhance the understanding about the effects of resilience and JE on this outcome. Lastly, there may be daily fluctuations in employees’ personal resources or the willingness to display CPERF. Therefore, in future studies, the participants can be invited to complete the surveys and a diary booklet to overcome such a limitation.